Skip to content

Instantly share code, notes, and snippets.

@stas-sl

stas-sl/mixed_14

Last active April 29, 2016 06:35
Show Gist options
  • Save stas-sl/5363abc18478dc5ad148886ab85bc676 to your computer and use it in GitHub Desktop.
Save stas-sl/5363abc18478dc5ad148886ab85bc676 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
mixed_14
This file has been truncated, but you can view the full file.
I0428 23:20:07.606009 6470 solver.cpp:280] Solving mixed_lstm
I0428 23:20:07.606020 6470 solver.cpp:281] Learning Rate Policy: fixed
I0428 23:20:07.969547 6470 solver.cpp:229] Iteration 0, loss = 27.5204
I0428 23:20:07.969617 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0
I0428 23:20:07.969635 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:20:07.969650 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:20:07.969666 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:20:07.969679 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:20:07.969691 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:20:07.969703 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0
I0428 23:20:07.969715 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0
I0428 23:20:07.969727 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0
I0428 23:20:07.969738 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0
I0428 23:20:07.969749 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0
I0428 23:20:07.969760 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0
I0428 23:20:07.969771 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0
I0428 23:20:07.969782 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0
I0428 23:20:07.969794 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0
I0428 23:20:07.969806 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0
I0428 23:20:07.969817 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0
I0428 23:20:07.969830 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 0
I0428 23:20:07.969841 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 0
I0428 23:20:07.969853 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 0
I0428 23:20:07.969864 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 0
I0428 23:20:07.969877 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 0
I0428 23:20:07.969887 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 0
I0428 23:20:07.969898 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0
I0428 23:20:07.969910 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.0408163
I0428 23:20:07.969930 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 4.30912 (* 0.3 = 1.29274 loss)
I0428 23:20:07.969946 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 4.27239 (* 0.3 = 1.28172 loss)
I0428 23:20:07.969993 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 4.32027 (* 0.0272727 = 0.117826 loss)
I0428 23:20:07.970008 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 4.31317 (* 0.0272727 = 0.117632 loss)
I0428 23:20:07.970022 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 4.3468 (* 0.0272727 = 0.118549 loss)
I0428 23:20:07.970037 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 4.33026 (* 0.0272727 = 0.118098 loss)
I0428 23:20:07.970052 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 4.26225 (* 0.0272727 = 0.116243 loss)
I0428 23:20:07.970065 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 4.24146 (* 0.0272727 = 0.115676 loss)
I0428 23:20:07.970078 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 4.25958 (* 0.0272727 = 0.11617 loss)
I0428 23:20:07.970093 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 4.33531 (* 0.0272727 = 0.118236 loss)
I0428 23:20:07.970106 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 4.29808 (* 0.0272727 = 0.11722 loss)
I0428 23:20:07.970120 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 4.42314 (* 0.0272727 = 0.120631 loss)
I0428 23:20:07.970134 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 4.27542 (* 0.0272727 = 0.116602 loss)
I0428 23:20:07.970147 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 4.32215 (* 0.0272727 = 0.117877 loss)
I0428 23:20:07.970160 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 4.30477 (* 0.0272727 = 0.117403 loss)
I0428 23:20:07.970175 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 4.23641 (* 0.0272727 = 0.115539 loss)
I0428 23:20:07.970187 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 4.37085 (* 0.0272727 = 0.119205 loss)
I0428 23:20:07.970201 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 4.27191 (* 0.0272727 = 0.116507 loss)
I0428 23:20:07.970216 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 4.35655 (* 0.0272727 = 0.118815 loss)
I0428 23:20:07.970228 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 4.32555 (* 0.0272727 = 0.11797 loss)
I0428 23:20:07.970242 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 4.30402 (* 0.0272727 = 0.117382 loss)
I0428 23:20:07.970255 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 4.23968 (* 0.0272727 = 0.115628 loss)
I0428 23:20:07.970268 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 4.37734 (* 0.0272727 = 0.119382 loss)
I0428 23:20:07.970283 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 4.33401 (* 0.0272727 = 0.1182 loss)
I0428 23:20:07.970294 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0
I0428 23:20:07.970306 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:20:07.970317 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:20:07.970329 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:20:07.970340 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:20:07.970350 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0
I0428 23:20:07.970361 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0
I0428 23:20:07.970372 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0
I0428 23:20:07.970384 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0
I0428 23:20:07.970396 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0
I0428 23:20:07.970407 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0
I0428 23:20:07.970417 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0
I0428 23:20:07.970428 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0
I0428 23:20:07.970438 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0
I0428 23:20:07.970451 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.125
I0428 23:20:07.970489 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0
I0428 23:20:07.970506 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0
I0428 23:20:07.970517 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 0
I0428 23:20:07.970530 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 0
I0428 23:20:07.970541 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 0
I0428 23:20:07.970551 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 0
I0428 23:20:07.970562 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 0
I0428 23:20:07.970573 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 0
I0428 23:20:07.970585 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0
I0428 23:20:07.970597 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.0816327
I0428 23:20:07.970610 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 4.30289 (* 0.3 = 1.29087 loss)
I0428 23:20:07.970625 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 4.29786 (* 0.3 = 1.28936 loss)
I0428 23:20:07.970638 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 4.30896 (* 0.0272727 = 0.117517 loss)
I0428 23:20:07.970652 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 4.29805 (* 0.0272727 = 0.117219 loss)
I0428 23:20:07.970666 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 4.30547 (* 0.0272727 = 0.117422 loss)
I0428 23:20:07.970680 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 4.29951 (* 0.0272727 = 0.117259 loss)
I0428 23:20:07.970693 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 4.31063 (* 0.0272727 = 0.117563 loss)
I0428 23:20:07.970707 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 4.3006 (* 0.0272727 = 0.117289 loss)
I0428 23:20:07.970726 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 4.29969 (* 0.0272727 = 0.117264 loss)
I0428 23:20:07.970739 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 4.30456 (* 0.0272727 = 0.117397 loss)
I0428 23:20:07.970752 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 4.2943 (* 0.0272727 = 0.117117 loss)
I0428 23:20:07.970767 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 4.32127 (* 0.0272727 = 0.117853 loss)
I0428 23:20:07.970779 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 4.31074 (* 0.0272727 = 0.117566 loss)
I0428 23:20:07.970793 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 4.31946 (* 0.0272727 = 0.117803 loss)
I0428 23:20:07.970806 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 4.32002 (* 0.0272727 = 0.117819 loss)
I0428 23:20:07.970820 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 4.30346 (* 0.0272727 = 0.117367 loss)
I0428 23:20:07.970834 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 4.31112 (* 0.0272727 = 0.117576 loss)
I0428 23:20:07.970847 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 4.30817 (* 0.0272727 = 0.117496 loss)
I0428 23:20:07.970860 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 4.30268 (* 0.0272727 = 0.117346 loss)
I0428 23:20:07.970875 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 4.31651 (* 0.0272727 = 0.117723 loss)
I0428 23:20:07.970887 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 4.31943 (* 0.0272727 = 0.117803 loss)
I0428 23:20:07.970901 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 4.30984 (* 0.0272727 = 0.117541 loss)
I0428 23:20:07.970914 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 4.29864 (* 0.0272727 = 0.117236 loss)
I0428 23:20:07.970927 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 4.30363 (* 0.0272727 = 0.117372 loss)
I0428 23:20:07.970939 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0204082
I0428 23:20:07.970950 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:20:07.970974 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:20:07.970988 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:20:07.971000 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0428 23:20:07.971011 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0
I0428 23:20:07.971024 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0
I0428 23:20:07.971035 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0
I0428 23:20:07.971045 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0
I0428 23:20:07.971056 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0
I0428 23:20:07.971067 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0
I0428 23:20:07.971078 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0
I0428 23:20:07.971091 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0
I0428 23:20:07.971101 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0
I0428 23:20:07.971112 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0
I0428 23:20:07.971123 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0
I0428 23:20:07.971134 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0
I0428 23:20:07.971146 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 0
I0428 23:20:07.971158 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 0
I0428 23:20:07.971168 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 0
I0428 23:20:07.971179 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 0
I0428 23:20:07.971190 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 0
I0428 23:20:07.971202 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 0
I0428 23:20:07.971213 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.00568182
I0428 23:20:07.971225 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.142857
I0428 23:20:07.971238 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 4.2888 (* 1 = 4.2888 loss)
I0428 23:20:07.971252 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 4.29579 (* 1 = 4.29579 loss)
I0428 23:20:07.971266 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 4.30503 (* 0.0909091 = 0.391366 loss)
I0428 23:20:07.971279 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 4.30741 (* 0.0909091 = 0.391583 loss)
I0428 23:20:07.971293 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 4.29836 (* 0.0909091 = 0.39076 loss)
I0428 23:20:07.971307 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 4.30572 (* 0.0909091 = 0.391429 loss)
I0428 23:20:07.971320 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 4.30305 (* 0.0909091 = 0.391187 loss)
I0428 23:20:07.971333 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 4.30141 (* 0.0909091 = 0.391037 loss)
I0428 23:20:07.971346 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 4.29977 (* 0.0909091 = 0.390888 loss)
I0428 23:20:07.971360 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 4.30218 (* 0.0909091 = 0.391108 loss)
I0428 23:20:07.971374 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 4.30199 (* 0.0909091 = 0.39109 loss)
I0428 23:20:07.971388 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 4.30803 (* 0.0909091 = 0.391639 loss)
I0428 23:20:07.971401 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 4.31052 (* 0.0909091 = 0.391865 loss)
I0428 23:20:07.971415 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 4.2978 (* 0.0909091 = 0.390709 loss)
I0428 23:20:07.971429 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 4.31069 (* 0.0909091 = 0.391881 loss)
I0428 23:20:07.971452 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 4.30932 (* 0.0909091 = 0.391756 loss)
I0428 23:20:07.971480 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 4.30178 (* 0.0909091 = 0.391071 loss)
I0428 23:20:07.971498 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 4.31597 (* 0.0909091 = 0.392361 loss)
I0428 23:20:07.971511 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 4.30662 (* 0.0909091 = 0.391511 loss)
I0428 23:20:07.971525 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 4.30325 (* 0.0909091 = 0.391204 loss)
I0428 23:20:07.971539 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 4.30492 (* 0.0909091 = 0.391356 loss)
I0428 23:20:07.971552 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 4.30636 (* 0.0909091 = 0.391487 loss)
I0428 23:20:07.971566 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 4.30215 (* 0.0909091 = 0.391105 loss)
I0428 23:20:07.971580 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 4.30551 (* 0.0909091 = 0.39141 loss)
I0428 23:20:07.971591 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:20:07.971602 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:20:07.971613 6470 solver.cpp:245] Train net output #149: total_confidence = 4.23318e-41
I0428 23:20:07.971626 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 1.12833e-41
I0428 23:20:07.971647 6470 sgd_solver.cpp:106] Iteration 0, lr = 0.01
I0428 23:20:10.758203 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 75.6498 > 30) by scale factor 0.396564
I0428 23:20:11.040961 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 116.532 > 30) by scale factor 0.257441
I0428 23:20:11.323570 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 87.41 > 30) by scale factor 0.34321
I0428 23:20:11.606402 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 42.7893 > 30) by scale factor 0.70111
I0428 23:22:24.491296 6470 solver.cpp:229] Iteration 500, loss = 13.7255
I0428 23:22:24.491765 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0196078
I0428 23:22:24.491786 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:22:24.491799 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:22:24.491811 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:22:24.491823 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0428 23:22:24.491835 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:22:24.491847 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:22:24.491859 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0428 23:22:24.491870 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0428 23:22:24.491883 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:22:24.491894 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:22:24.491906 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:22:24.491917 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:22:24.491930 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:22:24.491940 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:22:24.491951 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:22:24.491963 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:22:24.491974 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:22:24.491986 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:22:24.491997 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:22:24.492008 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:22:24.492020 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:22:24.492032 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:22:24.492043 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.704545
I0428 23:22:24.492054 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.156863
I0428 23:22:24.492070 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.97133 (* 0.3 = 1.1914 loss)
I0428 23:22:24.492085 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.63436 (* 0.3 = 0.490308 loss)
I0428 23:22:24.492100 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 4.00461 (* 0.0272727 = 0.109217 loss)
I0428 23:22:24.492113 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 4.48519 (* 0.0272727 = 0.122323 loss)
I0428 23:22:24.492126 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.84218 (* 0.0272727 = 0.104787 loss)
I0428 23:22:24.492141 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 4.0466 (* 0.0272727 = 0.110362 loss)
I0428 23:22:24.492154 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 4.023 (* 0.0272727 = 0.109718 loss)
I0428 23:22:24.492167 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.20951 (* 0.0272727 = 0.0875322 loss)
I0428 23:22:24.492182 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.72416 (* 0.0272727 = 0.0470224 loss)
I0428 23:22:24.492194 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.60653 (* 0.0272727 = 0.0165417 loss)
I0428 23:22:24.492208 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.741586 (* 0.0272727 = 0.0202251 loss)
I0428 23:22:24.492221 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.589667 (* 0.0272727 = 0.0160818 loss)
I0428 23:22:24.492235 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.631163 (* 0.0272727 = 0.0172135 loss)
I0428 23:22:24.492250 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.045589 (* 0.0272727 = 0.00124334 loss)
I0428 23:22:24.492264 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0404192 (* 0.0272727 = 0.00110234 loss)
I0428 23:22:24.492293 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0243041 (* 0.0272727 = 0.000662839 loss)
I0428 23:22:24.492308 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0268729 (* 0.0272727 = 0.000732896 loss)
I0428 23:22:24.492326 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0245886 (* 0.0272727 = 0.000670598 loss)
I0428 23:22:24.492341 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0196394 (* 0.0272727 = 0.000535621 loss)
I0428 23:22:24.492354 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0159518 (* 0.0272727 = 0.00043505 loss)
I0428 23:22:24.492367 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0119358 (* 0.0272727 = 0.000325522 loss)
I0428 23:22:24.492382 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0126653 (* 0.0272727 = 0.000345417 loss)
I0428 23:22:24.492395 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0189898 (* 0.0272727 = 0.000517902 loss)
I0428 23:22:24.492409 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0168041 (* 0.0272727 = 0.000458295 loss)
I0428 23:22:24.492421 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0196078
I0428 23:22:24.492434 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:22:24.492445 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:22:24.492456 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:22:24.492467 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:22:24.492480 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0428 23:22:24.492491 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0428 23:22:24.492503 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0428 23:22:24.492514 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0428 23:22:24.492527 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:22:24.492538 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:22:24.492549 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:22:24.492561 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:22:24.492573 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:22:24.492584 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:22:24.492595 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:22:24.492606 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:22:24.492617 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:22:24.492630 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:22:24.492640 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:22:24.492651 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:22:24.492662 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:22:24.492674 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:22:24.492686 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.715909
I0428 23:22:24.492697 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.0784314
I0428 23:22:24.492710 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 4.11491 (* 0.3 = 1.23447 loss)
I0428 23:22:24.492724 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.38756 (* 0.3 = 0.416268 loss)
I0428 23:22:24.492738 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.87037 (* 0.0272727 = 0.105555 loss)
I0428 23:22:24.492753 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 4.07053 (* 0.0272727 = 0.111014 loss)
I0428 23:22:24.492780 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.74107 (* 0.0272727 = 0.102029 loss)
I0428 23:22:24.492795 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 4.08892 (* 0.0272727 = 0.111516 loss)
I0428 23:22:24.492810 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.95515 (* 0.0272727 = 0.107868 loss)
I0428 23:22:24.492822 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.7082 (* 0.0272727 = 0.101133 loss)
I0428 23:22:24.492836 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.85309 (* 0.0272727 = 0.0505389 loss)
I0428 23:22:24.492849 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.758647 (* 0.0272727 = 0.0206904 loss)
I0428 23:22:24.492863 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.731381 (* 0.0272727 = 0.0199468 loss)
I0428 23:22:24.492877 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.760651 (* 0.0272727 = 0.020745 loss)
I0428 23:22:24.492890 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.954672 (* 0.0272727 = 0.0260365 loss)
I0428 23:22:24.492905 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0720547 (* 0.0272727 = 0.00196513 loss)
I0428 23:22:24.492919 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0548565 (* 0.0272727 = 0.00149609 loss)
I0428 23:22:24.492933 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0508263 (* 0.0272727 = 0.00138617 loss)
I0428 23:22:24.492946 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0467454 (* 0.0272727 = 0.00127487 loss)
I0428 23:22:24.492959 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0485418 (* 0.0272727 = 0.00132387 loss)
I0428 23:22:24.492974 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0304565 (* 0.0272727 = 0.000830632 loss)
I0428 23:22:24.492987 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0333753 (* 0.0272727 = 0.000910236 loss)
I0428 23:22:24.493001 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0203908 (* 0.0272727 = 0.000556114 loss)
I0428 23:22:24.493015 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0256688 (* 0.0272727 = 0.000700059 loss)
I0428 23:22:24.493028 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0250073 (* 0.0272727 = 0.000682018 loss)
I0428 23:22:24.493042 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0282523 (* 0.0272727 = 0.000770518 loss)
I0428 23:22:24.493054 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0
I0428 23:22:24.493067 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0428 23:22:24.493077 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:22:24.493088 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:22:24.493100 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:22:24.493111 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:22:24.493124 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:22:24.493134 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0428 23:22:24.493146 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0428 23:22:24.493157 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:22:24.493170 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:22:24.493181 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:22:24.493192 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:22:24.493204 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:22:24.493216 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:22:24.493227 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:22:24.493238 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:22:24.493259 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:22:24.493273 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:22:24.493283 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:22:24.493294 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:22:24.493306 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:22:24.493317 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:22:24.493332 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.710227
I0428 23:22:24.493345 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.0980392
I0428 23:22:24.493358 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.91882 (* 1 = 3.91882 loss)
I0428 23:22:24.493374 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.31203 (* 1 = 1.31203 loss)
I0428 23:22:24.493389 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.57427 (* 0.0909091 = 0.324933 loss)
I0428 23:22:24.493402 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.82416 (* 0.0909091 = 0.347651 loss)
I0428 23:22:24.493417 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.8046 (* 0.0909091 = 0.345873 loss)
I0428 23:22:24.493429 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.97741 (* 0.0909091 = 0.361583 loss)
I0428 23:22:24.493443 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.63629 (* 0.0909091 = 0.330572 loss)
I0428 23:22:24.493456 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 3.51965 (* 0.0909091 = 0.319968 loss)
I0428 23:22:24.493470 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.77731 (* 0.0909091 = 0.161574 loss)
I0428 23:22:24.493484 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.649032 (* 0.0909091 = 0.0590029 loss)
I0428 23:22:24.493497 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.729689 (* 0.0909091 = 0.0663354 loss)
I0428 23:22:24.493510 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.594953 (* 0.0909091 = 0.0540866 loss)
I0428 23:22:24.493525 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.598133 (* 0.0909091 = 0.0543757 loss)
I0428 23:22:24.493538 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0696603 (* 0.0909091 = 0.00633276 loss)
I0428 23:22:24.493551 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.0606917 (* 0.0909091 = 0.00551743 loss)
I0428 23:22:24.493566 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.045786 (* 0.0909091 = 0.00416237 loss)
I0428 23:22:24.493578 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0334759 (* 0.0909091 = 0.00304326 loss)
I0428 23:22:24.493592 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.019903 (* 0.0909091 = 0.00180936 loss)
I0428 23:22:24.493605 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0105132 (* 0.0909091 = 0.000955745 loss)
I0428 23:22:24.493619 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00827597 (* 0.0909091 = 0.000752361 loss)
I0428 23:22:24.493633 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0069585 (* 0.0909091 = 0.00063259 loss)
I0428 23:22:24.493646 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00642097 (* 0.0909091 = 0.000583725 loss)
I0428 23:22:24.493660 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00478001 (* 0.0909091 = 0.000434546 loss)
I0428 23:22:24.493674 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.003894 (* 0.0909091 = 0.000354 loss)
I0428 23:22:24.493685 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:22:24.493697 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:22:24.493708 6470 solver.cpp:245] Train net output #149: total_confidence = 1.20836e-08
I0428 23:22:24.493729 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 1.41271e-07
I0428 23:22:24.493743 6470 sgd_solver.cpp:106] Iteration 500, lr = 0.01
I0428 23:24:40.885253 6470 solver.cpp:229] Iteration 1000, loss = 12.2843
I0428 23:24:40.885419 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0754717
I0428 23:24:40.885439 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:24:40.885453 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:24:40.885465 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0428 23:24:40.885478 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:24:40.885489 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0428 23:24:40.885501 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0428 23:24:40.885514 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0428 23:24:40.885525 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:24:40.885537 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:24:40.885548 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:24:40.885560 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:24:40.885571 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0428 23:24:40.885584 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0428 23:24:40.885596 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0428 23:24:40.885607 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0428 23:24:40.885620 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0428 23:24:40.885632 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:24:40.885643 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:24:40.885654 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:24:40.885666 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:24:40.885678 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:24:40.885689 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:24:40.885700 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.698864
I0428 23:24:40.885712 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.0943396
I0428 23:24:40.885728 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.69237 (* 0.3 = 1.10771 loss)
I0428 23:24:40.885743 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.4698 (* 0.3 = 0.44094 loss)
I0428 23:24:40.885757 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.3959 (* 0.0272727 = 0.0926154 loss)
I0428 23:24:40.885771 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.41614 (* 0.0272727 = 0.0931675 loss)
I0428 23:24:40.885785 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.63239 (* 0.0272727 = 0.0990652 loss)
I0428 23:24:40.885799 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.4296 (* 0.0272727 = 0.0935346 loss)
I0428 23:24:40.885813 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.60033 (* 0.0272727 = 0.0981909 loss)
I0428 23:24:40.885828 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.60393 (* 0.0272727 = 0.0710163 loss)
I0428 23:24:40.885840 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.23612 (* 0.0272727 = 0.0609852 loss)
I0428 23:24:40.885854 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.55786 (* 0.0272727 = 0.0424872 loss)
I0428 23:24:40.885867 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.660334 (* 0.0272727 = 0.0180091 loss)
I0428 23:24:40.885882 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.659155 (* 0.0272727 = 0.0179769 loss)
I0428 23:24:40.885895 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.807591 (* 0.0272727 = 0.0220252 loss)
I0428 23:24:40.885910 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.724918 (* 0.0272727 = 0.0197705 loss)
I0428 23:24:40.885944 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.795347 (* 0.0272727 = 0.0216913 loss)
I0428 23:24:40.885960 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.812133 (* 0.0272727 = 0.0221491 loss)
I0428 23:24:40.885974 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.835301 (* 0.0272727 = 0.0227809 loss)
I0428 23:24:40.885988 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.923264 (* 0.0272727 = 0.0251799 loss)
I0428 23:24:40.886003 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00658959 (* 0.0272727 = 0.000179716 loss)
I0428 23:24:40.886016 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00950247 (* 0.0272727 = 0.000259158 loss)
I0428 23:24:40.886030 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00425098 (* 0.0272727 = 0.000115936 loss)
I0428 23:24:40.886044 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00950112 (* 0.0272727 = 0.000259121 loss)
I0428 23:24:40.886057 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00408015 (* 0.0272727 = 0.000111277 loss)
I0428 23:24:40.886071 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00543728 (* 0.0272727 = 0.000148289 loss)
I0428 23:24:40.886083 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0377358
I0428 23:24:40.886096 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:24:40.886106 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:24:40.886117 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:24:40.886129 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:24:40.886140 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0428 23:24:40.886152 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0428 23:24:40.886163 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0428 23:24:40.886178 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:24:40.886189 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:24:40.886200 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:24:40.886212 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:24:40.886224 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0428 23:24:40.886235 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0428 23:24:40.886247 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0428 23:24:40.886258 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0428 23:24:40.886270 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0428 23:24:40.886281 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:24:40.886292 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:24:40.886304 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:24:40.886318 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:24:40.886330 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:24:40.886342 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:24:40.886353 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.6875
I0428 23:24:40.886364 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.0754717
I0428 23:24:40.886379 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.72129 (* 0.3 = 1.11639 loss)
I0428 23:24:40.886392 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.48905 (* 0.3 = 0.446715 loss)
I0428 23:24:40.886406 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.27433 (* 0.0272727 = 0.0893 loss)
I0428 23:24:40.886420 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.56853 (* 0.0272727 = 0.0973236 loss)
I0428 23:24:40.886448 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.64339 (* 0.0272727 = 0.0993651 loss)
I0428 23:24:40.886463 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.65958 (* 0.0272727 = 0.0998066 loss)
I0428 23:24:40.886476 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.73208 (* 0.0272727 = 0.101784 loss)
I0428 23:24:40.886490 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.87589 (* 0.0272727 = 0.0784333 loss)
I0428 23:24:40.886503 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.6816 (* 0.0272727 = 0.0731344 loss)
I0428 23:24:40.886517 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.85245 (* 0.0272727 = 0.0505215 loss)
I0428 23:24:40.886530 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.656493 (* 0.0272727 = 0.0179044 loss)
I0428 23:24:40.886544 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.951379 (* 0.0272727 = 0.0259467 loss)
I0428 23:24:40.886557 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.887421 (* 0.0272727 = 0.0242024 loss)
I0428 23:24:40.886571 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.902598 (* 0.0272727 = 0.0246163 loss)
I0428 23:24:40.886585 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.865719 (* 0.0272727 = 0.0236105 loss)
I0428 23:24:40.886600 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.878432 (* 0.0272727 = 0.0239572 loss)
I0428 23:24:40.886612 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 1.11451 (* 0.0272727 = 0.0303958 loss)
I0428 23:24:40.886626 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 1.16812 (* 0.0272727 = 0.0318578 loss)
I0428 23:24:40.886641 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00264039 (* 0.0272727 = 7.20106e-05 loss)
I0428 23:24:40.886653 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00361687 (* 0.0272727 = 9.86418e-05 loss)
I0428 23:24:40.886667 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00244664 (* 0.0272727 = 6.67265e-05 loss)
I0428 23:24:40.886680 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00292428 (* 0.0272727 = 7.97531e-05 loss)
I0428 23:24:40.886694 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00466941 (* 0.0272727 = 0.000127348 loss)
I0428 23:24:40.886708 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00234803 (* 0.0272727 = 6.40373e-05 loss)
I0428 23:24:40.886720 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0188679
I0428 23:24:40.886732 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:24:40.886744 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:24:40.886754 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:24:40.886766 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:24:40.886777 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0428 23:24:40.886790 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0428 23:24:40.886801 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0428 23:24:40.886811 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:24:40.886823 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:24:40.886834 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:24:40.886845 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:24:40.886857 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0428 23:24:40.886868 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0428 23:24:40.886880 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0428 23:24:40.886891 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0428 23:24:40.886904 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0428 23:24:40.886924 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:24:40.886936 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:24:40.886947 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:24:40.886960 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:24:40.886970 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:24:40.886982 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:24:40.886993 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.681818
I0428 23:24:40.887006 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.0943396
I0428 23:24:40.887019 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.56899 (* 1 = 3.56899 loss)
I0428 23:24:40.887032 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.35729 (* 1 = 1.35729 loss)
I0428 23:24:40.887047 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.12001 (* 0.0909091 = 0.283637 loss)
I0428 23:24:40.887060 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.60163 (* 0.0909091 = 0.327421 loss)
I0428 23:24:40.887073 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.80779 (* 0.0909091 = 0.346163 loss)
I0428 23:24:40.887086 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.43265 (* 0.0909091 = 0.312059 loss)
I0428 23:24:40.887100 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.41721 (* 0.0909091 = 0.310655 loss)
I0428 23:24:40.887114 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.74802 (* 0.0909091 = 0.24982 loss)
I0428 23:24:40.887127 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.32348 (* 0.0909091 = 0.211225 loss)
I0428 23:24:40.887140 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.70865 (* 0.0909091 = 0.155332 loss)
I0428 23:24:40.887153 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.458388 (* 0.0909091 = 0.0416716 loss)
I0428 23:24:40.887167 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.759029 (* 0.0909091 = 0.0690027 loss)
I0428 23:24:40.887181 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.714936 (* 0.0909091 = 0.0649942 loss)
I0428 23:24:40.887194 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.813032 (* 0.0909091 = 0.073912 loss)
I0428 23:24:40.887207 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.720083 (* 0.0909091 = 0.0654621 loss)
I0428 23:24:40.887222 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.933381 (* 0.0909091 = 0.0848528 loss)
I0428 23:24:40.887234 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.829861 (* 0.0909091 = 0.0754419 loss)
I0428 23:24:40.887248 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 1.05239 (* 0.0909091 = 0.0956716 loss)
I0428 23:24:40.887262 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000693933 (* 0.0909091 = 6.30848e-05 loss)
I0428 23:24:40.887275 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000219698 (* 0.0909091 = 1.99725e-05 loss)
I0428 23:24:40.887290 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000114593 (* 0.0909091 = 1.04176e-05 loss)
I0428 23:24:40.887303 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000110911 (* 0.0909091 = 1.00828e-05 loss)
I0428 23:24:40.887317 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 6.59238e-05 (* 0.0909091 = 5.99307e-06 loss)
I0428 23:24:40.887331 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 5.28077e-05 (* 0.0909091 = 4.8007e-06 loss)
I0428 23:24:40.887342 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:24:40.887354 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:24:40.887367 6470 solver.cpp:245] Train net output #149: total_confidence = 2.14522e-07
I0428 23:24:40.887389 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 1.89347e-05
I0428 23:24:40.887403 6470 sgd_solver.cpp:106] Iteration 1000, lr = 0.01
I0428 23:26:57.277094 6470 solver.cpp:229] Iteration 1500, loss = 11.7686
I0428 23:26:57.277273 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0196078
I0428 23:26:57.277294 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:26:57.277307 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:26:57.277321 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:26:57.277333 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0428 23:26:57.277345 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:26:57.277357 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0428 23:26:57.277369 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0428 23:26:57.277380 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:26:57.277392 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:26:57.277405 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:26:57.277416 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:26:57.277428 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:26:57.277439 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:26:57.277451 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:26:57.277462 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:26:57.277473 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:26:57.277485 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:26:57.277496 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:26:57.277508 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:26:57.277519 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:26:57.277530 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:26:57.277541 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:26:57.277552 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.715909
I0428 23:26:57.277565 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.0588235
I0428 23:26:57.277580 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.65922 (* 0.3 = 1.09777 loss)
I0428 23:26:57.277595 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.22539 (* 0.3 = 0.367616 loss)
I0428 23:26:57.277609 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.6967 (* 0.0272727 = 0.100819 loss)
I0428 23:26:57.277622 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.84407 (* 0.0272727 = 0.104838 loss)
I0428 23:26:57.277637 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.94156 (* 0.0272727 = 0.107497 loss)
I0428 23:26:57.277650 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.40193 (* 0.0272727 = 0.0927799 loss)
I0428 23:26:57.277664 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.92095 (* 0.0272727 = 0.106935 loss)
I0428 23:26:57.277678 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.92901 (* 0.0272727 = 0.0798821 loss)
I0428 23:26:57.277691 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.17575 (* 0.0272727 = 0.0593386 loss)
I0428 23:26:57.277705 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.19352 (* 0.0272727 = 0.0325504 loss)
I0428 23:26:57.277719 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.953325 (* 0.0272727 = 0.0259998 loss)
I0428 23:26:57.277732 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.839821 (* 0.0272727 = 0.0229042 loss)
I0428 23:26:57.277746 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.196403 (* 0.0272727 = 0.00535645 loss)
I0428 23:26:57.277760 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.203995 (* 0.0272727 = 0.00556349 loss)
I0428 23:26:57.277775 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.138306 (* 0.0272727 = 0.00377197 loss)
I0428 23:26:57.277808 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.140965 (* 0.0272727 = 0.00384449 loss)
I0428 23:26:57.277824 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.10652 (* 0.0272727 = 0.00290509 loss)
I0428 23:26:57.277837 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.101664 (* 0.0272727 = 0.00277266 loss)
I0428 23:26:57.277851 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0712393 (* 0.0272727 = 0.00194289 loss)
I0428 23:26:57.277865 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0488875 (* 0.0272727 = 0.00133329 loss)
I0428 23:26:57.277879 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.046075 (* 0.0272727 = 0.00125659 loss)
I0428 23:26:57.277894 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0436903 (* 0.0272727 = 0.00119155 loss)
I0428 23:26:57.277907 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0466078 (* 0.0272727 = 0.00127112 loss)
I0428 23:26:57.277921 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0657621 (* 0.0272727 = 0.00179351 loss)
I0428 23:26:57.277933 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0196078
I0428 23:26:57.277945 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:26:57.277957 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:26:57.277968 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:26:57.277979 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:26:57.277992 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0428 23:26:57.278003 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0428 23:26:57.278015 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0428 23:26:57.278026 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:26:57.278038 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:26:57.278049 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:26:57.278061 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:26:57.278074 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:26:57.278084 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:26:57.278095 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:26:57.278106 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:26:57.278118 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:26:57.278129 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:26:57.278141 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:26:57.278151 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:26:57.278162 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:26:57.278173 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:26:57.278185 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:26:57.278197 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.710227
I0428 23:26:57.278208 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.117647
I0428 23:26:57.278221 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.63761 (* 0.3 = 1.09128 loss)
I0428 23:26:57.278235 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.21955 (* 0.3 = 0.365866 loss)
I0428 23:26:57.278249 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.79285 (* 0.0272727 = 0.103441 loss)
I0428 23:26:57.278262 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.87545 (* 0.0272727 = 0.105694 loss)
I0428 23:26:57.278276 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.7914 (* 0.0272727 = 0.103402 loss)
I0428 23:26:57.278304 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.60417 (* 0.0272727 = 0.0982955 loss)
I0428 23:26:57.278319 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 4.00495 (* 0.0272727 = 0.109226 loss)
I0428 23:26:57.278333 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.88559 (* 0.0272727 = 0.078698 loss)
I0428 23:26:57.278347 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.333 (* 0.0272727 = 0.0636273 loss)
I0428 23:26:57.278359 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.06086 (* 0.0272727 = 0.0289324 loss)
I0428 23:26:57.278376 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.624356 (* 0.0272727 = 0.0170279 loss)
I0428 23:26:57.278390 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.82094 (* 0.0272727 = 0.0223893 loss)
I0428 23:26:57.278404 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.143859 (* 0.0272727 = 0.00392342 loss)
I0428 23:26:57.278419 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.120377 (* 0.0272727 = 0.00328301 loss)
I0428 23:26:57.278432 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.107514 (* 0.0272727 = 0.00293221 loss)
I0428 23:26:57.278445 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.095388 (* 0.0272727 = 0.00260149 loss)
I0428 23:26:57.278460 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0740629 (* 0.0272727 = 0.0020199 loss)
I0428 23:26:57.278475 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0556117 (* 0.0272727 = 0.00151668 loss)
I0428 23:26:57.278487 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0367325 (* 0.0272727 = 0.00100179 loss)
I0428 23:26:57.278501 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0405299 (* 0.0272727 = 0.00110536 loss)
I0428 23:26:57.278515 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0244139 (* 0.0272727 = 0.000665834 loss)
I0428 23:26:57.278529 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.031133 (* 0.0272727 = 0.000849083 loss)
I0428 23:26:57.278542 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0320145 (* 0.0272727 = 0.000873121 loss)
I0428 23:26:57.278556 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0361682 (* 0.0272727 = 0.000986406 loss)
I0428 23:26:57.278568 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0588235
I0428 23:26:57.278580 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:26:57.278591 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:26:57.278602 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:26:57.278614 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0428 23:26:57.278625 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:26:57.278637 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0428 23:26:57.278648 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.5
I0428 23:26:57.278661 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:26:57.278671 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:26:57.278683 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:26:57.278694 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:26:57.278705 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:26:57.278717 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:26:57.278728 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:26:57.278739 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:26:57.278750 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:26:57.278771 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:26:57.278784 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:26:57.278795 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:26:57.278807 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:26:57.278818 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:26:57.278831 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:26:57.278841 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.721591
I0428 23:26:57.278853 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.196078
I0428 23:26:57.278867 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.54885 (* 1 = 3.54885 loss)
I0428 23:26:57.278880 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.13541 (* 1 = 1.13541 loss)
I0428 23:26:57.278894 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.23987 (* 0.0909091 = 0.294534 loss)
I0428 23:26:57.278908 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.81178 (* 0.0909091 = 0.346526 loss)
I0428 23:26:57.278921 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.32748 (* 0.0909091 = 0.302498 loss)
I0428 23:26:57.278935 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.31983 (* 0.0909091 = 0.301803 loss)
I0428 23:26:57.278949 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.61704 (* 0.0909091 = 0.328822 loss)
I0428 23:26:57.278962 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.58148 (* 0.0909091 = 0.23468 loss)
I0428 23:26:57.278975 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.13251 (* 0.0909091 = 0.193864 loss)
I0428 23:26:57.278990 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.947919 (* 0.0909091 = 0.0861745 loss)
I0428 23:26:57.279003 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.53412 (* 0.0909091 = 0.0485564 loss)
I0428 23:26:57.279016 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.637363 (* 0.0909091 = 0.0579421 loss)
I0428 23:26:57.279031 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.152743 (* 0.0909091 = 0.0138857 loss)
I0428 23:26:57.279043 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.143952 (* 0.0909091 = 0.0130865 loss)
I0428 23:26:57.279057 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.126802 (* 0.0909091 = 0.0115274 loss)
I0428 23:26:57.279072 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.102559 (* 0.0909091 = 0.00932355 loss)
I0428 23:26:57.279084 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0873258 (* 0.0909091 = 0.00793871 loss)
I0428 23:26:57.279098 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0470149 (* 0.0909091 = 0.00427408 loss)
I0428 23:26:57.279112 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.013873 (* 0.0909091 = 0.00126119 loss)
I0428 23:26:57.279126 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00808362 (* 0.0909091 = 0.000734875 loss)
I0428 23:26:57.279139 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0039794 (* 0.0909091 = 0.000361764 loss)
I0428 23:26:57.279153 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00324214 (* 0.0909091 = 0.00029474 loss)
I0428 23:26:57.279167 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00242322 (* 0.0909091 = 0.000220293 loss)
I0428 23:26:57.279181 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00205515 (* 0.0909091 = 0.000186832 loss)
I0428 23:26:57.279192 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:26:57.279204 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:26:57.279216 6470 solver.cpp:245] Train net output #149: total_confidence = 3.53364e-05
I0428 23:26:57.279235 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 2.37431e-05
I0428 23:26:57.279249 6470 sgd_solver.cpp:106] Iteration 1500, lr = 0.01
I0428 23:29:13.606395 6470 solver.cpp:229] Iteration 2000, loss = 11.3375
I0428 23:29:13.606523 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0441176
I0428 23:29:13.606542 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:29:13.606559 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:29:13.606570 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:29:13.606582 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0428 23:29:13.606595 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0428 23:29:13.606606 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:29:13.606623 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.25
I0428 23:29:13.606636 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.5
I0428 23:29:13.606647 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.5
I0428 23:29:13.606659 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.75
I0428 23:29:13.606672 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.75
I0428 23:29:13.606683 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0428 23:29:13.606695 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0428 23:29:13.606708 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0428 23:29:13.606719 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0428 23:29:13.606731 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0428 23:29:13.606744 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 0.875
I0428 23:29:13.606755 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 0.875
I0428 23:29:13.606767 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:29:13.606780 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:29:13.606791 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:29:13.606803 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:29:13.606814 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.625
I0428 23:29:13.606827 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.191176
I0428 23:29:13.606842 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.50384 (* 0.3 = 1.05115 loss)
I0428 23:29:13.606856 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.48933 (* 0.3 = 0.4468 loss)
I0428 23:29:13.606870 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.2783 (* 0.0272727 = 0.0894082 loss)
I0428 23:29:13.606884 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.48196 (* 0.0272727 = 0.0949625 loss)
I0428 23:29:13.606899 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.2695 (* 0.0272727 = 0.0891683 loss)
I0428 23:29:13.606912 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.24544 (* 0.0272727 = 0.0885121 loss)
I0428 23:29:13.606926 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.41006 (* 0.0272727 = 0.0657288 loss)
I0428 23:29:13.606940 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.07978 (* 0.0272727 = 0.083994 loss)
I0428 23:29:13.606953 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 3.65372 (* 0.0272727 = 0.0996469 loss)
I0428 23:29:13.606967 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.21219 (* 0.0272727 = 0.0603324 loss)
I0428 23:29:13.606981 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 2.0269 (* 0.0272727 = 0.055279 loss)
I0428 23:29:13.606994 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 1.10316 (* 0.0272727 = 0.0300861 loss)
I0428 23:29:13.607008 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 1.37785 (* 0.0272727 = 0.0375778 loss)
I0428 23:29:13.607023 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.593265 (* 0.0272727 = 0.0161799 loss)
I0428 23:29:13.607055 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.643384 (* 0.0272727 = 0.0175468 loss)
I0428 23:29:13.607071 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.648581 (* 0.0272727 = 0.0176886 loss)
I0428 23:29:13.607085 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.670373 (* 0.0272727 = 0.0182829 loss)
I0428 23:29:13.607100 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.739769 (* 0.0272727 = 0.0201755 loss)
I0428 23:29:13.607112 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.71031 (* 0.0272727 = 0.0193721 loss)
I0428 23:29:13.607126 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.797156 (* 0.0272727 = 0.0217406 loss)
I0428 23:29:13.607141 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0156263 (* 0.0272727 = 0.000426172 loss)
I0428 23:29:13.607156 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0200752 (* 0.0272727 = 0.000547506 loss)
I0428 23:29:13.607169 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0208122 (* 0.0272727 = 0.000567605 loss)
I0428 23:29:13.607183 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0164287 (* 0.0272727 = 0.000448056 loss)
I0428 23:29:13.607195 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0294118
I0428 23:29:13.607208 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:29:13.607219 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:29:13.607230 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0428 23:29:13.607242 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:29:13.607254 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:29:13.607265 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0428 23:29:13.607276 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.25
I0428 23:29:13.607288 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.5
I0428 23:29:13.607300 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.5
I0428 23:29:13.607312 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.75
I0428 23:29:13.607321 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.75
I0428 23:29:13.607328 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0428 23:29:13.607336 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0428 23:29:13.607348 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0428 23:29:13.607360 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0428 23:29:13.607372 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0428 23:29:13.607383 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 0.875
I0428 23:29:13.607395 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 0.875
I0428 23:29:13.607406 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:29:13.607417 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:29:13.607429 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:29:13.607440 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:29:13.607451 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.625
I0428 23:29:13.607463 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.147059
I0428 23:29:13.607496 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.57371 (* 0.3 = 1.07211 loss)
I0428 23:29:13.607511 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.47239 (* 0.3 = 0.441718 loss)
I0428 23:29:13.607524 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.80328 (* 0.0272727 = 0.103726 loss)
I0428 23:29:13.607538 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.21443 (* 0.0272727 = 0.0876664 loss)
I0428 23:29:13.607564 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.40637 (* 0.0272727 = 0.092901 loss)
I0428 23:29:13.607579 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.4358 (* 0.0272727 = 0.0937037 loss)
I0428 23:29:13.607592 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.5637 (* 0.0272727 = 0.0699191 loss)
I0428 23:29:13.607609 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.11163 (* 0.0272727 = 0.0848625 loss)
I0428 23:29:13.607623 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 3.31405 (* 0.0272727 = 0.0903831 loss)
I0428 23:29:13.607637 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.42064 (* 0.0272727 = 0.0660173 loss)
I0428 23:29:13.607650 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 2.03771 (* 0.0272727 = 0.0555739 loss)
I0428 23:29:13.607667 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 1.26108 (* 0.0272727 = 0.0343932 loss)
I0428 23:29:13.607682 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 1.40796 (* 0.0272727 = 0.038399 loss)
I0428 23:29:13.607697 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.572197 (* 0.0272727 = 0.0156054 loss)
I0428 23:29:13.607709 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.758381 (* 0.0272727 = 0.0206831 loss)
I0428 23:29:13.607723 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.691624 (* 0.0272727 = 0.0188625 loss)
I0428 23:29:13.607738 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.819984 (* 0.0272727 = 0.0223632 loss)
I0428 23:29:13.607750 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.917788 (* 0.0272727 = 0.0250306 loss)
I0428 23:29:13.607764 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 1.10125 (* 0.0272727 = 0.030034 loss)
I0428 23:29:13.607777 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 1.07297 (* 0.0272727 = 0.0292628 loss)
I0428 23:29:13.607791 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00454333 (* 0.0272727 = 0.000123909 loss)
I0428 23:29:13.607805 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0050418 (* 0.0272727 = 0.000137504 loss)
I0428 23:29:13.607818 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00336276 (* 0.0272727 = 9.17116e-05 loss)
I0428 23:29:13.607832 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00349435 (* 0.0272727 = 9.53005e-05 loss)
I0428 23:29:13.607844 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0882353
I0428 23:29:13.607856 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0428 23:29:13.607868 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:29:13.607879 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:29:13.607892 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0428 23:29:13.607903 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0428 23:29:13.607914 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:29:13.607925 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.25
I0428 23:29:13.607938 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.5
I0428 23:29:13.607949 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.5
I0428 23:29:13.607960 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.75
I0428 23:29:13.607971 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0428 23:29:13.607983 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0428 23:29:13.607995 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0428 23:29:13.608006 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0428 23:29:13.608017 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0428 23:29:13.608029 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0428 23:29:13.608050 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 0.875
I0428 23:29:13.608063 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 0.875
I0428 23:29:13.608075 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:29:13.608086 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:29:13.608098 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:29:13.608109 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:29:13.608120 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.647727
I0428 23:29:13.608132 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.205882
I0428 23:29:13.608146 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.67476 (* 1 = 3.67476 loss)
I0428 23:29:13.608160 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.47733 (* 1 = 1.47733 loss)
I0428 23:29:13.608173 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.02257 (* 0.0909091 = 0.274779 loss)
I0428 23:29:13.608186 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.2418 (* 0.0909091 = 0.294709 loss)
I0428 23:29:13.608201 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.15312 (* 0.0909091 = 0.286647 loss)
I0428 23:29:13.608214 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.8542 (* 0.0909091 = 0.259473 loss)
I0428 23:29:13.608227 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.29887 (* 0.0909091 = 0.208988 loss)
I0428 23:29:13.608240 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.87549 (* 0.0909091 = 0.261408 loss)
I0428 23:29:13.608254 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 3.47113 (* 0.0909091 = 0.315557 loss)
I0428 23:29:13.608268 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 2.05185 (* 0.0909091 = 0.186532 loss)
I0428 23:29:13.608281 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.85296 (* 0.0909091 = 0.168451 loss)
I0428 23:29:13.608294 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 1.07261 (* 0.0909091 = 0.09751 loss)
I0428 23:29:13.608309 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 1.53493 (* 0.0909091 = 0.139539 loss)
I0428 23:29:13.608321 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.698243 (* 0.0909091 = 0.0634766 loss)
I0428 23:29:13.608335 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.821895 (* 0.0909091 = 0.0747177 loss)
I0428 23:29:13.608348 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.68204 (* 0.0909091 = 0.0620037 loss)
I0428 23:29:13.608362 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.842586 (* 0.0909091 = 0.0765987 loss)
I0428 23:29:13.608376 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.88794 (* 0.0909091 = 0.0807218 loss)
I0428 23:29:13.608389 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.938502 (* 0.0909091 = 0.0853184 loss)
I0428 23:29:13.608403 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 1.11695 (* 0.0909091 = 0.10154 loss)
I0428 23:29:13.608417 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00190727 (* 0.0909091 = 0.000173388 loss)
I0428 23:29:13.608431 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.0011848 (* 0.0909091 = 0.000107709 loss)
I0428 23:29:13.608445 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00082242 (* 0.0909091 = 7.47655e-05 loss)
I0428 23:29:13.608459 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000609769 (* 0.0909091 = 5.54335e-05 loss)
I0428 23:29:13.608471 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:29:13.608484 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:29:13.608494 6470 solver.cpp:245] Train net output #149: total_confidence = 7.60891e-05
I0428 23:29:13.608515 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000171432
I0428 23:29:13.608530 6470 sgd_solver.cpp:106] Iteration 2000, lr = 0.01
I0428 23:29:22.127594 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 39.9585 > 30) by scale factor 0.750778
I0428 23:31:30.002750 6470 solver.cpp:229] Iteration 2500, loss = 11.1428
I0428 23:31:30.002931 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12766
I0428 23:31:30.002951 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:31:30.002965 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:31:30.002977 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:31:30.002990 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:31:30.003001 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0428 23:31:30.003013 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:31:30.003024 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0428 23:31:30.003036 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0428 23:31:30.003049 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0428 23:31:30.003060 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0428 23:31:30.003072 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:31:30.003083 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:31:30.003094 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:31:30.003106 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:31:30.003118 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:31:30.003129 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:31:30.003141 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:31:30.003154 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:31:30.003165 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:31:30.003176 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:31:30.003188 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:31:30.003201 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:31:30.003211 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.767045
I0428 23:31:30.003223 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.234043
I0428 23:31:30.003240 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.54002 (* 0.3 = 1.06201 loss)
I0428 23:31:30.003255 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.02858 (* 0.3 = 0.308574 loss)
I0428 23:31:30.003269 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.50366 (* 0.0272727 = 0.0955543 loss)
I0428 23:31:30.003283 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.63367 (* 0.0272727 = 0.0991 loss)
I0428 23:31:30.003298 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 4.13191 (* 0.0272727 = 0.112689 loss)
I0428 23:31:30.003314 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.54021 (* 0.0272727 = 0.0965511 loss)
I0428 23:31:30.003329 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.42464 (* 0.0272727 = 0.0933992 loss)
I0428 23:31:30.003343 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.26654 (* 0.0272727 = 0.0890874 loss)
I0428 23:31:30.003357 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.48768 (* 0.0272727 = 0.040573 loss)
I0428 23:31:30.003371 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.882903 (* 0.0272727 = 0.0240792 loss)
I0428 23:31:30.003386 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.153542 (* 0.0272727 = 0.00418751 loss)
I0428 23:31:30.003399 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.131763 (* 0.0272727 = 0.00359354 loss)
I0428 23:31:30.003413 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0932384 (* 0.0272727 = 0.00254286 loss)
I0428 23:31:30.003427 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0967612 (* 0.0272727 = 0.00263894 loss)
I0428 23:31:30.003442 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0689358 (* 0.0272727 = 0.00188007 loss)
I0428 23:31:30.003495 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0384951 (* 0.0272727 = 0.00104987 loss)
I0428 23:31:30.003512 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0360356 (* 0.0272727 = 0.00098279 loss)
I0428 23:31:30.003526 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0209151 (* 0.0272727 = 0.000570413 loss)
I0428 23:31:30.003540 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00903733 (* 0.0272727 = 0.000246473 loss)
I0428 23:31:30.003554 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0126951 (* 0.0272727 = 0.00034623 loss)
I0428 23:31:30.003568 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00619636 (* 0.0272727 = 0.000168992 loss)
I0428 23:31:30.003582 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00947337 (* 0.0272727 = 0.000258365 loss)
I0428 23:31:30.003597 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00697526 (* 0.0272727 = 0.000190234 loss)
I0428 23:31:30.003610 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0045945 (* 0.0272727 = 0.000125305 loss)
I0428 23:31:30.003623 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0851064
I0428 23:31:30.003635 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0428 23:31:30.003646 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:31:30.003659 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:31:30.003669 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:31:30.003681 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:31:30.003690 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0428 23:31:30.003697 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0428 23:31:30.003705 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0428 23:31:30.003717 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0428 23:31:30.003729 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0428 23:31:30.003741 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:31:30.003752 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:31:30.003763 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:31:30.003774 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:31:30.003787 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:31:30.003798 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:31:30.003808 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:31:30.003819 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:31:30.003831 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:31:30.003842 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:31:30.003854 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:31:30.003865 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:31:30.003876 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.75
I0428 23:31:30.003888 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.255319
I0428 23:31:30.003902 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.71052 (* 0.3 = 1.11316 loss)
I0428 23:31:30.003916 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.14186 (* 0.3 = 0.342559 loss)
I0428 23:31:30.003929 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.65429 (* 0.0272727 = 0.0996626 loss)
I0428 23:31:30.003949 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.61827 (* 0.0272727 = 0.09868 loss)
I0428 23:31:30.003963 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 4.07127 (* 0.0272727 = 0.111035 loss)
I0428 23:31:30.003989 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.44314 (* 0.0272727 = 0.0939038 loss)
I0428 23:31:30.004004 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.64533 (* 0.0272727 = 0.099418 loss)
I0428 23:31:30.004019 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.39125 (* 0.0272727 = 0.0924886 loss)
I0428 23:31:30.004031 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.02882 (* 0.0272727 = 0.0553314 loss)
I0428 23:31:30.004045 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.939796 (* 0.0272727 = 0.0256308 loss)
I0428 23:31:30.004060 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.248103 (* 0.0272727 = 0.00676645 loss)
I0428 23:31:30.004073 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.176967 (* 0.0272727 = 0.00482638 loss)
I0428 23:31:30.004087 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.153796 (* 0.0272727 = 0.00419443 loss)
I0428 23:31:30.004101 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.124343 (* 0.0272727 = 0.00339117 loss)
I0428 23:31:30.004114 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.101384 (* 0.0272727 = 0.00276501 loss)
I0428 23:31:30.004129 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0728929 (* 0.0272727 = 0.00198799 loss)
I0428 23:31:30.004142 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0585406 (* 0.0272727 = 0.00159656 loss)
I0428 23:31:30.004156 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0286963 (* 0.0272727 = 0.000782627 loss)
I0428 23:31:30.004169 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0126827 (* 0.0272727 = 0.000345892 loss)
I0428 23:31:30.004184 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0102714 (* 0.0272727 = 0.000280128 loss)
I0428 23:31:30.004197 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00851225 (* 0.0272727 = 0.000232152 loss)
I0428 23:31:30.004210 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00949507 (* 0.0272727 = 0.000258956 loss)
I0428 23:31:30.004225 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0100533 (* 0.0272727 = 0.00027418 loss)
I0428 23:31:30.004238 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0091593 (* 0.0272727 = 0.000249799 loss)
I0428 23:31:30.004251 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.106383
I0428 23:31:30.004262 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:31:30.004274 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0428 23:31:30.004286 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:31:30.004297 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:31:30.004308 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0428 23:31:30.004320 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:31:30.004331 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0428 23:31:30.004343 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0428 23:31:30.004354 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0428 23:31:30.004369 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0428 23:31:30.004380 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:31:30.004392 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:31:30.004403 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:31:30.004415 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:31:30.004426 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:31:30.004437 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:31:30.004458 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:31:30.004472 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:31:30.004482 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:31:30.004494 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:31:30.004505 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:31:30.004518 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:31:30.004528 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0428 23:31:30.004540 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.212766
I0428 23:31:30.004554 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.47048 (* 1 = 3.47048 loss)
I0428 23:31:30.004567 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.08853 (* 1 = 1.08853 loss)
I0428 23:31:30.004582 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.25098 (* 0.0909091 = 0.295544 loss)
I0428 23:31:30.004595 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.23015 (* 0.0909091 = 0.29365 loss)
I0428 23:31:30.004609 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.76975 (* 0.0909091 = 0.342705 loss)
I0428 23:31:30.004622 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.43087 (* 0.0909091 = 0.311897 loss)
I0428 23:31:30.004637 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.00947 (* 0.0909091 = 0.273588 loss)
I0428 23:31:30.004649 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 3.05594 (* 0.0909091 = 0.277813 loss)
I0428 23:31:30.004663 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.43358 (* 0.0909091 = 0.130326 loss)
I0428 23:31:30.004676 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.785253 (* 0.0909091 = 0.0713867 loss)
I0428 23:31:30.004690 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.134355 (* 0.0909091 = 0.0122141 loss)
I0428 23:31:30.004704 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.0730648 (* 0.0909091 = 0.00664226 loss)
I0428 23:31:30.004719 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.054676 (* 0.0909091 = 0.00497055 loss)
I0428 23:31:30.004731 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0396049 (* 0.0909091 = 0.00360045 loss)
I0428 23:31:30.004745 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.0347742 (* 0.0909091 = 0.00316129 loss)
I0428 23:31:30.004760 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0274896 (* 0.0909091 = 0.00249905 loss)
I0428 23:31:30.004773 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0220973 (* 0.0909091 = 0.00200885 loss)
I0428 23:31:30.004787 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0150306 (* 0.0909091 = 0.00136642 loss)
I0428 23:31:30.004801 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00611716 (* 0.0909091 = 0.000556106 loss)
I0428 23:31:30.004814 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00246447 (* 0.0909091 = 0.000224043 loss)
I0428 23:31:30.004828 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00086928 (* 0.0909091 = 7.90254e-05 loss)
I0428 23:31:30.004842 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000505727 (* 0.0909091 = 4.59752e-05 loss)
I0428 23:31:30.004856 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000296704 (* 0.0909091 = 2.69731e-05 loss)
I0428 23:31:30.004870 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000201173 (* 0.0909091 = 1.82885e-05 loss)
I0428 23:31:30.004883 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:31:30.004894 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:31:30.004905 6470 solver.cpp:245] Train net output #149: total_confidence = 4.81259e-07
I0428 23:31:30.004926 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 8.49046e-06
I0428 23:31:30.004941 6470 sgd_solver.cpp:106] Iteration 2500, lr = 0.01
I0428 23:33:46.523020 6470 solver.cpp:229] Iteration 3000, loss = 10.8447
I0428 23:33:46.523159 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0943396
I0428 23:33:46.523177 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:33:46.523190 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:33:46.523202 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:33:46.523213 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0428 23:33:46.523226 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:33:46.523237 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.125
I0428 23:33:46.523249 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0428 23:33:46.523262 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:33:46.523273 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:33:46.523284 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:33:46.523296 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:33:46.523308 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:33:46.523319 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:33:46.523330 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:33:46.523342 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:33:46.523353 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:33:46.523365 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:33:46.523375 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:33:46.523391 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:33:46.523403 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:33:46.523416 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:33:46.523427 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:33:46.523437 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.727273
I0428 23:33:46.523450 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.226415
I0428 23:33:46.523478 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.22643 (* 0.3 = 0.96793 loss)
I0428 23:33:46.523496 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.04491 (* 0.3 = 0.313474 loss)
I0428 23:33:46.523510 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.28914 (* 0.0272727 = 0.0897038 loss)
I0428 23:33:46.523524 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.17152 (* 0.0272727 = 0.0864959 loss)
I0428 23:33:46.523538 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.54595 (* 0.0272727 = 0.0967076 loss)
I0428 23:33:46.523552 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.05942 (* 0.0272727 = 0.0834388 loss)
I0428 23:33:46.523566 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.09002 (* 0.0272727 = 0.0842732 loss)
I0428 23:33:46.523582 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 4.34214 (* 0.0272727 = 0.118422 loss)
I0428 23:33:46.523597 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.80795 (* 0.0272727 = 0.0493077 loss)
I0428 23:33:46.523610 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.87665 (* 0.0272727 = 0.0511814 loss)
I0428 23:33:46.523624 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.444598 (* 0.0272727 = 0.0121254 loss)
I0428 23:33:46.523638 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.482754 (* 0.0272727 = 0.013166 loss)
I0428 23:33:46.523651 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.440617 (* 0.0272727 = 0.0120168 loss)
I0428 23:33:46.523665 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.223471 (* 0.0272727 = 0.00609465 loss)
I0428 23:33:46.523679 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.232748 (* 0.0272727 = 0.00634768 loss)
I0428 23:33:46.523715 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.197353 (* 0.0272727 = 0.00538235 loss)
I0428 23:33:46.523730 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.147032 (* 0.0272727 = 0.00400995 loss)
I0428 23:33:46.523743 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.149984 (* 0.0272727 = 0.00409048 loss)
I0428 23:33:46.523757 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0959204 (* 0.0272727 = 0.00261601 loss)
I0428 23:33:46.523772 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.091085 (* 0.0272727 = 0.00248414 loss)
I0428 23:33:46.523785 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0739093 (* 0.0272727 = 0.00201571 loss)
I0428 23:33:46.523799 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0567868 (* 0.0272727 = 0.00154873 loss)
I0428 23:33:46.523813 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0694178 (* 0.0272727 = 0.00189321 loss)
I0428 23:33:46.523826 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0641564 (* 0.0272727 = 0.00174972 loss)
I0428 23:33:46.523839 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0943396
I0428 23:33:46.523850 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:33:46.523862 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:33:46.523874 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0428 23:33:46.523885 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:33:46.523897 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:33:46.523908 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0428 23:33:46.523921 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0428 23:33:46.523932 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:33:46.523943 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:33:46.523955 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:33:46.523967 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:33:46.523978 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:33:46.523989 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:33:46.524000 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:33:46.524011 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:33:46.524022 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:33:46.524034 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:33:46.524045 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:33:46.524056 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:33:46.524067 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:33:46.524078 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:33:46.524090 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:33:46.524101 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.727273
I0428 23:33:46.524112 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.320755
I0428 23:33:46.524127 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.24617 (* 0.3 = 0.97385 loss)
I0428 23:33:46.524139 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.01692 (* 0.3 = 0.305076 loss)
I0428 23:33:46.524153 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.17704 (* 0.0272727 = 0.0866465 loss)
I0428 23:33:46.524166 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.30241 (* 0.0272727 = 0.0900656 loss)
I0428 23:33:46.524190 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.39627 (* 0.0272727 = 0.0926254 loss)
I0428 23:33:46.524205 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.03811 (* 0.0272727 = 0.0828577 loss)
I0428 23:33:46.524219 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.34958 (* 0.0272727 = 0.0913523 loss)
I0428 23:33:46.524233 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 4.80856 (* 0.0272727 = 0.131142 loss)
I0428 23:33:46.524246 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.97617 (* 0.0272727 = 0.0538955 loss)
I0428 23:33:46.524260 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.76248 (* 0.0272727 = 0.0480677 loss)
I0428 23:33:46.524273 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.499159 (* 0.0272727 = 0.0136134 loss)
I0428 23:33:46.524287 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.555828 (* 0.0272727 = 0.0151589 loss)
I0428 23:33:46.524301 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.516537 (* 0.0272727 = 0.0140874 loss)
I0428 23:33:46.524314 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0747535 (* 0.0272727 = 0.00203873 loss)
I0428 23:33:46.524327 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0657345 (* 0.0272727 = 0.00179276 loss)
I0428 23:33:46.524341 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0597536 (* 0.0272727 = 0.00162964 loss)
I0428 23:33:46.524354 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0385114 (* 0.0272727 = 0.00105031 loss)
I0428 23:33:46.524369 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0373327 (* 0.0272727 = 0.00101816 loss)
I0428 23:33:46.524381 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0155712 (* 0.0272727 = 0.000424669 loss)
I0428 23:33:46.524395 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0136851 (* 0.0272727 = 0.000373231 loss)
I0428 23:33:46.524410 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00977954 (* 0.0272727 = 0.000266715 loss)
I0428 23:33:46.524423 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0115144 (* 0.0272727 = 0.000314028 loss)
I0428 23:33:46.524437 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00859932 (* 0.0272727 = 0.000234527 loss)
I0428 23:33:46.524453 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0111144 (* 0.0272727 = 0.00030312 loss)
I0428 23:33:46.524466 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.132075
I0428 23:33:46.524478 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:33:46.524489 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.25
I0428 23:33:46.524502 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:33:46.524513 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.5
I0428 23:33:46.524523 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0428 23:33:46.524535 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.125
I0428 23:33:46.524546 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0428 23:33:46.524557 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:33:46.524569 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:33:46.524580 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:33:46.524592 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:33:46.524603 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:33:46.524615 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:33:46.524628 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:33:46.524641 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:33:46.524652 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:33:46.524672 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:33:46.524684 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:33:46.524696 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:33:46.524708 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:33:46.524719 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:33:46.524730 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:33:46.524742 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.738636
I0428 23:33:46.524754 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.264151
I0428 23:33:46.524767 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.09744 (* 1 = 3.09744 loss)
I0428 23:33:46.524780 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.99883 (* 1 = 0.99883 loss)
I0428 23:33:46.524794 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.96224 (* 0.0909091 = 0.269294 loss)
I0428 23:33:46.524807 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.84176 (* 0.0909091 = 0.258341 loss)
I0428 23:33:46.524821 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.32738 (* 0.0909091 = 0.302489 loss)
I0428 23:33:46.524834 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.63062 (* 0.0909091 = 0.239147 loss)
I0428 23:33:46.524847 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.2244 (* 0.0909091 = 0.293127 loss)
I0428 23:33:46.524860 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 4.08086 (* 0.0909091 = 0.370987 loss)
I0428 23:33:46.524873 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.96717 (* 0.0909091 = 0.178833 loss)
I0428 23:33:46.524888 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.45457 (* 0.0909091 = 0.132234 loss)
I0428 23:33:46.524900 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.48294 (* 0.0909091 = 0.0439036 loss)
I0428 23:33:46.524914 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.467097 (* 0.0909091 = 0.0424634 loss)
I0428 23:33:46.524929 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.414348 (* 0.0909091 = 0.037668 loss)
I0428 23:33:46.524941 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.163676 (* 0.0909091 = 0.0148796 loss)
I0428 23:33:46.524955 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.149646 (* 0.0909091 = 0.0136042 loss)
I0428 23:33:46.524968 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.12832 (* 0.0909091 = 0.0116654 loss)
I0428 23:33:46.524982 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0981126 (* 0.0909091 = 0.00891933 loss)
I0428 23:33:46.524996 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0730514 (* 0.0909091 = 0.00664104 loss)
I0428 23:33:46.525009 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0367902 (* 0.0909091 = 0.00334456 loss)
I0428 23:33:46.525023 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0297066 (* 0.0909091 = 0.0027006 loss)
I0428 23:33:46.525038 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.017892 (* 0.0909091 = 0.00162655 loss)
I0428 23:33:46.525051 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.0140433 (* 0.0909091 = 0.00127667 loss)
I0428 23:33:46.525065 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0112081 (* 0.0909091 = 0.00101892 loss)
I0428 23:33:46.525079 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.0102917 (* 0.0909091 = 0.000935614 loss)
I0428 23:33:46.525090 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:33:46.525102 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:33:46.525113 6470 solver.cpp:245] Train net output #149: total_confidence = 2.51854e-05
I0428 23:33:46.525133 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00015453
I0428 23:33:46.525147 6470 sgd_solver.cpp:106] Iteration 3000, lr = 0.01
I0428 23:36:03.070695 6470 solver.cpp:229] Iteration 3500, loss = 10.6133
I0428 23:36:03.070845 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12
I0428 23:36:03.070865 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0428 23:36:03.070879 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:36:03.070893 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0428 23:36:03.070904 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:36:03.070915 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0428 23:36:03.070927 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0428 23:36:03.070940 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0428 23:36:03.070951 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:36:03.070963 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:36:03.070974 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:36:03.070986 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:36:03.070998 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0428 23:36:03.071010 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0428 23:36:03.071022 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:36:03.071033 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:36:03.071045 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:36:03.071058 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:36:03.071069 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:36:03.071080 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:36:03.071091 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:36:03.071104 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:36:03.071115 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:36:03.071125 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.727273
I0428 23:36:03.071137 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.34
I0428 23:36:03.071153 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.036 (* 0.3 = 0.9108 loss)
I0428 23:36:03.071168 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.05841 (* 0.3 = 0.317524 loss)
I0428 23:36:03.071182 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.66961 (* 0.0272727 = 0.0728074 loss)
I0428 23:36:03.071195 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.38738 (* 0.0272727 = 0.0923832 loss)
I0428 23:36:03.071209 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.24168 (* 0.0272727 = 0.0884094 loss)
I0428 23:36:03.071223 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.77905 (* 0.0272727 = 0.103065 loss)
I0428 23:36:03.071238 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.4842 (* 0.0272727 = 0.067751 loss)
I0428 23:36:03.071250 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.45245 (* 0.0272727 = 0.0668849 loss)
I0428 23:36:03.071264 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.16392 (* 0.0272727 = 0.0317432 loss)
I0428 23:36:03.071277 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.24524 (* 0.0272727 = 0.0339612 loss)
I0428 23:36:03.071291 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.504786 (* 0.0272727 = 0.0137669 loss)
I0428 23:36:03.071305 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.727607 (* 0.0272727 = 0.0198438 loss)
I0428 23:36:03.071322 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.903365 (* 0.0272727 = 0.0246372 loss)
I0428 23:36:03.071336 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.648257 (* 0.0272727 = 0.0176797 loss)
I0428 23:36:03.071351 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.80856 (* 0.0272727 = 0.0220516 loss)
I0428 23:36:03.071384 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0475212 (* 0.0272727 = 0.00129603 loss)
I0428 23:36:03.071399 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0239579 (* 0.0272727 = 0.000653398 loss)
I0428 23:36:03.071413 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0184555 (* 0.0272727 = 0.000503332 loss)
I0428 23:36:03.071427 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0125045 (* 0.0272727 = 0.000341031 loss)
I0428 23:36:03.071441 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00916689 (* 0.0272727 = 0.000250006 loss)
I0428 23:36:03.071455 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0103475 (* 0.0272727 = 0.000282205 loss)
I0428 23:36:03.071486 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00619668 (* 0.0272727 = 0.000169 loss)
I0428 23:36:03.071502 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00464366 (* 0.0272727 = 0.000126645 loss)
I0428 23:36:03.071517 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00486562 (* 0.0272727 = 0.000132699 loss)
I0428 23:36:03.071529 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.08
I0428 23:36:03.071542 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:36:03.071553 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:36:03.071565 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0428 23:36:03.071576 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:36:03.071588 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0428 23:36:03.071599 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0428 23:36:03.071610 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0428 23:36:03.071622 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:36:03.071633 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:36:03.071645 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:36:03.071656 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:36:03.071665 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0428 23:36:03.071672 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0428 23:36:03.071684 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:36:03.071696 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:36:03.071707 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:36:03.071718 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:36:03.071729 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:36:03.071740 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:36:03.071751 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:36:03.071763 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:36:03.071774 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:36:03.071785 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.6875
I0428 23:36:03.071796 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.28
I0428 23:36:03.071810 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.15651 (* 0.3 = 0.946954 loss)
I0428 23:36:03.071823 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.16087 (* 0.3 = 0.348261 loss)
I0428 23:36:03.071837 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.80782 (* 0.0272727 = 0.0765768 loss)
I0428 23:36:03.071851 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.17336 (* 0.0272727 = 0.0865463 loss)
I0428 23:36:03.071880 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.20398 (* 0.0272727 = 0.0873814 loss)
I0428 23:36:03.071895 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.76282 (* 0.0272727 = 0.102622 loss)
I0428 23:36:03.071908 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.91947 (* 0.0272727 = 0.0796218 loss)
I0428 23:36:03.071923 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.92288 (* 0.0272727 = 0.0797149 loss)
I0428 23:36:03.071935 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.16415 (* 0.0272727 = 0.0317496 loss)
I0428 23:36:03.071949 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.06488 (* 0.0272727 = 0.0290421 loss)
I0428 23:36:03.071962 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.470804 (* 0.0272727 = 0.0128401 loss)
I0428 23:36:03.071975 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.794918 (* 0.0272727 = 0.0216796 loss)
I0428 23:36:03.071990 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.777073 (* 0.0272727 = 0.0211929 loss)
I0428 23:36:03.072002 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.80218 (* 0.0272727 = 0.0218776 loss)
I0428 23:36:03.072016 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.760585 (* 0.0272727 = 0.0207432 loss)
I0428 23:36:03.072031 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0183587 (* 0.0272727 = 0.000500693 loss)
I0428 23:36:03.072044 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0126035 (* 0.0272727 = 0.000343733 loss)
I0428 23:36:03.072058 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00870858 (* 0.0272727 = 0.000237507 loss)
I0428 23:36:03.072072 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00329021 (* 0.0272727 = 8.97331e-05 loss)
I0428 23:36:03.072084 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00320998 (* 0.0272727 = 8.75449e-05 loss)
I0428 23:36:03.072098 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00185798 (* 0.0272727 = 5.06722e-05 loss)
I0428 23:36:03.072113 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00217927 (* 0.0272727 = 5.94346e-05 loss)
I0428 23:36:03.072126 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00224157 (* 0.0272727 = 6.11337e-05 loss)
I0428 23:36:03.072139 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00174127 (* 0.0272727 = 4.74893e-05 loss)
I0428 23:36:03.072151 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.08
I0428 23:36:03.072163 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:36:03.072175 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0428 23:36:03.072186 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:36:03.072197 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:36:03.072208 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0428 23:36:03.072221 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0428 23:36:03.072232 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0428 23:36:03.072243 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:36:03.072254 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:36:03.072266 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:36:03.072278 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:36:03.072288 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0428 23:36:03.072300 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0428 23:36:03.072311 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:36:03.072322 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:36:03.072335 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:36:03.072355 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:36:03.072370 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:36:03.072381 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:36:03.072393 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:36:03.072404 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:36:03.072415 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:36:03.072427 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.738636
I0428 23:36:03.072439 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.28
I0428 23:36:03.072453 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.10185 (* 1 = 3.10185 loss)
I0428 23:36:03.072466 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.970785 (* 1 = 0.970785 loss)
I0428 23:36:03.072480 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.92871 (* 0.0909091 = 0.266247 loss)
I0428 23:36:03.072494 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.46108 (* 0.0909091 = 0.314644 loss)
I0428 23:36:03.072507 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.95934 (* 0.0909091 = 0.269031 loss)
I0428 23:36:03.072520 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 4.36014 (* 0.0909091 = 0.396377 loss)
I0428 23:36:03.072535 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.39976 (* 0.0909091 = 0.21816 loss)
I0428 23:36:03.072547 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.16767 (* 0.0909091 = 0.197061 loss)
I0428 23:36:03.072561 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.14162 (* 0.0909091 = 0.103784 loss)
I0428 23:36:03.072576 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.09472 (* 0.0909091 = 0.0995196 loss)
I0428 23:36:03.072588 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.465748 (* 0.0909091 = 0.0423407 loss)
I0428 23:36:03.072602 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.715547 (* 0.0909091 = 0.0650497 loss)
I0428 23:36:03.072615 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.765537 (* 0.0909091 = 0.0695943 loss)
I0428 23:36:03.072629 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.535386 (* 0.0909091 = 0.0486715 loss)
I0428 23:36:03.072643 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.661974 (* 0.0909091 = 0.0601794 loss)
I0428 23:36:03.072656 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0135286 (* 0.0909091 = 0.00122988 loss)
I0428 23:36:03.072670 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0110501 (* 0.0909091 = 0.00100455 loss)
I0428 23:36:03.072685 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00497884 (* 0.0909091 = 0.000452622 loss)
I0428 23:36:03.072697 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00509184 (* 0.0909091 = 0.000462895 loss)
I0428 23:36:03.072711 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00522835 (* 0.0909091 = 0.000475305 loss)
I0428 23:36:03.072726 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00339104 (* 0.0909091 = 0.000308276 loss)
I0428 23:36:03.072738 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00292894 (* 0.0909091 = 0.000266268 loss)
I0428 23:36:03.072752 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00181888 (* 0.0909091 = 0.000165353 loss)
I0428 23:36:03.072767 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00151915 (* 0.0909091 = 0.000138104 loss)
I0428 23:36:03.072778 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:36:03.072789 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:36:03.072800 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000330788
I0428 23:36:03.072821 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00955349
I0428 23:36:03.072835 6470 sgd_solver.cpp:106] Iteration 3500, lr = 0.01
I0428 23:38:19.143426 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 32.8872 > 30) by scale factor 0.912208
I0428 23:38:19.907325 6470 solver.cpp:229] Iteration 4000, loss = 10.4427
I0428 23:38:19.907400 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0377358
I0428 23:38:19.907418 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:38:19.907431 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:38:19.907444 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0428 23:38:19.907455 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0428 23:38:19.907480 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:38:19.907495 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.125
I0428 23:38:19.907506 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0428 23:38:19.907518 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:38:19.907531 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0428 23:38:19.907541 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0428 23:38:19.907552 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:38:19.907564 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:38:19.907578 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:38:19.907590 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:38:19.907603 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:38:19.907613 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:38:19.907625 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:38:19.907636 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:38:19.907649 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:38:19.907660 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:38:19.907672 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:38:19.907683 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:38:19.907696 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.710227
I0428 23:38:19.907707 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.245283
I0428 23:38:19.907723 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.06499 (* 0.3 = 0.919497 loss)
I0428 23:38:19.907738 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.06002 (* 0.3 = 0.318005 loss)
I0428 23:38:19.907752 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.0058 (* 0.0272727 = 0.0819764 loss)
I0428 23:38:19.907766 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.17308 (* 0.0272727 = 0.0865387 loss)
I0428 23:38:19.907779 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.07515 (* 0.0272727 = 0.0838678 loss)
I0428 23:38:19.907796 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.21753 (* 0.0272727 = 0.0877507 loss)
I0428 23:38:19.907811 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.09696 (* 0.0272727 = 0.0844626 loss)
I0428 23:38:19.907825 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.58572 (* 0.0272727 = 0.0977925 loss)
I0428 23:38:19.907838 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.29566 (* 0.0272727 = 0.062609 loss)
I0428 23:38:19.907852 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.42838 (* 0.0272727 = 0.0389559 loss)
I0428 23:38:19.907866 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.196681 (* 0.0272727 = 0.00536404 loss)
I0428 23:38:19.907881 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.17353 (* 0.0272727 = 0.00473265 loss)
I0428 23:38:19.907893 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.15312 (* 0.0272727 = 0.00417599 loss)
I0428 23:38:19.907907 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0632366 (* 0.0272727 = 0.00172463 loss)
I0428 23:38:19.907958 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0741323 (* 0.0272727 = 0.00202179 loss)
I0428 23:38:19.907973 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0435899 (* 0.0272727 = 0.00118881 loss)
I0428 23:38:19.907986 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0385319 (* 0.0272727 = 0.00105087 loss)
I0428 23:38:19.908000 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.017419 (* 0.0272727 = 0.000475064 loss)
I0428 23:38:19.908015 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00932098 (* 0.0272727 = 0.000254209 loss)
I0428 23:38:19.908028 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00647661 (* 0.0272727 = 0.000176635 loss)
I0428 23:38:19.908041 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00393809 (* 0.0272727 = 0.000107402 loss)
I0428 23:38:19.908056 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00248326 (* 0.0272727 = 6.77254e-05 loss)
I0428 23:38:19.908069 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.003227 (* 0.0272727 = 8.8009e-05 loss)
I0428 23:38:19.908083 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00231603 (* 0.0272727 = 6.31646e-05 loss)
I0428 23:38:19.908095 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0
I0428 23:38:19.908107 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:38:19.908118 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:38:19.908129 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:38:19.908140 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:38:19.908152 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0428 23:38:19.908164 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0428 23:38:19.908174 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.375
I0428 23:38:19.908186 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:38:19.908198 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0428 23:38:19.908210 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0428 23:38:19.908221 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:38:19.908232 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:38:19.908243 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:38:19.908254 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:38:19.908265 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:38:19.908277 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:38:19.908288 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:38:19.908295 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:38:19.908303 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:38:19.908310 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:38:19.908324 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:38:19.908334 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:38:19.908345 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.698864
I0428 23:38:19.908357 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.113208
I0428 23:38:19.908371 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.22134 (* 0.3 = 0.966401 loss)
I0428 23:38:19.908385 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.05857 (* 0.3 = 0.317571 loss)
I0428 23:38:19.908398 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.13858 (* 0.0272727 = 0.0855977 loss)
I0428 23:38:19.908412 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.31976 (* 0.0272727 = 0.090539 loss)
I0428 23:38:19.908437 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.17773 (* 0.0272727 = 0.0866655 loss)
I0428 23:38:19.908450 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.2722 (* 0.0272727 = 0.0892418 loss)
I0428 23:38:19.908464 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.25374 (* 0.0272727 = 0.0887383 loss)
I0428 23:38:19.908478 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.78771 (* 0.0272727 = 0.103301 loss)
I0428 23:38:19.908491 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.64287 (* 0.0272727 = 0.0720783 loss)
I0428 23:38:19.908504 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.80925 (* 0.0272727 = 0.0493433 loss)
I0428 23:38:19.908519 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.116165 (* 0.0272727 = 0.00316812 loss)
I0428 23:38:19.908532 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.0537414 (* 0.0272727 = 0.00146568 loss)
I0428 23:38:19.908546 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0536398 (* 0.0272727 = 0.0014629 loss)
I0428 23:38:19.908560 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0358684 (* 0.0272727 = 0.00097823 loss)
I0428 23:38:19.908574 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0209881 (* 0.0272727 = 0.000572402 loss)
I0428 23:38:19.908588 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0209053 (* 0.0272727 = 0.000570145 loss)
I0428 23:38:19.908602 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0112758 (* 0.0272727 = 0.000307522 loss)
I0428 23:38:19.908615 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00727887 (* 0.0272727 = 0.000198515 loss)
I0428 23:38:19.908632 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00214705 (* 0.0272727 = 5.8556e-05 loss)
I0428 23:38:19.908646 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00210394 (* 0.0272727 = 5.73802e-05 loss)
I0428 23:38:19.908659 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00120714 (* 0.0272727 = 3.2922e-05 loss)
I0428 23:38:19.908674 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000906643 (* 0.0272727 = 2.47266e-05 loss)
I0428 23:38:19.908686 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000804059 (* 0.0272727 = 2.19289e-05 loss)
I0428 23:38:19.908700 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00095026 (* 0.0272727 = 2.59162e-05 loss)
I0428 23:38:19.908711 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0566038
I0428 23:38:19.908723 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0428 23:38:19.908735 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:38:19.908746 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:38:19.908757 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:38:19.908768 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:38:19.908779 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.125
I0428 23:38:19.908792 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0428 23:38:19.908802 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:38:19.908813 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0428 23:38:19.908825 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0428 23:38:19.908836 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:38:19.908851 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:38:19.908864 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:38:19.908874 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:38:19.908885 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:38:19.908906 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:38:19.908920 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:38:19.908931 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:38:19.908941 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:38:19.908953 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:38:19.908964 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:38:19.908977 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:38:19.908987 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.715909
I0428 23:38:19.908999 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.169811
I0428 23:38:19.909013 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.15795 (* 1 = 3.15795 loss)
I0428 23:38:19.909026 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.05526 (* 1 = 1.05526 loss)
I0428 23:38:19.909040 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.76934 (* 0.0909091 = 0.251758 loss)
I0428 23:38:19.909054 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.02805 (* 0.0909091 = 0.275277 loss)
I0428 23:38:19.909067 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.37138 (* 0.0909091 = 0.306489 loss)
I0428 23:38:19.909080 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.44983 (* 0.0909091 = 0.313621 loss)
I0428 23:38:19.909095 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.76098 (* 0.0909091 = 0.250998 loss)
I0428 23:38:19.909108 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 3.22423 (* 0.0909091 = 0.293112 loss)
I0428 23:38:19.909121 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.50863 (* 0.0909091 = 0.228058 loss)
I0428 23:38:19.909135 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.24733 (* 0.0909091 = 0.113394 loss)
I0428 23:38:19.909148 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.0528727 (* 0.0909091 = 0.00480661 loss)
I0428 23:38:19.909162 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.0303373 (* 0.0909091 = 0.00275793 loss)
I0428 23:38:19.909176 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0243676 (* 0.0909091 = 0.00221524 loss)
I0428 23:38:19.909189 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0181491 (* 0.0909091 = 0.00164992 loss)
I0428 23:38:19.909204 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.0151853 (* 0.0909091 = 0.00138048 loss)
I0428 23:38:19.909216 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0136078 (* 0.0909091 = 0.00123708 loss)
I0428 23:38:19.909230 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0126477 (* 0.0909091 = 0.00114979 loss)
I0428 23:38:19.909245 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0128755 (* 0.0909091 = 0.0011705 loss)
I0428 23:38:19.909258 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0108401 (* 0.0909091 = 0.000985459 loss)
I0428 23:38:19.909272 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00925457 (* 0.0909091 = 0.000841325 loss)
I0428 23:38:19.909286 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00571262 (* 0.0909091 = 0.000519329 loss)
I0428 23:38:19.909299 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00381014 (* 0.0909091 = 0.000346376 loss)
I0428 23:38:19.909313 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0024264 (* 0.0909091 = 0.000220581 loss)
I0428 23:38:19.909327 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00181046 (* 0.0909091 = 0.000164587 loss)
I0428 23:38:19.909338 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:38:19.909350 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:38:19.909369 6470 solver.cpp:245] Train net output #149: total_confidence = 7.34608e-06
I0428 23:38:19.909382 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000268142
I0428 23:38:19.909395 6470 sgd_solver.cpp:106] Iteration 4000, lr = 0.01
I0428 23:40:10.789736 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.3854 > 30) by scale factor 0.955857
I0428 23:40:36.427985 6470 solver.cpp:229] Iteration 4500, loss = 10.3899
I0428 23:40:36.428066 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.031746
I0428 23:40:36.428083 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:40:36.428097 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:40:36.428109 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:40:36.428122 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:40:36.428133 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:40:36.428144 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:40:36.428156 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0428 23:40:36.428169 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:40:36.428180 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0428 23:40:36.428191 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:40:36.428203 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:40:36.428215 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0428 23:40:36.428227 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0428 23:40:36.428238 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0428 23:40:36.428251 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0428 23:40:36.428262 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0428 23:40:36.428274 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 0.875
I0428 23:40:36.428287 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:40:36.428298 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:40:36.428309 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:40:36.428321 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:40:36.428333 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:40:36.428344 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.647727
I0428 23:40:36.428356 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.253968
I0428 23:40:36.428372 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.45562 (* 0.3 = 1.03668 loss)
I0428 23:40:36.428386 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.30216 (* 0.3 = 0.390648 loss)
I0428 23:40:36.428401 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.20943 (* 0.0272727 = 0.0875299 loss)
I0428 23:40:36.428416 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.66799 (* 0.0272727 = 0.100036 loss)
I0428 23:40:36.428429 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.46568 (* 0.0272727 = 0.0945184 loss)
I0428 23:40:36.428442 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.48158 (* 0.0272727 = 0.0949522 loss)
I0428 23:40:36.428457 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.41051 (* 0.0272727 = 0.093014 loss)
I0428 23:40:36.428470 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.35149 (* 0.0272727 = 0.0914043 loss)
I0428 23:40:36.428484 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.05764 (* 0.0272727 = 0.0561174 loss)
I0428 23:40:36.428498 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.989459 (* 0.0272727 = 0.0269852 loss)
I0428 23:40:36.428513 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.811434 (* 0.0272727 = 0.02213 loss)
I0428 23:40:36.428525 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.418519 (* 0.0272727 = 0.0114141 loss)
I0428 23:40:36.428539 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.44867 (* 0.0272727 = 0.0122365 loss)
I0428 23:40:36.428553 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.552404 (* 0.0272727 = 0.0150656 loss)
I0428 23:40:36.428606 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.52976 (* 0.0272727 = 0.014448 loss)
I0428 23:40:36.428624 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.531046 (* 0.0272727 = 0.0144831 loss)
I0428 23:40:36.428639 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.584078 (* 0.0272727 = 0.0159294 loss)
I0428 23:40:36.428653 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.879429 (* 0.0272727 = 0.0239844 loss)
I0428 23:40:36.428666 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.890432 (* 0.0272727 = 0.0242845 loss)
I0428 23:40:36.428681 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0137878 (* 0.0272727 = 0.00037603 loss)
I0428 23:40:36.428695 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.010042 (* 0.0272727 = 0.000273874 loss)
I0428 23:40:36.428709 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00932139 (* 0.0272727 = 0.00025422 loss)
I0428 23:40:36.428724 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0112831 (* 0.0272727 = 0.000307721 loss)
I0428 23:40:36.428737 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00830374 (* 0.0272727 = 0.000226466 loss)
I0428 23:40:36.428750 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0793651
I0428 23:40:36.428761 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:40:36.428772 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:40:36.428784 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:40:36.428796 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:40:36.428807 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0
I0428 23:40:36.428819 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0428 23:40:36.428830 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0428 23:40:36.428843 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:40:36.428853 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0428 23:40:36.428865 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:40:36.428876 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:40:36.428889 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0428 23:40:36.428900 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0428 23:40:36.428911 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0428 23:40:36.428922 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0428 23:40:36.428938 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0428 23:40:36.428951 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 0.875
I0428 23:40:36.428962 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:40:36.428973 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:40:36.428985 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:40:36.428997 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:40:36.429008 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:40:36.429019 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.664773
I0428 23:40:36.429031 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.238095
I0428 23:40:36.429045 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.23755 (* 0.3 = 0.971264 loss)
I0428 23:40:36.429060 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.30594 (* 0.3 = 0.391781 loss)
I0428 23:40:36.429070 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.19987 (* 0.0272727 = 0.0872691 loss)
I0428 23:40:36.429100 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.56225 (* 0.0272727 = 0.0971522 loss)
I0428 23:40:36.429116 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.35875 (* 0.0272727 = 0.0916022 loss)
I0428 23:40:36.429131 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.55649 (* 0.0272727 = 0.0969951 loss)
I0428 23:40:36.429143 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.58451 (* 0.0272727 = 0.0977594 loss)
I0428 23:40:36.429157 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.7331 (* 0.0272727 = 0.101812 loss)
I0428 23:40:36.429174 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.41454 (* 0.0272727 = 0.0658511 loss)
I0428 23:40:36.429203 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.07817 (* 0.0272727 = 0.0294046 loss)
I0428 23:40:36.429227 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.930588 (* 0.0272727 = 0.0253797 loss)
I0428 23:40:36.429244 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.551434 (* 0.0272727 = 0.0150391 loss)
I0428 23:40:36.429257 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.518757 (* 0.0272727 = 0.0141479 loss)
I0428 23:40:36.429271 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.55713 (* 0.0272727 = 0.0151945 loss)
I0428 23:40:36.429285 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.505306 (* 0.0272727 = 0.0137811 loss)
I0428 23:40:36.429298 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.587315 (* 0.0272727 = 0.0160177 loss)
I0428 23:40:36.429312 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.489905 (* 0.0272727 = 0.013361 loss)
I0428 23:40:36.429325 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.655143 (* 0.0272727 = 0.0178675 loss)
I0428 23:40:36.429339 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.70913 (* 0.0272727 = 0.0193399 loss)
I0428 23:40:36.429353 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0514931 (* 0.0272727 = 0.00140436 loss)
I0428 23:40:36.429368 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0328104 (* 0.0272727 = 0.000894828 loss)
I0428 23:40:36.429380 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0307489 (* 0.0272727 = 0.000838607 loss)
I0428 23:40:36.429394 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0318894 (* 0.0272727 = 0.000869711 loss)
I0428 23:40:36.429409 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0313198 (* 0.0272727 = 0.000854177 loss)
I0428 23:40:36.429420 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0793651
I0428 23:40:36.429431 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:40:36.429443 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:40:36.429455 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:40:36.429466 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:40:36.429476 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:40:36.429488 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.125
I0428 23:40:36.429500 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0428 23:40:36.429512 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:40:36.429523 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0428 23:40:36.429534 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:40:36.429545 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:40:36.429558 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0428 23:40:36.429569 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0428 23:40:36.429580 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0428 23:40:36.429591 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0428 23:40:36.429613 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0428 23:40:36.429626 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 0.875
I0428 23:40:36.429638 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:40:36.429649 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:40:36.429661 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:40:36.429677 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:40:36.429688 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:40:36.429700 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.670455
I0428 23:40:36.429713 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.269841
I0428 23:40:36.429725 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.19234 (* 1 = 3.19234 loss)
I0428 23:40:36.429739 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.22256 (* 1 = 1.22256 loss)
I0428 23:40:36.429754 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.90558 (* 0.0909091 = 0.264144 loss)
I0428 23:40:36.429767 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.4626 (* 0.0909091 = 0.314782 loss)
I0428 23:40:36.429780 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.23302 (* 0.0909091 = 0.293911 loss)
I0428 23:40:36.429795 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.04967 (* 0.0909091 = 0.277243 loss)
I0428 23:40:36.429807 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.10598 (* 0.0909091 = 0.282362 loss)
I0428 23:40:36.429821 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 3.38499 (* 0.0909091 = 0.307726 loss)
I0428 23:40:36.429834 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.93362 (* 0.0909091 = 0.175784 loss)
I0428 23:40:36.429848 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.06699 (* 0.0909091 = 0.0969989 loss)
I0428 23:40:36.429862 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.865713 (* 0.0909091 = 0.0787011 loss)
I0428 23:40:36.429877 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.376081 (* 0.0909091 = 0.0341892 loss)
I0428 23:40:36.429889 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.392858 (* 0.0909091 = 0.0357143 loss)
I0428 23:40:36.429903 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.393843 (* 0.0909091 = 0.0358039 loss)
I0428 23:40:36.429916 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.37762 (* 0.0909091 = 0.0343291 loss)
I0428 23:40:36.429930 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.611994 (* 0.0909091 = 0.0556358 loss)
I0428 23:40:36.429944 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.469895 (* 0.0909091 = 0.0427177 loss)
I0428 23:40:36.429957 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.719909 (* 0.0909091 = 0.0654463 loss)
I0428 23:40:36.429971 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.806898 (* 0.0909091 = 0.0733544 loss)
I0428 23:40:36.429989 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00134504 (* 0.0909091 = 0.000122277 loss)
I0428 23:40:36.430004 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000750702 (* 0.0909091 = 6.82457e-05 loss)
I0428 23:40:36.430018 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000387948 (* 0.0909091 = 3.5268e-05 loss)
I0428 23:40:36.430032 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000296055 (* 0.0909091 = 2.69141e-05 loss)
I0428 23:40:36.430047 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000211924 (* 0.0909091 = 1.92658e-05 loss)
I0428 23:40:36.430058 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:40:36.430069 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:40:36.430090 6470 solver.cpp:245] Train net output #149: total_confidence = 2.05411e-07
I0428 23:40:36.430104 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 2.49432e-06
I0428 23:40:36.430117 6470 sgd_solver.cpp:106] Iteration 4500, lr = 0.01
I0428 23:42:52.552018 6470 solver.cpp:338] Iteration 5000, Testing net (#0)
I0428 23:43:34.220815 6470 solver.cpp:393] Test loss: 9.34922
I0428 23:43:34.220945 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0745373
I0428 23:43:34.220964 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.103
I0428 23:43:34.220978 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.098
I0428 23:43:34.220989 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.075
I0428 23:43:34.221001 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.16
I0428 23:43:34.221014 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.328
I0428 23:43:34.221025 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.493
I0428 23:43:34.221036 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.746
I0428 23:43:34.221047 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.919
I0428 23:43:34.221060 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.992
I0428 23:43:34.221071 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.999
I0428 23:43:34.221082 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 1
I0428 23:43:34.221093 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 1
I0428 23:43:34.221104 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 1
I0428 23:43:34.221115 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 1
I0428 23:43:34.221127 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 1
I0428 23:43:34.221138 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 1
I0428 23:43:34.221148 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 1
I0428 23:43:34.221159 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 1
I0428 23:43:34.221170 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0428 23:43:34.221181 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0428 23:43:34.221192 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0428 23:43:34.221204 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0428 23:43:34.221215 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.765819
I0428 23:43:34.221226 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.215553
I0428 23:43:34.221241 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.32663 (* 0.3 = 0.997989 loss)
I0428 23:43:34.221256 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.890529 (* 0.3 = 0.267159 loss)
I0428 23:43:34.221269 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 3.05251 (* 0.0272727 = 0.0832503 loss)
I0428 23:43:34.221282 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 3.20021 (* 0.0272727 = 0.0872785 loss)
I0428 23:43:34.221295 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 3.29335 (* 0.0272727 = 0.0898185 loss)
I0428 23:43:34.221308 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 3.1002 (* 0.0272727 = 0.0845509 loss)
I0428 23:43:34.221325 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.64696 (* 0.0272727 = 0.0721898 loss)
I0428 23:43:34.221338 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.17786 (* 0.0272727 = 0.0593961 loss)
I0428 23:43:34.221352 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.30334 (* 0.0272727 = 0.0355456 loss)
I0428 23:43:34.221365 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.550317 (* 0.0272727 = 0.0150087 loss)
I0428 23:43:34.221379 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.0752755 (* 0.0272727 = 0.00205297 loss)
I0428 23:43:34.221392 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.0440914 (* 0.0272727 = 0.00120249 loss)
I0428 23:43:34.221405 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.0322531 (* 0.0272727 = 0.000879631 loss)
I0428 23:43:34.221420 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.0234507 (* 0.0272727 = 0.000639564 loss)
I0428 23:43:34.221432 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.0170194 (* 0.0272727 = 0.000464164 loss)
I0428 23:43:34.221464 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.0143451 (* 0.0272727 = 0.00039123 loss)
I0428 23:43:34.221478 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.0114019 (* 0.0272727 = 0.000310961 loss)
I0428 23:43:34.221493 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.00834557 (* 0.0272727 = 0.000227606 loss)
I0428 23:43:34.221505 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.00528081 (* 0.0272727 = 0.000144022 loss)
I0428 23:43:34.221519 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.00430858 (* 0.0272727 = 0.000117507 loss)
I0428 23:43:34.221532 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00353596 (* 0.0272727 = 9.64353e-05 loss)
I0428 23:43:34.221546 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.00340661 (* 0.0272727 = 9.29074e-05 loss)
I0428 23:43:34.221560 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.00288206 (* 0.0272727 = 7.86015e-05 loss)
I0428 23:43:34.221572 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.00269066 (* 0.0272727 = 7.33817e-05 loss)
I0428 23:43:34.221585 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0756968
I0428 23:43:34.221596 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.1
I0428 23:43:34.221606 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.098
I0428 23:43:34.221618 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.066
I0428 23:43:34.221629 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.163
I0428 23:43:34.221640 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.327
I0428 23:43:34.221652 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.494
I0428 23:43:34.221663 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.746
I0428 23:43:34.221674 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.92
I0428 23:43:34.221685 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.995
I0428 23:43:34.221696 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.999
I0428 23:43:34.221709 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 1
I0428 23:43:34.221719 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 1
I0428 23:43:34.221730 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 1
I0428 23:43:34.221740 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 1
I0428 23:43:34.221751 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 1
I0428 23:43:34.221762 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 1
I0428 23:43:34.221773 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 1
I0428 23:43:34.221784 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 1
I0428 23:43:34.221796 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0428 23:43:34.221807 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0428 23:43:34.221817 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0428 23:43:34.221827 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0428 23:43:34.221838 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.76491
I0428 23:43:34.221849 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.216426
I0428 23:43:34.221863 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.30403 (* 0.3 = 0.991209 loss)
I0428 23:43:34.221876 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.895554 (* 0.3 = 0.268666 loss)
I0428 23:43:34.221891 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 3.07195 (* 0.0272727 = 0.0837804 loss)
I0428 23:43:34.221904 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 3.2068 (* 0.0272727 = 0.0874581 loss)
I0428 23:43:34.221917 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.28989 (* 0.0272727 = 0.0897243 loss)
I0428 23:43:34.221941 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 3.10242 (* 0.0272727 = 0.0846116 loss)
I0428 23:43:34.221956 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.68937 (* 0.0272727 = 0.0733465 loss)
I0428 23:43:34.221972 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.21457 (* 0.0272727 = 0.0603974 loss)
I0428 23:43:34.221987 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.33566 (* 0.0272727 = 0.0364271 loss)
I0428 23:43:34.221999 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.584912 (* 0.0272727 = 0.0159522 loss)
I0428 23:43:34.222012 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.085381 (* 0.0272727 = 0.00232857 loss)
I0428 23:43:34.222026 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.0529955 (* 0.0272727 = 0.00144533 loss)
I0428 23:43:34.222039 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.0395948 (* 0.0272727 = 0.00107986 loss)
I0428 23:43:34.222054 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.0310241 (* 0.0272727 = 0.000846111 loss)
I0428 23:43:34.222066 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.0239738 (* 0.0272727 = 0.000653832 loss)
I0428 23:43:34.222079 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.0204946 (* 0.0272727 = 0.000558943 loss)
I0428 23:43:34.222093 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.0158099 (* 0.0272727 = 0.000431179 loss)
I0428 23:43:34.222106 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.0124 (* 0.0272727 = 0.000338181 loss)
I0428 23:43:34.222120 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.00904643 (* 0.0272727 = 0.000246721 loss)
I0428 23:43:34.222133 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.00831051 (* 0.0272727 = 0.00022665 loss)
I0428 23:43:34.222146 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.00716676 (* 0.0272727 = 0.000195457 loss)
I0428 23:43:34.222159 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00676805 (* 0.0272727 = 0.000184583 loss)
I0428 23:43:34.222173 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.00627934 (* 0.0272727 = 0.000171255 loss)
I0428 23:43:34.222187 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.00661475 (* 0.0272727 = 0.000180402 loss)
I0428 23:43:34.222198 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.0823972
I0428 23:43:34.222208 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.112
I0428 23:43:34.222219 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.085
I0428 23:43:34.222230 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.081
I0428 23:43:34.222241 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.149
I0428 23:43:34.222252 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.333
I0428 23:43:34.222264 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.489
I0428 23:43:34.222275 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.743
I0428 23:43:34.222286 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.916
I0428 23:43:34.222297 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.99
I0428 23:43:34.222308 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.999
I0428 23:43:34.222319 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 1
I0428 23:43:34.222331 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 1
I0428 23:43:34.222342 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 1
I0428 23:43:34.222352 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 1
I0428 23:43:34.222364 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 1
I0428 23:43:34.222376 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 1
I0428 23:43:34.222388 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 1
I0428 23:43:34.222407 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 1
I0428 23:43:34.222419 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0428 23:43:34.222430 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0428 23:43:34.222441 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0428 23:43:34.222452 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0428 23:43:34.222463 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.764683
I0428 23:43:34.222475 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.237405
I0428 23:43:34.222487 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 3.14908 (* 1 = 3.14908 loss)
I0428 23:43:34.222501 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.866275 (* 1 = 0.866275 loss)
I0428 23:43:34.222514 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.9925 (* 0.0909091 = 0.272045 loss)
I0428 23:43:34.222527 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 3.13872 (* 0.0909091 = 0.285338 loss)
I0428 23:43:34.222540 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 3.20759 (* 0.0909091 = 0.291599 loss)
I0428 23:43:34.222553 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 3.03945 (* 0.0909091 = 0.276313 loss)
I0428 23:43:34.222566 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.58682 (* 0.0909091 = 0.235166 loss)
I0428 23:43:34.222579 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.12415 (* 0.0909091 = 0.193105 loss)
I0428 23:43:34.222592 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.22742 (* 0.0909091 = 0.111584 loss)
I0428 23:43:34.222605 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.500683 (* 0.0909091 = 0.0455167 loss)
I0428 23:43:34.222618 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.075512 (* 0.0909091 = 0.00686473 loss)
I0428 23:43:34.222631 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.0452636 (* 0.0909091 = 0.00411487 loss)
I0428 23:43:34.222645 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.0330721 (* 0.0909091 = 0.00300655 loss)
I0428 23:43:34.222657 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.0221688 (* 0.0909091 = 0.00201535 loss)
I0428 23:43:34.222671 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.0176837 (* 0.0909091 = 0.00160761 loss)
I0428 23:43:34.222684 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.0141807 (* 0.0909091 = 0.00128915 loss)
I0428 23:43:34.222697 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.0109792 (* 0.0909091 = 0.000998111 loss)
I0428 23:43:34.222710 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.00911073 (* 0.0909091 = 0.000828249 loss)
I0428 23:43:34.222723 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.00726279 (* 0.0909091 = 0.000660253 loss)
I0428 23:43:34.222736 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.00619772 (* 0.0909091 = 0.000563429 loss)
I0428 23:43:34.222749 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00598993 (* 0.0909091 = 0.000544539 loss)
I0428 23:43:34.222762 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.00520173 (* 0.0909091 = 0.000472884 loss)
I0428 23:43:34.222775 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.0047041 (* 0.0909091 = 0.000427645 loss)
I0428 23:43:34.222789 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.00425583 (* 0.0909091 = 0.000386894 loss)
I0428 23:43:34.222800 6470 solver.cpp:406] Test net output #147: total_accuracy = 0.001
I0428 23:43:34.222811 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0
I0428 23:43:34.222822 6470 solver.cpp:406] Test net output #149: total_confidence = 0.000149364
I0428 23:43:34.222833 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.000414384
I0428 23:43:34.222854 6470 solver.cpp:338] Iteration 5000, Testing net (#1)
I0428 23:44:15.699610 6470 solver.cpp:393] Test loss: 9.88118
I0428 23:44:15.699754 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0736422
I0428 23:44:15.699774 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.105
I0428 23:44:15.699789 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.088
I0428 23:44:15.699800 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.106
I0428 23:44:15.699812 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.194
I0428 23:44:15.699823 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.328
I0428 23:44:15.699836 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.422
I0428 23:44:15.699847 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.659
I0428 23:44:15.699858 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.803
I0428 23:44:15.699869 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.885
I0428 23:44:15.699880 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.911
I0428 23:44:15.699892 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 0.923
I0428 23:44:15.699903 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 0.941
I0428 23:44:15.699915 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 0.955
I0428 23:44:15.699926 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 0.962
I0428 23:44:15.699939 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 0.976
I0428 23:44:15.699951 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 0.987
I0428 23:44:15.699972 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 0.996
I0428 23:44:15.699995 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 0.999
I0428 23:44:15.700008 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0428 23:44:15.700019 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0428 23:44:15.700031 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0428 23:44:15.700042 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0428 23:44:15.700052 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.733273
I0428 23:44:15.700063 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.225634
I0428 23:44:15.700079 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.26797 (* 0.3 = 0.980391 loss)
I0428 23:44:15.700093 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 1.00944 (* 0.3 = 0.302832 loss)
I0428 23:44:15.700108 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 3.03046 (* 0.0272727 = 0.0826488 loss)
I0428 23:44:15.700121 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 3.17968 (* 0.0272727 = 0.0867184 loss)
I0428 23:44:15.700134 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 3.24684 (* 0.0272727 = 0.0885502 loss)
I0428 23:44:15.700148 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 3.0535 (* 0.0272727 = 0.0832772 loss)
I0428 23:44:15.700161 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.65231 (* 0.0272727 = 0.0723358 loss)
I0428 23:44:15.700175 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.39765 (* 0.0272727 = 0.0653905 loss)
I0428 23:44:15.700187 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.58859 (* 0.0272727 = 0.0433252 loss)
I0428 23:44:15.700201 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.941553 (* 0.0272727 = 0.0256787 loss)
I0428 23:44:15.700213 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.519616 (* 0.0272727 = 0.0141713 loss)
I0428 23:44:15.700227 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.427745 (* 0.0272727 = 0.0116658 loss)
I0428 23:44:15.700240 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.384878 (* 0.0272727 = 0.0104967 loss)
I0428 23:44:15.700254 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.311705 (* 0.0272727 = 0.00850106 loss)
I0428 23:44:15.700268 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.246705 (* 0.0272727 = 0.00672833 loss)
I0428 23:44:15.700301 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.217122 (* 0.0272727 = 0.0059215 loss)
I0428 23:44:15.700320 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.153347 (* 0.0272727 = 0.00418218 loss)
I0428 23:44:15.700333 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.0900425 (* 0.0272727 = 0.0024557 loss)
I0428 23:44:15.700347 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.0369176 (* 0.0272727 = 0.00100684 loss)
I0428 23:44:15.700361 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.0122805 (* 0.0272727 = 0.000334922 loss)
I0428 23:44:15.700374 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00475166 (* 0.0272727 = 0.000129591 loss)
I0428 23:44:15.700388 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.00443055 (* 0.0272727 = 0.000120833 loss)
I0428 23:44:15.700402 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.00373153 (* 0.0272727 = 0.000101769 loss)
I0428 23:44:15.700415 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.0034512 (* 0.0272727 = 9.41237e-05 loss)
I0428 23:44:15.700428 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0733038
I0428 23:44:15.700439 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.108
I0428 23:44:15.700448 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.083
I0428 23:44:15.700454 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.075
I0428 23:44:15.700461 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.173
I0428 23:44:15.700469 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.322
I0428 23:44:15.700481 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.427
I0428 23:44:15.700492 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.659
I0428 23:44:15.700503 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.803
I0428 23:44:15.700515 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.882
I0428 23:44:15.700526 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.911
I0428 23:44:15.700537 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 0.923
I0428 23:44:15.700548 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 0.941
I0428 23:44:15.700559 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 0.955
I0428 23:44:15.700570 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 0.962
I0428 23:44:15.700582 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 0.976
I0428 23:44:15.700592 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 0.987
I0428 23:44:15.700604 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 0.996
I0428 23:44:15.700615 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 0.999
I0428 23:44:15.700626 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0428 23:44:15.700637 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0428 23:44:15.700649 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0428 23:44:15.700659 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0428 23:44:15.700670 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.731909
I0428 23:44:15.700681 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.215833
I0428 23:44:15.700695 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.26207 (* 0.3 = 0.97862 loss)
I0428 23:44:15.700708 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 1.01816 (* 0.3 = 0.305448 loss)
I0428 23:44:15.700721 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 3.04729 (* 0.0272727 = 0.0831079 loss)
I0428 23:44:15.700736 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 3.20655 (* 0.0272727 = 0.0874514 loss)
I0428 23:44:15.700759 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.24883 (* 0.0272727 = 0.0886045 loss)
I0428 23:44:15.700778 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 3.06882 (* 0.0272727 = 0.0836952 loss)
I0428 23:44:15.700790 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.70897 (* 0.0272727 = 0.073881 loss)
I0428 23:44:15.700804 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.42889 (* 0.0272727 = 0.0662424 loss)
I0428 23:44:15.700817 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.60052 (* 0.0272727 = 0.0436506 loss)
I0428 23:44:15.700830 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.967348 (* 0.0272727 = 0.0263822 loss)
I0428 23:44:15.700844 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.528639 (* 0.0272727 = 0.0144174 loss)
I0428 23:44:15.700857 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.435329 (* 0.0272727 = 0.0118726 loss)
I0428 23:44:15.700870 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.391257 (* 0.0272727 = 0.0106706 loss)
I0428 23:44:15.700884 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.319631 (* 0.0272727 = 0.00871722 loss)
I0428 23:44:15.700897 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.254192 (* 0.0272727 = 0.00693252 loss)
I0428 23:44:15.700911 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.22242 (* 0.0272727 = 0.00606601 loss)
I0428 23:44:15.700924 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.160062 (* 0.0272727 = 0.00436533 loss)
I0428 23:44:15.700937 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.0963658 (* 0.0272727 = 0.00262816 loss)
I0428 23:44:15.700951 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.0392417 (* 0.0272727 = 0.00107023 loss)
I0428 23:44:15.700964 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.0154227 (* 0.0272727 = 0.000420619 loss)
I0428 23:44:15.700978 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.00773817 (* 0.0272727 = 0.000211041 loss)
I0428 23:44:15.700990 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00692739 (* 0.0272727 = 0.000188929 loss)
I0428 23:44:15.701004 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.00618781 (* 0.0272727 = 0.000168758 loss)
I0428 23:44:15.701017 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.00650157 (* 0.0272727 = 0.000177315 loss)
I0428 23:44:15.701030 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.0823733
I0428 23:44:15.701040 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.118
I0428 23:44:15.701052 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.084
I0428 23:44:15.701063 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.088
I0428 23:44:15.701074 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.181
I0428 23:44:15.701086 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.335
I0428 23:44:15.701097 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.439
I0428 23:44:15.701107 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.658
I0428 23:44:15.701118 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.808
I0428 23:44:15.701130 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.889
I0428 23:44:15.701141 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.913
I0428 23:44:15.701153 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.923
I0428 23:44:15.701164 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 0.941
I0428 23:44:15.701174 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 0.955
I0428 23:44:15.701185 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 0.962
I0428 23:44:15.701196 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 0.976
I0428 23:44:15.701207 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 0.987
I0428 23:44:15.701227 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 0.996
I0428 23:44:15.701241 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 0.999
I0428 23:44:15.701251 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0428 23:44:15.701262 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0428 23:44:15.701273 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0428 23:44:15.701284 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0428 23:44:15.701295 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.731682
I0428 23:44:15.701306 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.25206
I0428 23:44:15.701320 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 3.11938 (* 1 = 3.11938 loss)
I0428 23:44:15.701333 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.98351 (* 1 = 0.98351 loss)
I0428 23:44:15.701347 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.9358 (* 0.0909091 = 0.266891 loss)
I0428 23:44:15.701360 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 3.1214 (* 0.0909091 = 0.283764 loss)
I0428 23:44:15.701376 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 3.17299 (* 0.0909091 = 0.288454 loss)
I0428 23:44:15.701390 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.97632 (* 0.0909091 = 0.270574 loss)
I0428 23:44:15.701402 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.57701 (* 0.0909091 = 0.234273 loss)
I0428 23:44:15.701416 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.30246 (* 0.0909091 = 0.209314 loss)
I0428 23:44:15.701428 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.518 (* 0.0909091 = 0.138 loss)
I0428 23:44:15.701442 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.875416 (* 0.0909091 = 0.0795833 loss)
I0428 23:44:15.701455 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.49223 (* 0.0909091 = 0.0447482 loss)
I0428 23:44:15.701468 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.407575 (* 0.0909091 = 0.0370523 loss)
I0428 23:44:15.701481 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.361388 (* 0.0909091 = 0.0328534 loss)
I0428 23:44:15.701494 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.28041 (* 0.0909091 = 0.0254919 loss)
I0428 23:44:15.701508 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.223168 (* 0.0909091 = 0.020288 loss)
I0428 23:44:15.701521 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.198095 (* 0.0909091 = 0.0180086 loss)
I0428 23:44:15.701534 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.141138 (* 0.0909091 = 0.0128308 loss)
I0428 23:44:15.701547 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.0835278 (* 0.0909091 = 0.00759344 loss)
I0428 23:44:15.701561 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.0353673 (* 0.0909091 = 0.00321521 loss)
I0428 23:44:15.701575 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.0141826 (* 0.0909091 = 0.00128933 loss)
I0428 23:44:15.701587 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00712328 (* 0.0909091 = 0.00064757 loss)
I0428 23:44:15.701601 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.0057363 (* 0.0909091 = 0.000521482 loss)
I0428 23:44:15.701614 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.00498689 (* 0.0909091 = 0.000453354 loss)
I0428 23:44:15.701627 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.00445051 (* 0.0909091 = 0.000404592 loss)
I0428 23:44:15.701638 6470 solver.cpp:406] Test net output #147: total_accuracy = 0
I0428 23:44:15.701649 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0
I0428 23:44:15.701660 6470 solver.cpp:406] Test net output #149: total_confidence = 0.000128226
I0428 23:44:15.701680 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.00035519
I0428 23:44:15.881873 6470 solver.cpp:229] Iteration 5000, loss = 10.2869
I0428 23:44:15.881956 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0684932
I0428 23:44:15.881974 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:44:15.881988 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:44:15.881999 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:44:15.882011 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0428 23:44:15.882024 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:44:15.882035 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.125
I0428 23:44:15.882047 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0428 23:44:15.882060 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.5
I0428 23:44:15.882071 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.625
I0428 23:44:15.882083 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.625
I0428 23:44:15.882094 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.625
I0428 23:44:15.882107 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.75
I0428 23:44:15.882129 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.75
I0428 23:44:15.882153 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.75
I0428 23:44:15.882171 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.75
I0428 23:44:15.882184 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0428 23:44:15.882196 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:44:15.882208 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:44:15.882220 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:44:15.882231 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:44:15.882243 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:44:15.882256 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:44:15.882266 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.613636
I0428 23:44:15.882278 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.219178
I0428 23:44:15.882294 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.21006 (* 0.3 = 0.963018 loss)
I0428 23:44:15.882308 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.43774 (* 0.3 = 0.431323 loss)
I0428 23:44:15.882323 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.17363 (* 0.0272727 = 0.0865536 loss)
I0428 23:44:15.882336 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.81665 (* 0.0272727 = 0.0768176 loss)
I0428 23:44:15.882349 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.1335 (* 0.0272727 = 0.085459 loss)
I0428 23:44:15.882364 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.44412 (* 0.0272727 = 0.0939306 loss)
I0428 23:44:15.882377 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.83395 (* 0.0272727 = 0.0772897 loss)
I0428 23:44:15.882390 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.03963 (* 0.0272727 = 0.082899 loss)
I0428 23:44:15.882405 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.6618 (* 0.0272727 = 0.0725944 loss)
I0428 23:44:15.882418 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.03877 (* 0.0272727 = 0.0556029 loss)
I0428 23:44:15.882431 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.6442 (* 0.0272727 = 0.0448419 loss)
I0428 23:44:15.882444 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 1.34393 (* 0.0272727 = 0.0366528 loss)
I0428 23:44:15.882458 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 1.21981 (* 0.0272727 = 0.0332676 loss)
I0428 23:44:15.882508 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 1.28554 (* 0.0272727 = 0.0350601 loss)
I0428 23:44:15.882522 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 1.53342 (* 0.0272727 = 0.0418206 loss)
I0428 23:44:15.882536 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 1.46011 (* 0.0272727 = 0.0398211 loss)
I0428 23:44:15.882553 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 1.72119 (* 0.0272727 = 0.0469415 loss)
I0428 23:44:15.882566 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.860166 (* 0.0272727 = 0.0234591 loss)
I0428 23:44:15.882581 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00299028 (* 0.0272727 = 8.15532e-05 loss)
I0428 23:44:15.882596 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00348509 (* 0.0272727 = 9.5048e-05 loss)
I0428 23:44:15.882609 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00220218 (* 0.0272727 = 6.00593e-05 loss)
I0428 23:44:15.882622 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000876269 (* 0.0272727 = 2.38982e-05 loss)
I0428 23:44:15.882637 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000610233 (* 0.0272727 = 1.66427e-05 loss)
I0428 23:44:15.882650 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0011324 (* 0.0272727 = 3.08838e-05 loss)
I0428 23:44:15.882663 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0410959
I0428 23:44:15.882674 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0428 23:44:15.882688 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.375
I0428 23:44:15.882699 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:44:15.882710 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:44:15.882722 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0428 23:44:15.882735 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0428 23:44:15.882745 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.375
I0428 23:44:15.882757 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.5
I0428 23:44:15.882768 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.625
I0428 23:44:15.882781 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.625
I0428 23:44:15.882791 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.625
I0428 23:44:15.882803 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.75
I0428 23:44:15.882815 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.75
I0428 23:44:15.882827 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.75
I0428 23:44:15.882838 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.75
I0428 23:44:15.882850 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0428 23:44:15.882861 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:44:15.882874 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:44:15.882884 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:44:15.882895 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:44:15.882906 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:44:15.882918 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:44:15.882930 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.590909
I0428 23:44:15.882941 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.260274
I0428 23:44:15.882956 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.18998 (* 0.3 = 0.956995 loss)
I0428 23:44:15.882968 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.46601 (* 0.3 = 0.439803 loss)
I0428 23:44:15.882982 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.8612 (* 0.0272727 = 0.0780328 loss)
I0428 23:44:15.883011 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.99268 (* 0.0272727 = 0.0816185 loss)
I0428 23:44:15.883026 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.12942 (* 0.0272727 = 0.0853478 loss)
I0428 23:44:15.883040 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.14153 (* 0.0272727 = 0.085678 loss)
I0428 23:44:15.883054 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.93572 (* 0.0272727 = 0.080065 loss)
I0428 23:44:15.883067 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.7297 (* 0.0272727 = 0.0744464 loss)
I0428 23:44:15.883080 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.66723 (* 0.0272727 = 0.0727426 loss)
I0428 23:44:15.883095 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.27 (* 0.0272727 = 0.0619092 loss)
I0428 23:44:15.883107 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.76486 (* 0.0272727 = 0.0481325 loss)
I0428 23:44:15.883121 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 1.44581 (* 0.0272727 = 0.0394312 loss)
I0428 23:44:15.883134 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 1.49913 (* 0.0272727 = 0.0408854 loss)
I0428 23:44:15.883147 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 1.17407 (* 0.0272727 = 0.03202 loss)
I0428 23:44:15.883162 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 1.40494 (* 0.0272727 = 0.0383164 loss)
I0428 23:44:15.883174 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 1.3776 (* 0.0272727 = 0.0375708 loss)
I0428 23:44:15.883188 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 1.47636 (* 0.0272727 = 0.0402643 loss)
I0428 23:44:15.883200 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.854643 (* 0.0272727 = 0.0233084 loss)
I0428 23:44:15.883214 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0149152 (* 0.0272727 = 0.000406778 loss)
I0428 23:44:15.883229 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0131839 (* 0.0272727 = 0.000359561 loss)
I0428 23:44:15.883242 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0055695 (* 0.0272727 = 0.000151895 loss)
I0428 23:44:15.883255 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00328956 (* 0.0272727 = 8.97154e-05 loss)
I0428 23:44:15.883270 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0029361 (* 0.0272727 = 8.00754e-05 loss)
I0428 23:44:15.883283 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00391448 (* 0.0272727 = 0.000106759 loss)
I0428 23:44:15.883296 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0410959
I0428 23:44:15.883307 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:44:15.883319 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.375
I0428 23:44:15.883330 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:44:15.883342 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0428 23:44:15.883353 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0428 23:44:15.883365 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:44:15.883378 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0428 23:44:15.883389 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.5
I0428 23:44:15.883400 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.625
I0428 23:44:15.883412 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.625
I0428 23:44:15.883424 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.625
I0428 23:44:15.883435 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.75
I0428 23:44:15.883446 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.75
I0428 23:44:15.883458 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.75
I0428 23:44:15.883496 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.75
I0428 23:44:15.883509 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0428 23:44:15.883522 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:44:15.883533 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:44:15.883544 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:44:15.883556 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:44:15.883568 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:44:15.883579 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:44:15.883590 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.596591
I0428 23:44:15.883605 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.342466
I0428 23:44:15.883618 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.18972 (* 1 = 3.18972 loss)
I0428 23:44:15.883632 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.41807 (* 1 = 1.41807 loss)
I0428 23:44:15.883646 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.96741 (* 0.0909091 = 0.269764 loss)
I0428 23:44:15.883661 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.97875 (* 0.0909091 = 0.270796 loss)
I0428 23:44:15.883673 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.20454 (* 0.0909091 = 0.291322 loss)
I0428 23:44:15.883687 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.99169 (* 0.0909091 = 0.271972 loss)
I0428 23:44:15.883700 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.31024 (* 0.0909091 = 0.210022 loss)
I0428 23:44:15.883714 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.43061 (* 0.0909091 = 0.220965 loss)
I0428 23:44:15.883728 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.22219 (* 0.0909091 = 0.202017 loss)
I0428 23:44:15.883741 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 2.03251 (* 0.0909091 = 0.184773 loss)
I0428 23:44:15.883754 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.70865 (* 0.0909091 = 0.155332 loss)
I0428 23:44:15.883769 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 1.18023 (* 0.0909091 = 0.107294 loss)
I0428 23:44:15.883781 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 1.33918 (* 0.0909091 = 0.121744 loss)
I0428 23:44:15.883795 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 1.34908 (* 0.0909091 = 0.122643 loss)
I0428 23:44:15.883808 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 1.56643 (* 0.0909091 = 0.142402 loss)
I0428 23:44:15.883821 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 1.43995 (* 0.0909091 = 0.130904 loss)
I0428 23:44:15.883834 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 1.77406 (* 0.0909091 = 0.161278 loss)
I0428 23:44:15.883848 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.764145 (* 0.0909091 = 0.0694677 loss)
I0428 23:44:15.883862 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00154001 (* 0.0909091 = 0.000140001 loss)
I0428 23:44:15.883877 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000494604 (* 0.0909091 = 4.4964e-05 loss)
I0428 23:44:15.883890 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000295981 (* 0.0909091 = 2.69073e-05 loss)
I0428 23:44:15.883904 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000136615 (* 0.0909091 = 1.24195e-05 loss)
I0428 23:44:15.883918 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 8.88666e-05 (* 0.0909091 = 8.07878e-06 loss)
I0428 23:44:15.883931 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 5.47141e-05 (* 0.0909091 = 4.97401e-06 loss)
I0428 23:44:15.883944 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:44:15.883965 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:44:15.883977 6470 solver.cpp:245] Train net output #149: total_confidence = 1.2164e-06
I0428 23:44:15.883990 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 7.52123e-06
I0428 23:44:15.884002 6470 sgd_solver.cpp:106] Iteration 5000, lr = 0.01
I0428 23:46:32.558919 6470 solver.cpp:229] Iteration 5500, loss = 10.1735
I0428 23:46:32.559087 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0980392
I0428 23:46:32.559109 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:46:32.559123 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:46:32.559134 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:46:32.559146 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0428 23:46:32.559159 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0428 23:46:32.559170 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:46:32.559181 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0428 23:46:32.559193 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0428 23:46:32.559206 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0428 23:46:32.559217 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:46:32.559229 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0428 23:46:32.559242 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:46:32.559252 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:46:32.559264 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:46:32.559275 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:46:32.559288 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:46:32.559298 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:46:32.559310 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:46:32.559325 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:46:32.559337 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:46:32.559348 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:46:32.559360 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:46:32.559371 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.727273
I0428 23:46:32.559383 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.215686
I0428 23:46:32.559399 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.12654 (* 0.3 = 0.937961 loss)
I0428 23:46:32.559413 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.09712 (* 0.3 = 0.329136 loss)
I0428 23:46:32.559427 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.07026 (* 0.0272727 = 0.0837345 loss)
I0428 23:46:32.559448 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.63833 (* 0.0272727 = 0.0992272 loss)
I0428 23:46:32.559489 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.6887 (* 0.0272727 = 0.100601 loss)
I0428 23:46:32.559506 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.01667 (* 0.0272727 = 0.0822728 loss)
I0428 23:46:32.559520 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.30564 (* 0.0272727 = 0.0901539 loss)
I0428 23:46:32.559535 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.01929 (* 0.0272727 = 0.0823442 loss)
I0428 23:46:32.559548 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.38774 (* 0.0272727 = 0.0378475 loss)
I0428 23:46:32.559561 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.912622 (* 0.0272727 = 0.0248897 loss)
I0428 23:46:32.559576 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.487757 (* 0.0272727 = 0.0133025 loss)
I0428 23:46:32.559589 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.564984 (* 0.0272727 = 0.0154087 loss)
I0428 23:46:32.559603 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.750194 (* 0.0272727 = 0.0204598 loss)
I0428 23:46:32.559617 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.179608 (* 0.0272727 = 0.0048984 loss)
I0428 23:46:32.559630 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.163863 (* 0.0272727 = 0.004469 loss)
I0428 23:46:32.559669 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.11116 (* 0.0272727 = 0.00303164 loss)
I0428 23:46:32.559685 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0686005 (* 0.0272727 = 0.00187092 loss)
I0428 23:46:32.559700 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0669118 (* 0.0272727 = 0.00182487 loss)
I0428 23:46:32.559715 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0502339 (* 0.0272727 = 0.00137002 loss)
I0428 23:46:32.559728 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0378863 (* 0.0272727 = 0.00103326 loss)
I0428 23:46:32.559742 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0187948 (* 0.0272727 = 0.000512585 loss)
I0428 23:46:32.559757 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0193006 (* 0.0272727 = 0.000526379 loss)
I0428 23:46:32.559772 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0137487 (* 0.0272727 = 0.000374964 loss)
I0428 23:46:32.559784 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.017605 (* 0.0272727 = 0.000480136 loss)
I0428 23:46:32.559804 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0980392
I0428 23:46:32.559823 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:46:32.559835 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.375
I0428 23:46:32.559846 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:46:32.559859 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:46:32.559870 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:46:32.559881 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0428 23:46:32.559893 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0428 23:46:32.559906 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0428 23:46:32.559916 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0428 23:46:32.559928 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:46:32.559939 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0428 23:46:32.559952 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:46:32.559962 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:46:32.559974 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:46:32.559985 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:46:32.559998 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:46:32.560009 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:46:32.560019 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:46:32.560031 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:46:32.560042 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:46:32.560055 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:46:32.560065 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:46:32.560077 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.721591
I0428 23:46:32.560088 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.294118
I0428 23:46:32.560107 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.12576 (* 0.3 = 0.937728 loss)
I0428 23:46:32.560122 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.06132 (* 0.3 = 0.318395 loss)
I0428 23:46:32.560135 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.14273 (* 0.0272727 = 0.0857109 loss)
I0428 23:46:32.560149 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.55109 (* 0.0272727 = 0.096848 loss)
I0428 23:46:32.560175 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.44414 (* 0.0272727 = 0.0939312 loss)
I0428 23:46:32.560190 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.21361 (* 0.0272727 = 0.0876439 loss)
I0428 23:46:32.560204 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.29606 (* 0.0272727 = 0.0898924 loss)
I0428 23:46:32.560217 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.87707 (* 0.0272727 = 0.0784655 loss)
I0428 23:46:32.560230 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.46119 (* 0.0272727 = 0.0398506 loss)
I0428 23:46:32.560243 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.829689 (* 0.0272727 = 0.0226279 loss)
I0428 23:46:32.560257 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.504032 (* 0.0272727 = 0.0137463 loss)
I0428 23:46:32.560271 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.59327 (* 0.0272727 = 0.0161801 loss)
I0428 23:46:32.560284 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.661153 (* 0.0272727 = 0.0180315 loss)
I0428 23:46:32.560299 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.181174 (* 0.0272727 = 0.00494111 loss)
I0428 23:46:32.560312 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.163502 (* 0.0272727 = 0.00445915 loss)
I0428 23:46:32.560326 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.146584 (* 0.0272727 = 0.00399773 loss)
I0428 23:46:32.560340 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.122664 (* 0.0272727 = 0.00334538 loss)
I0428 23:46:32.560353 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0694155 (* 0.0272727 = 0.00189315 loss)
I0428 23:46:32.560369 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0403527 (* 0.0272727 = 0.00110053 loss)
I0428 23:46:32.560384 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0352697 (* 0.0272727 = 0.000961901 loss)
I0428 23:46:32.560398 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0195524 (* 0.0272727 = 0.000533247 loss)
I0428 23:46:32.560412 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0181205 (* 0.0272727 = 0.000494195 loss)
I0428 23:46:32.560425 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0171942 (* 0.0272727 = 0.000468933 loss)
I0428 23:46:32.560441 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0131426 (* 0.0272727 = 0.000358434 loss)
I0428 23:46:32.560452 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.117647
I0428 23:46:32.560464 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:46:32.560475 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:46:32.560487 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:46:32.560498 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0428 23:46:32.560510 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:46:32.560523 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:46:32.560534 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0428 23:46:32.560545 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0428 23:46:32.560556 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0428 23:46:32.560569 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:46:32.560580 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0428 23:46:32.560591 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:46:32.560602 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:46:32.560614 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:46:32.560626 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:46:32.560636 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:46:32.560657 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:46:32.560670 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:46:32.560683 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:46:32.560693 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:46:32.560704 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:46:32.560716 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:46:32.560727 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.727273
I0428 23:46:32.560739 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.333333
I0428 23:46:32.560753 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.01298 (* 1 = 3.01298 loss)
I0428 23:46:32.560767 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.05134 (* 1 = 1.05134 loss)
I0428 23:46:32.560781 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.96729 (* 0.0909091 = 0.269753 loss)
I0428 23:46:32.560796 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.45169 (* 0.0909091 = 0.31379 loss)
I0428 23:46:32.560808 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.46823 (* 0.0909091 = 0.315293 loss)
I0428 23:46:32.560822 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.10525 (* 0.0909091 = 0.282295 loss)
I0428 23:46:32.560835 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.04493 (* 0.0909091 = 0.276812 loss)
I0428 23:46:32.560849 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.68384 (* 0.0909091 = 0.243986 loss)
I0428 23:46:32.560863 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.11551 (* 0.0909091 = 0.10141 loss)
I0428 23:46:32.560876 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.709498 (* 0.0909091 = 0.0644999 loss)
I0428 23:46:32.560889 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.468002 (* 0.0909091 = 0.0425456 loss)
I0428 23:46:32.560904 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.504287 (* 0.0909091 = 0.0458442 loss)
I0428 23:46:32.560917 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.62706 (* 0.0909091 = 0.0570055 loss)
I0428 23:46:32.560930 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.177562 (* 0.0909091 = 0.016142 loss)
I0428 23:46:32.560945 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.151636 (* 0.0909091 = 0.0137851 loss)
I0428 23:46:32.560957 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.121049 (* 0.0909091 = 0.0110045 loss)
I0428 23:46:32.560971 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0921042 (* 0.0909091 = 0.00837311 loss)
I0428 23:46:32.560989 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0701487 (* 0.0909091 = 0.00637715 loss)
I0428 23:46:32.561003 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0527418 (* 0.0909091 = 0.00479471 loss)
I0428 23:46:32.561017 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0355498 (* 0.0909091 = 0.0032318 loss)
I0428 23:46:32.561031 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0275462 (* 0.0909091 = 0.0025042 loss)
I0428 23:46:32.561044 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.0179894 (* 0.0909091 = 0.0016354 loss)
I0428 23:46:32.561058 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0154543 (* 0.0909091 = 0.00140493 loss)
I0428 23:46:32.561075 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.0126348 (* 0.0909091 = 0.00114862 loss)
I0428 23:46:32.561096 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:46:32.561110 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:46:32.561131 6470 solver.cpp:245] Train net output #149: total_confidence = 1.01709e-06
I0428 23:46:32.561157 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 9.26749e-07
I0428 23:46:32.561172 6470 sgd_solver.cpp:106] Iteration 5500, lr = 0.01
I0428 23:48:49.132174 6470 solver.cpp:229] Iteration 6000, loss = 10.1561
I0428 23:48:49.132313 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0571429
I0428 23:48:49.132333 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0428 23:48:49.132345 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0428 23:48:49.132357 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:48:49.132369 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0428 23:48:49.132381 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0428 23:48:49.132392 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.75
I0428 23:48:49.132405 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 1
I0428 23:48:49.132416 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0428 23:48:49.132428 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0428 23:48:49.132439 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0428 23:48:49.132452 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:48:49.132462 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:48:49.132473 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:48:49.132485 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:48:49.132496 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:48:49.132508 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:48:49.132519 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:48:49.132531 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:48:49.132542 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:48:49.132553 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:48:49.132565 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:48:49.132576 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:48:49.132588 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.784091
I0428 23:48:49.132601 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.257143
I0428 23:48:49.132616 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.28831 (* 0.3 = 0.986494 loss)
I0428 23:48:49.132630 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.842169 (* 0.3 = 0.252651 loss)
I0428 23:48:49.132647 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.26514 (* 0.0272727 = 0.0890491 loss)
I0428 23:48:49.132673 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.33652 (* 0.0272727 = 0.0909959 loss)
I0428 23:48:49.132690 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.59156 (* 0.0272727 = 0.0979515 loss)
I0428 23:48:49.132704 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.16532 (* 0.0272727 = 0.086327 loss)
I0428 23:48:49.132717 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.26076 (* 0.0272727 = 0.061657 loss)
I0428 23:48:49.132731 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.43631 (* 0.0272727 = 0.0391721 loss)
I0428 23:48:49.132745 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.517966 (* 0.0272727 = 0.0141263 loss)
I0428 23:48:49.132758 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.108775 (* 0.0272727 = 0.00296659 loss)
I0428 23:48:49.132772 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0226739 (* 0.0272727 = 0.00061838 loss)
I0428 23:48:49.132786 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0209022 (* 0.0272727 = 0.000570061 loss)
I0428 23:48:49.132800 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0194107 (* 0.0272727 = 0.000529384 loss)
I0428 23:48:49.132813 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0108825 (* 0.0272727 = 0.000296796 loss)
I0428 23:48:49.132827 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0143992 (* 0.0272727 = 0.000392706 loss)
I0428 23:48:49.132861 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00764699 (* 0.0272727 = 0.000208554 loss)
I0428 23:48:49.132876 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00966025 (* 0.0272727 = 0.000263461 loss)
I0428 23:48:49.132890 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00819274 (* 0.0272727 = 0.000223438 loss)
I0428 23:48:49.132904 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0094109 (* 0.0272727 = 0.000256661 loss)
I0428 23:48:49.132917 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00603071 (* 0.0272727 = 0.000164474 loss)
I0428 23:48:49.132931 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0112492 (* 0.0272727 = 0.000306796 loss)
I0428 23:48:49.132946 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00612928 (* 0.0272727 = 0.000167162 loss)
I0428 23:48:49.132958 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00288777 (* 0.0272727 = 7.87574e-05 loss)
I0428 23:48:49.132975 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00327855 (* 0.0272727 = 8.9415e-05 loss)
I0428 23:48:49.132995 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0285714
I0428 23:48:49.133008 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0428 23:48:49.133020 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:48:49.133031 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:48:49.133044 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:48:49.133054 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0428 23:48:49.133066 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.75
I0428 23:48:49.133077 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 1
I0428 23:48:49.133090 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0428 23:48:49.133100 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0428 23:48:49.133111 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0428 23:48:49.133122 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:48:49.133133 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:48:49.133146 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:48:49.133157 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:48:49.133167 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:48:49.133178 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:48:49.133189 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:48:49.133201 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:48:49.133213 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:48:49.133224 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:48:49.133234 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:48:49.133245 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:48:49.133256 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.772727
I0428 23:48:49.133268 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.142857
I0428 23:48:49.133282 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.42896 (* 0.3 = 1.02869 loss)
I0428 23:48:49.133296 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.9251 (* 0.3 = 0.27753 loss)
I0428 23:48:49.133309 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.27683 (* 0.0272727 = 0.089368 loss)
I0428 23:48:49.133329 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.88002 (* 0.0272727 = 0.105819 loss)
I0428 23:48:49.133355 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.77982 (* 0.0272727 = 0.103086 loss)
I0428 23:48:49.133370 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.1275 (* 0.0272727 = 0.0852953 loss)
I0428 23:48:49.133383 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.38124 (* 0.0272727 = 0.0649428 loss)
I0428 23:48:49.133397 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.55102 (* 0.0272727 = 0.0423005 loss)
I0428 23:48:49.133410 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.383998 (* 0.0272727 = 0.0104727 loss)
I0428 23:48:49.133424 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.0528126 (* 0.0272727 = 0.00144034 loss)
I0428 23:48:49.133438 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.0252331 (* 0.0272727 = 0.000688174 loss)
I0428 23:48:49.133451 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.0164998 (* 0.0272727 = 0.000449994 loss)
I0428 23:48:49.133466 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0157034 (* 0.0272727 = 0.000428274 loss)
I0428 23:48:49.133478 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00816431 (* 0.0272727 = 0.000222663 loss)
I0428 23:48:49.133492 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0061849 (* 0.0272727 = 0.000168679 loss)
I0428 23:48:49.133507 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00747873 (* 0.0272727 = 0.000203965 loss)
I0428 23:48:49.133520 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00348179 (* 0.0272727 = 9.49579e-05 loss)
I0428 23:48:49.133533 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00193994 (* 0.0272727 = 5.29074e-05 loss)
I0428 23:48:49.133548 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00298637 (* 0.0272727 = 8.14464e-05 loss)
I0428 23:48:49.133561 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00356418 (* 0.0272727 = 9.7205e-05 loss)
I0428 23:48:49.133574 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00239528 (* 0.0272727 = 6.53257e-05 loss)
I0428 23:48:49.133589 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00176589 (* 0.0272727 = 4.81605e-05 loss)
I0428 23:48:49.133601 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00168422 (* 0.0272727 = 4.59333e-05 loss)
I0428 23:48:49.133615 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0015883 (* 0.0272727 = 4.33174e-05 loss)
I0428 23:48:49.133627 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0857143
I0428 23:48:49.133640 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:48:49.133651 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:48:49.133662 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:48:49.133673 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0428 23:48:49.133684 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0428 23:48:49.133697 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.75
I0428 23:48:49.133708 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 1
I0428 23:48:49.133718 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0428 23:48:49.133730 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0428 23:48:49.133741 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0428 23:48:49.133752 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:48:49.133764 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:48:49.133774 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:48:49.133786 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:48:49.133801 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:48:49.133823 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:48:49.133857 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:48:49.133882 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:48:49.133898 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:48:49.133909 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:48:49.133921 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:48:49.133932 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:48:49.133944 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.795455
I0428 23:48:49.133956 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.0857143
I0428 23:48:49.133970 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.24046 (* 1 = 3.24046 loss)
I0428 23:48:49.133985 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.797568 (* 1 = 0.797568 loss)
I0428 23:48:49.133998 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.15028 (* 0.0909091 = 0.286389 loss)
I0428 23:48:49.134012 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.32942 (* 0.0909091 = 0.302675 loss)
I0428 23:48:49.134026 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.42716 (* 0.0909091 = 0.31156 loss)
I0428 23:48:49.134039 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.66169 (* 0.0909091 = 0.241972 loss)
I0428 23:48:49.134052 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.6917 (* 0.0909091 = 0.153791 loss)
I0428 23:48:49.134066 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.22937 (* 0.0909091 = 0.111761 loss)
I0428 23:48:49.134079 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.512098 (* 0.0909091 = 0.0465544 loss)
I0428 23:48:49.134093 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.0455109 (* 0.0909091 = 0.00413735 loss)
I0428 23:48:49.134107 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.00916841 (* 0.0909091 = 0.000833492 loss)
I0428 23:48:49.134121 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.00952032 (* 0.0909091 = 0.000865484 loss)
I0428 23:48:49.134135 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0102429 (* 0.0909091 = 0.000931174 loss)
I0428 23:48:49.134150 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0037023 (* 0.0909091 = 0.000336573 loss)
I0428 23:48:49.134163 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00264373 (* 0.0909091 = 0.000240339 loss)
I0428 23:48:49.134176 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00189184 (* 0.0909091 = 0.000171985 loss)
I0428 23:48:49.134191 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00110749 (* 0.0909091 = 0.000100681 loss)
I0428 23:48:49.134204 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000418653 (* 0.0909091 = 3.80593e-05 loss)
I0428 23:48:49.134218 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000299761 (* 0.0909091 = 2.7251e-05 loss)
I0428 23:48:49.134229 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000238371 (* 0.0909091 = 2.16701e-05 loss)
I0428 23:48:49.134238 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000196177 (* 0.0909091 = 1.78342e-05 loss)
I0428 23:48:49.134253 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000177269 (* 0.0909091 = 1.61154e-05 loss)
I0428 23:48:49.134268 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000172439 (* 0.0909091 = 1.56763e-05 loss)
I0428 23:48:49.134281 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000131624 (* 0.0909091 = 1.19658e-05 loss)
I0428 23:48:49.134292 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:48:49.134305 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:48:49.134315 6470 solver.cpp:245] Train net output #149: total_confidence = 2.94732e-05
I0428 23:48:49.134337 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 9.80327e-05
I0428 23:48:49.134356 6470 sgd_solver.cpp:106] Iteration 6000, lr = 0.01
I0428 23:51:05.752809 6470 solver.cpp:229] Iteration 6500, loss = 10.0433
I0428 23:51:05.752961 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.15
I0428 23:51:05.752980 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:51:05.752993 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:51:05.753005 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:51:05.753017 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0428 23:51:05.753029 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0428 23:51:05.753041 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.75
I0428 23:51:05.753053 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.875
I0428 23:51:05.753065 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0428 23:51:05.753077 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0428 23:51:05.753089 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0428 23:51:05.753100 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:51:05.753113 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:51:05.753123 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:51:05.753135 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:51:05.753146 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:51:05.753159 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:51:05.753170 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:51:05.753181 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:51:05.753192 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:51:05.753204 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:51:05.753216 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:51:05.753228 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:51:05.753239 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.789773
I0428 23:51:05.753252 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.325
I0428 23:51:05.753268 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.17952 (* 0.3 = 0.953856 loss)
I0428 23:51:05.753281 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.84696 (* 0.3 = 0.254088 loss)
I0428 23:51:05.753295 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.19265 (* 0.0272727 = 0.0870724 loss)
I0428 23:51:05.753316 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.38552 (* 0.0272727 = 0.0923323 loss)
I0428 23:51:05.753341 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.24869 (* 0.0272727 = 0.0886006 loss)
I0428 23:51:05.753357 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.68646 (* 0.0272727 = 0.073267 loss)
I0428 23:51:05.753371 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.96822 (* 0.0272727 = 0.0809515 loss)
I0428 23:51:05.753384 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.42686 (* 0.0272727 = 0.0389145 loss)
I0428 23:51:05.753398 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.925911 (* 0.0272727 = 0.0252521 loss)
I0428 23:51:05.753413 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.125397 (* 0.0272727 = 0.00341991 loss)
I0428 23:51:05.753427 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0147172 (* 0.0272727 = 0.000401378 loss)
I0428 23:51:05.753442 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0106546 (* 0.0272727 = 0.000290579 loss)
I0428 23:51:05.753455 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.00684459 (* 0.0272727 = 0.000186671 loss)
I0428 23:51:05.753470 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0078946 (* 0.0272727 = 0.000215307 loss)
I0428 23:51:05.753484 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0028991 (* 0.0272727 = 7.90663e-05 loss)
I0428 23:51:05.753520 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00248942 (* 0.0272727 = 6.78932e-05 loss)
I0428 23:51:05.753545 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00182876 (* 0.0272727 = 4.98752e-05 loss)
I0428 23:51:05.753561 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00123973 (* 0.0272727 = 3.38107e-05 loss)
I0428 23:51:05.753574 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0022791 (* 0.0272727 = 6.21572e-05 loss)
I0428 23:51:05.753588 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0013876 (* 0.0272727 = 3.78436e-05 loss)
I0428 23:51:05.753602 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0004376 (* 0.0272727 = 1.19345e-05 loss)
I0428 23:51:05.753618 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000519936 (* 0.0272727 = 1.41801e-05 loss)
I0428 23:51:05.753630 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000561934 (* 0.0272727 = 1.53255e-05 loss)
I0428 23:51:05.753644 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00029091 (* 0.0272727 = 7.9339e-06 loss)
I0428 23:51:05.753656 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.1
I0428 23:51:05.753669 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0428 23:51:05.753680 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0428 23:51:05.753691 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:51:05.753702 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0428 23:51:05.753715 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:51:05.753726 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.625
I0428 23:51:05.753737 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.875
I0428 23:51:05.753748 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0428 23:51:05.753759 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0428 23:51:05.753772 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0428 23:51:05.753782 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:51:05.753793 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:51:05.753805 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:51:05.753816 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:51:05.753828 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:51:05.753839 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:51:05.753850 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:51:05.753861 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:51:05.753872 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:51:05.753883 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:51:05.753895 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:51:05.753906 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:51:05.753917 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.778409
I0428 23:51:05.753929 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.25
I0428 23:51:05.753942 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.15232 (* 0.3 = 0.945697 loss)
I0428 23:51:05.753957 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.883783 (* 0.3 = 0.265135 loss)
I0428 23:51:05.753969 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.96016 (* 0.0272727 = 0.0807318 loss)
I0428 23:51:05.753988 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.35826 (* 0.0272727 = 0.0915889 loss)
I0428 23:51:05.754014 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.65082 (* 0.0272727 = 0.0995677 loss)
I0428 23:51:05.754029 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.7633 (* 0.0272727 = 0.0753627 loss)
I0428 23:51:05.754042 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.82029 (* 0.0272727 = 0.0769171 loss)
I0428 23:51:05.754055 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.71492 (* 0.0272727 = 0.0467704 loss)
I0428 23:51:05.754070 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.04856 (* 0.0272727 = 0.0285971 loss)
I0428 23:51:05.754083 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.084704 (* 0.0272727 = 0.00231011 loss)
I0428 23:51:05.754097 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.00311918 (* 0.0272727 = 8.50685e-05 loss)
I0428 23:51:05.754112 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.00076121 (* 0.0272727 = 2.07603e-05 loss)
I0428 23:51:05.754125 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.000718611 (* 0.0272727 = 1.95985e-05 loss)
I0428 23:51:05.754138 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.000571064 (* 0.0272727 = 1.55745e-05 loss)
I0428 23:51:05.754153 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.000625257 (* 0.0272727 = 1.70525e-05 loss)
I0428 23:51:05.754165 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.000492131 (* 0.0272727 = 1.34217e-05 loss)
I0428 23:51:05.754179 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00025922 (* 0.0272727 = 7.06964e-06 loss)
I0428 23:51:05.754192 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00023401 (* 0.0272727 = 6.38209e-06 loss)
I0428 23:51:05.754206 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.000125308 (* 0.0272727 = 3.41749e-06 loss)
I0428 23:51:05.754220 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 8.7387e-05 (* 0.0272727 = 2.38328e-06 loss)
I0428 23:51:05.754233 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 9.32554e-05 (* 0.0272727 = 2.54333e-06 loss)
I0428 23:51:05.754246 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 5.37409e-05 (* 0.0272727 = 1.46566e-06 loss)
I0428 23:51:05.754261 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 2.74789e-05 (* 0.0272727 = 7.49426e-07 loss)
I0428 23:51:05.754274 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 2.98191e-05 (* 0.0272727 = 8.13247e-07 loss)
I0428 23:51:05.754286 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.125
I0428 23:51:05.754297 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:51:05.754309 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:51:05.754320 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:51:05.754331 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0428 23:51:05.754343 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0428 23:51:05.754354 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.75
I0428 23:51:05.754369 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.875
I0428 23:51:05.754381 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0428 23:51:05.754392 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0428 23:51:05.754405 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0428 23:51:05.754415 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:51:05.754426 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:51:05.754438 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:51:05.754446 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:51:05.754453 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:51:05.754465 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:51:05.754487 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:51:05.754499 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:51:05.754511 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:51:05.754523 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:51:05.754534 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:51:05.754544 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:51:05.754560 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.772727
I0428 23:51:05.754580 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.2
I0428 23:51:05.754595 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.99399 (* 1 = 2.99399 loss)
I0428 23:51:05.754617 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.880702 (* 1 = 0.880702 loss)
I0428 23:51:05.754633 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.8576 (* 0.0909091 = 0.259782 loss)
I0428 23:51:05.754647 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.31949 (* 0.0909091 = 0.301771 loss)
I0428 23:51:05.754659 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.31842 (* 0.0909091 = 0.301675 loss)
I0428 23:51:05.754673 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.66925 (* 0.0909091 = 0.242659 loss)
I0428 23:51:05.754685 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.52366 (* 0.0909091 = 0.229424 loss)
I0428 23:51:05.754698 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.50008 (* 0.0909091 = 0.136371 loss)
I0428 23:51:05.754712 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.879751 (* 0.0909091 = 0.0799774 loss)
I0428 23:51:05.754725 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.147988 (* 0.0909091 = 0.0134535 loss)
I0428 23:51:05.754739 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.0134541 (* 0.0909091 = 0.0012231 loss)
I0428 23:51:05.754752 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.00294294 (* 0.0909091 = 0.00026754 loss)
I0428 23:51:05.754766 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00122199 (* 0.0909091 = 0.00011109 loss)
I0428 23:51:05.754779 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.000850718 (* 0.0909091 = 7.7338e-05 loss)
I0428 23:51:05.754793 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.000855894 (* 0.0909091 = 7.78085e-05 loss)
I0428 23:51:05.754806 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.000776078 (* 0.0909091 = 7.05525e-05 loss)
I0428 23:51:05.754820 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000787858 (* 0.0909091 = 7.16235e-05 loss)
I0428 23:51:05.754833 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000897582 (* 0.0909091 = 8.15984e-05 loss)
I0428 23:51:05.754847 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00144203 (* 0.0909091 = 0.000131093 loss)
I0428 23:51:05.754860 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00154105 (* 0.0909091 = 0.000140096 loss)
I0428 23:51:05.754873 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00169842 (* 0.0909091 = 0.000154402 loss)
I0428 23:51:05.754886 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00182451 (* 0.0909091 = 0.000165864 loss)
I0428 23:51:05.754900 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00161378 (* 0.0909091 = 0.000146707 loss)
I0428 23:51:05.754914 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00145719 (* 0.0909091 = 0.000132472 loss)
I0428 23:51:05.754925 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:51:05.754935 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:51:05.754946 6470 solver.cpp:245] Train net output #149: total_confidence = 8.84912e-05
I0428 23:51:05.754968 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000516477
I0428 23:51:05.754982 6470 sgd_solver.cpp:106] Iteration 6500, lr = 0.01
I0428 23:53:22.355398 6470 solver.cpp:229] Iteration 7000, loss = 10.0216
I0428 23:53:22.355561 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.107692
I0428 23:53:22.355581 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0428 23:53:22.355593 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0428 23:53:22.355605 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:53:22.355618 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:53:22.355629 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:53:22.355641 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0428 23:53:22.355654 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0428 23:53:22.355664 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.5
I0428 23:53:22.355676 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.5
I0428 23:53:22.355689 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.625
I0428 23:53:22.355700 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.75
I0428 23:53:22.355711 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.75
I0428 23:53:22.355723 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0428 23:53:22.355736 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:53:22.355747 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:53:22.355758 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:53:22.355770 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:53:22.355782 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:53:22.355793 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:53:22.355804 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:53:22.355816 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:53:22.355828 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:53:22.355839 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.664773
I0428 23:53:22.355850 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.292308
I0428 23:53:22.355866 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.18445 (* 0.3 = 0.955335 loss)
I0428 23:53:22.355880 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.37324 (* 0.3 = 0.411972 loss)
I0428 23:53:22.355895 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.04514 (* 0.0272727 = 0.0830492 loss)
I0428 23:53:22.355908 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.95146 (* 0.0272727 = 0.0804944 loss)
I0428 23:53:22.355922 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.38483 (* 0.0272727 = 0.0923135 loss)
I0428 23:53:22.355937 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.31185 (* 0.0272727 = 0.0903232 loss)
I0428 23:53:22.355949 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.19285 (* 0.0272727 = 0.0870779 loss)
I0428 23:53:22.355963 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.37151 (* 0.0272727 = 0.0646776 loss)
I0428 23:53:22.355976 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.70353 (* 0.0272727 = 0.04646 loss)
I0428 23:53:22.355990 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.33482 (* 0.0272727 = 0.0636769 loss)
I0428 23:53:22.356003 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.62425 (* 0.0272727 = 0.0442978 loss)
I0428 23:53:22.356017 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 1.45063 (* 0.0272727 = 0.0395626 loss)
I0428 23:53:22.356030 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 1.12545 (* 0.0272727 = 0.0306941 loss)
I0428 23:53:22.356045 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 1.19786 (* 0.0272727 = 0.0326689 loss)
I0428 23:53:22.356057 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.740659 (* 0.0272727 = 0.0201998 loss)
I0428 23:53:22.356093 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.187241 (* 0.0272727 = 0.00510656 loss)
I0428 23:53:22.356109 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0997756 (* 0.0272727 = 0.00272115 loss)
I0428 23:53:22.356122 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.108367 (* 0.0272727 = 0.00295545 loss)
I0428 23:53:22.356137 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0402258 (* 0.0272727 = 0.00109707 loss)
I0428 23:53:22.356150 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0211381 (* 0.0272727 = 0.000576493 loss)
I0428 23:53:22.356165 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0194344 (* 0.0272727 = 0.000530029 loss)
I0428 23:53:22.356179 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00860097 (* 0.0272727 = 0.000234572 loss)
I0428 23:53:22.356194 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00739055 (* 0.0272727 = 0.00020156 loss)
I0428 23:53:22.356207 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00646871 (* 0.0272727 = 0.000176419 loss)
I0428 23:53:22.356220 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0769231
I0428 23:53:22.356230 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:53:22.356242 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.375
I0428 23:53:22.356254 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0428 23:53:22.356266 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0428 23:53:22.356276 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0428 23:53:22.356288 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0428 23:53:22.356299 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0428 23:53:22.356313 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.5
I0428 23:53:22.356326 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.625
I0428 23:53:22.356338 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.5
I0428 23:53:22.356349 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.75
I0428 23:53:22.356361 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.75
I0428 23:53:22.356374 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0428 23:53:22.356384 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:53:22.356396 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:53:22.356408 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:53:22.356420 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:53:22.356431 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:53:22.356442 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:53:22.356453 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:53:22.356464 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:53:22.356475 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:53:22.356487 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.653409
I0428 23:53:22.356498 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.261538
I0428 23:53:22.356511 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.13687 (* 0.3 = 0.941062 loss)
I0428 23:53:22.356525 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.3297 (* 0.3 = 0.39891 loss)
I0428 23:53:22.356539 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.99909 (* 0.0272727 = 0.0817933 loss)
I0428 23:53:22.356552 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.83602 (* 0.0272727 = 0.077346 loss)
I0428 23:53:22.356581 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.46467 (* 0.0272727 = 0.094491 loss)
I0428 23:53:22.356597 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.4056 (* 0.0272727 = 0.09288 loss)
I0428 23:53:22.356611 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.14826 (* 0.0272727 = 0.0858618 loss)
I0428 23:53:22.356624 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.21585 (* 0.0272727 = 0.0604324 loss)
I0428 23:53:22.356638 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.75414 (* 0.0272727 = 0.0478401 loss)
I0428 23:53:22.356652 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.15541 (* 0.0272727 = 0.058784 loss)
I0428 23:53:22.356665 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.55894 (* 0.0272727 = 0.0425165 loss)
I0428 23:53:22.356678 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 1.57166 (* 0.0272727 = 0.0428634 loss)
I0428 23:53:22.356693 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 1.38101 (* 0.0272727 = 0.0376638 loss)
I0428 23:53:22.356705 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 1.09842 (* 0.0272727 = 0.029957 loss)
I0428 23:53:22.356719 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.772273 (* 0.0272727 = 0.021062 loss)
I0428 23:53:22.356732 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.138092 (* 0.0272727 = 0.00376615 loss)
I0428 23:53:22.356746 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.122706 (* 0.0272727 = 0.00334652 loss)
I0428 23:53:22.356760 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0539772 (* 0.0272727 = 0.00147211 loss)
I0428 23:53:22.356773 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0288303 (* 0.0272727 = 0.000786281 loss)
I0428 23:53:22.356787 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0125673 (* 0.0272727 = 0.000342746 loss)
I0428 23:53:22.356801 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0120496 (* 0.0272727 = 0.000328625 loss)
I0428 23:53:22.356814 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00759963 (* 0.0272727 = 0.000207263 loss)
I0428 23:53:22.356828 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00550696 (* 0.0272727 = 0.00015019 loss)
I0428 23:53:22.356842 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00539471 (* 0.0272727 = 0.000147128 loss)
I0428 23:53:22.356853 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.153846
I0428 23:53:22.356865 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0428 23:53:22.356878 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0428 23:53:22.356889 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:53:22.356900 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:53:22.356911 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0428 23:53:22.356923 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0428 23:53:22.356935 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0428 23:53:22.356946 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.5
I0428 23:53:22.356957 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.625
I0428 23:53:22.356969 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.625
I0428 23:53:22.356981 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0428 23:53:22.356992 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.75
I0428 23:53:22.357004 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0428 23:53:22.357017 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:53:22.357028 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:53:22.357038 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:53:22.357060 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:53:22.357074 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:53:22.357084 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:53:22.357096 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:53:22.357107 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:53:22.357118 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:53:22.357131 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.676136
I0428 23:53:22.357141 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.230769
I0428 23:53:22.357156 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.06836 (* 1 = 3.06836 loss)
I0428 23:53:22.357169 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.31649 (* 1 = 1.31649 loss)
I0428 23:53:22.357183 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.57167 (* 0.0909091 = 0.233788 loss)
I0428 23:53:22.357197 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.78359 (* 0.0909091 = 0.253053 loss)
I0428 23:53:22.357210 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.28724 (* 0.0909091 = 0.29884 loss)
I0428 23:53:22.357224 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.21648 (* 0.0909091 = 0.292408 loss)
I0428 23:53:22.357237 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.05717 (* 0.0909091 = 0.277925 loss)
I0428 23:53:22.357251 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.02446 (* 0.0909091 = 0.184042 loss)
I0428 23:53:22.357264 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.74795 (* 0.0909091 = 0.158905 loss)
I0428 23:53:22.357278 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 2.18622 (* 0.0909091 = 0.198748 loss)
I0428 23:53:22.357291 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.57275 (* 0.0909091 = 0.142977 loss)
I0428 23:53:22.357306 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 1.40893 (* 0.0909091 = 0.128085 loss)
I0428 23:53:22.357318 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 1.25372 (* 0.0909091 = 0.113974 loss)
I0428 23:53:22.357332 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 1.05731 (* 0.0909091 = 0.0961194 loss)
I0428 23:53:22.357345 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.592604 (* 0.0909091 = 0.0538731 loss)
I0428 23:53:22.357359 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.152704 (* 0.0909091 = 0.0138822 loss)
I0428 23:53:22.357375 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.119939 (* 0.0909091 = 0.0109035 loss)
I0428 23:53:22.357389 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0655086 (* 0.0909091 = 0.00595533 loss)
I0428 23:53:22.357403 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0245937 (* 0.0909091 = 0.00223579 loss)
I0428 23:53:22.357417 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0126908 (* 0.0909091 = 0.00115371 loss)
I0428 23:53:22.357430 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00673835 (* 0.0909091 = 0.000612578 loss)
I0428 23:53:22.357445 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00348362 (* 0.0909091 = 0.000316692 loss)
I0428 23:53:22.357458 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00102744 (* 0.0909091 = 9.34032e-05 loss)
I0428 23:53:22.357472 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000544615 (* 0.0909091 = 4.95105e-05 loss)
I0428 23:53:22.357484 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:53:22.357496 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:53:22.357506 6470 solver.cpp:245] Train net output #149: total_confidence = 5.43264e-07
I0428 23:53:22.357527 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 3.0289e-05
I0428 23:53:22.357542 6470 sgd_solver.cpp:106] Iteration 7000, lr = 0.01
I0428 23:55:38.864153 6470 solver.cpp:229] Iteration 7500, loss = 9.89484
I0428 23:55:38.864270 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0535714
I0428 23:55:38.864290 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0428 23:55:38.864303 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.375
I0428 23:55:38.864318 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0428 23:55:38.864331 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:55:38.864342 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0428 23:55:38.864354 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:55:38.864367 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0428 23:55:38.864378 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0428 23:55:38.864390 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0428 23:55:38.864401 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0428 23:55:38.864413 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:55:38.864424 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:55:38.864435 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:55:38.864447 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:55:38.864459 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:55:38.864470 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:55:38.864481 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:55:38.864492 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:55:38.864503 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:55:38.864516 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:55:38.864526 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:55:38.864537 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:55:38.864548 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.693182
I0428 23:55:38.864560 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.160714
I0428 23:55:38.864575 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.3005 (* 0.3 = 0.99015 loss)
I0428 23:55:38.864589 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.12366 (* 0.3 = 0.337098 loss)
I0428 23:55:38.864603 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.1886 (* 0.0272727 = 0.0869619 loss)
I0428 23:55:38.864617 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.205 (* 0.0272727 = 0.0874091 loss)
I0428 23:55:38.864631 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.40635 (* 0.0272727 = 0.0929004 loss)
I0428 23:55:38.864645 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.2353 (* 0.0272727 = 0.0882355 loss)
I0428 23:55:38.864658 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.06796 (* 0.0272727 = 0.0836716 loss)
I0428 23:55:38.864671 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.58987 (* 0.0272727 = 0.0706329 loss)
I0428 23:55:38.864686 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.71901 (* 0.0272727 = 0.046882 loss)
I0428 23:55:38.864698 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.8771 (* 0.0272727 = 0.0511937 loss)
I0428 23:55:38.864711 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.05577 (* 0.0272727 = 0.0287937 loss)
I0428 23:55:38.864725 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.501235 (* 0.0272727 = 0.0136701 loss)
I0428 23:55:38.864740 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0565181 (* 0.0272727 = 0.0015414 loss)
I0428 23:55:38.864754 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0260214 (* 0.0272727 = 0.000709674 loss)
I0428 23:55:38.864768 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.027296 (* 0.0272727 = 0.000744437 loss)
I0428 23:55:38.864799 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.022344 (* 0.0272727 = 0.000609382 loss)
I0428 23:55:38.864814 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0147002 (* 0.0272727 = 0.000400914 loss)
I0428 23:55:38.864828 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00813598 (* 0.0272727 = 0.00022189 loss)
I0428 23:55:38.864842 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0116215 (* 0.0272727 = 0.00031695 loss)
I0428 23:55:38.864856 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00696466 (* 0.0272727 = 0.000189945 loss)
I0428 23:55:38.864869 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00748091 (* 0.0272727 = 0.000204025 loss)
I0428 23:55:38.864882 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00668045 (* 0.0272727 = 0.000182194 loss)
I0428 23:55:38.864897 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00650788 (* 0.0272727 = 0.000177488 loss)
I0428 23:55:38.864909 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00526662 (* 0.0272727 = 0.000143635 loss)
I0428 23:55:38.864922 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0357143
I0428 23:55:38.864933 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:55:38.864944 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:55:38.864956 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0428 23:55:38.864967 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0428 23:55:38.864979 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0
I0428 23:55:38.864990 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0428 23:55:38.865001 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0428 23:55:38.865013 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0428 23:55:38.865025 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0428 23:55:38.865036 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0428 23:55:38.865047 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:55:38.865058 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:55:38.865069 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:55:38.865080 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:55:38.865092 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:55:38.865103 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:55:38.865113 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:55:38.865125 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:55:38.865136 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:55:38.865147 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:55:38.865159 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:55:38.865170 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:55:38.865180 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.6875
I0428 23:55:38.865191 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.178571
I0428 23:55:38.865206 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.30371 (* 0.3 = 0.991112 loss)
I0428 23:55:38.865219 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.12688 (* 0.3 = 0.338063 loss)
I0428 23:55:38.865232 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.6047 (* 0.0272727 = 0.0983101 loss)
I0428 23:55:38.865245 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.40791 (* 0.0272727 = 0.0929431 loss)
I0428 23:55:38.865269 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.76361 (* 0.0272727 = 0.102644 loss)
I0428 23:55:38.865284 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.59383 (* 0.0272727 = 0.0980135 loss)
I0428 23:55:38.865303 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.29994 (* 0.0272727 = 0.0899983 loss)
I0428 23:55:38.865315 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.78904 (* 0.0272727 = 0.0760646 loss)
I0428 23:55:38.865329 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.09233 (* 0.0272727 = 0.0570636 loss)
I0428 23:55:38.865342 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.24566 (* 0.0272727 = 0.0612452 loss)
I0428 23:55:38.865356 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.10307 (* 0.0272727 = 0.0300836 loss)
I0428 23:55:38.865372 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.515356 (* 0.0272727 = 0.0140552 loss)
I0428 23:55:38.865386 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0467422 (* 0.0272727 = 0.00127479 loss)
I0428 23:55:38.865399 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0293695 (* 0.0272727 = 0.000800987 loss)
I0428 23:55:38.865413 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0223463 (* 0.0272727 = 0.000609443 loss)
I0428 23:55:38.865427 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0176463 (* 0.0272727 = 0.000481262 loss)
I0428 23:55:38.865440 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0155089 (* 0.0272727 = 0.000422971 loss)
I0428 23:55:38.865454 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00904274 (* 0.0272727 = 0.00024662 loss)
I0428 23:55:38.865468 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00453788 (* 0.0272727 = 0.00012376 loss)
I0428 23:55:38.865481 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00230733 (* 0.0272727 = 6.29271e-05 loss)
I0428 23:55:38.865495 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00145243 (* 0.0272727 = 3.96118e-05 loss)
I0428 23:55:38.865509 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00118718 (* 0.0272727 = 3.23777e-05 loss)
I0428 23:55:38.865522 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000979683 (* 0.0272727 = 2.67186e-05 loss)
I0428 23:55:38.865536 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000866975 (* 0.0272727 = 2.36448e-05 loss)
I0428 23:55:38.865547 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0357143
I0428 23:55:38.865559 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0428 23:55:38.865571 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:55:38.865581 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0428 23:55:38.865592 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0428 23:55:38.865604 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0
I0428 23:55:38.865615 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:55:38.865627 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.5
I0428 23:55:38.865638 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0428 23:55:38.865649 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0428 23:55:38.865660 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0428 23:55:38.865671 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:55:38.865682 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:55:38.865694 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:55:38.865705 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:55:38.865716 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:55:38.865727 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:55:38.865747 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:55:38.865761 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:55:38.865772 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:55:38.865782 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:55:38.865794 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:55:38.865805 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:55:38.865816 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.6875
I0428 23:55:38.865828 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.214286
I0428 23:55:38.865839 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.26177 (* 1 = 3.26177 loss)
I0428 23:55:38.865847 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.08319 (* 1 = 1.08319 loss)
I0428 23:55:38.865861 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.50747 (* 0.0909091 = 0.318861 loss)
I0428 23:55:38.865875 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.27604 (* 0.0909091 = 0.297822 loss)
I0428 23:55:38.865890 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.2794 (* 0.0909091 = 0.298127 loss)
I0428 23:55:38.865902 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.44665 (* 0.0909091 = 0.313332 loss)
I0428 23:55:38.865916 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.05363 (* 0.0909091 = 0.277602 loss)
I0428 23:55:38.865929 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.73967 (* 0.0909091 = 0.249061 loss)
I0428 23:55:38.865942 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.97211 (* 0.0909091 = 0.179283 loss)
I0428 23:55:38.865955 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.74772 (* 0.0909091 = 0.158884 loss)
I0428 23:55:38.865968 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.02783 (* 0.0909091 = 0.0934392 loss)
I0428 23:55:38.865981 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.577704 (* 0.0909091 = 0.0525186 loss)
I0428 23:55:38.865995 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0087275 (* 0.0909091 = 0.000793409 loss)
I0428 23:55:38.866009 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00627306 (* 0.0909091 = 0.000570279 loss)
I0428 23:55:38.866022 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00610439 (* 0.0909091 = 0.000554945 loss)
I0428 23:55:38.866036 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00317895 (* 0.0909091 = 0.000288995 loss)
I0428 23:55:38.866050 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00240166 (* 0.0909091 = 0.000218333 loss)
I0428 23:55:38.866062 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00228661 (* 0.0909091 = 0.000207874 loss)
I0428 23:55:38.866076 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00252365 (* 0.0909091 = 0.000229423 loss)
I0428 23:55:38.866089 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00221153 (* 0.0909091 = 0.000201048 loss)
I0428 23:55:38.866102 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00147467 (* 0.0909091 = 0.000134061 loss)
I0428 23:55:38.866116 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00123624 (* 0.0909091 = 0.000112386 loss)
I0428 23:55:38.866128 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00056596 (* 0.0909091 = 5.14509e-05 loss)
I0428 23:55:38.866142 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000410994 (* 0.0909091 = 3.73631e-05 loss)
I0428 23:55:38.866153 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:55:38.866164 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:55:38.866175 6470 solver.cpp:245] Train net output #149: total_confidence = 5.69677e-07
I0428 23:55:38.866195 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 8.20917e-06
I0428 23:55:38.866209 6470 sgd_solver.cpp:106] Iteration 7500, lr = 0.01
I0428 23:57:55.577915 6470 solver.cpp:229] Iteration 8000, loss = 9.86118
I0428 23:57:55.578080 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.117647
I0428 23:57:55.578101 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0428 23:57:55.578114 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.25
I0428 23:57:55.578126 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0428 23:57:55.578137 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0428 23:57:55.578150 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0428 23:57:55.578161 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0428 23:57:55.578173 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0428 23:57:55.578184 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0428 23:57:55.578197 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0428 23:57:55.578208 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0428 23:57:55.578219 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0428 23:57:55.578232 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0428 23:57:55.578243 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0428 23:57:55.578253 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0428 23:57:55.578265 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0428 23:57:55.578277 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0428 23:57:55.578289 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0428 23:57:55.578299 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0428 23:57:55.578315 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0428 23:57:55.578326 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0428 23:57:55.578338 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0428 23:57:55.578349 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0428 23:57:55.578361 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.732955
I0428 23:57:55.578373 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.333333
I0428 23:57:55.578389 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.80152 (* 0.3 = 0.840457 loss)
I0428 23:57:55.578403 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.89349 (* 0.3 = 0.268047 loss)
I0428 23:57:55.578418 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.91155 (* 0.0272727 = 0.0794058 loss)
I0428 23:57:55.578430 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.92743 (* 0.0272727 = 0.0798389 loss)
I0428 23:57:55.578444 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.79179 (* 0.0272727 = 0.0761399 loss)
I0428 23:57:55.578459 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.3868 (* 0.0272727 = 0.0923673 loss)
I0428 23:57:55.578471 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.2588 (* 0.0272727 = 0.0616037 loss)
I0428 23:57:55.578485 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.42289 (* 0.0272727 = 0.0660789 loss)
I0428 23:57:55.578498 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.51759 (* 0.0272727 = 0.0686615 loss)
I0428 23:57:55.578512 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.15739 (* 0.0272727 = 0.0315652 loss)
I0428 23:57:55.578526 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0950478 (* 0.0272727 = 0.00259221 loss)
I0428 23:57:55.578541 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0650917 (* 0.0272727 = 0.00177523 loss)
I0428 23:57:55.578553 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0493394 (* 0.0272727 = 0.00134562 loss)
I0428 23:57:55.578567 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.029434 (* 0.0272727 = 0.000802745 loss)
I0428 23:57:55.578582 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0212084 (* 0.0272727 = 0.000578411 loss)
I0428 23:57:55.578615 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0084558 (* 0.0272727 = 0.000230613 loss)
I0428 23:57:55.578630 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00610768 (* 0.0272727 = 0.000166573 loss)
I0428 23:57:55.578644 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00417864 (* 0.0272727 = 0.000113963 loss)
I0428 23:57:55.578658 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00179383 (* 0.0272727 = 4.89227e-05 loss)
I0428 23:57:55.578671 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000723894 (* 0.0272727 = 1.97426e-05 loss)
I0428 23:57:55.578685 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000574226 (* 0.0272727 = 1.56607e-05 loss)
I0428 23:57:55.578699 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000539896 (* 0.0272727 = 1.47244e-05 loss)
I0428 23:57:55.578712 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000493125 (* 0.0272727 = 1.34489e-05 loss)
I0428 23:57:55.578727 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000298821 (* 0.0272727 = 8.14966e-06 loss)
I0428 23:57:55.578737 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.117647
I0428 23:57:55.578749 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0428 23:57:55.578761 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0428 23:57:55.578773 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0428 23:57:55.578784 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0428 23:57:55.578796 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0428 23:57:55.578809 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0428 23:57:55.578820 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.375
I0428 23:57:55.578832 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0428 23:57:55.578845 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0428 23:57:55.578855 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0428 23:57:55.578866 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0428 23:57:55.578877 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0428 23:57:55.578889 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0428 23:57:55.578901 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0428 23:57:55.578912 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0428 23:57:55.578922 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0428 23:57:55.578933 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0428 23:57:55.578944 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0428 23:57:55.578956 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0428 23:57:55.578968 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0428 23:57:55.578979 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0428 23:57:55.578989 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0428 23:57:55.579001 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.732955
I0428 23:57:55.579012 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.313726
I0428 23:57:55.579026 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.87684 (* 0.3 = 0.863053 loss)
I0428 23:57:55.579040 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.941062 (* 0.3 = 0.282319 loss)
I0428 23:57:55.579053 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.98776 (* 0.0272727 = 0.0814845 loss)
I0428 23:57:55.579066 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.86391 (* 0.0272727 = 0.0781066 loss)
I0428 23:57:55.579095 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.99825 (* 0.0272727 = 0.0817706 loss)
I0428 23:57:55.579110 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.54683 (* 0.0272727 = 0.0967318 loss)
I0428 23:57:55.579124 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.31295 (* 0.0272727 = 0.0630805 loss)
I0428 23:57:55.579138 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.60833 (* 0.0272727 = 0.0711363 loss)
I0428 23:57:55.579151 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.26606 (* 0.0272727 = 0.0618016 loss)
I0428 23:57:55.579164 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.03329 (* 0.0272727 = 0.0281806 loss)
I0428 23:57:55.579179 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.270005 (* 0.0272727 = 0.00736377 loss)
I0428 23:57:55.579192 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.177421 (* 0.0272727 = 0.00483877 loss)
I0428 23:57:55.579205 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.120036 (* 0.0272727 = 0.00327372 loss)
I0428 23:57:55.579219 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0938492 (* 0.0272727 = 0.00255952 loss)
I0428 23:57:55.579233 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0907947 (* 0.0272727 = 0.00247622 loss)
I0428 23:57:55.579247 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0488362 (* 0.0272727 = 0.0013319 loss)
I0428 23:57:55.579262 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0347492 (* 0.0272727 = 0.000947706 loss)
I0428 23:57:55.579274 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0216117 (* 0.0272727 = 0.00058941 loss)
I0428 23:57:55.579288 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0263293 (* 0.0272727 = 0.000718072 loss)
I0428 23:57:55.579301 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.025904 (* 0.0272727 = 0.000706472 loss)
I0428 23:57:55.579315 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.013413 (* 0.0272727 = 0.000365808 loss)
I0428 23:57:55.579329 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0191698 (* 0.0272727 = 0.000522812 loss)
I0428 23:57:55.579342 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00902349 (* 0.0272727 = 0.000246095 loss)
I0428 23:57:55.579355 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0147952 (* 0.0272727 = 0.000403505 loss)
I0428 23:57:55.579370 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0784314
I0428 23:57:55.579382 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0428 23:57:55.579394 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0428 23:57:55.579406 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0428 23:57:55.579417 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0428 23:57:55.579428 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0428 23:57:55.579440 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0428 23:57:55.579452 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0428 23:57:55.579463 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0428 23:57:55.579489 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0428 23:57:55.579502 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0428 23:57:55.579514 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0428 23:57:55.579524 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0428 23:57:55.579535 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0428 23:57:55.579547 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0428 23:57:55.579558 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0428 23:57:55.579569 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0428 23:57:55.579593 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0428 23:57:55.579607 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0428 23:57:55.579617 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0428 23:57:55.579629 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0428 23:57:55.579640 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0428 23:57:55.579653 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0428 23:57:55.579663 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.732955
I0428 23:57:55.579675 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.313726
I0428 23:57:55.579689 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.79885 (* 1 = 2.79885 loss)
I0428 23:57:55.579702 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.872807 (* 1 = 0.872807 loss)
I0428 23:57:55.579716 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.07154 (* 0.0909091 = 0.279231 loss)
I0428 23:57:55.579730 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.78619 (* 0.0909091 = 0.25329 loss)
I0428 23:57:55.579744 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.78144 (* 0.0909091 = 0.252859 loss)
I0428 23:57:55.579757 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.19295 (* 0.0909091 = 0.290268 loss)
I0428 23:57:55.579771 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.28483 (* 0.0909091 = 0.207712 loss)
I0428 23:57:55.579784 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.62301 (* 0.0909091 = 0.238455 loss)
I0428 23:57:55.579798 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.11807 (* 0.0909091 = 0.192552 loss)
I0428 23:57:55.579812 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.974745 (* 0.0909091 = 0.0886132 loss)
I0428 23:57:55.579825 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.155419 (* 0.0909091 = 0.014129 loss)
I0428 23:57:55.579838 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.0748154 (* 0.0909091 = 0.0068014 loss)
I0428 23:57:55.579854 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0280972 (* 0.0909091 = 0.00255429 loss)
I0428 23:57:55.579866 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0134302 (* 0.0909091 = 0.00122092 loss)
I0428 23:57:55.579880 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.011743 (* 0.0909091 = 0.00106755 loss)
I0428 23:57:55.579895 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00924432 (* 0.0909091 = 0.000840393 loss)
I0428 23:57:55.579907 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00582971 (* 0.0909091 = 0.000529973 loss)
I0428 23:57:55.579921 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00465876 (* 0.0909091 = 0.000423524 loss)
I0428 23:57:55.579934 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00475564 (* 0.0909091 = 0.000432331 loss)
I0428 23:57:55.579948 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00486547 (* 0.0909091 = 0.000442316 loss)
I0428 23:57:55.579962 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00464079 (* 0.0909091 = 0.00042189 loss)
I0428 23:57:55.579975 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00401717 (* 0.0909091 = 0.000365197 loss)
I0428 23:57:55.579989 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00255996 (* 0.0909091 = 0.000232723 loss)
I0428 23:57:55.580003 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.0020989 (* 0.0909091 = 0.00019081 loss)
I0428 23:57:55.580014 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0428 23:57:55.580026 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0428 23:57:55.580037 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000104017
I0428 23:57:55.580057 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00139913
I0428 23:57:55.580071 6470 sgd_solver.cpp:106] Iteration 8000, lr = 0.01
I0429 00:00:12.204144 6470 solver.cpp:229] Iteration 8500, loss = 9.74572
I0429 00:00:12.204303 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0425532
I0429 00:00:12.204325 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:00:12.204339 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.25
I0429 00:00:12.204352 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:00:12.204365 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:00:12.204376 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:00:12.204388 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.625
I0429 00:00:12.204401 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:00:12.204411 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 00:00:12.204423 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:00:12.204435 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:00:12.204447 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:00:12.204459 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:00:12.204471 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:00:12.204483 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:00:12.204494 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:00:12.204506 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:00:12.204519 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:00:12.204530 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:00:12.204541 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:00:12.204553 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:00:12.204565 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:00:12.204579 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:00:12.204601 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.738636
I0429 00:00:12.204622 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.234043
I0429 00:00:12.204651 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.91402 (* 0.3 = 0.874205 loss)
I0429 00:00:12.204676 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.865218 (* 0.3 = 0.259565 loss)
I0429 00:00:12.204699 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.94313 (* 0.0272727 = 0.0802672 loss)
I0429 00:00:12.204725 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.93089 (* 0.0272727 = 0.0799334 loss)
I0429 00:00:12.204749 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.81304 (* 0.0272727 = 0.0767193 loss)
I0429 00:00:12.204764 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.49816 (* 0.0272727 = 0.0681315 loss)
I0429 00:00:12.204778 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.2615 (* 0.0272727 = 0.0616774 loss)
I0429 00:00:12.204792 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.90044 (* 0.0272727 = 0.0518303 loss)
I0429 00:00:12.204805 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.66789 (* 0.0272727 = 0.045488 loss)
I0429 00:00:12.204820 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.357118 (* 0.0272727 = 0.00973958 loss)
I0429 00:00:12.204834 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.402323 (* 0.0272727 = 0.0109725 loss)
I0429 00:00:12.204849 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.35636 (* 0.0272727 = 0.00971891 loss)
I0429 00:00:12.204862 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.623823 (* 0.0272727 = 0.0170134 loss)
I0429 00:00:12.204876 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.061689 (* 0.0272727 = 0.00168243 loss)
I0429 00:00:12.204910 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.053182 (* 0.0272727 = 0.00145042 loss)
I0429 00:00:12.204926 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0554907 (* 0.0272727 = 0.00151338 loss)
I0429 00:00:12.204939 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0502708 (* 0.0272727 = 0.00137102 loss)
I0429 00:00:12.204953 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0325388 (* 0.0272727 = 0.000887422 loss)
I0429 00:00:12.204967 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0118611 (* 0.0272727 = 0.000323485 loss)
I0429 00:00:12.204982 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00661923 (* 0.0272727 = 0.000180525 loss)
I0429 00:00:12.204994 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00226538 (* 0.0272727 = 6.17832e-05 loss)
I0429 00:00:12.205008 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.003739 (* 0.0272727 = 0.000101973 loss)
I0429 00:00:12.205023 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00197702 (* 0.0272727 = 5.39187e-05 loss)
I0429 00:00:12.205035 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00137372 (* 0.0272727 = 3.74651e-05 loss)
I0429 00:00:12.205047 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0851064
I0429 00:00:12.205060 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:00:12.205070 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:00:12.205082 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:00:12.205090 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:00:12.205103 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:00:12.205116 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:00:12.205137 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:00:12.205162 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 00:00:12.205188 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:00:12.205215 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:00:12.205231 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:00:12.205243 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:00:12.205255 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:00:12.205266 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:00:12.205277 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:00:12.205289 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:00:12.205301 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:00:12.205317 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:00:12.205328 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:00:12.205340 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:00:12.205351 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:00:12.205365 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:00:12.205377 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.744318
I0429 00:00:12.205389 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.255319
I0429 00:00:12.205404 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.93426 (* 0.3 = 0.880279 loss)
I0429 00:00:12.205417 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.863202 (* 0.3 = 0.258961 loss)
I0429 00:00:12.205431 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.93692 (* 0.0272727 = 0.0800978 loss)
I0429 00:00:12.205445 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.1719 (* 0.0272727 = 0.0865063 loss)
I0429 00:00:12.205471 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.9317 (* 0.0272727 = 0.0799554 loss)
I0429 00:00:12.205487 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.45634 (* 0.0272727 = 0.066991 loss)
I0429 00:00:12.205500 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.28076 (* 0.0272727 = 0.0622025 loss)
I0429 00:00:12.205514 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.04482 (* 0.0272727 = 0.0557679 loss)
I0429 00:00:12.205528 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.92606 (* 0.0272727 = 0.0525288 loss)
I0429 00:00:12.205541 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.38689 (* 0.0272727 = 0.0105515 loss)
I0429 00:00:12.205555 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.259134 (* 0.0272727 = 0.00706728 loss)
I0429 00:00:12.205569 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.252171 (* 0.0272727 = 0.0068774 loss)
I0429 00:00:12.205584 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.765629 (* 0.0272727 = 0.0208808 loss)
I0429 00:00:12.205597 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0289518 (* 0.0272727 = 0.000789594 loss)
I0429 00:00:12.205611 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0160402 (* 0.0272727 = 0.00043746 loss)
I0429 00:00:12.205624 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00620877 (* 0.0272727 = 0.00016933 loss)
I0429 00:00:12.205638 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00574238 (* 0.0272727 = 0.00015661 loss)
I0429 00:00:12.205652 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00109414 (* 0.0272727 = 2.98403e-05 loss)
I0429 00:00:12.205665 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.000688546 (* 0.0272727 = 1.87785e-05 loss)
I0429 00:00:12.205679 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000267185 (* 0.0272727 = 7.28686e-06 loss)
I0429 00:00:12.205693 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000196574 (* 0.0272727 = 5.36111e-06 loss)
I0429 00:00:12.205706 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000178407 (* 0.0272727 = 4.86565e-06 loss)
I0429 00:00:12.205720 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000291445 (* 0.0272727 = 7.94851e-06 loss)
I0429 00:00:12.205734 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000228583 (* 0.0272727 = 6.23409e-06 loss)
I0429 00:00:12.205746 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0638298
I0429 00:00:12.205757 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:00:12.205770 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:00:12.205780 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:00:12.205792 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:00:12.205803 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:00:12.205816 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:00:12.205826 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:00:12.205838 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:00:12.205849 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:00:12.205862 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:00:12.205873 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:00:12.205884 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:00:12.205895 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:00:12.205906 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:00:12.205917 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:00:12.205929 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:00:12.205950 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:00:12.205962 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:00:12.205973 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:00:12.205984 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:00:12.205996 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:00:12.206007 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:00:12.206018 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0429 00:00:12.206029 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.361702
I0429 00:00:12.206043 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.91845 (* 1 = 2.91845 loss)
I0429 00:00:12.206056 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.818389 (* 1 = 0.818389 loss)
I0429 00:00:12.206070 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.96322 (* 0.0909091 = 0.269384 loss)
I0429 00:00:12.206084 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.02397 (* 0.0909091 = 0.274907 loss)
I0429 00:00:12.206097 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.84555 (* 0.0909091 = 0.258686 loss)
I0429 00:00:12.206110 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.48745 (* 0.0909091 = 0.226132 loss)
I0429 00:00:12.206125 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.15072 (* 0.0909091 = 0.19552 loss)
I0429 00:00:12.206137 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.15662 (* 0.0909091 = 0.196056 loss)
I0429 00:00:12.206151 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.6086 (* 0.0909091 = 0.146236 loss)
I0429 00:00:12.206164 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.416533 (* 0.0909091 = 0.0378666 loss)
I0429 00:00:12.206177 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.352606 (* 0.0909091 = 0.0320551 loss)
I0429 00:00:12.206192 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.388621 (* 0.0909091 = 0.0353291 loss)
I0429 00:00:12.206204 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.794881 (* 0.0909091 = 0.0722619 loss)
I0429 00:00:12.206218 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00499932 (* 0.0909091 = 0.000454484 loss)
I0429 00:00:12.206233 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00337147 (* 0.0909091 = 0.000306498 loss)
I0429 00:00:12.206246 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00231451 (* 0.0909091 = 0.00021041 loss)
I0429 00:00:12.206259 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00181021 (* 0.0909091 = 0.000164564 loss)
I0429 00:00:12.206274 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00111137 (* 0.0909091 = 0.000101034 loss)
I0429 00:00:12.206287 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00119443 (* 0.0909091 = 0.000108584 loss)
I0429 00:00:12.206300 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000648799 (* 0.0909091 = 5.89817e-05 loss)
I0429 00:00:12.206315 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000531575 (* 0.0909091 = 4.8325e-05 loss)
I0429 00:00:12.206327 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000394551 (* 0.0909091 = 3.58682e-05 loss)
I0429 00:00:12.206341 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000150864 (* 0.0909091 = 1.37149e-05 loss)
I0429 00:00:12.206356 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 9.05532e-05 (* 0.0909091 = 8.23211e-06 loss)
I0429 00:00:12.206370 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:00:12.206382 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:00:12.206403 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000673672
I0429 00:00:12.206418 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00404892
I0429 00:00:12.206432 6470 sgd_solver.cpp:106] Iteration 8500, lr = 0.01
I0429 00:02:28.744500 6470 solver.cpp:229] Iteration 9000, loss = 9.72699
I0429 00:02:28.744663 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0526316
I0429 00:02:28.744683 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:02:28.744696 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:02:28.744709 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:02:28.744720 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0429 00:02:28.744732 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.625
I0429 00:02:28.744745 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:02:28.744756 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:02:28.744767 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:02:28.744779 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:02:28.744791 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.75
I0429 00:02:28.744803 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.75
I0429 00:02:28.744814 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:02:28.744827 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:02:28.744838 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:02:28.744850 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:02:28.744863 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:02:28.744874 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:02:28.744885 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:02:28.744896 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:02:28.744909 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:02:28.744920 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:02:28.744931 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:02:28.744942 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.6875
I0429 00:02:28.744954 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.22807
I0429 00:02:28.744971 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.25715 (* 0.3 = 0.977144 loss)
I0429 00:02:28.744984 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.17188 (* 0.3 = 0.351563 loss)
I0429 00:02:28.744998 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.71132 (* 0.0272727 = 0.101218 loss)
I0429 00:02:28.745012 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.47011 (* 0.0272727 = 0.0946393 loss)
I0429 00:02:28.745026 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.86003 (* 0.0272727 = 0.105273 loss)
I0429 00:02:28.745039 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.87723 (* 0.0272727 = 0.105743 loss)
I0429 00:02:28.745057 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.01325 (* 0.0272727 = 0.0549068 loss)
I0429 00:02:28.745071 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.84056 (* 0.0272727 = 0.0501971 loss)
I0429 00:02:28.745085 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.62567 (* 0.0272727 = 0.0443363 loss)
I0429 00:02:28.745098 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.13447 (* 0.0272727 = 0.0309402 loss)
I0429 00:02:28.745111 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.0372 (* 0.0272727 = 0.0282873 loss)
I0429 00:02:28.745124 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.611209 (* 0.0272727 = 0.0166693 loss)
I0429 00:02:28.745138 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.82482 (* 0.0272727 = 0.0224951 loss)
I0429 00:02:28.745152 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.628257 (* 0.0272727 = 0.0171343 loss)
I0429 00:02:28.745168 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.558911 (* 0.0272727 = 0.015243 loss)
I0429 00:02:28.745218 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.360196 (* 0.0272727 = 0.00982353 loss)
I0429 00:02:28.745235 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.373798 (* 0.0272727 = 0.0101945 loss)
I0429 00:02:28.745250 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0517624 (* 0.0272727 = 0.0014117 loss)
I0429 00:02:28.745265 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0132971 (* 0.0272727 = 0.000362649 loss)
I0429 00:02:28.745278 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00524177 (* 0.0272727 = 0.000142957 loss)
I0429 00:02:28.745292 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00216175 (* 0.0272727 = 5.89568e-05 loss)
I0429 00:02:28.745306 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00247397 (* 0.0272727 = 6.74718e-05 loss)
I0429 00:02:28.745321 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00150407 (* 0.0272727 = 4.10201e-05 loss)
I0429 00:02:28.745334 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000668725 (* 0.0272727 = 1.8238e-05 loss)
I0429 00:02:28.745347 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.105263
I0429 00:02:28.745358 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:02:28.745370 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:02:28.745381 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:02:28.745393 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0429 00:02:28.745404 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 00:02:28.745414 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:02:28.745420 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:02:28.745434 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 00:02:28.745445 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:02:28.745456 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.75
I0429 00:02:28.745467 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.75
I0429 00:02:28.745479 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:02:28.745491 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:02:28.745502 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:02:28.745514 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:02:28.745525 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:02:28.745537 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:02:28.745548 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:02:28.745559 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:02:28.745573 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:02:28.745585 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:02:28.745596 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:02:28.745607 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.704545
I0429 00:02:28.745620 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.263158
I0429 00:02:28.745632 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.21416 (* 0.3 = 0.964247 loss)
I0429 00:02:28.745646 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.15561 (* 0.3 = 0.346684 loss)
I0429 00:02:28.745661 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.60266 (* 0.0272727 = 0.0982545 loss)
I0429 00:02:28.745673 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.61279 (* 0.0272727 = 0.0985305 loss)
I0429 00:02:28.745699 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.50106 (* 0.0272727 = 0.0954835 loss)
I0429 00:02:28.745714 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.75305 (* 0.0272727 = 0.102356 loss)
I0429 00:02:28.745728 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.30448 (* 0.0272727 = 0.0628493 loss)
I0429 00:02:28.745741 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.86726 (* 0.0272727 = 0.0509252 loss)
I0429 00:02:28.745755 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.64643 (* 0.0272727 = 0.0449027 loss)
I0429 00:02:28.745769 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.20916 (* 0.0272727 = 0.0329771 loss)
I0429 00:02:28.745782 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.00849 (* 0.0272727 = 0.0275042 loss)
I0429 00:02:28.745795 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.803894 (* 0.0272727 = 0.0219244 loss)
I0429 00:02:28.745810 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.806429 (* 0.0272727 = 0.0219935 loss)
I0429 00:02:28.745823 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.578875 (* 0.0272727 = 0.0157875 loss)
I0429 00:02:28.745836 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.638709 (* 0.0272727 = 0.0174193 loss)
I0429 00:02:28.745851 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.520451 (* 0.0272727 = 0.0141941 loss)
I0429 00:02:28.745863 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.504485 (* 0.0272727 = 0.0137587 loss)
I0429 00:02:28.745877 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.104254 (* 0.0272727 = 0.00284329 loss)
I0429 00:02:28.745890 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0258134 (* 0.0272727 = 0.000704002 loss)
I0429 00:02:28.745904 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0107609 (* 0.0272727 = 0.000293478 loss)
I0429 00:02:28.745918 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00880632 (* 0.0272727 = 0.000240172 loss)
I0429 00:02:28.745931 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00448162 (* 0.0272727 = 0.000122226 loss)
I0429 00:02:28.745944 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00494941 (* 0.0272727 = 0.000134984 loss)
I0429 00:02:28.745959 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00452516 (* 0.0272727 = 0.000123413 loss)
I0429 00:02:28.745970 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.157895
I0429 00:02:28.745981 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0429 00:02:28.745993 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:02:28.746004 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:02:28.746016 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0429 00:02:28.746027 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:02:28.746039 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:02:28.746050 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:02:28.746062 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:02:28.746073 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:02:28.746085 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:02:28.746099 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0429 00:02:28.746111 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:02:28.746124 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:02:28.746135 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:02:28.746146 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:02:28.746158 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:02:28.746179 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:02:28.746192 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:02:28.746204 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:02:28.746215 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:02:28.746227 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:02:28.746238 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:02:28.746249 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.715909
I0429 00:02:28.746261 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.350877
I0429 00:02:28.746275 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.04654 (* 1 = 3.04654 loss)
I0429 00:02:28.746289 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.09001 (* 1 = 1.09001 loss)
I0429 00:02:28.746302 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.18617 (* 0.0909091 = 0.289652 loss)
I0429 00:02:28.746316 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.36629 (* 0.0909091 = 0.306027 loss)
I0429 00:02:28.746330 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.28589 (* 0.0909091 = 0.298717 loss)
I0429 00:02:28.746343 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.60115 (* 0.0909091 = 0.327378 loss)
I0429 00:02:28.746356 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.18694 (* 0.0909091 = 0.198813 loss)
I0429 00:02:28.746369 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.93313 (* 0.0909091 = 0.175739 loss)
I0429 00:02:28.746383 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.50427 (* 0.0909091 = 0.136752 loss)
I0429 00:02:28.746397 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.13976 (* 0.0909091 = 0.103614 loss)
I0429 00:02:28.746409 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.924496 (* 0.0909091 = 0.0840451 loss)
I0429 00:02:28.746423 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.620386 (* 0.0909091 = 0.0563988 loss)
I0429 00:02:28.746436 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.762158 (* 0.0909091 = 0.0692871 loss)
I0429 00:02:28.746450 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.527646 (* 0.0909091 = 0.0479678 loss)
I0429 00:02:28.746464 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.481789 (* 0.0909091 = 0.043799 loss)
I0429 00:02:28.746477 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.398929 (* 0.0909091 = 0.0362662 loss)
I0429 00:02:28.746490 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.384391 (* 0.0909091 = 0.0349447 loss)
I0429 00:02:28.746505 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0953544 (* 0.0909091 = 0.00866858 loss)
I0429 00:02:28.746518 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.026752 (* 0.0909091 = 0.002432 loss)
I0429 00:02:28.746532 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0144424 (* 0.0909091 = 0.00131295 loss)
I0429 00:02:28.746546 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.011539 (* 0.0909091 = 0.001049 loss)
I0429 00:02:28.746561 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00857339 (* 0.0909091 = 0.000779399 loss)
I0429 00:02:28.746574 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00459288 (* 0.0909091 = 0.000417535 loss)
I0429 00:02:28.746587 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00329543 (* 0.0909091 = 0.000299584 loss)
I0429 00:02:28.746599 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:02:28.746611 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:02:28.746624 6470 solver.cpp:245] Train net output #149: total_confidence = 5.35199e-05
I0429 00:02:28.746646 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00012025
I0429 00:02:28.746660 6470 sgd_solver.cpp:106] Iteration 9000, lr = 0.01
I0429 00:04:45.219805 6470 solver.cpp:229] Iteration 9500, loss = 9.78929
I0429 00:04:45.219956 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0754717
I0429 00:04:45.219976 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 00:04:45.219990 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:04:45.220001 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:04:45.220013 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 00:04:45.220026 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.625
I0429 00:04:45.220037 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:04:45.220049 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:04:45.220060 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:04:45.220072 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:04:45.220084 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:04:45.220095 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:04:45.220108 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:04:45.220119 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:04:45.220131 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:04:45.220142 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:04:45.220155 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:04:45.220166 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:04:45.220177 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:04:45.220190 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:04:45.220201 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:04:45.220211 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:04:45.220223 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:04:45.220234 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.721591
I0429 00:04:45.220247 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.264151
I0429 00:04:45.220263 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.16894 (* 0.3 = 0.950683 loss)
I0429 00:04:45.220276 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.999014 (* 0.3 = 0.299704 loss)
I0429 00:04:45.220290 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.35208 (* 0.0272727 = 0.0914203 loss)
I0429 00:04:45.220304 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.71855 (* 0.0272727 = 0.0741423 loss)
I0429 00:04:45.220320 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.29971 (* 0.0272727 = 0.0899922 loss)
I0429 00:04:45.220335 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.72666 (* 0.0272727 = 0.0743635 loss)
I0429 00:04:45.220348 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.22735 (* 0.0272727 = 0.0607458 loss)
I0429 00:04:45.220363 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.73478 (* 0.0272727 = 0.0473123 loss)
I0429 00:04:45.220376 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.3079 (* 0.0272727 = 0.0356699 loss)
I0429 00:04:45.220391 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.166 (* 0.0272727 = 0.0317999 loss)
I0429 00:04:45.220404 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.00033 (* 0.0272727 = 0.0272818 loss)
I0429 00:04:45.220417 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.453293 (* 0.0272727 = 0.0123625 loss)
I0429 00:04:45.220432 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.498029 (* 0.0272727 = 0.0135826 loss)
I0429 00:04:45.220445 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.410593 (* 0.0272727 = 0.011198 loss)
I0429 00:04:45.220479 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.333851 (* 0.0272727 = 0.00910502 loss)
I0429 00:04:45.220494 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.451117 (* 0.0272727 = 0.0123032 loss)
I0429 00:04:45.220507 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.701495 (* 0.0272727 = 0.0191317 loss)
I0429 00:04:45.220522 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0782538 (* 0.0272727 = 0.00213419 loss)
I0429 00:04:45.220535 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0212574 (* 0.0272727 = 0.000579747 loss)
I0429 00:04:45.220549 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0106205 (* 0.0272727 = 0.000289649 loss)
I0429 00:04:45.220563 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00637615 (* 0.0272727 = 0.000173895 loss)
I0429 00:04:45.220577 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00591791 (* 0.0272727 = 0.000161398 loss)
I0429 00:04:45.220590 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00807331 (* 0.0272727 = 0.000220181 loss)
I0429 00:04:45.220603 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00532448 (* 0.0272727 = 0.000145213 loss)
I0429 00:04:45.220615 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.132075
I0429 00:04:45.220628 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:04:45.220639 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:04:45.220650 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.25
I0429 00:04:45.220662 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.375
I0429 00:04:45.220674 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.625
I0429 00:04:45.220685 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:04:45.220697 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:04:45.220710 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:04:45.220721 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:04:45.220732 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:04:45.220743 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:04:45.220755 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:04:45.220767 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:04:45.220778 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:04:45.220790 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:04:45.220801 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:04:45.220813 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:04:45.220824 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:04:45.220835 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:04:45.220847 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:04:45.220859 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:04:45.220870 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:04:45.220880 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.727273
I0429 00:04:45.220892 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.283019
I0429 00:04:45.220906 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.12955 (* 0.3 = 0.938866 loss)
I0429 00:04:45.220921 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.06414 (* 0.3 = 0.319241 loss)
I0429 00:04:45.220934 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.44446 (* 0.0272727 = 0.0939397 loss)
I0429 00:04:45.220947 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.94465 (* 0.0272727 = 0.0803086 loss)
I0429 00:04:45.220975 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.15235 (* 0.0272727 = 0.0859731 loss)
I0429 00:04:45.220991 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.81822 (* 0.0272727 = 0.0768606 loss)
I0429 00:04:45.221004 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 1.92927 (* 0.0272727 = 0.0526164 loss)
I0429 00:04:45.221019 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.02151 (* 0.0272727 = 0.0551322 loss)
I0429 00:04:45.221031 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.24338 (* 0.0272727 = 0.0339104 loss)
I0429 00:04:45.221045 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.17045 (* 0.0272727 = 0.0319213 loss)
I0429 00:04:45.221060 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.991928 (* 0.0272727 = 0.0270526 loss)
I0429 00:04:45.221072 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.471314 (* 0.0272727 = 0.012854 loss)
I0429 00:04:45.221086 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.448967 (* 0.0272727 = 0.0122446 loss)
I0429 00:04:45.221101 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.441813 (* 0.0272727 = 0.0120494 loss)
I0429 00:04:45.221113 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.366769 (* 0.0272727 = 0.0100028 loss)
I0429 00:04:45.221127 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.645458 (* 0.0272727 = 0.0176034 loss)
I0429 00:04:45.221140 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.563043 (* 0.0272727 = 0.0153557 loss)
I0429 00:04:45.221154 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0339999 (* 0.0272727 = 0.000927271 loss)
I0429 00:04:45.221168 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00667001 (* 0.0272727 = 0.000181909 loss)
I0429 00:04:45.221181 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0041973 (* 0.0272727 = 0.000114472 loss)
I0429 00:04:45.221195 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00315898 (* 0.0272727 = 8.61539e-05 loss)
I0429 00:04:45.221209 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00448487 (* 0.0272727 = 0.000122315 loss)
I0429 00:04:45.221222 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00181149 (* 0.0272727 = 4.94044e-05 loss)
I0429 00:04:45.221236 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0031023 (* 0.0272727 = 8.46081e-05 loss)
I0429 00:04:45.221252 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0377358
I0429 00:04:45.221261 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:04:45.221272 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:04:45.221284 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:04:45.221295 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.5
I0429 00:04:45.221307 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.625
I0429 00:04:45.221319 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:04:45.221330 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:04:45.221343 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:04:45.221354 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:04:45.221367 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:04:45.221379 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:04:45.221391 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:04:45.221402 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:04:45.221415 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:04:45.221426 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:04:45.221446 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:04:45.221459 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:04:45.221472 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:04:45.221482 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:04:45.221493 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:04:45.221504 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:04:45.221516 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:04:45.221527 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.704545
I0429 00:04:45.221539 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.283019
I0429 00:04:45.221552 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.03583 (* 1 = 3.03583 loss)
I0429 00:04:45.221566 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.973128 (* 1 = 0.973128 loss)
I0429 00:04:45.221580 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.17231 (* 0.0909091 = 0.288392 loss)
I0429 00:04:45.221593 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.90588 (* 0.0909091 = 0.264171 loss)
I0429 00:04:45.221606 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.08908 (* 0.0909091 = 0.280826 loss)
I0429 00:04:45.221619 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.44678 (* 0.0909091 = 0.222435 loss)
I0429 00:04:45.221632 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.84771 (* 0.0909091 = 0.167974 loss)
I0429 00:04:45.221645 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.89128 (* 0.0909091 = 0.171935 loss)
I0429 00:04:45.221659 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.21553 (* 0.0909091 = 0.110503 loss)
I0429 00:04:45.221673 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.08229 (* 0.0909091 = 0.0983904 loss)
I0429 00:04:45.221685 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.01501 (* 0.0909091 = 0.092274 loss)
I0429 00:04:45.221699 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.397122 (* 0.0909091 = 0.036102 loss)
I0429 00:04:45.221712 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.577921 (* 0.0909091 = 0.0525383 loss)
I0429 00:04:45.221726 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.320578 (* 0.0909091 = 0.0291435 loss)
I0429 00:04:45.221740 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.293177 (* 0.0909091 = 0.0266524 loss)
I0429 00:04:45.221752 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.449953 (* 0.0909091 = 0.0409048 loss)
I0429 00:04:45.221766 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.577848 (* 0.0909091 = 0.0525316 loss)
I0429 00:04:45.221781 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0555538 (* 0.0909091 = 0.00505035 loss)
I0429 00:04:45.221793 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00331422 (* 0.0909091 = 0.000301293 loss)
I0429 00:04:45.221807 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0013046 (* 0.0909091 = 0.0001186 loss)
I0429 00:04:45.221822 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000798188 (* 0.0909091 = 7.25626e-05 loss)
I0429 00:04:45.221835 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000632443 (* 0.0909091 = 5.74948e-05 loss)
I0429 00:04:45.221849 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000494041 (* 0.0909091 = 4.49128e-05 loss)
I0429 00:04:45.221864 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00038808 (* 0.0909091 = 3.528e-05 loss)
I0429 00:04:45.221876 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:04:45.221887 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:04:45.221899 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00013581
I0429 00:04:45.221920 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000149046
I0429 00:04:45.221933 6470 sgd_solver.cpp:106] Iteration 9500, lr = 0.01
I0429 00:07:01.657374 6470 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm14_bn_iter_10000.caffemodel
I0429 00:07:03.879348 6470 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm14_bn_iter_10000.solverstate
I0429 00:07:04.967103 6470 solver.cpp:338] Iteration 10000, Testing net (#0)
I0429 00:07:46.210314 6470 solver.cpp:393] Test loss: 8.84569
I0429 00:07:46.210464 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0830441
I0429 00:07:46.210484 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.139
I0429 00:07:46.210496 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.114
I0429 00:07:46.210508 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.083
I0429 00:07:46.210520 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.176
I0429 00:07:46.210532 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.352
I0429 00:07:46.210544 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.489
I0429 00:07:46.210556 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.74
I0429 00:07:46.210567 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.916
I0429 00:07:46.210579 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.994
I0429 00:07:46.210590 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.999
I0429 00:07:46.210602 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 1
I0429 00:07:46.210613 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 1
I0429 00:07:46.210624 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 1
I0429 00:07:46.210636 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 1
I0429 00:07:46.210647 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 1
I0429 00:07:46.210659 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 1
I0429 00:07:46.210670 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 1
I0429 00:07:46.210681 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 1
I0429 00:07:46.210692 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:07:46.210703 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:07:46.210714 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:07:46.210726 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:07:46.210738 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.768865
I0429 00:07:46.210749 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.248957
I0429 00:07:46.210765 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.11786 (* 0.3 = 0.935359 loss)
I0429 00:07:46.210779 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.826366 (* 0.3 = 0.24791 loss)
I0429 00:07:46.210793 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.83924 (* 0.0272727 = 0.0774338 loss)
I0429 00:07:46.210806 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 3.01158 (* 0.0272727 = 0.082134 loss)
I0429 00:07:46.210820 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 3.10936 (* 0.0272727 = 0.0848008 loss)
I0429 00:07:46.210834 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.90371 (* 0.0272727 = 0.079192 loss)
I0429 00:07:46.210847 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.4807 (* 0.0272727 = 0.0676554 loss)
I0429 00:07:46.210860 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.00469 (* 0.0272727 = 0.0546735 loss)
I0429 00:07:46.210875 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.12244 (* 0.0272727 = 0.030612 loss)
I0429 00:07:46.210896 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.454181 (* 0.0272727 = 0.0123868 loss)
I0429 00:07:46.210918 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.0533118 (* 0.0272727 = 0.00145396 loss)
I0429 00:07:46.210933 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.0225386 (* 0.0272727 = 0.000614689 loss)
I0429 00:07:46.210947 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.0103002 (* 0.0272727 = 0.000280913 loss)
I0429 00:07:46.210960 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.00825207 (* 0.0272727 = 0.000225057 loss)
I0429 00:07:46.210974 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.00650849 (* 0.0272727 = 0.000177504 loss)
I0429 00:07:46.211007 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.00454435 (* 0.0272727 = 0.000123937 loss)
I0429 00:07:46.211022 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.00329466 (* 0.0272727 = 8.98543e-05 loss)
I0429 00:07:46.211036 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.00186652 (* 0.0272727 = 5.09052e-05 loss)
I0429 00:07:46.211050 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.00116331 (* 0.0272727 = 3.17267e-05 loss)
I0429 00:07:46.211063 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.000856757 (* 0.0272727 = 2.33661e-05 loss)
I0429 00:07:46.211076 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.000602962 (* 0.0272727 = 1.64444e-05 loss)
I0429 00:07:46.211091 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.000551901 (* 0.0272727 = 1.50518e-05 loss)
I0429 00:07:46.211104 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.000385206 (* 0.0272727 = 1.05056e-05 loss)
I0429 00:07:46.211117 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.000342768 (* 0.0272727 = 9.34821e-06 loss)
I0429 00:07:46.211129 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0799103
I0429 00:07:46.211141 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.134
I0429 00:07:46.211153 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.092
I0429 00:07:46.211164 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.075
I0429 00:07:46.211175 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.172
I0429 00:07:46.211187 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.335
I0429 00:07:46.211195 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.493
I0429 00:07:46.211207 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.743
I0429 00:07:46.211220 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.916
I0429 00:07:46.211230 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.994
I0429 00:07:46.211242 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.999
I0429 00:07:46.211253 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 1
I0429 00:07:46.211264 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 1
I0429 00:07:46.211275 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 1
I0429 00:07:46.211287 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 1
I0429 00:07:46.211297 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 1
I0429 00:07:46.211316 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 1
I0429 00:07:46.211339 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 1
I0429 00:07:46.211360 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 1
I0429 00:07:46.211382 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:07:46.211396 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:07:46.211407 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:07:46.211418 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:07:46.211429 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.768592
I0429 00:07:46.211441 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.231764
I0429 00:07:46.211454 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.14108 (* 0.3 = 0.942325 loss)
I0429 00:07:46.211484 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.834491 (* 0.3 = 0.250347 loss)
I0429 00:07:46.211500 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.86525 (* 0.0272727 = 0.0781431 loss)
I0429 00:07:46.211513 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 3.04162 (* 0.0272727 = 0.0829533 loss)
I0429 00:07:46.211527 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.12229 (* 0.0272727 = 0.0851533 loss)
I0429 00:07:46.211555 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.92809 (* 0.0272727 = 0.079857 loss)
I0429 00:07:46.211570 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.51018 (* 0.0272727 = 0.0684593 loss)
I0429 00:07:46.211582 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.04405 (* 0.0272727 = 0.0557468 loss)
I0429 00:07:46.211596 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.16149 (* 0.0272727 = 0.031677 loss)
I0429 00:07:46.211609 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.472029 (* 0.0272727 = 0.0128735 loss)
I0429 00:07:46.211623 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.0591623 (* 0.0272727 = 0.00161352 loss)
I0429 00:07:46.211637 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.0263341 (* 0.0272727 = 0.000718203 loss)
I0429 00:07:46.211650 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.0137438 (* 0.0272727 = 0.000374832 loss)
I0429 00:07:46.211664 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.011338 (* 0.0272727 = 0.000309217 loss)
I0429 00:07:46.211678 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.00906166 (* 0.0272727 = 0.000247136 loss)
I0429 00:07:46.211691 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.00658404 (* 0.0272727 = 0.000179565 loss)
I0429 00:07:46.211704 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.0052407 (* 0.0272727 = 0.000142928 loss)
I0429 00:07:46.211717 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.00339156 (* 0.0272727 = 9.24972e-05 loss)
I0429 00:07:46.211730 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.00241807 (* 0.0272727 = 6.59473e-05 loss)
I0429 00:07:46.211745 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.00143411 (* 0.0272727 = 3.91122e-05 loss)
I0429 00:07:46.211757 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.00104611 (* 0.0272727 = 2.85304e-05 loss)
I0429 00:07:46.211771 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.000901004 (* 0.0272727 = 2.45728e-05 loss)
I0429 00:07:46.211784 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.000779661 (* 0.0272727 = 2.12635e-05 loss)
I0429 00:07:46.211797 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000638715 (* 0.0272727 = 1.74195e-05 loss)
I0429 00:07:46.211809 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.0958432
I0429 00:07:46.211820 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.142
I0429 00:07:46.211832 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.106
I0429 00:07:46.211843 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.108
I0429 00:07:46.211854 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.173
I0429 00:07:46.211865 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.331
I0429 00:07:46.211876 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.494
I0429 00:07:46.211887 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.743
I0429 00:07:46.211899 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.918
I0429 00:07:46.211910 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.995
I0429 00:07:46.211921 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.999
I0429 00:07:46.211932 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 1
I0429 00:07:46.211943 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 1
I0429 00:07:46.211954 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 1
I0429 00:07:46.211964 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 1
I0429 00:07:46.211976 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 1
I0429 00:07:46.211985 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 1
I0429 00:07:46.212007 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 1
I0429 00:07:46.212019 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 1
I0429 00:07:46.212030 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:07:46.212041 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:07:46.212052 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:07:46.212064 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:07:46.212074 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.764047
I0429 00:07:46.212085 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.271637
I0429 00:07:46.212098 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.99949 (* 1 = 2.99949 loss)
I0429 00:07:46.212111 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.847276 (* 1 = 0.847276 loss)
I0429 00:07:46.212124 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.79512 (* 0.0909091 = 0.254102 loss)
I0429 00:07:46.212138 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 3.0027 (* 0.0909091 = 0.272972 loss)
I0429 00:07:46.212152 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 3.06069 (* 0.0909091 = 0.278245 loss)
I0429 00:07:46.212164 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.8927 (* 0.0909091 = 0.262973 loss)
I0429 00:07:46.212177 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.46916 (* 0.0909091 = 0.224469 loss)
I0429 00:07:46.212190 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.01254 (* 0.0909091 = 0.182958 loss)
I0429 00:07:46.212203 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.12619 (* 0.0909091 = 0.102381 loss)
I0429 00:07:46.212216 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.451402 (* 0.0909091 = 0.0410365 loss)
I0429 00:07:46.212229 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.0645672 (* 0.0909091 = 0.00586974 loss)
I0429 00:07:46.212242 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.0278258 (* 0.0909091 = 0.00252962 loss)
I0429 00:07:46.212255 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.0121634 (* 0.0909091 = 0.00110577 loss)
I0429 00:07:46.212270 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.00963499 (* 0.0909091 = 0.000875908 loss)
I0429 00:07:46.212282 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.00795027 (* 0.0909091 = 0.000722752 loss)
I0429 00:07:46.212296 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.00572871 (* 0.0909091 = 0.000520792 loss)
I0429 00:07:46.212308 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.00480405 (* 0.0909091 = 0.000436732 loss)
I0429 00:07:46.212322 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.00273814 (* 0.0909091 = 0.000248922 loss)
I0429 00:07:46.212335 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.00237226 (* 0.0909091 = 0.00021566 loss)
I0429 00:07:46.212348 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.001806 (* 0.0909091 = 0.000164182 loss)
I0429 00:07:46.212362 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00145967 (* 0.0909091 = 0.000132697 loss)
I0429 00:07:46.212378 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.0012926 (* 0.0909091 = 0.000117509 loss)
I0429 00:07:46.212391 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.00100982 (* 0.0909091 = 9.18019e-05 loss)
I0429 00:07:46.212404 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.000810855 (* 0.0909091 = 7.37141e-05 loss)
I0429 00:07:46.212416 6470 solver.cpp:406] Test net output #147: total_accuracy = 0
I0429 00:07:46.212427 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.001
I0429 00:07:46.212438 6470 solver.cpp:406] Test net output #149: total_confidence = 0.000238908
I0429 00:07:46.212458 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.000782293
I0429 00:07:46.212472 6470 solver.cpp:338] Iteration 10000, Testing net (#1)
I0429 00:08:27.106183 6470 solver.cpp:393] Test loss: 9.34185
I0429 00:08:27.106333 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0894392
I0429 00:08:27.106354 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.134
I0429 00:08:27.106367 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.108
I0429 00:08:27.106380 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.088
I0429 00:08:27.106392 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.192
I0429 00:08:27.106405 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.345
I0429 00:08:27.106415 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.432
I0429 00:08:27.106427 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.658
I0429 00:08:27.106439 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.803
I0429 00:08:27.106451 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.891
I0429 00:08:27.106462 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.912
I0429 00:08:27.106473 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 0.924
I0429 00:08:27.106485 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 0.942
I0429 00:08:27.106497 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 0.956
I0429 00:08:27.106508 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 0.964
I0429 00:08:27.106519 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 0.978
I0429 00:08:27.106531 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 0.989
I0429 00:08:27.106542 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 0.996
I0429 00:08:27.106554 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 0.999
I0429 00:08:27.106565 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:08:27.106577 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:08:27.106588 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:08:27.106600 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:08:27.106611 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.738455
I0429 00:08:27.106623 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.259062
I0429 00:08:27.106638 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.07643 (* 0.3 = 0.922928 loss)
I0429 00:08:27.106654 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.938655 (* 0.3 = 0.281596 loss)
I0429 00:08:27.106668 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.79595 (* 0.0272727 = 0.0762533 loss)
I0429 00:08:27.106681 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 2.98406 (* 0.0272727 = 0.0813833 loss)
I0429 00:08:27.106695 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 3.0488 (* 0.0272727 = 0.0831491 loss)
I0429 00:08:27.106709 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.8606 (* 0.0272727 = 0.0780164 loss)
I0429 00:08:27.106722 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.44347 (* 0.0272727 = 0.06664 loss)
I0429 00:08:27.106735 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.20321 (* 0.0272727 = 0.0600874 loss)
I0429 00:08:27.106750 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.43514 (* 0.0272727 = 0.0391401 loss)
I0429 00:08:27.106762 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.842766 (* 0.0272727 = 0.0229845 loss)
I0429 00:08:27.106776 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.455368 (* 0.0272727 = 0.0124191 loss)
I0429 00:08:27.106789 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.379619 (* 0.0272727 = 0.0103532 loss)
I0429 00:08:27.106803 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.335706 (* 0.0272727 = 0.00915561 loss)
I0429 00:08:27.106817 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.267723 (* 0.0272727 = 0.00730152 loss)
I0429 00:08:27.106830 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.200889 (* 0.0272727 = 0.00547878 loss)
I0429 00:08:27.106864 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.180692 (* 0.0272727 = 0.00492797 loss)
I0429 00:08:27.106879 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.127282 (* 0.0272727 = 0.00347132 loss)
I0429 00:08:27.106894 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.0767401 (* 0.0272727 = 0.00209291 loss)
I0429 00:08:27.106907 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.0340804 (* 0.0272727 = 0.000929465 loss)
I0429 00:08:27.106921 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.00993404 (* 0.0272727 = 0.000270928 loss)
I0429 00:08:27.106935 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00139224 (* 0.0272727 = 3.79703e-05 loss)
I0429 00:08:27.106948 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.0010111 (* 0.0272727 = 2.75754e-05 loss)
I0429 00:08:27.106962 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.000688317 (* 0.0272727 = 1.87723e-05 loss)
I0429 00:08:27.106976 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.000619119 (* 0.0272727 = 1.68851e-05 loss)
I0429 00:08:27.106988 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0844472
I0429 00:08:27.106999 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.144
I0429 00:08:27.107012 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.095
I0429 00:08:27.107023 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.083
I0429 00:08:27.107034 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.19
I0429 00:08:27.107045 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.341
I0429 00:08:27.107056 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.429
I0429 00:08:27.107069 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.659
I0429 00:08:27.107079 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.803
I0429 00:08:27.107090 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.892
I0429 00:08:27.107102 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.912
I0429 00:08:27.107113 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 0.924
I0429 00:08:27.107125 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 0.942
I0429 00:08:27.107136 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 0.956
I0429 00:08:27.107147 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 0.964
I0429 00:08:27.107158 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 0.978
I0429 00:08:27.107169 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 0.989
I0429 00:08:27.107180 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 0.996
I0429 00:08:27.107192 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 0.999
I0429 00:08:27.107203 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:08:27.107214 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:08:27.107225 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:08:27.107236 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:08:27.107247 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.73741
I0429 00:08:27.107259 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.249421
I0429 00:08:27.107271 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.09851 (* 0.3 = 0.929554 loss)
I0429 00:08:27.107285 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.946801 (* 0.3 = 0.28404 loss)
I0429 00:08:27.107298 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.82356 (* 0.0272727 = 0.0770061 loss)
I0429 00:08:27.107316 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 3.02292 (* 0.0272727 = 0.0824433 loss)
I0429 00:08:27.107341 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.08883 (* 0.0272727 = 0.0842408 loss)
I0429 00:08:27.107355 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.88988 (* 0.0272727 = 0.0788148 loss)
I0429 00:08:27.107369 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.4728 (* 0.0272727 = 0.0674399 loss)
I0429 00:08:27.107383 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.24757 (* 0.0272727 = 0.0612974 loss)
I0429 00:08:27.107395 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.47544 (* 0.0272727 = 0.0402394 loss)
I0429 00:08:27.107409 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.860983 (* 0.0272727 = 0.0234814 loss)
I0429 00:08:27.107422 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.466428 (* 0.0272727 = 0.0127208 loss)
I0429 00:08:27.107435 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.385312 (* 0.0272727 = 0.0105085 loss)
I0429 00:08:27.107448 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.339786 (* 0.0272727 = 0.00926688 loss)
I0429 00:08:27.107462 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.280022 (* 0.0272727 = 0.00763696 loss)
I0429 00:08:27.107492 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.205582 (* 0.0272727 = 0.00560677 loss)
I0429 00:08:27.107503 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.177512 (* 0.0272727 = 0.00484123 loss)
I0429 00:08:27.107512 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.129398 (* 0.0272727 = 0.00352904 loss)
I0429 00:08:27.107527 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.0772335 (* 0.0272727 = 0.00210637 loss)
I0429 00:08:27.107542 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.0367279 (* 0.0272727 = 0.00100167 loss)
I0429 00:08:27.107554 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.00986757 (* 0.0272727 = 0.000269115 loss)
I0429 00:08:27.107568 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.0020341 (* 0.0272727 = 5.54753e-05 loss)
I0429 00:08:27.107583 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00145693 (* 0.0272727 = 3.97346e-05 loss)
I0429 00:08:27.107596 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.00109288 (* 0.0272727 = 2.98059e-05 loss)
I0429 00:08:27.107609 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000919084 (* 0.0272727 = 2.50659e-05 loss)
I0429 00:08:27.107621 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.101048
I0429 00:08:27.107632 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.151
I0429 00:08:27.107645 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.11
I0429 00:08:27.107656 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.1
I0429 00:08:27.107666 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.191
I0429 00:08:27.107677 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.339
I0429 00:08:27.107689 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.431
I0429 00:08:27.107700 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.66
I0429 00:08:27.107712 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.806
I0429 00:08:27.107722 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.892
I0429 00:08:27.107733 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.913
I0429 00:08:27.107744 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.925
I0429 00:08:27.107755 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 0.941
I0429 00:08:27.107766 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 0.955
I0429 00:08:27.107777 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 0.964
I0429 00:08:27.107789 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 0.978
I0429 00:08:27.107800 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 0.989
I0429 00:08:27.107822 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 0.996
I0429 00:08:27.107834 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 0.999
I0429 00:08:27.107846 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:08:27.107857 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:08:27.107868 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:08:27.107879 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:08:27.107889 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.735001
I0429 00:08:27.107902 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.287917
I0429 00:08:27.107914 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.9671 (* 1 = 2.9671 loss)
I0429 00:08:27.107928 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.957839 (* 1 = 0.957839 loss)
I0429 00:08:27.107941 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.75618 (* 0.0909091 = 0.250562 loss)
I0429 00:08:27.107954 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 2.9856 (* 0.0909091 = 0.271418 loss)
I0429 00:08:27.107967 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 3.04872 (* 0.0909091 = 0.277156 loss)
I0429 00:08:27.107980 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.84777 (* 0.0909091 = 0.258888 loss)
I0429 00:08:27.107993 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.41587 (* 0.0909091 = 0.219625 loss)
I0429 00:08:27.108006 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.17799 (* 0.0909091 = 0.197999 loss)
I0429 00:08:27.108019 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.44243 (* 0.0909091 = 0.13113 loss)
I0429 00:08:27.108033 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.830835 (* 0.0909091 = 0.0755304 loss)
I0429 00:08:27.108047 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.450059 (* 0.0909091 = 0.0409145 loss)
I0429 00:08:27.108059 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.362533 (* 0.0909091 = 0.0329575 loss)
I0429 00:08:27.108073 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.318609 (* 0.0909091 = 0.0289645 loss)
I0429 00:08:27.108086 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.251715 (* 0.0909091 = 0.0228832 loss)
I0429 00:08:27.108099 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.191714 (* 0.0909091 = 0.0174285 loss)
I0429 00:08:27.108114 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.165871 (* 0.0909091 = 0.0150792 loss)
I0429 00:08:27.108126 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.118204 (* 0.0909091 = 0.0107459 loss)
I0429 00:08:27.108139 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.068576 (* 0.0909091 = 0.00623418 loss)
I0429 00:08:27.108152 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.0327362 (* 0.0909091 = 0.00297602 loss)
I0429 00:08:27.108166 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.0093182 (* 0.0909091 = 0.000847109 loss)
I0429 00:08:27.108180 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00325215 (* 0.0909091 = 0.00029565 loss)
I0429 00:08:27.108193 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.00204709 (* 0.0909091 = 0.000186099 loss)
I0429 00:08:27.108206 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.00131044 (* 0.0909091 = 0.000119131 loss)
I0429 00:08:27.108219 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.00096497 (* 0.0909091 = 8.77246e-05 loss)
I0429 00:08:27.108232 6470 solver.cpp:406] Test net output #147: total_accuracy = 0
I0429 00:08:27.108242 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.001
I0429 00:08:27.108253 6470 solver.cpp:406] Test net output #149: total_confidence = 0.000238399
I0429 00:08:27.108274 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.000733547
I0429 00:08:27.289371 6470 solver.cpp:229] Iteration 10000, loss = 9.65647
I0429 00:08:27.289460 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12069
I0429 00:08:27.289489 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:08:27.289505 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:08:27.289520 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.25
I0429 00:08:27.289532 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 00:08:27.289544 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0429 00:08:27.289556 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:08:27.289568 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0429 00:08:27.289579 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0429 00:08:27.289592 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:08:27.289603 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:08:27.289614 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:08:27.289626 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:08:27.289638 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:08:27.289650 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:08:27.289662 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:08:27.289674 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:08:27.289685 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:08:27.289697 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:08:27.289710 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:08:27.289721 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:08:27.289732 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:08:27.289744 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:08:27.289755 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.693182
I0429 00:08:27.289767 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.362069
I0429 00:08:27.289783 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.77822 (* 0.3 = 0.833467 loss)
I0429 00:08:27.289798 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.03241 (* 0.3 = 0.309722 loss)
I0429 00:08:27.289813 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.72936 (* 0.0272727 = 0.0744372 loss)
I0429 00:08:27.289827 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.66385 (* 0.0272727 = 0.0726504 loss)
I0429 00:08:27.289841 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.7913 (* 0.0272727 = 0.0761263 loss)
I0429 00:08:27.289855 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.40595 (* 0.0272727 = 0.0656169 loss)
I0429 00:08:27.289868 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.65507 (* 0.0272727 = 0.072411 loss)
I0429 00:08:27.289882 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.29879 (* 0.0272727 = 0.0626943 loss)
I0429 00:08:27.289896 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.2018 (* 0.0272727 = 0.0600492 loss)
I0429 00:08:27.289911 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.00865 (* 0.0272727 = 0.0547815 loss)
I0429 00:08:27.289923 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.745229 (* 0.0272727 = 0.0203244 loss)
I0429 00:08:27.289937 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.626817 (* 0.0272727 = 0.017095 loss)
I0429 00:08:27.289952 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.490151 (* 0.0272727 = 0.0133677 loss)
I0429 00:08:27.290000 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.440242 (* 0.0272727 = 0.0120066 loss)
I0429 00:08:27.290015 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.572087 (* 0.0272727 = 0.0156024 loss)
I0429 00:08:27.290030 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.539561 (* 0.0272727 = 0.0147153 loss)
I0429 00:08:27.290045 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0328652 (* 0.0272727 = 0.000896324 loss)
I0429 00:08:27.290058 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0157516 (* 0.0272727 = 0.000429589 loss)
I0429 00:08:27.290076 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0092107 (* 0.0272727 = 0.000251201 loss)
I0429 00:08:27.290091 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00693414 (* 0.0272727 = 0.000189113 loss)
I0429 00:08:27.290104 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00565899 (* 0.0272727 = 0.000154336 loss)
I0429 00:08:27.290118 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0043402 (* 0.0272727 = 0.000118369 loss)
I0429 00:08:27.290132 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00488901 (* 0.0272727 = 0.000133337 loss)
I0429 00:08:27.290148 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00308603 (* 0.0272727 = 8.41644e-05 loss)
I0429 00:08:27.290159 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0689655
I0429 00:08:27.290171 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:08:27.290182 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:08:27.290194 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:08:27.290206 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:08:27.290217 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 00:08:27.290230 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 00:08:27.290241 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0429 00:08:27.290253 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 00:08:27.290264 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:08:27.290277 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:08:27.290287 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:08:27.290299 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:08:27.290310 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:08:27.290323 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:08:27.290334 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:08:27.290345 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:08:27.290356 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:08:27.290369 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:08:27.290380 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:08:27.290390 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:08:27.290402 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:08:27.290413 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:08:27.290426 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.676136
I0429 00:08:27.290437 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.275862
I0429 00:08:27.290451 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.04324 (* 0.3 = 0.912971 loss)
I0429 00:08:27.290464 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.13876 (* 0.3 = 0.341628 loss)
I0429 00:08:27.290489 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.90579 (* 0.0272727 = 0.0792489 loss)
I0429 00:08:27.290504 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.77412 (* 0.0272727 = 0.0756577 loss)
I0429 00:08:27.290518 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.82962 (* 0.0272727 = 0.0771716 loss)
I0429 00:08:27.290532 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.92629 (* 0.0272727 = 0.0798078 loss)
I0429 00:08:27.290546 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.78647 (* 0.0272727 = 0.0759947 loss)
I0429 00:08:27.290560 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.30085 (* 0.0272727 = 0.0627504 loss)
I0429 00:08:27.290576 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.21159 (* 0.0272727 = 0.060316 loss)
I0429 00:08:27.290591 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.76245 (* 0.0272727 = 0.0480667 loss)
I0429 00:08:27.290603 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.888502 (* 0.0272727 = 0.0242319 loss)
I0429 00:08:27.290617 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.697583 (* 0.0272727 = 0.019025 loss)
I0429 00:08:27.290630 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.608282 (* 0.0272727 = 0.0165895 loss)
I0429 00:08:27.290644 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.591482 (* 0.0272727 = 0.0161313 loss)
I0429 00:08:27.290658 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.62936 (* 0.0272727 = 0.0171644 loss)
I0429 00:08:27.290671 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.640018 (* 0.0272727 = 0.017455 loss)
I0429 00:08:27.290685 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0274646 (* 0.0272727 = 0.000749036 loss)
I0429 00:08:27.290699 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0148461 (* 0.0272727 = 0.000404893 loss)
I0429 00:08:27.290714 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0133494 (* 0.0272727 = 0.000364076 loss)
I0429 00:08:27.290727 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00821688 (* 0.0272727 = 0.000224097 loss)
I0429 00:08:27.290741 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00784198 (* 0.0272727 = 0.000213872 loss)
I0429 00:08:27.290755 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00641953 (* 0.0272727 = 0.000175078 loss)
I0429 00:08:27.290768 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00656962 (* 0.0272727 = 0.000179171 loss)
I0429 00:08:27.290782 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00414393 (* 0.0272727 = 0.000113016 loss)
I0429 00:08:27.290794 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.172414
I0429 00:08:27.290807 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:08:27.290817 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:08:27.290829 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:08:27.290841 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:08:27.290849 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:08:27.290858 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:08:27.290865 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0429 00:08:27.290877 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0429 00:08:27.290889 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:08:27.290901 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:08:27.290912 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:08:27.290925 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:08:27.290935 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:08:27.290956 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:08:27.290971 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:08:27.290982 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:08:27.290992 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:08:27.291004 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:08:27.291015 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:08:27.291026 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:08:27.291038 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:08:27.291049 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:08:27.291060 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.704545
I0429 00:08:27.291071 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.396552
I0429 00:08:27.291085 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.79 (* 1 = 2.79 loss)
I0429 00:08:27.291098 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.08503 (* 1 = 1.08503 loss)
I0429 00:08:27.291112 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.78218 (* 0.0909091 = 0.252926 loss)
I0429 00:08:27.291129 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.71125 (* 0.0909091 = 0.246477 loss)
I0429 00:08:27.291143 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.81708 (* 0.0909091 = 0.256098 loss)
I0429 00:08:27.291157 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.79428 (* 0.0909091 = 0.254025 loss)
I0429 00:08:27.291170 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.41494 (* 0.0909091 = 0.21954 loss)
I0429 00:08:27.291184 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.23258 (* 0.0909091 = 0.202961 loss)
I0429 00:08:27.291198 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.16972 (* 0.0909091 = 0.197247 loss)
I0429 00:08:27.291210 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.73385 (* 0.0909091 = 0.157623 loss)
I0429 00:08:27.291224 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.743701 (* 0.0909091 = 0.0676092 loss)
I0429 00:08:27.291239 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.762533 (* 0.0909091 = 0.0693212 loss)
I0429 00:08:27.291251 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.568333 (* 0.0909091 = 0.0516667 loss)
I0429 00:08:27.291265 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.432174 (* 0.0909091 = 0.0392885 loss)
I0429 00:08:27.291278 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.533323 (* 0.0909091 = 0.0484839 loss)
I0429 00:08:27.291292 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.534597 (* 0.0909091 = 0.0485998 loss)
I0429 00:08:27.291306 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0659478 (* 0.0909091 = 0.00599526 loss)
I0429 00:08:27.291319 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0258204 (* 0.0909091 = 0.00234731 loss)
I0429 00:08:27.291333 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0206594 (* 0.0909091 = 0.00187812 loss)
I0429 00:08:27.291347 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0129311 (* 0.0909091 = 0.00117556 loss)
I0429 00:08:27.291360 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00908755 (* 0.0909091 = 0.000826141 loss)
I0429 00:08:27.291374 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00681928 (* 0.0909091 = 0.000619934 loss)
I0429 00:08:27.291388 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00587587 (* 0.0909091 = 0.00053417 loss)
I0429 00:08:27.291402 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00477314 (* 0.0909091 = 0.000433921 loss)
I0429 00:08:27.291424 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:08:27.291437 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:08:27.291448 6470 solver.cpp:245] Train net output #149: total_confidence = 2.46442e-07
I0429 00:08:27.291460 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 8.05034e-06
I0429 00:08:27.291488 6470 sgd_solver.cpp:106] Iteration 10000, lr = 0.01
I0429 00:10:43.966632 6470 solver.cpp:229] Iteration 10500, loss = 9.74112
I0429 00:10:43.966795 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0307692
I0429 00:10:43.966825 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:10:43.966845 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:10:43.966858 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:10:43.966871 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 00:10:43.966881 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 00:10:43.966893 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:10:43.966905 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0429 00:10:43.966917 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.5
I0429 00:10:43.966928 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.5
I0429 00:10:43.966940 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.625
I0429 00:10:43.966953 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.75
I0429 00:10:43.966964 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:10:43.966975 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:10:43.966987 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:10:43.967000 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:10:43.967011 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:10:43.967023 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:10:43.967034 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:10:43.967046 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:10:43.967058 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:10:43.967069 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:10:43.967082 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:10:43.967092 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.636364
I0429 00:10:43.967104 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.169231
I0429 00:10:43.967120 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.28959 (* 0.3 = 0.986877 loss)
I0429 00:10:43.967134 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.27407 (* 0.3 = 0.38222 loss)
I0429 00:10:43.967149 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.27291 (* 0.0272727 = 0.0892612 loss)
I0429 00:10:43.967162 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.04639 (* 0.0272727 = 0.0830834 loss)
I0429 00:10:43.967175 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.40127 (* 0.0272727 = 0.0927618 loss)
I0429 00:10:43.967190 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.82496 (* 0.0272727 = 0.0770443 loss)
I0429 00:10:43.967203 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.86148 (* 0.0272727 = 0.0780403 loss)
I0429 00:10:43.967217 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.6745 (* 0.0272727 = 0.0729409 loss)
I0429 00:10:43.967231 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.0344 (* 0.0272727 = 0.0554836 loss)
I0429 00:10:43.967245 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.11361 (* 0.0272727 = 0.0576438 loss)
I0429 00:10:43.967258 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 2.19577 (* 0.0272727 = 0.0598845 loss)
I0429 00:10:43.967272 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 1.61176 (* 0.0272727 = 0.043957 loss)
I0429 00:10:43.967286 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 1.0494 (* 0.0272727 = 0.0286201 loss)
I0429 00:10:43.967299 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.544954 (* 0.0272727 = 0.0148624 loss)
I0429 00:10:43.967316 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.48037 (* 0.0272727 = 0.013101 loss)
I0429 00:10:43.967350 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.640983 (* 0.0272727 = 0.0174814 loss)
I0429 00:10:43.967365 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.628009 (* 0.0272727 = 0.0171275 loss)
I0429 00:10:43.967381 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.114374 (* 0.0272727 = 0.00311929 loss)
I0429 00:10:43.967394 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0544193 (* 0.0272727 = 0.00148416 loss)
I0429 00:10:43.967408 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0596806 (* 0.0272727 = 0.00162765 loss)
I0429 00:10:43.967422 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0453629 (* 0.0272727 = 0.00123717 loss)
I0429 00:10:43.967437 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.020077 (* 0.0272727 = 0.000547556 loss)
I0429 00:10:43.967450 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.0110383 (* 0.0272727 = 0.000301045 loss)
I0429 00:10:43.967478 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0126476 (* 0.0272727 = 0.000344934 loss)
I0429 00:10:43.967494 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0307692
I0429 00:10:43.967507 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:10:43.967519 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:10:43.967530 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:10:43.967542 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.375
I0429 00:10:43.967555 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0429 00:10:43.967566 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0429 00:10:43.967577 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0429 00:10:43.967589 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.5
I0429 00:10:43.967602 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.5
I0429 00:10:43.967612 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.625
I0429 00:10:43.967624 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.75
I0429 00:10:43.967636 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:10:43.967648 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:10:43.967659 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:10:43.967671 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:10:43.967684 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:10:43.967694 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:10:43.967705 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:10:43.967717 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:10:43.967728 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:10:43.967741 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:10:43.967751 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:10:43.967763 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.642045
I0429 00:10:43.967775 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.138462
I0429 00:10:43.967789 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.35601 (* 0.3 = 1.0068 loss)
I0429 00:10:43.967803 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.28932 (* 0.3 = 0.386795 loss)
I0429 00:10:43.967816 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.38966 (* 0.0272727 = 0.0924453 loss)
I0429 00:10:43.967829 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.92254 (* 0.0272727 = 0.0797056 loss)
I0429 00:10:43.967856 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.21753 (* 0.0272727 = 0.0877508 loss)
I0429 00:10:43.967875 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.88996 (* 0.0272727 = 0.0788172 loss)
I0429 00:10:43.967888 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.03561 (* 0.0272727 = 0.0827895 loss)
I0429 00:10:43.967902 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.42537 (* 0.0272727 = 0.0661466 loss)
I0429 00:10:43.967916 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.84618 (* 0.0272727 = 0.0503505 loss)
I0429 00:10:43.967931 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.27879 (* 0.0272727 = 0.0621489 loss)
I0429 00:10:43.967944 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 2.49889 (* 0.0272727 = 0.0681514 loss)
I0429 00:10:43.967958 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 1.65702 (* 0.0272727 = 0.0451914 loss)
I0429 00:10:43.967972 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 1.33159 (* 0.0272727 = 0.036316 loss)
I0429 00:10:43.967986 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.504519 (* 0.0272727 = 0.0137596 loss)
I0429 00:10:43.967999 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.481922 (* 0.0272727 = 0.0131433 loss)
I0429 00:10:43.968014 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.590013 (* 0.0272727 = 0.0160913 loss)
I0429 00:10:43.968027 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.689074 (* 0.0272727 = 0.0187929 loss)
I0429 00:10:43.968041 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0302049 (* 0.0272727 = 0.000823771 loss)
I0429 00:10:43.968055 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00801489 (* 0.0272727 = 0.000218588 loss)
I0429 00:10:43.968070 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0064441 (* 0.0272727 = 0.000175748 loss)
I0429 00:10:43.968083 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0050536 (* 0.0272727 = 0.000137825 loss)
I0429 00:10:43.968097 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00579128 (* 0.0272727 = 0.000157944 loss)
I0429 00:10:43.968111 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00443145 (* 0.0272727 = 0.000120858 loss)
I0429 00:10:43.968124 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00335893 (* 0.0272727 = 9.16073e-05 loss)
I0429 00:10:43.968137 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0615385
I0429 00:10:43.968149 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0429 00:10:43.968160 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:10:43.968173 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:10:43.968184 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 00:10:43.968196 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 00:10:43.968209 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:10:43.968219 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.5
I0429 00:10:43.968231 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.5
I0429 00:10:43.968242 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.5
I0429 00:10:43.968255 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.625
I0429 00:10:43.968266 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0429 00:10:43.968277 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:10:43.968289 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:10:43.968302 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:10:43.968313 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:10:43.968324 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:10:43.968346 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:10:43.968359 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:10:43.968374 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:10:43.968386 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:10:43.968397 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:10:43.968410 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:10:43.968421 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.647727
I0429 00:10:43.968433 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.230769
I0429 00:10:43.968447 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.25575 (* 1 = 3.25575 loss)
I0429 00:10:43.968461 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.27101 (* 1 = 1.27101 loss)
I0429 00:10:43.968474 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.26476 (* 0.0909091 = 0.296796 loss)
I0429 00:10:43.968488 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.81752 (* 0.0909091 = 0.256138 loss)
I0429 00:10:43.968502 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.29372 (* 0.0909091 = 0.299429 loss)
I0429 00:10:43.968516 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.91812 (* 0.0909091 = 0.265284 loss)
I0429 00:10:43.968530 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.86702 (* 0.0909091 = 0.260638 loss)
I0429 00:10:43.968544 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.51064 (* 0.0909091 = 0.22824 loss)
I0429 00:10:43.968557 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.66612 (* 0.0909091 = 0.151465 loss)
I0429 00:10:43.968570 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.9631 (* 0.0909091 = 0.178463 loss)
I0429 00:10:43.968585 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 2.09775 (* 0.0909091 = 0.190705 loss)
I0429 00:10:43.968598 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 1.50634 (* 0.0909091 = 0.13694 loss)
I0429 00:10:43.968612 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 1.0581 (* 0.0909091 = 0.0961911 loss)
I0429 00:10:43.968626 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.577487 (* 0.0909091 = 0.0524988 loss)
I0429 00:10:43.968639 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.54414 (* 0.0909091 = 0.0494673 loss)
I0429 00:10:43.968653 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.63395 (* 0.0909091 = 0.0576319 loss)
I0429 00:10:43.968667 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.670789 (* 0.0909091 = 0.0609809 loss)
I0429 00:10:43.968680 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0284599 (* 0.0909091 = 0.00258727 loss)
I0429 00:10:43.968694 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.018848 (* 0.0909091 = 0.00171345 loss)
I0429 00:10:43.968708 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0164199 (* 0.0909091 = 0.00149271 loss)
I0429 00:10:43.968722 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0152534 (* 0.0909091 = 0.00138667 loss)
I0429 00:10:43.968736 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.013816 (* 0.0909091 = 0.001256 loss)
I0429 00:10:43.968750 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0112686 (* 0.0909091 = 0.00102442 loss)
I0429 00:10:43.968765 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.010138 (* 0.0909091 = 0.000921633 loss)
I0429 00:10:43.968776 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:10:43.968787 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:10:43.968799 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00131495
I0429 00:10:43.968821 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00172045
I0429 00:10:43.968832 6470 sgd_solver.cpp:106] Iteration 10500, lr = 0.01
I0429 00:13:00.469928 6470 solver.cpp:229] Iteration 11000, loss = 9.67309
I0429 00:13:00.470065 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0983607
I0429 00:13:00.470087 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:13:00.470099 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:13:00.470111 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:13:00.470124 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:13:00.470135 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 00:13:00.470147 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:13:00.470160 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:13:00.470171 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:13:00.470183 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:13:00.470194 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:13:00.470206 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:13:00.470218 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:13:00.470229 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:13:00.470242 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:13:00.470253 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:13:00.470265 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0429 00:13:00.470276 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 0.875
I0429 00:13:00.470288 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 0.875
I0429 00:13:00.470300 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:13:00.470314 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:13:00.470326 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:13:00.470338 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:13:00.470350 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.681818
I0429 00:13:00.470362 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.344262
I0429 00:13:00.470377 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.9404 (* 0.3 = 0.882119 loss)
I0429 00:13:00.470392 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.11732 (* 0.3 = 0.335196 loss)
I0429 00:13:00.470407 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.92983 (* 0.0272727 = 0.0799043 loss)
I0429 00:13:00.470420 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.7631 (* 0.0272727 = 0.0753574 loss)
I0429 00:13:00.470434 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.93184 (* 0.0272727 = 0.0799594 loss)
I0429 00:13:00.470448 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.92631 (* 0.0272727 = 0.0798084 loss)
I0429 00:13:00.470461 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.79294 (* 0.0272727 = 0.0761711 loss)
I0429 00:13:00.470475 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.37262 (* 0.0272727 = 0.0647078 loss)
I0429 00:13:00.470489 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.59099 (* 0.0272727 = 0.0433907 loss)
I0429 00:13:00.470501 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.206 (* 0.0272727 = 0.0328909 loss)
I0429 00:13:00.470515 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.518163 (* 0.0272727 = 0.0141317 loss)
I0429 00:13:00.470528 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.620987 (* 0.0272727 = 0.016936 loss)
I0429 00:13:00.470542 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.513972 (* 0.0272727 = 0.0140174 loss)
I0429 00:13:00.470556 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.666817 (* 0.0272727 = 0.0181859 loss)
I0429 00:13:00.470589 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.713477 (* 0.0272727 = 0.0194585 loss)
I0429 00:13:00.470604 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.923217 (* 0.0272727 = 0.0251786 loss)
I0429 00:13:00.470618 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.661196 (* 0.0272727 = 0.0180326 loss)
I0429 00:13:00.470631 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.817856 (* 0.0272727 = 0.0223052 loss)
I0429 00:13:00.470644 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 1.0714 (* 0.0272727 = 0.0292201 loss)
I0429 00:13:00.470659 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.958761 (* 0.0272727 = 0.026148 loss)
I0429 00:13:00.470672 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000758058 (* 0.0272727 = 2.06743e-05 loss)
I0429 00:13:00.470686 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000931832 (* 0.0272727 = 2.54136e-05 loss)
I0429 00:13:00.470700 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000492955 (* 0.0272727 = 1.34442e-05 loss)
I0429 00:13:00.470713 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000396092 (* 0.0272727 = 1.08025e-05 loss)
I0429 00:13:00.470726 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0983607
I0429 00:13:00.470737 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:13:00.470748 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.375
I0429 00:13:00.470760 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:13:00.470772 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0429 00:13:00.470783 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:13:00.470794 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 00:13:00.470806 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:13:00.470818 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:13:00.470829 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:13:00.470840 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:13:00.470852 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:13:00.470865 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:13:00.470875 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:13:00.470887 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:13:00.470899 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:13:00.470911 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0429 00:13:00.470922 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 0.875
I0429 00:13:00.470933 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 0.875
I0429 00:13:00.470945 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:13:00.470957 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:13:00.470968 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:13:00.470979 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:13:00.470990 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.6875
I0429 00:13:00.471001 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.311475
I0429 00:13:00.471015 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.9468 (* 0.3 = 0.88404 loss)
I0429 00:13:00.471029 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.1023 (* 0.3 = 0.330689 loss)
I0429 00:13:00.471043 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.1068 (* 0.0272727 = 0.084731 loss)
I0429 00:13:00.471056 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.77319 (* 0.0272727 = 0.0756324 loss)
I0429 00:13:00.471084 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.05126 (* 0.0272727 = 0.0832162 loss)
I0429 00:13:00.471101 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.87199 (* 0.0272727 = 0.078327 loss)
I0429 00:13:00.471113 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.62142 (* 0.0272727 = 0.0714932 loss)
I0429 00:13:00.471127 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.35809 (* 0.0272727 = 0.0643116 loss)
I0429 00:13:00.471140 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.35601 (* 0.0272727 = 0.0369821 loss)
I0429 00:13:00.471153 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.05391 (* 0.0272727 = 0.0287431 loss)
I0429 00:13:00.471168 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.483142 (* 0.0272727 = 0.0131766 loss)
I0429 00:13:00.471180 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.47549 (* 0.0272727 = 0.0129679 loss)
I0429 00:13:00.471194 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.377772 (* 0.0272727 = 0.0103029 loss)
I0429 00:13:00.471207 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.447159 (* 0.0272727 = 0.0121953 loss)
I0429 00:13:00.471221 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.577889 (* 0.0272727 = 0.0157606 loss)
I0429 00:13:00.471235 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.77449 (* 0.0272727 = 0.0211225 loss)
I0429 00:13:00.471248 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.514396 (* 0.0272727 = 0.014029 loss)
I0429 00:13:00.471261 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.590468 (* 0.0272727 = 0.0161037 loss)
I0429 00:13:00.471276 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.817512 (* 0.0272727 = 0.0222958 loss)
I0429 00:13:00.471289 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.903504 (* 0.0272727 = 0.024641 loss)
I0429 00:13:00.471303 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00467132 (* 0.0272727 = 0.0001274 loss)
I0429 00:13:00.471318 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00210313 (* 0.0272727 = 5.73582e-05 loss)
I0429 00:13:00.471331 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00145439 (* 0.0272727 = 3.96653e-05 loss)
I0429 00:13:00.471345 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00114841 (* 0.0272727 = 3.13204e-05 loss)
I0429 00:13:00.471357 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.163934
I0429 00:13:00.471371 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:13:00.471385 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:13:00.471395 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.25
I0429 00:13:00.471407 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:13:00.471418 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0429 00:13:00.471431 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:13:00.471441 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:13:00.471452 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:13:00.471475 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:13:00.471492 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:13:00.471503 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:13:00.471515 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:13:00.471526 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:13:00.471539 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:13:00.471549 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:13:00.471576 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0429 00:13:00.471590 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 0.875
I0429 00:13:00.471601 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 0.875
I0429 00:13:00.471612 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:13:00.471624 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:13:00.471635 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:13:00.471647 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:13:00.471658 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.698864
I0429 00:13:00.471669 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.360656
I0429 00:13:00.471683 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.86729 (* 1 = 2.86729 loss)
I0429 00:13:00.471698 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.10183 (* 1 = 1.10183 loss)
I0429 00:13:00.471710 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.74956 (* 0.0909091 = 0.24996 loss)
I0429 00:13:00.471724 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.77779 (* 0.0909091 = 0.252527 loss)
I0429 00:13:00.471736 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.84034 (* 0.0909091 = 0.258213 loss)
I0429 00:13:00.471750 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.09655 (* 0.0909091 = 0.281504 loss)
I0429 00:13:00.471763 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.76235 (* 0.0909091 = 0.251123 loss)
I0429 00:13:00.471776 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.17054 (* 0.0909091 = 0.197321 loss)
I0429 00:13:00.471789 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.37507 (* 0.0909091 = 0.125007 loss)
I0429 00:13:00.471802 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.15043 (* 0.0909091 = 0.104584 loss)
I0429 00:13:00.471817 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.41687 (* 0.0909091 = 0.0378973 loss)
I0429 00:13:00.471829 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.458481 (* 0.0909091 = 0.0416801 loss)
I0429 00:13:00.471843 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.390871 (* 0.0909091 = 0.0355337 loss)
I0429 00:13:00.471856 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.450656 (* 0.0909091 = 0.0409687 loss)
I0429 00:13:00.471869 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.511574 (* 0.0909091 = 0.0465067 loss)
I0429 00:13:00.471884 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.728296 (* 0.0909091 = 0.0662087 loss)
I0429 00:13:00.471896 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.427423 (* 0.0909091 = 0.0388566 loss)
I0429 00:13:00.471909 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.58365 (* 0.0909091 = 0.0530591 loss)
I0429 00:13:00.471923 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.716082 (* 0.0909091 = 0.0650983 loss)
I0429 00:13:00.471936 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.949569 (* 0.0909091 = 0.0863244 loss)
I0429 00:13:00.471951 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00091317 (* 0.0909091 = 8.30155e-05 loss)
I0429 00:13:00.471964 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000389604 (* 0.0909091 = 3.54185e-05 loss)
I0429 00:13:00.471977 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000111183 (* 0.0909091 = 1.01075e-05 loss)
I0429 00:13:00.471992 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 6.62442e-05 (* 0.0909091 = 6.0222e-06 loss)
I0429 00:13:00.472003 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:13:00.472014 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:13:00.472034 6470 solver.cpp:245] Train net output #149: total_confidence = 9.51203e-07
I0429 00:13:00.472048 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 1.08569e-05
I0429 00:13:00.472060 6470 sgd_solver.cpp:106] Iteration 11000, lr = 0.01
I0429 00:15:17.228919 6470 solver.cpp:229] Iteration 11500, loss = 9.6639
I0429 00:15:17.229079 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0754717
I0429 00:15:17.229099 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:15:17.229113 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:15:17.229126 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:15:17.229138 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0429 00:15:17.229149 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 00:15:17.229161 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:15:17.229173 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:15:17.229185 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 00:15:17.229197 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:15:17.229209 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:15:17.229220 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:15:17.229233 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:15:17.229244 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:15:17.229256 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:15:17.229269 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:15:17.229280 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:15:17.229292 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:15:17.229303 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:15:17.229317 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:15:17.229329 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:15:17.229341 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:15:17.229353 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:15:17.229364 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.715909
I0429 00:15:17.229377 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.226415
I0429 00:15:17.229392 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.16331 (* 0.3 = 0.948992 loss)
I0429 00:15:17.229406 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.00216 (* 0.3 = 0.300649 loss)
I0429 00:15:17.229420 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.8648 (* 0.0272727 = 0.0781308 loss)
I0429 00:15:17.229434 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.30516 (* 0.0272727 = 0.0901409 loss)
I0429 00:15:17.229449 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.86752 (* 0.0272727 = 0.105478 loss)
I0429 00:15:17.229462 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.52496 (* 0.0272727 = 0.0961354 loss)
I0429 00:15:17.229475 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.8594 (* 0.0272727 = 0.0779835 loss)
I0429 00:15:17.229490 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.09878 (* 0.0272727 = 0.0572396 loss)
I0429 00:15:17.229503 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.7848 (* 0.0272727 = 0.0486764 loss)
I0429 00:15:17.229517 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.518911 (* 0.0272727 = 0.0141521 loss)
I0429 00:15:17.229532 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.348845 (* 0.0272727 = 0.00951395 loss)
I0429 00:15:17.229545 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.924215 (* 0.0272727 = 0.0252059 loss)
I0429 00:15:17.229559 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.882625 (* 0.0272727 = 0.0240716 loss)
I0429 00:15:17.229573 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.540195 (* 0.0272727 = 0.0147326 loss)
I0429 00:15:17.229609 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.698612 (* 0.0272727 = 0.019053 loss)
I0429 00:15:17.229624 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00277991 (* 0.0272727 = 7.58158e-05 loss)
I0429 00:15:17.229638 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00219218 (* 0.0272727 = 5.97868e-05 loss)
I0429 00:15:17.229652 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.000756188 (* 0.0272727 = 2.06233e-05 loss)
I0429 00:15:17.229666 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00031835 (* 0.0272727 = 8.68228e-06 loss)
I0429 00:15:17.229681 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000145119 (* 0.0272727 = 3.9578e-06 loss)
I0429 00:15:17.229694 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 5.20262e-05 (* 0.0272727 = 1.4189e-06 loss)
I0429 00:15:17.229708 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 3.16069e-05 (* 0.0272727 = 8.62005e-07 loss)
I0429 00:15:17.229722 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 2.53483e-05 (* 0.0272727 = 6.91319e-07 loss)
I0429 00:15:17.229737 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 1.46481e-05 (* 0.0272727 = 3.99494e-07 loss)
I0429 00:15:17.229748 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0943396
I0429 00:15:17.229760 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:15:17.229771 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:15:17.229784 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:15:17.229794 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0429 00:15:17.229806 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0429 00:15:17.229818 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0429 00:15:17.229830 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:15:17.229838 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 00:15:17.229846 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:15:17.229854 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:15:17.229866 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:15:17.229878 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:15:17.229889 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:15:17.229902 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:15:17.229912 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:15:17.229923 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:15:17.229935 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:15:17.229946 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:15:17.229957 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:15:17.229969 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:15:17.229980 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:15:17.229991 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:15:17.230003 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.721591
I0429 00:15:17.230015 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.169811
I0429 00:15:17.230028 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.30733 (* 0.3 = 0.992199 loss)
I0429 00:15:17.230042 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.05045 (* 0.3 = 0.315135 loss)
I0429 00:15:17.230056 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.72231 (* 0.0272727 = 0.0742448 loss)
I0429 00:15:17.230069 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.77583 (* 0.0272727 = 0.102977 loss)
I0429 00:15:17.230098 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.92952 (* 0.0272727 = 0.107169 loss)
I0429 00:15:17.230113 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.64899 (* 0.0272727 = 0.0995179 loss)
I0429 00:15:17.230126 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.81535 (* 0.0272727 = 0.0767822 loss)
I0429 00:15:17.230139 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.34444 (* 0.0272727 = 0.0639392 loss)
I0429 00:15:17.230154 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.61267 (* 0.0272727 = 0.0439818 loss)
I0429 00:15:17.230166 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.495911 (* 0.0272727 = 0.0135248 loss)
I0429 00:15:17.230180 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.440158 (* 0.0272727 = 0.0120043 loss)
I0429 00:15:17.230195 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.705826 (* 0.0272727 = 0.0192498 loss)
I0429 00:15:17.230208 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.862637 (* 0.0272727 = 0.0235265 loss)
I0429 00:15:17.230221 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.596735 (* 0.0272727 = 0.0162746 loss)
I0429 00:15:17.230235 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.660803 (* 0.0272727 = 0.0180219 loss)
I0429 00:15:17.230249 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0161854 (* 0.0272727 = 0.00044142 loss)
I0429 00:15:17.230263 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00897345 (* 0.0272727 = 0.000244731 loss)
I0429 00:15:17.230276 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00441654 (* 0.0272727 = 0.000120451 loss)
I0429 00:15:17.230290 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00176291 (* 0.0272727 = 4.80794e-05 loss)
I0429 00:15:17.230304 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00223156 (* 0.0272727 = 6.08609e-05 loss)
I0429 00:15:17.230319 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00104617 (* 0.0272727 = 2.85318e-05 loss)
I0429 00:15:17.230332 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00068877 (* 0.0272727 = 1.87846e-05 loss)
I0429 00:15:17.230346 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000394996 (* 0.0272727 = 1.07726e-05 loss)
I0429 00:15:17.230360 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000383515 (* 0.0272727 = 1.04595e-05 loss)
I0429 00:15:17.230375 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0377358
I0429 00:15:17.230387 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0429 00:15:17.230399 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:15:17.230411 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:15:17.230422 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:15:17.230433 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 00:15:17.230445 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:15:17.230456 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:15:17.230468 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:15:17.230479 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:15:17.230491 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:15:17.230502 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:15:17.230515 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:15:17.230525 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:15:17.230537 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:15:17.230548 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:15:17.230569 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:15:17.230582 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:15:17.230594 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:15:17.230605 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:15:17.230618 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:15:17.230628 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:15:17.230640 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:15:17.230651 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.698864
I0429 00:15:17.230664 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.226415
I0429 00:15:17.230676 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.25174 (* 1 = 3.25174 loss)
I0429 00:15:17.230690 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.06822 (* 1 = 1.06822 loss)
I0429 00:15:17.230705 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.65136 (* 0.0909091 = 0.241032 loss)
I0429 00:15:17.230718 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.88911 (* 0.0909091 = 0.353555 loss)
I0429 00:15:17.230731 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 4.00871 (* 0.0909091 = 0.364429 loss)
I0429 00:15:17.230746 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.54347 (* 0.0909091 = 0.322134 loss)
I0429 00:15:17.230758 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.77699 (* 0.0909091 = 0.252453 loss)
I0429 00:15:17.230772 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.2902 (* 0.0909091 = 0.2082 loss)
I0429 00:15:17.230787 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.69343 (* 0.0909091 = 0.153948 loss)
I0429 00:15:17.230799 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.564132 (* 0.0909091 = 0.0512847 loss)
I0429 00:15:17.230813 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.365373 (* 0.0909091 = 0.0332158 loss)
I0429 00:15:17.230828 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.632208 (* 0.0909091 = 0.0574735 loss)
I0429 00:15:17.230840 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.725761 (* 0.0909091 = 0.0659782 loss)
I0429 00:15:17.230854 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.515257 (* 0.0909091 = 0.0468416 loss)
I0429 00:15:17.230868 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.552009 (* 0.0909091 = 0.0501826 loss)
I0429 00:15:17.230882 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0156753 (* 0.0909091 = 0.00142503 loss)
I0429 00:15:17.230896 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.010693 (* 0.0909091 = 0.000972092 loss)
I0429 00:15:17.230911 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0066936 (* 0.0909091 = 0.000608509 loss)
I0429 00:15:17.230923 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00408307 (* 0.0909091 = 0.000371188 loss)
I0429 00:15:17.230937 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00180477 (* 0.0909091 = 0.00016407 loss)
I0429 00:15:17.230952 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00126726 (* 0.0909091 = 0.000115206 loss)
I0429 00:15:17.230965 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000599698 (* 0.0909091 = 5.4518e-05 loss)
I0429 00:15:17.230979 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000245178 (* 0.0909091 = 2.22889e-05 loss)
I0429 00:15:17.230993 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00012832 (* 0.0909091 = 1.16655e-05 loss)
I0429 00:15:17.231005 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:15:17.231016 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:15:17.231039 6470 solver.cpp:245] Train net output #149: total_confidence = 1.24263e-05
I0429 00:15:17.231050 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000101448
I0429 00:15:17.231063 6470 sgd_solver.cpp:106] Iteration 11500, lr = 0.01
I0429 00:17:33.820783 6470 solver.cpp:229] Iteration 12000, loss = 9.5611
I0429 00:17:33.820957 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0816327
I0429 00:17:33.820977 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 00:17:33.820991 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:17:33.821003 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:17:33.821015 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 00:17:33.821027 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:17:33.821039 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0429 00:17:33.821050 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:17:33.821063 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 00:17:33.821075 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:17:33.821086 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:17:33.821099 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:17:33.821110 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:17:33.821121 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:17:33.821133 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:17:33.821146 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:17:33.821157 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:17:33.821171 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:17:33.821182 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:17:33.821193 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:17:33.821205 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:17:33.821218 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:17:33.821228 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:17:33.821240 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.744318
I0429 00:17:33.821252 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.204082
I0429 00:17:33.821269 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.11868 (* 0.3 = 0.935605 loss)
I0429 00:17:33.821282 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.956516 (* 0.3 = 0.286955 loss)
I0429 00:17:33.821297 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.62578 (* 0.0272727 = 0.0716122 loss)
I0429 00:17:33.821313 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.02752 (* 0.0272727 = 0.0825688 loss)
I0429 00:17:33.821328 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.91046 (* 0.0272727 = 0.0793761 loss)
I0429 00:17:33.821342 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.16276 (* 0.0272727 = 0.0862571 loss)
I0429 00:17:33.821357 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.01536 (* 0.0272727 = 0.082237 loss)
I0429 00:17:33.821370 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.80066 (* 0.0272727 = 0.0763815 loss)
I0429 00:17:33.821383 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.59752 (* 0.0272727 = 0.0435687 loss)
I0429 00:17:33.821398 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.502523 (* 0.0272727 = 0.0137052 loss)
I0429 00:17:33.821411 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.163086 (* 0.0272727 = 0.0044478 loss)
I0429 00:17:33.821426 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0638541 (* 0.0272727 = 0.00174148 loss)
I0429 00:17:33.821440 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.035975 (* 0.0272727 = 0.000981136 loss)
I0429 00:17:33.821455 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0163235 (* 0.0272727 = 0.000445186 loss)
I0429 00:17:33.821470 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0113238 (* 0.0272727 = 0.00030883 loss)
I0429 00:17:33.821503 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00622052 (* 0.0272727 = 0.00016965 loss)
I0429 00:17:33.821519 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0066757 (* 0.0272727 = 0.000182065 loss)
I0429 00:17:33.821533 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00601437 (* 0.0272727 = 0.000164028 loss)
I0429 00:17:33.821547 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00427672 (* 0.0272727 = 0.000116638 loss)
I0429 00:17:33.821563 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00269806 (* 0.0272727 = 7.35835e-05 loss)
I0429 00:17:33.821576 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00187084 (* 0.0272727 = 5.10229e-05 loss)
I0429 00:17:33.821590 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00194339 (* 0.0272727 = 5.30014e-05 loss)
I0429 00:17:33.821604 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00120808 (* 0.0272727 = 3.29478e-05 loss)
I0429 00:17:33.821617 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00101199 (* 0.0272727 = 2.75998e-05 loss)
I0429 00:17:33.821630 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0612245
I0429 00:17:33.821641 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:17:33.821653 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:17:33.821666 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.25
I0429 00:17:33.821676 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:17:33.821688 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0429 00:17:33.821699 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.25
I0429 00:17:33.821712 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:17:33.821723 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 00:17:33.821734 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:17:33.821746 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:17:33.821758 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:17:33.821769 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:17:33.821781 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:17:33.821792 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:17:33.821804 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:17:33.821816 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:17:33.821825 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:17:33.821836 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:17:33.821848 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:17:33.821859 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:17:33.821871 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:17:33.821882 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:17:33.821894 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.738636
I0429 00:17:33.821905 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.265306
I0429 00:17:33.821919 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.15282 (* 0.3 = 0.945847 loss)
I0429 00:17:33.821933 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.965768 (* 0.3 = 0.28973 loss)
I0429 00:17:33.821950 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.69077 (* 0.0272727 = 0.0733847 loss)
I0429 00:17:33.821969 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.28713 (* 0.0272727 = 0.0896489 loss)
I0429 00:17:33.822003 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.21837 (* 0.0272727 = 0.0877738 loss)
I0429 00:17:33.822031 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.37152 (* 0.0272727 = 0.0919505 loss)
I0429 00:17:33.822054 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.98835 (* 0.0272727 = 0.0815005 loss)
I0429 00:17:33.822072 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.6471 (* 0.0272727 = 0.0721936 loss)
I0429 00:17:33.822088 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.58134 (* 0.0272727 = 0.0431273 loss)
I0429 00:17:33.822103 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.547125 (* 0.0272727 = 0.0149216 loss)
I0429 00:17:33.822116 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.0795185 (* 0.0272727 = 0.00216869 loss)
I0429 00:17:33.822130 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.032345 (* 0.0272727 = 0.000882137 loss)
I0429 00:17:33.822144 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0124414 (* 0.0272727 = 0.000339312 loss)
I0429 00:17:33.822159 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00611343 (* 0.0272727 = 0.00016673 loss)
I0429 00:17:33.822172 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00423296 (* 0.0272727 = 0.000115444 loss)
I0429 00:17:33.822186 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00282025 (* 0.0272727 = 7.6916e-05 loss)
I0429 00:17:33.822201 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0023425 (* 0.0272727 = 6.38863e-05 loss)
I0429 00:17:33.822213 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00124234 (* 0.0272727 = 3.38821e-05 loss)
I0429 00:17:33.822227 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00120241 (* 0.0272727 = 3.27931e-05 loss)
I0429 00:17:33.822242 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000814094 (* 0.0272727 = 2.22026e-05 loss)
I0429 00:17:33.822255 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000665899 (* 0.0272727 = 1.81609e-05 loss)
I0429 00:17:33.822268 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000236741 (* 0.0272727 = 6.45658e-06 loss)
I0429 00:17:33.822283 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000478517 (* 0.0272727 = 1.30505e-05 loss)
I0429 00:17:33.822296 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000132252 (* 0.0272727 = 3.60688e-06 loss)
I0429 00:17:33.822309 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0612245
I0429 00:17:33.822320 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0
I0429 00:17:33.822332 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:17:33.822343 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:17:33.822355 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0429 00:17:33.822370 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 00:17:33.822381 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.25
I0429 00:17:33.822393 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:17:33.822405 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:17:33.822417 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:17:33.822428 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:17:33.822439 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:17:33.822451 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:17:33.822463 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:17:33.822474 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:17:33.822485 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:17:33.822496 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:17:33.822520 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:17:33.822532 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:17:33.822543 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:17:33.822556 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:17:33.822567 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:17:33.822578 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:17:33.822589 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.732955
I0429 00:17:33.822602 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.265306
I0429 00:17:33.822615 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.11004 (* 1 = 3.11004 loss)
I0429 00:17:33.822629 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.969641 (* 1 = 0.969641 loss)
I0429 00:17:33.822643 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.67578 (* 0.0909091 = 0.243253 loss)
I0429 00:17:33.822657 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.30524 (* 0.0909091 = 0.300476 loss)
I0429 00:17:33.822670 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.03882 (* 0.0909091 = 0.276257 loss)
I0429 00:17:33.822684 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.64408 (* 0.0909091 = 0.33128 loss)
I0429 00:17:33.822697 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.0348 (* 0.0909091 = 0.275891 loss)
I0429 00:17:33.822710 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.63127 (* 0.0909091 = 0.239206 loss)
I0429 00:17:33.822724 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.60066 (* 0.0909091 = 0.145514 loss)
I0429 00:17:33.822737 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.499804 (* 0.0909091 = 0.0454368 loss)
I0429 00:17:33.822751 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.093521 (* 0.0909091 = 0.00850191 loss)
I0429 00:17:33.822767 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.0562256 (* 0.0909091 = 0.00511142 loss)
I0429 00:17:33.822780 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0243382 (* 0.0909091 = 0.00221257 loss)
I0429 00:17:33.822793 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0169319 (* 0.0909091 = 0.00153926 loss)
I0429 00:17:33.822808 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00960984 (* 0.0909091 = 0.000873622 loss)
I0429 00:17:33.822821 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00729316 (* 0.0909091 = 0.000663015 loss)
I0429 00:17:33.822834 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00470855 (* 0.0909091 = 0.00042805 loss)
I0429 00:17:33.822849 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00484821 (* 0.0909091 = 0.000440746 loss)
I0429 00:17:33.822861 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00371034 (* 0.0909091 = 0.000337303 loss)
I0429 00:17:33.822875 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00261905 (* 0.0909091 = 0.000238095 loss)
I0429 00:17:33.822890 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00247169 (* 0.0909091 = 0.000224699 loss)
I0429 00:17:33.822903 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00238425 (* 0.0909091 = 0.00021675 loss)
I0429 00:17:33.822917 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00188168 (* 0.0909091 = 0.000171062 loss)
I0429 00:17:33.822932 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00145922 (* 0.0909091 = 0.000132656 loss)
I0429 00:17:33.822943 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:17:33.822954 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:17:33.822965 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000148044
I0429 00:17:33.822986 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00108652
I0429 00:17:33.823000 6470 sgd_solver.cpp:106] Iteration 12000, lr = 0.01
I0429 00:19:50.393777 6470 solver.cpp:229] Iteration 12500, loss = 9.53223
I0429 00:19:50.393946 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.126984
I0429 00:19:50.393965 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:19:50.393980 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:19:50.393992 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:19:50.394004 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0429 00:19:50.394016 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:19:50.394028 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0429 00:19:50.394040 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.25
I0429 00:19:50.394052 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0429 00:19:50.394064 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:19:50.394075 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.75
I0429 00:19:50.394088 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:19:50.394099 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:19:50.394111 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:19:50.394122 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:19:50.394134 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:19:50.394146 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:19:50.394158 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:19:50.394170 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:19:50.394181 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:19:50.394192 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:19:50.394204 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:19:50.394217 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:19:50.394227 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.647727
I0429 00:19:50.394239 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.31746
I0429 00:19:50.394255 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.04366 (* 0.3 = 0.913097 loss)
I0429 00:19:50.394269 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.32776 (* 0.3 = 0.398328 loss)
I0429 00:19:50.394284 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.97389 (* 0.0272727 = 0.0811062 loss)
I0429 00:19:50.394299 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.79685 (* 0.0272727 = 0.0762777 loss)
I0429 00:19:50.394315 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.11602 (* 0.0272727 = 0.0849824 loss)
I0429 00:19:50.394330 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.86487 (* 0.0272727 = 0.0781327 loss)
I0429 00:19:50.394343 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.98417 (* 0.0272727 = 0.0813864 loss)
I0429 00:19:50.394357 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 3.18155 (* 0.0272727 = 0.0867694 loss)
I0429 00:19:50.394371 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 3.17965 (* 0.0272727 = 0.0867178 loss)
I0429 00:19:50.394384 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.89497 (* 0.0272727 = 0.051681 loss)
I0429 00:19:50.394398 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.871004 (* 0.0272727 = 0.0237546 loss)
I0429 00:19:50.394412 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.954793 (* 0.0272727 = 0.0260398 loss)
I0429 00:19:50.394426 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.587088 (* 0.0272727 = 0.0160115 loss)
I0429 00:19:50.394440 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.512486 (* 0.0272727 = 0.0139769 loss)
I0429 00:19:50.394454 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.631893 (* 0.0272727 = 0.0172334 loss)
I0429 00:19:50.394490 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.185675 (* 0.0272727 = 0.00506386 loss)
I0429 00:19:50.394505 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.110726 (* 0.0272727 = 0.00301979 loss)
I0429 00:19:50.394520 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0484012 (* 0.0272727 = 0.00132003 loss)
I0429 00:19:50.394533 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0277407 (* 0.0272727 = 0.000756563 loss)
I0429 00:19:50.394547 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0121129 (* 0.0272727 = 0.000330351 loss)
I0429 00:19:50.394562 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00487813 (* 0.0272727 = 0.00013304 loss)
I0429 00:19:50.394575 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00463021 (* 0.0272727 = 0.000126278 loss)
I0429 00:19:50.394589 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00544067 (* 0.0272727 = 0.000148382 loss)
I0429 00:19:50.394603 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00248664 (* 0.0272727 = 6.78174e-05 loss)
I0429 00:19:50.394615 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.111111
I0429 00:19:50.394628 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:19:50.394639 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.25
I0429 00:19:50.394650 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.25
I0429 00:19:50.394662 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:19:50.394673 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 00:19:50.394685 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0429 00:19:50.394696 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.25
I0429 00:19:50.394708 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 00:19:50.394721 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:19:50.394734 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.75
I0429 00:19:50.394742 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:19:50.394754 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:19:50.394767 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:19:50.394779 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:19:50.394791 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:19:50.394803 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:19:50.394814 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:19:50.394825 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:19:50.394836 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:19:50.394847 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:19:50.394860 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:19:50.394870 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:19:50.394881 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.653409
I0429 00:19:50.394893 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.301587
I0429 00:19:50.394907 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.91232 (* 0.3 = 0.873695 loss)
I0429 00:19:50.394920 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.26442 (* 0.3 = 0.379325 loss)
I0429 00:19:50.394934 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.81826 (* 0.0272727 = 0.0768617 loss)
I0429 00:19:50.394948 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.72377 (* 0.0272727 = 0.0742845 loss)
I0429 00:19:50.394978 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.96466 (* 0.0272727 = 0.0808543 loss)
I0429 00:19:50.394992 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.00514 (* 0.0272727 = 0.0819583 loss)
I0429 00:19:50.395006 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.74134 (* 0.0272727 = 0.0747637 loss)
I0429 00:19:50.395020 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 3.04941 (* 0.0272727 = 0.0831658 loss)
I0429 00:19:50.395033 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.67863 (* 0.0272727 = 0.0730535 loss)
I0429 00:19:50.395047 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.8392 (* 0.0272727 = 0.05016 loss)
I0429 00:19:50.395061 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.961935 (* 0.0272727 = 0.0262346 loss)
I0429 00:19:50.395074 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.756049 (* 0.0272727 = 0.0206195 loss)
I0429 00:19:50.395088 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.566237 (* 0.0272727 = 0.0154428 loss)
I0429 00:19:50.395102 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.619708 (* 0.0272727 = 0.0169011 loss)
I0429 00:19:50.395115 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.678561 (* 0.0272727 = 0.0185062 loss)
I0429 00:19:50.395129 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.111387 (* 0.0272727 = 0.00303784 loss)
I0429 00:19:50.395143 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0657593 (* 0.0272727 = 0.00179343 loss)
I0429 00:19:50.395158 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0356879 (* 0.0272727 = 0.000973305 loss)
I0429 00:19:50.395170 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00951297 (* 0.0272727 = 0.000259445 loss)
I0429 00:19:50.395184 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00424945 (* 0.0272727 = 0.000115894 loss)
I0429 00:19:50.395198 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0026413 (* 0.0272727 = 7.20354e-05 loss)
I0429 00:19:50.395212 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00190016 (* 0.0272727 = 5.18227e-05 loss)
I0429 00:19:50.395226 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0011208 (* 0.0272727 = 3.05672e-05 loss)
I0429 00:19:50.395241 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000823326 (* 0.0272727 = 2.24543e-05 loss)
I0429 00:19:50.395252 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.126984
I0429 00:19:50.395264 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:19:50.395277 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:19:50.395287 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:19:50.395299 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:19:50.395310 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:19:50.395323 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.125
I0429 00:19:50.395334 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.25
I0429 00:19:50.395345 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0429 00:19:50.395357 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:19:50.395371 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.75
I0429 00:19:50.395383 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:19:50.395395 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:19:50.395407 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:19:50.395418 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:19:50.395429 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:19:50.395442 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:19:50.395462 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:19:50.395491 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:19:50.395504 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:19:50.395515 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:19:50.395527 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:19:50.395539 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:19:50.395550 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.642045
I0429 00:19:50.395561 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.31746
I0429 00:19:50.395576 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.87925 (* 1 = 2.87925 loss)
I0429 00:19:50.395589 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.31058 (* 1 = 1.31058 loss)
I0429 00:19:50.395602 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.84493 (* 0.0909091 = 0.25863 loss)
I0429 00:19:50.395617 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.71619 (* 0.0909091 = 0.246926 loss)
I0429 00:19:50.395630 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.99859 (* 0.0909091 = 0.272599 loss)
I0429 00:19:50.395644 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.99509 (* 0.0909091 = 0.272281 loss)
I0429 00:19:50.395658 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.68031 (* 0.0909091 = 0.243665 loss)
I0429 00:19:50.395671 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.92913 (* 0.0909091 = 0.266285 loss)
I0429 00:19:50.395684 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.82373 (* 0.0909091 = 0.256703 loss)
I0429 00:19:50.395699 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.83144 (* 0.0909091 = 0.166495 loss)
I0429 00:19:50.395711 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.892952 (* 0.0909091 = 0.0811775 loss)
I0429 00:19:50.395725 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.840467 (* 0.0909091 = 0.0764061 loss)
I0429 00:19:50.395738 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.535281 (* 0.0909091 = 0.0486619 loss)
I0429 00:19:50.395752 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.464316 (* 0.0909091 = 0.0422105 loss)
I0429 00:19:50.395766 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.612722 (* 0.0909091 = 0.055702 loss)
I0429 00:19:50.395781 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.106663 (* 0.0909091 = 0.00969664 loss)
I0429 00:19:50.395793 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0591 (* 0.0909091 = 0.00537273 loss)
I0429 00:19:50.395807 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0456373 (* 0.0909091 = 0.00414884 loss)
I0429 00:19:50.395822 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0220933 (* 0.0909091 = 0.00200848 loss)
I0429 00:19:50.395834 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0150698 (* 0.0909091 = 0.00136998 loss)
I0429 00:19:50.395849 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0115087 (* 0.0909091 = 0.00104624 loss)
I0429 00:19:50.395862 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00934005 (* 0.0909091 = 0.000849095 loss)
I0429 00:19:50.395876 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00392069 (* 0.0909091 = 0.000356427 loss)
I0429 00:19:50.395890 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00175994 (* 0.0909091 = 0.000159995 loss)
I0429 00:19:50.395902 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:19:50.395913 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:19:50.395925 6470 solver.cpp:245] Train net output #149: total_confidence = 1.29026e-06
I0429 00:19:50.395948 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 7.58217e-06
I0429 00:19:50.395962 6470 sgd_solver.cpp:106] Iteration 12500, lr = 0.01
I0429 00:22:07.195240 6470 solver.cpp:229] Iteration 13000, loss = 9.52544
I0429 00:22:07.195390 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.05
I0429 00:22:07.195411 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:22:07.195425 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:22:07.195436 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.25
I0429 00:22:07.195448 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:22:07.195461 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:22:07.195472 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:22:07.195484 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 1
I0429 00:22:07.195497 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 00:22:07.195524 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:22:07.195539 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:22:07.195550 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:22:07.195562 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:22:07.195574 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:22:07.195586 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:22:07.195597 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:22:07.195610 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:22:07.195621 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:22:07.195632 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:22:07.195644 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:22:07.195657 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:22:07.195667 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:22:07.195679 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:22:07.195690 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.772727
I0429 00:22:07.195703 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.175
I0429 00:22:07.195719 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.95739 (* 0.3 = 0.887216 loss)
I0429 00:22:07.195734 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.777252 (* 0.3 = 0.233176 loss)
I0429 00:22:07.195747 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.99736 (* 0.0272727 = 0.0817463 loss)
I0429 00:22:07.195761 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.93173 (* 0.0272727 = 0.0799564 loss)
I0429 00:22:07.195775 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.98496 (* 0.0272727 = 0.081408 loss)
I0429 00:22:07.195788 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.14275 (* 0.0272727 = 0.0857113 loss)
I0429 00:22:07.195802 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.18284 (* 0.0272727 = 0.059532 loss)
I0429 00:22:07.195816 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.85699 (* 0.0272727 = 0.0506452 loss)
I0429 00:22:07.195830 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.296958 (* 0.0272727 = 0.00809886 loss)
I0429 00:22:07.195844 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.075183 (* 0.0272727 = 0.00205045 loss)
I0429 00:22:07.195858 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0146036 (* 0.0272727 = 0.00039828 loss)
I0429 00:22:07.195873 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0141408 (* 0.0272727 = 0.000385659 loss)
I0429 00:22:07.195886 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.015663 (* 0.0272727 = 0.000427173 loss)
I0429 00:22:07.195900 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.009203 (* 0.0272727 = 0.000250991 loss)
I0429 00:22:07.195914 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.00821661 (* 0.0272727 = 0.000224089 loss)
I0429 00:22:07.195950 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00716049 (* 0.0272727 = 0.000195286 loss)
I0429 00:22:07.195965 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00424561 (* 0.0272727 = 0.000115789 loss)
I0429 00:22:07.195978 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00596333 (* 0.0272727 = 0.000162636 loss)
I0429 00:22:07.195992 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00518645 (* 0.0272727 = 0.000141449 loss)
I0429 00:22:07.196007 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00519357 (* 0.0272727 = 0.000141643 loss)
I0429 00:22:07.196019 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00460426 (* 0.0272727 = 0.000125571 loss)
I0429 00:22:07.196033 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0038706 (* 0.0272727 = 0.000105562 loss)
I0429 00:22:07.196048 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.003954 (* 0.0272727 = 0.000107836 loss)
I0429 00:22:07.196063 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00245016 (* 0.0272727 = 6.68226e-05 loss)
I0429 00:22:07.196074 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.075
I0429 00:22:07.196086 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:22:07.196099 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:22:07.196110 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:22:07.196120 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:22:07.196132 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:22:07.196144 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:22:07.196156 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 1
I0429 00:22:07.196168 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 00:22:07.196179 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:22:07.196190 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:22:07.196202 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:22:07.196213 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:22:07.196225 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:22:07.196236 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:22:07.196247 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:22:07.196259 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:22:07.196270 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:22:07.196281 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:22:07.196293 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:22:07.196305 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:22:07.196318 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:22:07.196331 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:22:07.196342 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.778409
I0429 00:22:07.196354 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.275
I0429 00:22:07.196367 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.95272 (* 0.3 = 0.885817 loss)
I0429 00:22:07.196382 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.782839 (* 0.3 = 0.234852 loss)
I0429 00:22:07.196395 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.89594 (* 0.0272727 = 0.0789801 loss)
I0429 00:22:07.196413 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.01292 (* 0.0272727 = 0.0821706 loss)
I0429 00:22:07.196439 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.48029 (* 0.0272727 = 0.094917 loss)
I0429 00:22:07.196454 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.04862 (* 0.0272727 = 0.0831442 loss)
I0429 00:22:07.196467 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.20156 (* 0.0272727 = 0.0600427 loss)
I0429 00:22:07.196480 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.93676 (* 0.0272727 = 0.0528206 loss)
I0429 00:22:07.196494 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.260277 (* 0.0272727 = 0.00709847 loss)
I0429 00:22:07.196508 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.0323848 (* 0.0272727 = 0.000883221 loss)
I0429 00:22:07.196522 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.00505257 (* 0.0272727 = 0.000137797 loss)
I0429 00:22:07.196537 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.00185387 (* 0.0272727 = 5.05602e-05 loss)
I0429 00:22:07.196550 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.00118423 (* 0.0272727 = 3.22972e-05 loss)
I0429 00:22:07.196564 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00147097 (* 0.0272727 = 4.01174e-05 loss)
I0429 00:22:07.196578 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.000624517 (* 0.0272727 = 1.70323e-05 loss)
I0429 00:22:07.196593 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.000403691 (* 0.0272727 = 1.10098e-05 loss)
I0429 00:22:07.196606 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.000192828 (* 0.0272727 = 5.25894e-06 loss)
I0429 00:22:07.196620 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.000238209 (* 0.0272727 = 6.49662e-06 loss)
I0429 00:22:07.196635 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.000223786 (* 0.0272727 = 6.10325e-06 loss)
I0429 00:22:07.196648 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000234055 (* 0.0272727 = 6.38331e-06 loss)
I0429 00:22:07.196662 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00013699 (* 0.0272727 = 3.73608e-06 loss)
I0429 00:22:07.196676 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000208736 (* 0.0272727 = 5.6928e-06 loss)
I0429 00:22:07.196691 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000200755 (* 0.0272727 = 5.47514e-06 loss)
I0429 00:22:07.196704 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 8.59494e-05 (* 0.0272727 = 2.34408e-06 loss)
I0429 00:22:07.196717 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.125
I0429 00:22:07.196728 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:22:07.196739 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:22:07.196751 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.25
I0429 00:22:07.196763 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:22:07.196774 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:22:07.196786 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:22:07.196799 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 1
I0429 00:22:07.196810 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0429 00:22:07.196821 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:22:07.196832 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:22:07.196844 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:22:07.196856 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:22:07.196866 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:22:07.196878 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:22:07.196889 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:22:07.196902 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:22:07.196923 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:22:07.196935 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:22:07.196948 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:22:07.196959 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:22:07.196970 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:22:07.196981 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:22:07.196993 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.795455
I0429 00:22:07.197005 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.175
I0429 00:22:07.197018 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.06391 (* 1 = 3.06391 loss)
I0429 00:22:07.197032 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.786865 (* 1 = 0.786865 loss)
I0429 00:22:07.197046 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.84248 (* 0.0909091 = 0.258407 loss)
I0429 00:22:07.197060 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.13743 (* 0.0909091 = 0.285221 loss)
I0429 00:22:07.197074 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.01068 (* 0.0909091 = 0.273698 loss)
I0429 00:22:07.197088 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.93938 (* 0.0909091 = 0.267217 loss)
I0429 00:22:07.197101 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.30406 (* 0.0909091 = 0.20946 loss)
I0429 00:22:07.197115 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.91755 (* 0.0909091 = 0.174323 loss)
I0429 00:22:07.197129 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.198609 (* 0.0909091 = 0.0180554 loss)
I0429 00:22:07.197142 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.00555029 (* 0.0909091 = 0.000504572 loss)
I0429 00:22:07.197156 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.000166003 (* 0.0909091 = 1.50912e-05 loss)
I0429 00:22:07.197170 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.000129869 (* 0.0909091 = 1.18063e-05 loss)
I0429 00:22:07.197185 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 8.52853e-05 (* 0.0909091 = 7.7532e-06 loss)
I0429 00:22:07.197198 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 5.10942e-05 (* 0.0909091 = 4.64493e-06 loss)
I0429 00:22:07.197212 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 3.55791e-05 (* 0.0909091 = 3.23446e-06 loss)
I0429 00:22:07.197227 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 2.52141e-05 (* 0.0909091 = 2.29219e-06 loss)
I0429 00:22:07.197239 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 2.02367e-05 (* 0.0909091 = 1.8397e-06 loss)
I0429 00:22:07.197253 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 2.17718e-05 (* 0.0909091 = 1.97925e-06 loss)
I0429 00:22:07.197268 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 2.26365e-05 (* 0.0909091 = 2.05786e-06 loss)
I0429 00:22:07.197280 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 2.286e-05 (* 0.0909091 = 2.07819e-06 loss)
I0429 00:22:07.197294 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 3.16309e-05 (* 0.0909091 = 2.87553e-06 loss)
I0429 00:22:07.197309 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 3.24952e-05 (* 0.0909091 = 2.95411e-06 loss)
I0429 00:22:07.197321 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 3.30615e-05 (* 0.0909091 = 3.00559e-06 loss)
I0429 00:22:07.197335 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 2.13248e-05 (* 0.0909091 = 1.93862e-06 loss)
I0429 00:22:07.197347 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:22:07.197360 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:22:07.197377 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00075519
I0429 00:22:07.197399 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00146274
I0429 00:22:07.197413 6470 sgd_solver.cpp:106] Iteration 13000, lr = 0.01
I0429 00:24:23.839275 6470 solver.cpp:229] Iteration 13500, loss = 9.46115
I0429 00:24:23.839452 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0972222
I0429 00:24:23.839476 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:24:23.839491 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:24:23.839504 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:24:23.839516 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:24:23.839529 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:24:23.839541 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:24:23.839553 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0429 00:24:23.839584 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.375
I0429 00:24:23.839599 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.625
I0429 00:24:23.839612 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.5
I0429 00:24:23.839623 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.5
I0429 00:24:23.839635 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.625
I0429 00:24:23.839648 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:24:23.839660 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:24:23.839673 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:24:23.839684 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0429 00:24:23.839696 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:24:23.839709 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:24:23.839720 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:24:23.839732 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:24:23.839743 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:24:23.839756 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:24:23.839767 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.596591
I0429 00:24:23.839779 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.25
I0429 00:24:23.839795 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.93215 (* 0.3 = 0.879644 loss)
I0429 00:24:23.839810 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.42908 (* 0.3 = 0.428725 loss)
I0429 00:24:23.839824 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.24769 (* 0.0272727 = 0.0885734 loss)
I0429 00:24:23.839838 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.10878 (* 0.0272727 = 0.084785 loss)
I0429 00:24:23.839853 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.11033 (* 0.0272727 = 0.0848272 loss)
I0429 00:24:23.839866 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.83821 (* 0.0272727 = 0.0774058 loss)
I0429 00:24:23.839880 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.4273 (* 0.0272727 = 0.066199 loss)
I0429 00:24:23.839895 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.17031 (* 0.0272727 = 0.0591904 loss)
I0429 00:24:23.839908 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.87492 (* 0.0272727 = 0.0511342 loss)
I0429 00:24:23.839921 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.44299 (* 0.0272727 = 0.066627 loss)
I0429 00:24:23.839936 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.27466 (* 0.0272727 = 0.0347634 loss)
I0429 00:24:23.839949 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 1.51683 (* 0.0272727 = 0.041368 loss)
I0429 00:24:23.839963 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 1.57943 (* 0.0272727 = 0.0430754 loss)
I0429 00:24:23.839977 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 1.49116 (* 0.0272727 = 0.040668 loss)
I0429 00:24:23.839990 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.427285 (* 0.0272727 = 0.0116532 loss)
I0429 00:24:23.840032 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.401501 (* 0.0272727 = 0.01095 loss)
I0429 00:24:23.840047 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.455994 (* 0.0272727 = 0.0124362 loss)
I0429 00:24:23.840062 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.57226 (* 0.0272727 = 0.0156071 loss)
I0429 00:24:23.840076 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0461344 (* 0.0272727 = 0.00125821 loss)
I0429 00:24:23.840091 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0335456 (* 0.0272727 = 0.000914881 loss)
I0429 00:24:23.840106 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0246069 (* 0.0272727 = 0.000671096 loss)
I0429 00:24:23.840119 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0144191 (* 0.0272727 = 0.000393248 loss)
I0429 00:24:23.840132 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00520877 (* 0.0272727 = 0.000142057 loss)
I0429 00:24:23.840147 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00688718 (* 0.0272727 = 0.000187832 loss)
I0429 00:24:23.840158 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.111111
I0429 00:24:23.840170 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:24:23.840183 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:24:23.840195 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:24:23.840206 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:24:23.840219 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 00:24:23.840230 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:24:23.840242 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.375
I0429 00:24:23.840255 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.375
I0429 00:24:23.840266 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.625
I0429 00:24:23.840278 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.5
I0429 00:24:23.840291 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.5
I0429 00:24:23.840302 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.625
I0429 00:24:23.840317 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:24:23.840328 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:24:23.840340 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:24:23.840353 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0429 00:24:23.840364 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:24:23.840375 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:24:23.840387 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:24:23.840399 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:24:23.840409 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:24:23.840421 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:24:23.840432 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.630682
I0429 00:24:23.840445 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.305556
I0429 00:24:23.840462 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.92356 (* 0.3 = 0.877067 loss)
I0429 00:24:23.840477 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.33651 (* 0.3 = 0.400953 loss)
I0429 00:24:23.840492 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.58348 (* 0.0272727 = 0.0977313 loss)
I0429 00:24:23.840505 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.26403 (* 0.0272727 = 0.0890191 loss)
I0429 00:24:23.840530 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.01982 (* 0.0272727 = 0.0823587 loss)
I0429 00:24:23.840545 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.93751 (* 0.0272727 = 0.080114 loss)
I0429 00:24:23.840559 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.29136 (* 0.0272727 = 0.0624916 loss)
I0429 00:24:23.840572 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.05857 (* 0.0272727 = 0.0561429 loss)
I0429 00:24:23.840586 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.81991 (* 0.0272727 = 0.049634 loss)
I0429 00:24:23.840600 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.41821 (* 0.0272727 = 0.0659511 loss)
I0429 00:24:23.840613 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.38371 (* 0.0272727 = 0.0377375 loss)
I0429 00:24:23.840627 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 1.43967 (* 0.0272727 = 0.0392636 loss)
I0429 00:24:23.840641 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 1.77369 (* 0.0272727 = 0.0483733 loss)
I0429 00:24:23.840656 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 1.86075 (* 0.0272727 = 0.0507478 loss)
I0429 00:24:23.840668 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.328142 (* 0.0272727 = 0.00894932 loss)
I0429 00:24:23.840682 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.53595 (* 0.0272727 = 0.0146168 loss)
I0429 00:24:23.840697 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.417708 (* 0.0272727 = 0.011392 loss)
I0429 00:24:23.840709 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.518266 (* 0.0272727 = 0.0141345 loss)
I0429 00:24:23.840724 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0171169 (* 0.0272727 = 0.000466825 loss)
I0429 00:24:23.840739 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00567791 (* 0.0272727 = 0.000154852 loss)
I0429 00:24:23.840752 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00312127 (* 0.0272727 = 8.51255e-05 loss)
I0429 00:24:23.840766 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00212105 (* 0.0272727 = 5.78467e-05 loss)
I0429 00:24:23.840780 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000485239 (* 0.0272727 = 1.32338e-05 loss)
I0429 00:24:23.840795 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000395204 (* 0.0272727 = 1.07783e-05 loss)
I0429 00:24:23.840806 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.152778
I0429 00:24:23.840818 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:24:23.840831 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:24:23.840842 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:24:23.840854 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:24:23.840865 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0429 00:24:23.840878 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:24:23.840889 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0429 00:24:23.840900 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.375
I0429 00:24:23.840912 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:24:23.840924 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.5
I0429 00:24:23.840935 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.5
I0429 00:24:23.840947 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.625
I0429 00:24:23.840960 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:24:23.840971 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:24:23.840982 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:24:23.840994 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0429 00:24:23.841014 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:24:23.841027 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:24:23.841039 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:24:23.841051 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:24:23.841063 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:24:23.841074 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:24:23.841086 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.642045
I0429 00:24:23.841099 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.305556
I0429 00:24:23.841112 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.87774 (* 1 = 2.87774 loss)
I0429 00:24:23.841126 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.29071 (* 1 = 1.29071 loss)
I0429 00:24:23.841140 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.2702 (* 0.0909091 = 0.297291 loss)
I0429 00:24:23.841155 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.06434 (* 0.0909091 = 0.278576 loss)
I0429 00:24:23.841168 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.85329 (* 0.0909091 = 0.25939 loss)
I0429 00:24:23.841182 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.86355 (* 0.0909091 = 0.260323 loss)
I0429 00:24:23.841195 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.33898 (* 0.0909091 = 0.212634 loss)
I0429 00:24:23.841209 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.85964 (* 0.0909091 = 0.169058 loss)
I0429 00:24:23.841223 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.82867 (* 0.0909091 = 0.166242 loss)
I0429 00:24:23.841236 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 2.07298 (* 0.0909091 = 0.188452 loss)
I0429 00:24:23.841250 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.17929 (* 0.0909091 = 0.107208 loss)
I0429 00:24:23.841264 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 1.40635 (* 0.0909091 = 0.12785 loss)
I0429 00:24:23.841279 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 1.44036 (* 0.0909091 = 0.130942 loss)
I0429 00:24:23.841291 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 1.64493 (* 0.0909091 = 0.149539 loss)
I0429 00:24:23.841305 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.406526 (* 0.0909091 = 0.0369569 loss)
I0429 00:24:23.841320 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.378304 (* 0.0909091 = 0.0343913 loss)
I0429 00:24:23.841332 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.36947 (* 0.0909091 = 0.0335882 loss)
I0429 00:24:23.841346 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.446705 (* 0.0909091 = 0.0406095 loss)
I0429 00:24:23.841361 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0296097 (* 0.0909091 = 0.00269179 loss)
I0429 00:24:23.841378 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0164969 (* 0.0909091 = 0.00149972 loss)
I0429 00:24:23.841392 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00765093 (* 0.0909091 = 0.000695539 loss)
I0429 00:24:23.841408 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00289661 (* 0.0909091 = 0.000263328 loss)
I0429 00:24:23.841421 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000747983 (* 0.0909091 = 6.79985e-05 loss)
I0429 00:24:23.841436 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000137039 (* 0.0909091 = 1.24581e-05 loss)
I0429 00:24:23.841449 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:24:23.841459 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:24:23.841471 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000178306
I0429 00:24:23.841492 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00126531
I0429 00:24:23.841511 6470 sgd_solver.cpp:106] Iteration 13500, lr = 0.01
I0429 00:26:40.349478 6470 solver.cpp:229] Iteration 14000, loss = 9.48927
I0429 00:26:40.349607 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.131579
I0429 00:26:40.349627 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0
I0429 00:26:40.349640 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:26:40.349653 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:26:40.349664 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:26:40.349676 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.625
I0429 00:26:40.349689 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.75
I0429 00:26:40.349701 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 00:26:40.349714 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 00:26:40.349725 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:26:40.349736 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:26:40.349748 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:26:40.349759 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:26:40.349771 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:26:40.349782 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:26:40.349794 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:26:40.349807 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:26:40.349817 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:26:40.349828 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:26:40.349840 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:26:40.349851 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:26:40.349864 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:26:40.349875 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:26:40.349886 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.795455
I0429 00:26:40.349898 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.236842
I0429 00:26:40.349913 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.08252 (* 0.3 = 0.924756 loss)
I0429 00:26:40.349928 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.795786 (* 0.3 = 0.238736 loss)
I0429 00:26:40.349942 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.86459 (* 0.0272727 = 0.0781251 loss)
I0429 00:26:40.349956 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.12185 (* 0.0272727 = 0.0851415 loss)
I0429 00:26:40.349969 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.96611 (* 0.0272727 = 0.0808939 loss)
I0429 00:26:40.349983 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.86049 (* 0.0272727 = 0.0780133 loss)
I0429 00:26:40.349997 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 1.36196 (* 0.0272727 = 0.0371443 loss)
I0429 00:26:40.350010 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 0.97576 (* 0.0272727 = 0.0266116 loss)
I0429 00:26:40.350024 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.02413 (* 0.0272727 = 0.0279308 loss)
I0429 00:26:40.350039 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.692851 (* 0.0272727 = 0.0188959 loss)
I0429 00:26:40.350054 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0171963 (* 0.0272727 = 0.000468989 loss)
I0429 00:26:40.350067 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0180631 (* 0.0272727 = 0.000492629 loss)
I0429 00:26:40.350081 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0120147 (* 0.0272727 = 0.000327674 loss)
I0429 00:26:40.350095 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.01038 (* 0.0272727 = 0.000283092 loss)
I0429 00:26:40.350108 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0101418 (* 0.0272727 = 0.000276595 loss)
I0429 00:26:40.350142 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00874567 (* 0.0272727 = 0.000238518 loss)
I0429 00:26:40.350157 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0111122 (* 0.0272727 = 0.00030306 loss)
I0429 00:26:40.350172 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00775868 (* 0.0272727 = 0.000211601 loss)
I0429 00:26:40.350185 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0110338 (* 0.0272727 = 0.000300922 loss)
I0429 00:26:40.350199 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0081213 (* 0.0272727 = 0.00022149 loss)
I0429 00:26:40.350214 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00862361 (* 0.0272727 = 0.000235189 loss)
I0429 00:26:40.350227 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00681895 (* 0.0272727 = 0.000185971 loss)
I0429 00:26:40.350240 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00811531 (* 0.0272727 = 0.000221327 loss)
I0429 00:26:40.350255 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00516496 (* 0.0272727 = 0.000140862 loss)
I0429 00:26:40.350266 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.131579
I0429 00:26:40.350278 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:26:40.350291 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:26:40.350302 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:26:40.350317 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:26:40.350329 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.625
I0429 00:26:40.350342 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.75
I0429 00:26:40.350353 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 00:26:40.350365 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 00:26:40.350378 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:26:40.350389 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:26:40.350400 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:26:40.350411 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:26:40.350422 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:26:40.350435 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:26:40.350445 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:26:40.350456 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:26:40.350468 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:26:40.350479 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:26:40.350492 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:26:40.350502 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:26:40.350514 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:26:40.350525 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:26:40.350536 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.801136
I0429 00:26:40.350548 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.315789
I0429 00:26:40.350563 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.90556 (* 0.3 = 0.871669 loss)
I0429 00:26:40.350575 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.733746 (* 0.3 = 0.220124 loss)
I0429 00:26:40.350589 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.53128 (* 0.0272727 = 0.0690349 loss)
I0429 00:26:40.350603 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.03554 (* 0.0272727 = 0.0827876 loss)
I0429 00:26:40.350630 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.85887 (* 0.0272727 = 0.0779693 loss)
I0429 00:26:40.350646 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.06293 (* 0.0272727 = 0.0835346 loss)
I0429 00:26:40.350659 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 1.19916 (* 0.0272727 = 0.0327044 loss)
I0429 00:26:40.350673 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 0.99402 (* 0.0272727 = 0.0271096 loss)
I0429 00:26:40.350687 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.851723 (* 0.0272727 = 0.0232288 loss)
I0429 00:26:40.350700 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.55729 (* 0.0272727 = 0.0151988 loss)
I0429 00:26:40.350713 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.0203519 (* 0.0272727 = 0.000555052 loss)
I0429 00:26:40.350728 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.0249662 (* 0.0272727 = 0.000680896 loss)
I0429 00:26:40.350741 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0191468 (* 0.0272727 = 0.000522186 loss)
I0429 00:26:40.350755 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0145743 (* 0.0272727 = 0.000397482 loss)
I0429 00:26:40.350769 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0141986 (* 0.0272727 = 0.000387235 loss)
I0429 00:26:40.350782 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0139175 (* 0.0272727 = 0.000379569 loss)
I0429 00:26:40.350796 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0111376 (* 0.0272727 = 0.000303753 loss)
I0429 00:26:40.350811 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0157485 (* 0.0272727 = 0.000429505 loss)
I0429 00:26:40.350823 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.015763 (* 0.0272727 = 0.000429901 loss)
I0429 00:26:40.350837 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0134476 (* 0.0272727 = 0.000366754 loss)
I0429 00:26:40.350852 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0164275 (* 0.0272727 = 0.000448024 loss)
I0429 00:26:40.350865 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.0169429 (* 0.0272727 = 0.00046208 loss)
I0429 00:26:40.350878 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.0123783 (* 0.0272727 = 0.000337589 loss)
I0429 00:26:40.350893 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.0183692 (* 0.0272727 = 0.000500979 loss)
I0429 00:26:40.350904 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0789474
I0429 00:26:40.350916 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0429 00:26:40.350929 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:26:40.350940 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:26:40.350951 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 00:26:40.350963 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.625
I0429 00:26:40.350975 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.875
I0429 00:26:40.350986 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 00:26:40.350998 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:26:40.351011 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:26:40.351022 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:26:40.351030 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:26:40.351037 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:26:40.351053 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:26:40.351065 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:26:40.351078 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:26:40.351089 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:26:40.351109 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:26:40.351122 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:26:40.351133 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:26:40.351145 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:26:40.351156 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:26:40.351167 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:26:40.351179 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.789773
I0429 00:26:40.351191 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.315789
I0429 00:26:40.351204 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.01207 (* 1 = 3.01207 loss)
I0429 00:26:40.351218 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.73057 (* 1 = 0.73057 loss)
I0429 00:26:40.351232 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.39359 (* 0.0909091 = 0.217599 loss)
I0429 00:26:40.351245 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.11064 (* 0.0909091 = 0.282786 loss)
I0429 00:26:40.351258 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.65569 (* 0.0909091 = 0.241426 loss)
I0429 00:26:40.351272 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.62864 (* 0.0909091 = 0.238967 loss)
I0429 00:26:40.351285 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.25698 (* 0.0909091 = 0.114271 loss)
I0429 00:26:40.351299 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.01536 (* 0.0909091 = 0.0923057 loss)
I0429 00:26:40.351312 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.01947 (* 0.0909091 = 0.0926789 loss)
I0429 00:26:40.351325 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.434143 (* 0.0909091 = 0.0394676 loss)
I0429 00:26:40.351339 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.00491455 (* 0.0909091 = 0.000446777 loss)
I0429 00:26:40.351353 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.00382693 (* 0.0909091 = 0.000347903 loss)
I0429 00:26:40.351369 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00325133 (* 0.0909091 = 0.000295575 loss)
I0429 00:26:40.351383 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00229361 (* 0.0909091 = 0.00020851 loss)
I0429 00:26:40.351397 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00149329 (* 0.0909091 = 0.000135754 loss)
I0429 00:26:40.351410 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00114677 (* 0.0909091 = 0.000104252 loss)
I0429 00:26:40.351424 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000941993 (* 0.0909091 = 8.56357e-05 loss)
I0429 00:26:40.351438 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000871213 (* 0.0909091 = 7.92012e-05 loss)
I0429 00:26:40.351451 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000956663 (* 0.0909091 = 8.69693e-05 loss)
I0429 00:26:40.351477 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000944733 (* 0.0909091 = 8.58848e-05 loss)
I0429 00:26:40.351495 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00105357 (* 0.0909091 = 9.57792e-05 loss)
I0429 00:26:40.351510 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00115393 (* 0.0909091 = 0.000104903 loss)
I0429 00:26:40.351523 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00136487 (* 0.0909091 = 0.000124079 loss)
I0429 00:26:40.351537 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00153034 (* 0.0909091 = 0.000139122 loss)
I0429 00:26:40.351549 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:26:40.351560 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:26:40.351583 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000460266
I0429 00:26:40.351595 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000523955
I0429 00:26:40.351608 6470 sgd_solver.cpp:106] Iteration 14000, lr = 0.01
I0429 00:28:56.947053 6470 solver.cpp:229] Iteration 14500, loss = 9.43688
I0429 00:28:56.947216 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0588235
I0429 00:28:56.947237 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:28:56.947250 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:28:56.947263 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:28:56.947274 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:28:56.947286 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:28:56.947299 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:28:56.947310 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0429 00:28:56.947325 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:28:56.947337 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:28:56.947350 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:28:56.947361 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:28:56.947373 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:28:56.947386 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:28:56.947397 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:28:56.947409 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:28:56.947422 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:28:56.947432 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:28:56.947444 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:28:56.947456 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:28:56.947484 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:28:56.947499 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:28:56.947511 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:28:56.947522 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.727273
I0429 00:28:56.947535 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.176471
I0429 00:28:56.947551 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.03932 (* 0.3 = 0.911796 loss)
I0429 00:28:56.947564 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.922619 (* 0.3 = 0.276786 loss)
I0429 00:28:56.947579 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.58173 (* 0.0272727 = 0.070411 loss)
I0429 00:28:56.947593 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.04873 (* 0.0272727 = 0.0831471 loss)
I0429 00:28:56.947607 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.08629 (* 0.0272727 = 0.0841714 loss)
I0429 00:28:56.947621 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.0012 (* 0.0272727 = 0.0818509 loss)
I0429 00:28:56.947634 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.35588 (* 0.0272727 = 0.0642513 loss)
I0429 00:28:56.947649 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.21094 (* 0.0272727 = 0.0602983 loss)
I0429 00:28:56.947661 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.98063 (* 0.0272727 = 0.0540171 loss)
I0429 00:28:56.947675 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.901682 (* 0.0272727 = 0.0245913 loss)
I0429 00:28:56.947690 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.42206 (* 0.0272727 = 0.0115107 loss)
I0429 00:28:56.947702 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.509003 (* 0.0272727 = 0.0138819 loss)
I0429 00:28:56.947716 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.433793 (* 0.0272727 = 0.0118307 loss)
I0429 00:28:56.947731 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.200515 (* 0.0272727 = 0.00546859 loss)
I0429 00:28:56.947744 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.147243 (* 0.0272727 = 0.00401571 loss)
I0429 00:28:56.947780 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.124467 (* 0.0272727 = 0.00339457 loss)
I0429 00:28:56.947796 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0951499 (* 0.0272727 = 0.002595 loss)
I0429 00:28:56.947810 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0585895 (* 0.0272727 = 0.00159789 loss)
I0429 00:28:56.947824 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0383429 (* 0.0272727 = 0.00104572 loss)
I0429 00:28:56.947839 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0248828 (* 0.0272727 = 0.000678621 loss)
I0429 00:28:56.947852 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0158762 (* 0.0272727 = 0.000432986 loss)
I0429 00:28:56.947866 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.0120338 (* 0.0272727 = 0.000328194 loss)
I0429 00:28:56.947880 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00710777 (* 0.0272727 = 0.000193848 loss)
I0429 00:28:56.947895 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.0045231 (* 0.0272727 = 0.000123357 loss)
I0429 00:28:56.947906 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0784314
I0429 00:28:56.947918 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0429 00:28:56.947931 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:28:56.947942 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:28:56.947953 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:28:56.947965 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:28:56.947978 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 00:28:56.947985 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0429 00:28:56.947999 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:28:56.948009 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:28:56.948021 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:28:56.948034 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:28:56.948045 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:28:56.948056 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:28:56.948067 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:28:56.948079 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:28:56.948091 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:28:56.948101 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:28:56.948113 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:28:56.948124 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:28:56.948137 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:28:56.948148 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:28:56.948158 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:28:56.948170 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.732955
I0429 00:28:56.948182 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.156863
I0429 00:28:56.948196 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.06745 (* 0.3 = 0.920234 loss)
I0429 00:28:56.948210 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.950092 (* 0.3 = 0.285028 loss)
I0429 00:28:56.948227 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.30676 (* 0.0272727 = 0.0629116 loss)
I0429 00:28:56.948242 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.10926 (* 0.0272727 = 0.084798 loss)
I0429 00:28:56.948267 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.05526 (* 0.0272727 = 0.0833254 loss)
I0429 00:28:56.948282 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.62129 (* 0.0272727 = 0.0714898 loss)
I0429 00:28:56.948297 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.34716 (* 0.0272727 = 0.0640135 loss)
I0429 00:28:56.948310 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.17833 (* 0.0272727 = 0.0594089 loss)
I0429 00:28:56.948323 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.95501 (* 0.0272727 = 0.0533186 loss)
I0429 00:28:56.948338 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.991264 (* 0.0272727 = 0.0270345 loss)
I0429 00:28:56.948351 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.335171 (* 0.0272727 = 0.00914103 loss)
I0429 00:28:56.948369 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.523089 (* 0.0272727 = 0.0142661 loss)
I0429 00:28:56.948382 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.470095 (* 0.0272727 = 0.0128208 loss)
I0429 00:28:56.948397 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.157441 (* 0.0272727 = 0.00429384 loss)
I0429 00:28:56.948411 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.135828 (* 0.0272727 = 0.00370439 loss)
I0429 00:28:56.948424 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.116012 (* 0.0272727 = 0.00316397 loss)
I0429 00:28:56.948438 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0717851 (* 0.0272727 = 0.00195778 loss)
I0429 00:28:56.948452 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0384765 (* 0.0272727 = 0.00104936 loss)
I0429 00:28:56.948467 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.036487 (* 0.0272727 = 0.000995101 loss)
I0429 00:28:56.948479 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0238385 (* 0.0272727 = 0.000650142 loss)
I0429 00:28:56.948493 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0105822 (* 0.0272727 = 0.000288606 loss)
I0429 00:28:56.948506 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00822407 (* 0.0272727 = 0.000224293 loss)
I0429 00:28:56.948520 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00236153 (* 0.0272727 = 6.44055e-05 loss)
I0429 00:28:56.948534 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00126143 (* 0.0272727 = 3.44026e-05 loss)
I0429 00:28:56.948546 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0588235
I0429 00:28:56.948559 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:28:56.948570 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:28:56.948581 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:28:56.948593 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.375
I0429 00:28:56.948606 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 00:28:56.948617 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:28:56.948628 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.5
I0429 00:28:56.948640 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:28:56.948652 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:28:56.948663 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:28:56.948675 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:28:56.948686 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:28:56.948698 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:28:56.948709 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:28:56.948721 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:28:56.948732 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:28:56.948753 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:28:56.948766 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:28:56.948778 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:28:56.948789 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:28:56.948801 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:28:56.948812 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:28:56.948824 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.727273
I0429 00:28:56.948837 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.313726
I0429 00:28:56.948850 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.92157 (* 1 = 2.92157 loss)
I0429 00:28:56.948863 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.897426 (* 1 = 0.897426 loss)
I0429 00:28:56.948878 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.20568 (* 0.0909091 = 0.200516 loss)
I0429 00:28:56.948891 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.87134 (* 0.0909091 = 0.261031 loss)
I0429 00:28:56.948904 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.9367 (* 0.0909091 = 0.266973 loss)
I0429 00:28:56.948918 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.38405 (* 0.0909091 = 0.216732 loss)
I0429 00:28:56.948932 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.17522 (* 0.0909091 = 0.197747 loss)
I0429 00:28:56.948945 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.04824 (* 0.0909091 = 0.186204 loss)
I0429 00:28:56.948959 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.7922 (* 0.0909091 = 0.162927 loss)
I0429 00:28:56.948972 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.869422 (* 0.0909091 = 0.0790384 loss)
I0429 00:28:56.948987 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.335995 (* 0.0909091 = 0.030545 loss)
I0429 00:28:56.949000 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.403702 (* 0.0909091 = 0.0367002 loss)
I0429 00:28:56.949014 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.360942 (* 0.0909091 = 0.0328129 loss)
I0429 00:28:56.949028 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.184487 (* 0.0909091 = 0.0167715 loss)
I0429 00:28:56.949041 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.112726 (* 0.0909091 = 0.0102478 loss)
I0429 00:28:56.949055 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0942155 (* 0.0909091 = 0.00856505 loss)
I0429 00:28:56.949069 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0739315 (* 0.0909091 = 0.00672104 loss)
I0429 00:28:56.949082 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0400927 (* 0.0909091 = 0.00364479 loss)
I0429 00:28:56.949096 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.017682 (* 0.0909091 = 0.00160745 loss)
I0429 00:28:56.949110 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00734831 (* 0.0909091 = 0.000668028 loss)
I0429 00:28:56.949123 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00268584 (* 0.0909091 = 0.000244168 loss)
I0429 00:28:56.949137 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00124858 (* 0.0909091 = 0.000113508 loss)
I0429 00:28:56.949151 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000474848 (* 0.0909091 = 4.3168e-05 loss)
I0429 00:28:56.949165 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000144402 (* 0.0909091 = 1.31275e-05 loss)
I0429 00:28:56.949177 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:28:56.949188 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:28:56.949199 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00214335
I0429 00:28:56.949220 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00628959
I0429 00:28:56.949234 6470 sgd_solver.cpp:106] Iteration 14500, lr = 0.01
I0429 00:31:13.293303 6470 solver.cpp:338] Iteration 15000, Testing net (#0)
I0429 00:31:54.239573 6470 solver.cpp:393] Test loss: 8.57977
I0429 00:31:54.239699 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0817191
I0429 00:31:54.239719 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.161
I0429 00:31:54.239733 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.101
I0429 00:31:54.239747 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.107
I0429 00:31:54.239758 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.185
I0429 00:31:54.239770 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.348
I0429 00:31:54.239781 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.499
I0429 00:31:54.239794 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.744
I0429 00:31:54.239805 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.921
I0429 00:31:54.239816 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.995
I0429 00:31:54.239828 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.999
I0429 00:31:54.239840 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 1
I0429 00:31:54.239851 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 1
I0429 00:31:54.239862 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 1
I0429 00:31:54.239874 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 1
I0429 00:31:54.239886 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 1
I0429 00:31:54.239897 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 1
I0429 00:31:54.239908 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 1
I0429 00:31:54.239919 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 1
I0429 00:31:54.239930 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:31:54.239941 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:31:54.239953 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:31:54.239964 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:31:54.239975 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.768365
I0429 00:31:54.239987 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.253969
I0429 00:31:54.240002 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.08309 (* 0.3 = 0.924928 loss)
I0429 00:31:54.240016 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.813043 (* 0.3 = 0.243913 loss)
I0429 00:31:54.240031 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.70303 (* 0.0272727 = 0.0737191 loss)
I0429 00:31:54.240043 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 2.94228 (* 0.0272727 = 0.0802439 loss)
I0429 00:31:54.240057 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 3.03733 (* 0.0272727 = 0.0828362 loss)
I0429 00:31:54.240070 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.81505 (* 0.0272727 = 0.0767741 loss)
I0429 00:31:54.240084 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.36635 (* 0.0272727 = 0.0645369 loss)
I0429 00:31:54.240097 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 1.90758 (* 0.0272727 = 0.0520249 loss)
I0429 00:31:54.240110 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.09426 (* 0.0272727 = 0.0298434 loss)
I0429 00:31:54.240123 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.440212 (* 0.0272727 = 0.0120058 loss)
I0429 00:31:54.240137 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.0531277 (* 0.0272727 = 0.00144894 loss)
I0429 00:31:54.240151 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.0248052 (* 0.0272727 = 0.000676505 loss)
I0429 00:31:54.240164 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.0119902 (* 0.0272727 = 0.000327006 loss)
I0429 00:31:54.240178 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.00818592 (* 0.0272727 = 0.000223252 loss)
I0429 00:31:54.240192 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.00577903 (* 0.0272727 = 0.00015761 loss)
I0429 00:31:54.240226 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.0047189 (* 0.0272727 = 0.000128697 loss)
I0429 00:31:54.240242 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.00356065 (* 0.0272727 = 9.71087e-05 loss)
I0429 00:31:54.240255 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.00280158 (* 0.0272727 = 7.64067e-05 loss)
I0429 00:31:54.240268 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.00218162 (* 0.0272727 = 5.94987e-05 loss)
I0429 00:31:54.240283 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.00188087 (* 0.0272727 = 5.12965e-05 loss)
I0429 00:31:54.240295 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00151544 (* 0.0272727 = 4.13301e-05 loss)
I0429 00:31:54.240309 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.00129803 (* 0.0272727 = 3.54009e-05 loss)
I0429 00:31:54.240327 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.00101486 (* 0.0272727 = 2.76779e-05 loss)
I0429 00:31:54.240341 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.000911883 (* 0.0272727 = 2.48695e-05 loss)
I0429 00:31:54.240353 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0814845
I0429 00:31:54.240365 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.159
I0429 00:31:54.240375 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.105
I0429 00:31:54.240387 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.099
I0429 00:31:54.240398 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.181
I0429 00:31:54.240411 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.343
I0429 00:31:54.240422 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.497
I0429 00:31:54.240433 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.744
I0429 00:31:54.240447 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.921
I0429 00:31:54.240458 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.995
I0429 00:31:54.240469 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.999
I0429 00:31:54.240480 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 1
I0429 00:31:54.240491 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 1
I0429 00:31:54.240502 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 1
I0429 00:31:54.240514 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 1
I0429 00:31:54.240525 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 1
I0429 00:31:54.240536 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 1
I0429 00:31:54.240546 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 1
I0429 00:31:54.240557 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 1
I0429 00:31:54.240568 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:31:54.240579 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:31:54.240591 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:31:54.240602 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:31:54.240613 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.769228
I0429 00:31:54.240624 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.249131
I0429 00:31:54.240638 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.08642 (* 0.3 = 0.925927 loss)
I0429 00:31:54.240650 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.811693 (* 0.3 = 0.243508 loss)
I0429 00:31:54.240664 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.72114 (* 0.0272727 = 0.0742128 loss)
I0429 00:31:54.240677 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 2.96549 (* 0.0272727 = 0.080877 loss)
I0429 00:31:54.240690 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.03921 (* 0.0272727 = 0.0828875 loss)
I0429 00:31:54.240718 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.82591 (* 0.0272727 = 0.0770701 loss)
I0429 00:31:54.240733 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.39185 (* 0.0272727 = 0.0652322 loss)
I0429 00:31:54.240747 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 1.91903 (* 0.0272727 = 0.0523371 loss)
I0429 00:31:54.240761 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.09484 (* 0.0272727 = 0.0298593 loss)
I0429 00:31:54.240773 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.442213 (* 0.0272727 = 0.0120604 loss)
I0429 00:31:54.240787 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.0526836 (* 0.0272727 = 0.00143683 loss)
I0429 00:31:54.240800 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.0223913 (* 0.0272727 = 0.000610671 loss)
I0429 00:31:54.240814 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.0111895 (* 0.0272727 = 0.000305169 loss)
I0429 00:31:54.240828 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.0079737 (* 0.0272727 = 0.000217464 loss)
I0429 00:31:54.240841 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.00550192 (* 0.0272727 = 0.000150052 loss)
I0429 00:31:54.240854 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.00407689 (* 0.0272727 = 0.000111188 loss)
I0429 00:31:54.240869 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.00334383 (* 0.0272727 = 9.11954e-05 loss)
I0429 00:31:54.240882 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.00285625 (* 0.0272727 = 7.78976e-05 loss)
I0429 00:31:54.240895 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.00241391 (* 0.0272727 = 6.5834e-05 loss)
I0429 00:31:54.240908 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.0018889 (* 0.0272727 = 5.15155e-05 loss)
I0429 00:31:54.240922 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.00143849 (* 0.0272727 = 3.92316e-05 loss)
I0429 00:31:54.240936 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00127269 (* 0.0272727 = 3.47096e-05 loss)
I0429 00:31:54.240949 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.00111806 (* 0.0272727 = 3.04925e-05 loss)
I0429 00:31:54.240962 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000937686 (* 0.0272727 = 2.55732e-05 loss)
I0429 00:31:54.240973 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.118157
I0429 00:31:54.240985 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.228
I0429 00:31:54.240996 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.112
I0429 00:31:54.241008 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.107
I0429 00:31:54.241019 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.183
I0429 00:31:54.241030 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.358
I0429 00:31:54.241041 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.501
I0429 00:31:54.241052 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.742
I0429 00:31:54.241065 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.918
I0429 00:31:54.241075 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.993
I0429 00:31:54.241086 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.998
I0429 00:31:54.241097 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.999
I0429 00:31:54.241108 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 1
I0429 00:31:54.241119 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 1
I0429 00:31:54.241130 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 1
I0429 00:31:54.241142 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 1
I0429 00:31:54.241153 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 1
I0429 00:31:54.241173 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 1
I0429 00:31:54.241185 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 1
I0429 00:31:54.241196 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:31:54.241207 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:31:54.241217 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:31:54.241228 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:31:54.241240 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.77441
I0429 00:31:54.241250 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.302498
I0429 00:31:54.241264 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.94072 (* 1 = 2.94072 loss)
I0429 00:31:54.241277 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.794416 (* 1 = 0.794416 loss)
I0429 00:31:54.241291 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.50049 (* 0.0909091 = 0.227318 loss)
I0429 00:31:54.241304 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 2.93049 (* 0.0909091 = 0.266408 loss)
I0429 00:31:54.241317 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 2.99532 (* 0.0909091 = 0.272302 loss)
I0429 00:31:54.241330 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.77938 (* 0.0909091 = 0.252671 loss)
I0429 00:31:54.241343 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.34981 (* 0.0909091 = 0.213619 loss)
I0429 00:31:54.241356 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 1.89068 (* 0.0909091 = 0.17188 loss)
I0429 00:31:54.241372 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.07687 (* 0.0909091 = 0.0978976 loss)
I0429 00:31:54.241385 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.430441 (* 0.0909091 = 0.039131 loss)
I0429 00:31:54.241400 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.054053 (* 0.0909091 = 0.00491391 loss)
I0429 00:31:54.241412 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.0245348 (* 0.0909091 = 0.00223044 loss)
I0429 00:31:54.241426 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.0130495 (* 0.0909091 = 0.00118632 loss)
I0429 00:31:54.241439 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.0109539 (* 0.0909091 = 0.000995812 loss)
I0429 00:31:54.241453 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.0077307 (* 0.0909091 = 0.000702791 loss)
I0429 00:31:54.241466 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.00587193 (* 0.0909091 = 0.000533812 loss)
I0429 00:31:54.241479 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.00426148 (* 0.0909091 = 0.000387408 loss)
I0429 00:31:54.241493 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.00305719 (* 0.0909091 = 0.000277926 loss)
I0429 00:31:54.241506 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.00209887 (* 0.0909091 = 0.000190806 loss)
I0429 00:31:54.241519 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.00153315 (* 0.0909091 = 0.000139377 loss)
I0429 00:31:54.241533 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00136645 (* 0.0909091 = 0.000124223 loss)
I0429 00:31:54.241546 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.00131566 (* 0.0909091 = 0.000119605 loss)
I0429 00:31:54.241559 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.00111136 (* 0.0909091 = 0.000101032 loss)
I0429 00:31:54.241574 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.000837647 (* 0.0909091 = 7.61498e-05 loss)
I0429 00:31:54.241585 6470 solver.cpp:406] Test net output #147: total_accuracy = 0.003
I0429 00:31:54.241596 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.004
I0429 00:31:54.241607 6470 solver.cpp:406] Test net output #149: total_confidence = 0.00100348
I0429 00:31:54.241628 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.0032246
I0429 00:31:54.241642 6470 solver.cpp:338] Iteration 15000, Testing net (#1)
I0429 00:32:35.471637 6470 solver.cpp:393] Test loss: 9.10628
I0429 00:32:35.471742 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.099163
I0429 00:32:35.471761 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.171
I0429 00:32:35.471774 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.124
I0429 00:32:35.471786 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.119
I0429 00:32:35.471798 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.191
I0429 00:32:35.471810 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.334
I0429 00:32:35.471822 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.442
I0429 00:32:35.471833 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.673
I0429 00:32:35.471845 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.82
I0429 00:32:35.471856 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.89
I0429 00:32:35.471868 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.912
I0429 00:32:35.471879 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 0.922
I0429 00:32:35.471891 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 0.941
I0429 00:32:35.471904 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 0.957
I0429 00:32:35.471915 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 0.96
I0429 00:32:35.471926 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 0.975
I0429 00:32:35.471938 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 0.986
I0429 00:32:35.471951 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 0.995
I0429 00:32:35.471961 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 0.998
I0429 00:32:35.471973 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:32:35.471984 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:32:35.471997 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:32:35.472007 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:32:35.472019 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.738137
I0429 00:32:35.472030 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.263655
I0429 00:32:35.472046 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 3.03431 (* 0.3 = 0.910294 loss)
I0429 00:32:35.472060 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.934986 (* 0.3 = 0.280496 loss)
I0429 00:32:35.472074 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.70362 (* 0.0272727 = 0.073735 loss)
I0429 00:32:35.472087 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 2.91053 (* 0.0272727 = 0.079378 loss)
I0429 00:32:35.472101 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 2.98436 (* 0.0272727 = 0.0813916 loss)
I0429 00:32:35.472115 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.79715 (* 0.0272727 = 0.076286 loss)
I0429 00:32:35.472128 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.37892 (* 0.0272727 = 0.0648796 loss)
I0429 00:32:35.472141 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.08781 (* 0.0272727 = 0.0569403 loss)
I0429 00:32:35.472154 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.33969 (* 0.0272727 = 0.036537 loss)
I0429 00:32:35.472167 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.769928 (* 0.0272727 = 0.020998 loss)
I0429 00:32:35.472182 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.463976 (* 0.0272727 = 0.0126539 loss)
I0429 00:32:35.472198 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.376383 (* 0.0272727 = 0.010265 loss)
I0429 00:32:35.472213 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.322319 (* 0.0272727 = 0.00879052 loss)
I0429 00:32:35.472228 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.255698 (* 0.0272727 = 0.00697357 loss)
I0429 00:32:35.472240 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.199106 (* 0.0272727 = 0.00543017 loss)
I0429 00:32:35.472275 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.1846 (* 0.0272727 = 0.00503455 loss)
I0429 00:32:35.472290 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.139431 (* 0.0272727 = 0.00380267 loss)
I0429 00:32:35.472304 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.0880639 (* 0.0272727 = 0.00240174 loss)
I0429 00:32:35.472318 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.039699 (* 0.0272727 = 0.0010827 loss)
I0429 00:32:35.472332 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.0201017 (* 0.0272727 = 0.000548229 loss)
I0429 00:32:35.472347 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00303088 (* 0.0272727 = 8.26605e-05 loss)
I0429 00:32:35.472359 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.0020976 (* 0.0272727 = 5.72072e-05 loss)
I0429 00:32:35.472378 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.00134421 (* 0.0272727 = 3.66604e-05 loss)
I0429 00:32:35.472391 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.00106698 (* 0.0272727 = 2.90995e-05 loss)
I0429 00:32:35.472404 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.0895392
I0429 00:32:35.472415 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.158
I0429 00:32:35.472426 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.11
I0429 00:32:35.472439 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.103
I0429 00:32:35.472450 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.185
I0429 00:32:35.472460 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.337
I0429 00:32:35.472472 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.44
I0429 00:32:35.472483 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.668
I0429 00:32:35.472496 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.817
I0429 00:32:35.472506 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.891
I0429 00:32:35.472518 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.912
I0429 00:32:35.472525 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 0.922
I0429 00:32:35.472533 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 0.94
I0429 00:32:35.472544 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 0.954
I0429 00:32:35.472556 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 0.96
I0429 00:32:35.472568 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 0.975
I0429 00:32:35.472579 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 0.986
I0429 00:32:35.472590 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 0.995
I0429 00:32:35.472601 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 0.998
I0429 00:32:35.472612 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:32:35.472623 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:32:35.472635 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:32:35.472645 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:32:35.472656 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.739137
I0429 00:32:35.472667 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.254093
I0429 00:32:35.472681 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 3.05245 (* 0.3 = 0.915735 loss)
I0429 00:32:35.472694 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.931523 (* 0.3 = 0.279457 loss)
I0429 00:32:35.472708 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.7188 (* 0.0272727 = 0.0741492 loss)
I0429 00:32:35.472721 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 2.93956 (* 0.0272727 = 0.0801698 loss)
I0429 00:32:35.472745 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 3.01105 (* 0.0272727 = 0.0821197 loss)
I0429 00:32:35.472760 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.82195 (* 0.0272727 = 0.0769623 loss)
I0429 00:32:35.472774 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.40199 (* 0.0272727 = 0.0655089 loss)
I0429 00:32:35.472787 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.10026 (* 0.0272727 = 0.0572798 loss)
I0429 00:32:35.472800 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.33366 (* 0.0272727 = 0.0363726 loss)
I0429 00:32:35.472813 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.779378 (* 0.0272727 = 0.0212558 loss)
I0429 00:32:35.472826 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.467378 (* 0.0272727 = 0.0127467 loss)
I0429 00:32:35.472839 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.370328 (* 0.0272727 = 0.0100999 loss)
I0429 00:32:35.472853 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.323208 (* 0.0272727 = 0.00881477 loss)
I0429 00:32:35.472867 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.257795 (* 0.0272727 = 0.00703077 loss)
I0429 00:32:35.472880 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.197239 (* 0.0272727 = 0.00537924 loss)
I0429 00:32:35.472894 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.179323 (* 0.0272727 = 0.00489063 loss)
I0429 00:32:35.472908 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.135965 (* 0.0272727 = 0.00370815 loss)
I0429 00:32:35.472921 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.0864312 (* 0.0272727 = 0.00235721 loss)
I0429 00:32:35.472934 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.0409782 (* 0.0272727 = 0.00111759 loss)
I0429 00:32:35.472949 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.0173397 (* 0.0272727 = 0.0004729 loss)
I0429 00:32:35.472962 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.00218404 (* 0.0272727 = 5.95648e-05 loss)
I0429 00:32:35.472975 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00156805 (* 0.0272727 = 4.27651e-05 loss)
I0429 00:32:35.472988 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.00101149 (* 0.0272727 = 2.75861e-05 loss)
I0429 00:32:35.473002 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000796584 (* 0.0272727 = 2.1725e-05 loss)
I0429 00:32:35.473013 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.122606
I0429 00:32:35.473026 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.225
I0429 00:32:35.473037 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.121
I0429 00:32:35.473048 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.111
I0429 00:32:35.473059 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.214
I0429 00:32:35.473070 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.347
I0429 00:32:35.473083 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.445
I0429 00:32:35.473093 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.674
I0429 00:32:35.473104 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.817
I0429 00:32:35.473115 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.896
I0429 00:32:35.473126 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.916
I0429 00:32:35.473139 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.922
I0429 00:32:35.473150 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 0.94
I0429 00:32:35.473161 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 0.957
I0429 00:32:35.473172 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 0.96
I0429 00:32:35.473183 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 0.973
I0429 00:32:35.473194 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 0.986
I0429 00:32:35.473215 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 0.995
I0429 00:32:35.473227 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 0.998
I0429 00:32:35.473239 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:32:35.473253 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:32:35.473265 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:32:35.473276 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:32:35.473287 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.739455
I0429 00:32:35.473299 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.315056
I0429 00:32:35.473312 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.91177 (* 1 = 2.91177 loss)
I0429 00:32:35.473325 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.925324 (* 1 = 0.925324 loss)
I0429 00:32:35.473340 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.56086 (* 0.0909091 = 0.232805 loss)
I0429 00:32:35.473352 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 2.90193 (* 0.0909091 = 0.263811 loss)
I0429 00:32:35.473366 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 2.96068 (* 0.0909091 = 0.269153 loss)
I0429 00:32:35.473379 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.7463 (* 0.0909091 = 0.249664 loss)
I0429 00:32:35.473392 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.33836 (* 0.0909091 = 0.212578 loss)
I0429 00:32:35.473405 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.05441 (* 0.0909091 = 0.186764 loss)
I0429 00:32:35.473420 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.33054 (* 0.0909091 = 0.120958 loss)
I0429 00:32:35.473434 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.768084 (* 0.0909091 = 0.0698258 loss)
I0429 00:32:35.473448 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.441143 (* 0.0909091 = 0.0401039 loss)
I0429 00:32:35.473461 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.35419 (* 0.0909091 = 0.0321991 loss)
I0429 00:32:35.473474 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.308753 (* 0.0909091 = 0.0280684 loss)
I0429 00:32:35.473489 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.245284 (* 0.0909091 = 0.0222986 loss)
I0429 00:32:35.473501 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.184369 (* 0.0909091 = 0.0167608 loss)
I0429 00:32:35.473515 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.173152 (* 0.0909091 = 0.015741 loss)
I0429 00:32:35.473527 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.128366 (* 0.0909091 = 0.0116697 loss)
I0429 00:32:35.473541 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.0814294 (* 0.0909091 = 0.00740267 loss)
I0429 00:32:35.473554 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.0353338 (* 0.0909091 = 0.00321217 loss)
I0429 00:32:35.473567 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.0149188 (* 0.0909091 = 0.00135625 loss)
I0429 00:32:35.473580 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00461198 (* 0.0909091 = 0.000419271 loss)
I0429 00:32:35.473593 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.00322091 (* 0.0909091 = 0.00029281 loss)
I0429 00:32:35.473608 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.00150952 (* 0.0909091 = 0.000137229 loss)
I0429 00:32:35.473620 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.000724544 (* 0.0909091 = 6.58677e-05 loss)
I0429 00:32:35.473631 6470 solver.cpp:406] Test net output #147: total_accuracy = 0
I0429 00:32:35.473642 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.002
I0429 00:32:35.473654 6470 solver.cpp:406] Test net output #149: total_confidence = 0.000963758
I0429 00:32:35.473675 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.0025363
I0429 00:32:35.653406 6470 solver.cpp:229] Iteration 15000, loss = 9.48239
I0429 00:32:35.653482 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0392157
I0429 00:32:35.653503 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:32:35.653522 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.25
I0429 00:32:35.653534 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.25
I0429 00:32:35.653547 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0429 00:32:35.653558 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 00:32:35.653570 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:32:35.653584 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:32:35.653595 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:32:35.653607 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:32:35.653620 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:32:35.653630 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:32:35.653642 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:32:35.653655 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:32:35.653666 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:32:35.653678 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:32:35.653690 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:32:35.653702 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:32:35.653713 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:32:35.653725 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:32:35.653738 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:32:35.653749 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:32:35.653761 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:32:35.653772 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.704545
I0429 00:32:35.653784 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.254902
I0429 00:32:35.653800 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.8882 (* 0.3 = 0.86646 loss)
I0429 00:32:35.653815 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.929206 (* 0.3 = 0.278762 loss)
I0429 00:32:35.653830 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.54609 (* 0.0272727 = 0.0694388 loss)
I0429 00:32:35.653843 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.8099 (* 0.0272727 = 0.0766336 loss)
I0429 00:32:35.653857 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.58839 (* 0.0272727 = 0.0705923 loss)
I0429 00:32:35.653872 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.23732 (* 0.0272727 = 0.0882905 loss)
I0429 00:32:35.653885 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.87753 (* 0.0272727 = 0.0784781 loss)
I0429 00:32:35.653903 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.49252 (* 0.0272727 = 0.0679777 loss)
I0429 00:32:35.653918 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.52469 (* 0.0272727 = 0.0415826 loss)
I0429 00:32:35.653931 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.974123 (* 0.0272727 = 0.026567 loss)
I0429 00:32:35.653945 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.06677 (* 0.0272727 = 0.0290938 loss)
I0429 00:32:35.653960 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.0128568 (* 0.0272727 = 0.00035064 loss)
I0429 00:32:35.653975 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.00584058 (* 0.0272727 = 0.000159289 loss)
I0429 00:32:35.654022 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.00393076 (* 0.0272727 = 0.000107203 loss)
I0429 00:32:35.654038 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.00186132 (* 0.0272727 = 5.07633e-05 loss)
I0429 00:32:35.654052 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00128648 (* 0.0272727 = 3.50857e-05 loss)
I0429 00:32:35.654067 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.000966743 (* 0.0272727 = 2.63657e-05 loss)
I0429 00:32:35.654080 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00110525 (* 0.0272727 = 3.01432e-05 loss)
I0429 00:32:35.654094 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.000608118 (* 0.0272727 = 1.6585e-05 loss)
I0429 00:32:35.654109 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000356403 (* 0.0272727 = 9.72007e-06 loss)
I0429 00:32:35.654122 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000180788 (* 0.0272727 = 4.93059e-06 loss)
I0429 00:32:35.654136 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00016413 (* 0.0272727 = 4.47628e-06 loss)
I0429 00:32:35.654150 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 8.42836e-05 (* 0.0272727 = 2.29864e-06 loss)
I0429 00:32:35.654165 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 7.61218e-05 (* 0.0272727 = 2.07605e-06 loss)
I0429 00:32:35.654176 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0392157
I0429 00:32:35.654188 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:32:35.654199 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:32:35.654211 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:32:35.654223 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0
I0429 00:32:35.654234 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0429 00:32:35.654247 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 00:32:35.654258 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:32:35.654270 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:32:35.654281 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:32:35.654294 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:32:35.654304 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:32:35.654316 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:32:35.654328 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:32:35.654336 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:32:35.654345 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:32:35.654352 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:32:35.654364 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:32:35.654376 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:32:35.654388 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:32:35.654399 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:32:35.654412 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:32:35.654422 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:32:35.654433 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.704545
I0429 00:32:35.654445 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.27451
I0429 00:32:35.654459 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.94929 (* 0.3 = 0.884787 loss)
I0429 00:32:35.654472 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.950387 (* 0.3 = 0.285116 loss)
I0429 00:32:35.654486 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.62836 (* 0.0272727 = 0.0716826 loss)
I0429 00:32:35.654511 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.68091 (* 0.0272727 = 0.0731156 loss)
I0429 00:32:35.654526 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.82013 (* 0.0272727 = 0.0769126 loss)
I0429 00:32:35.654541 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.29761 (* 0.0272727 = 0.0899349 loss)
I0429 00:32:35.654553 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.92594 (* 0.0272727 = 0.0797984 loss)
I0429 00:32:35.654570 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.43738 (* 0.0272727 = 0.066474 loss)
I0429 00:32:35.654584 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.41428 (* 0.0272727 = 0.0385711 loss)
I0429 00:32:35.654597 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.932908 (* 0.0272727 = 0.0254429 loss)
I0429 00:32:35.654611 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.929781 (* 0.0272727 = 0.0253577 loss)
I0429 00:32:35.654625 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.0374958 (* 0.0272727 = 0.00102261 loss)
I0429 00:32:35.654639 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0115314 (* 0.0272727 = 0.000314492 loss)
I0429 00:32:35.654654 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00969079 (* 0.0272727 = 0.000264294 loss)
I0429 00:32:35.654666 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00547488 (* 0.0272727 = 0.000149315 loss)
I0429 00:32:35.654680 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00352433 (* 0.0272727 = 9.61181e-05 loss)
I0429 00:32:35.654695 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00400037 (* 0.0272727 = 0.000109101 loss)
I0429 00:32:35.654707 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.003778 (* 0.0272727 = 0.000103036 loss)
I0429 00:32:35.654721 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00460598 (* 0.0272727 = 0.000125618 loss)
I0429 00:32:35.654734 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00253458 (* 0.0272727 = 6.91248e-05 loss)
I0429 00:32:35.654748 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00285862 (* 0.0272727 = 7.79624e-05 loss)
I0429 00:32:35.654762 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00117761 (* 0.0272727 = 3.21166e-05 loss)
I0429 00:32:35.654774 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000638901 (* 0.0272727 = 1.74246e-05 loss)
I0429 00:32:35.654788 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000624694 (* 0.0272727 = 1.70371e-05 loss)
I0429 00:32:35.654800 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0784314
I0429 00:32:35.654813 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:32:35.654824 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:32:35.654835 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.375
I0429 00:32:35.654847 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:32:35.654858 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 00:32:35.654870 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:32:35.654882 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:32:35.654893 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:32:35.654906 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:32:35.654917 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:32:35.654928 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:32:35.654939 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:32:35.654955 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:32:35.654978 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:32:35.654989 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:32:35.655001 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:32:35.655012 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:32:35.655025 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:32:35.655036 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:32:35.655047 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:32:35.655060 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:32:35.655071 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:32:35.655082 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.710227
I0429 00:32:35.655093 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.294118
I0429 00:32:35.655107 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.86137 (* 1 = 2.86137 loss)
I0429 00:32:35.655122 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.937289 (* 1 = 0.937289 loss)
I0429 00:32:35.655134 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.32907 (* 0.0909091 = 0.211733 loss)
I0429 00:32:35.655148 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.74816 (* 0.0909091 = 0.249833 loss)
I0429 00:32:35.655161 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.44003 (* 0.0909091 = 0.221821 loss)
I0429 00:32:35.655175 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.8918 (* 0.0909091 = 0.262891 loss)
I0429 00:32:35.655189 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.88077 (* 0.0909091 = 0.261888 loss)
I0429 00:32:35.655202 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.32779 (* 0.0909091 = 0.211617 loss)
I0429 00:32:35.655215 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.66603 (* 0.0909091 = 0.151457 loss)
I0429 00:32:35.655228 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.17069 (* 0.0909091 = 0.106427 loss)
I0429 00:32:35.655241 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 1.16556 (* 0.0909091 = 0.10596 loss)
I0429 00:32:35.655256 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.0185924 (* 0.0909091 = 0.00169022 loss)
I0429 00:32:35.655269 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00539901 (* 0.0909091 = 0.000490819 loss)
I0429 00:32:35.655282 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00389461 (* 0.0909091 = 0.000354056 loss)
I0429 00:32:35.655297 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00406394 (* 0.0909091 = 0.000369449 loss)
I0429 00:32:35.655309 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00350349 (* 0.0909091 = 0.000318499 loss)
I0429 00:32:35.655323 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00299328 (* 0.0909091 = 0.000272117 loss)
I0429 00:32:35.655338 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00249214 (* 0.0909091 = 0.000226559 loss)
I0429 00:32:35.655351 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0011099 (* 0.0909091 = 0.0001009 loss)
I0429 00:32:35.655365 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00098741 (* 0.0909091 = 8.97646e-05 loss)
I0429 00:32:35.655380 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.001039 (* 0.0909091 = 9.44547e-05 loss)
I0429 00:32:35.655392 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00102522 (* 0.0909091 = 9.32016e-05 loss)
I0429 00:32:35.655406 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00085394 (* 0.0909091 = 7.76309e-05 loss)
I0429 00:32:35.655421 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00052047 (* 0.0909091 = 4.73155e-05 loss)
I0429 00:32:35.655441 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:32:35.655454 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:32:35.655478 6470 solver.cpp:245] Train net output #149: total_confidence = 3.39573e-06
I0429 00:32:35.655493 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 2.50716e-05
I0429 00:32:35.655506 6470 sgd_solver.cpp:106] Iteration 15000, lr = 0.01
I0429 00:34:52.242475 6470 solver.cpp:229] Iteration 15500, loss = 9.36717
I0429 00:34:52.242650 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.129032
I0429 00:34:52.242671 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 00:34:52.242684 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.25
I0429 00:34:52.242697 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:34:52.242709 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:34:52.242720 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0
I0429 00:34:52.242733 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.625
I0429 00:34:52.242744 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0429 00:34:52.242756 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0429 00:34:52.242769 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 00:34:52.242779 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.75
I0429 00:34:52.242791 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:34:52.242804 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:34:52.242815 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:34:52.242827 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:34:52.242840 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:34:52.242851 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:34:52.242862 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:34:52.242873 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:34:52.242885 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:34:52.242897 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:34:52.242908 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:34:52.242919 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:34:52.242931 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.676136
I0429 00:34:52.242944 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.322581
I0429 00:34:52.242959 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.01285 (* 0.3 = 0.903855 loss)
I0429 00:34:52.242974 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.23836 (* 0.3 = 0.371507 loss)
I0429 00:34:52.242987 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.6339 (* 0.0272727 = 0.0718336 loss)
I0429 00:34:52.243001 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.44797 (* 0.0272727 = 0.0667629 loss)
I0429 00:34:52.243016 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.68611 (* 0.0272727 = 0.10053 loss)
I0429 00:34:52.243028 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.85841 (* 0.0272727 = 0.0779566 loss)
I0429 00:34:52.243042 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.31817 (* 0.0272727 = 0.0904955 loss)
I0429 00:34:52.243057 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.08369 (* 0.0272727 = 0.0568279 loss)
I0429 00:34:52.243069 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.82591 (* 0.0272727 = 0.0497976 loss)
I0429 00:34:52.243083 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.42222 (* 0.0272727 = 0.0387878 loss)
I0429 00:34:52.243098 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 1.04267 (* 0.0272727 = 0.0284365 loss)
I0429 00:34:52.243110 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.759262 (* 0.0272727 = 0.0207071 loss)
I0429 00:34:52.243124 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.775488 (* 0.0272727 = 0.0211497 loss)
I0429 00:34:52.243139 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.580999 (* 0.0272727 = 0.0158454 loss)
I0429 00:34:52.243152 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.694928 (* 0.0272727 = 0.0189526 loss)
I0429 00:34:52.243186 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.464111 (* 0.0272727 = 0.0126576 loss)
I0429 00:34:52.243201 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0702106 (* 0.0272727 = 0.00191484 loss)
I0429 00:34:52.243216 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0329783 (* 0.0272727 = 0.000899409 loss)
I0429 00:34:52.243230 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0158979 (* 0.0272727 = 0.000433578 loss)
I0429 00:34:52.243243 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00408017 (* 0.0272727 = 0.000111277 loss)
I0429 00:34:52.243257 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00227603 (* 0.0272727 = 6.20735e-05 loss)
I0429 00:34:52.243271 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00108021 (* 0.0272727 = 2.94602e-05 loss)
I0429 00:34:52.243285 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000613743 (* 0.0272727 = 1.67385e-05 loss)
I0429 00:34:52.243299 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000297882 (* 0.0272727 = 8.12406e-06 loss)
I0429 00:34:52.243314 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.112903
I0429 00:34:52.243327 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.375
I0429 00:34:52.243336 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.25
I0429 00:34:52.243343 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:34:52.243351 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:34:52.243363 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0
I0429 00:34:52.243374 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:34:52.243386 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:34:52.243398 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 00:34:52.243410 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 00:34:52.243422 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.75
I0429 00:34:52.243433 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:34:52.243445 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:34:52.243456 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:34:52.243489 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:34:52.243502 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:34:52.243515 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:34:52.243525 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:34:52.243537 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:34:52.243548 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:34:52.243559 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:34:52.243571 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:34:52.243582 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:34:52.243594 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.681818
I0429 00:34:52.243607 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.33871
I0429 00:34:52.243619 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.04351 (* 0.3 = 0.913053 loss)
I0429 00:34:52.243633 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.18475 (* 0.3 = 0.355426 loss)
I0429 00:34:52.243648 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.60764 (* 0.0272727 = 0.0711175 loss)
I0429 00:34:52.243665 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.43349 (* 0.0272727 = 0.0663678 loss)
I0429 00:34:52.243692 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.69809 (* 0.0272727 = 0.100857 loss)
I0429 00:34:52.243707 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.91529 (* 0.0272727 = 0.0795078 loss)
I0429 00:34:52.243721 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 3.55723 (* 0.0272727 = 0.0970154 loss)
I0429 00:34:52.243734 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.36087 (* 0.0272727 = 0.0643874 loss)
I0429 00:34:52.243748 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.56895 (* 0.0272727 = 0.0427895 loss)
I0429 00:34:52.243762 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.40335 (* 0.0272727 = 0.0382731 loss)
I0429 00:34:52.243775 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 1.01351 (* 0.0272727 = 0.0276413 loss)
I0429 00:34:52.243788 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.896985 (* 0.0272727 = 0.0244632 loss)
I0429 00:34:52.243803 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.766585 (* 0.0272727 = 0.0209069 loss)
I0429 00:34:52.243816 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.70588 (* 0.0272727 = 0.0192513 loss)
I0429 00:34:52.243830 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.732708 (* 0.0272727 = 0.0199829 loss)
I0429 00:34:52.243844 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.699092 (* 0.0272727 = 0.0190661 loss)
I0429 00:34:52.243857 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.126092 (* 0.0272727 = 0.00343886 loss)
I0429 00:34:52.243871 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0764048 (* 0.0272727 = 0.00208377 loss)
I0429 00:34:52.243885 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0415185 (* 0.0272727 = 0.00113232 loss)
I0429 00:34:52.243898 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0258817 (* 0.0272727 = 0.000705863 loss)
I0429 00:34:52.243912 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0208396 (* 0.0272727 = 0.000568354 loss)
I0429 00:34:52.243927 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00786287 (* 0.0272727 = 0.000214442 loss)
I0429 00:34:52.243939 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00885538 (* 0.0272727 = 0.00024151 loss)
I0429 00:34:52.243953 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00566766 (* 0.0272727 = 0.000154573 loss)
I0429 00:34:52.243965 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.177419
I0429 00:34:52.243978 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:34:52.243989 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:34:52.244000 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:34:52.244012 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.375
I0429 00:34:52.244024 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0
I0429 00:34:52.244035 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:34:52.244047 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:34:52.244058 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0429 00:34:52.244071 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 00:34:52.244082 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.75
I0429 00:34:52.244093 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0429 00:34:52.244105 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:34:52.244117 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:34:52.244128 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:34:52.244139 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:34:52.244151 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:34:52.244173 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:34:52.244185 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:34:52.244196 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:34:52.244207 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:34:52.244220 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:34:52.244230 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:34:52.244242 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.704545
I0429 00:34:52.244254 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.322581
I0429 00:34:52.244267 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.95191 (* 1 = 2.95191 loss)
I0429 00:34:52.244282 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.1649 (* 1 = 1.1649 loss)
I0429 00:34:52.244295 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.65476 (* 0.0909091 = 0.241342 loss)
I0429 00:34:52.244309 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.3683 (* 0.0909091 = 0.2153 loss)
I0429 00:34:52.244323 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.68942 (* 0.0909091 = 0.335402 loss)
I0429 00:34:52.244336 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.67366 (* 0.0909091 = 0.24306 loss)
I0429 00:34:52.244349 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.24792 (* 0.0909091 = 0.295266 loss)
I0429 00:34:52.244365 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.32839 (* 0.0909091 = 0.211672 loss)
I0429 00:34:52.244379 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.46645 (* 0.0909091 = 0.133313 loss)
I0429 00:34:52.244393 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.40217 (* 0.0909091 = 0.12747 loss)
I0429 00:34:52.244407 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.893894 (* 0.0909091 = 0.0812631 loss)
I0429 00:34:52.244421 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.756334 (* 0.0909091 = 0.0687576 loss)
I0429 00:34:52.244434 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.861415 (* 0.0909091 = 0.0783104 loss)
I0429 00:34:52.244448 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.591916 (* 0.0909091 = 0.0538105 loss)
I0429 00:34:52.244462 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.544148 (* 0.0909091 = 0.049468 loss)
I0429 00:34:52.244475 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.480357 (* 0.0909091 = 0.0436688 loss)
I0429 00:34:52.244489 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.100896 (* 0.0909091 = 0.00917232 loss)
I0429 00:34:52.244503 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0820612 (* 0.0909091 = 0.00746011 loss)
I0429 00:34:52.244518 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0592701 (* 0.0909091 = 0.00538819 loss)
I0429 00:34:52.244531 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0364621 (* 0.0909091 = 0.00331474 loss)
I0429 00:34:52.244545 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0200856 (* 0.0909091 = 0.00182596 loss)
I0429 00:34:52.244560 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.0105647 (* 0.0909091 = 0.000960423 loss)
I0429 00:34:52.244572 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0058973 (* 0.0909091 = 0.000536118 loss)
I0429 00:34:52.244586 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00244677 (* 0.0909091 = 0.000222434 loss)
I0429 00:34:52.244598 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:34:52.244609 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:34:52.244621 6470 solver.cpp:245] Train net output #149: total_confidence = 2.6212e-06
I0429 00:34:52.244642 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 1.99159e-06
I0429 00:34:52.244657 6470 sgd_solver.cpp:106] Iteration 15500, lr = 0.01
I0429 00:37:08.930845 6470 solver.cpp:229] Iteration 16000, loss = 9.42678
I0429 00:37:08.930977 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0714286
I0429 00:37:08.930996 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 00:37:08.931010 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:37:08.931022 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:37:08.931033 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:37:08.931046 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0429 00:37:08.931057 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.625
I0429 00:37:08.931071 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.875
I0429 00:37:08.931082 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 00:37:08.931094 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:37:08.931105 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:37:08.931118 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:37:08.931129 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:37:08.931141 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:37:08.931152 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:37:08.931164 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:37:08.931175 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:37:08.931187 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:37:08.931200 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:37:08.931210 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:37:08.931222 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:37:08.931233 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:37:08.931246 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:37:08.931257 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.767045
I0429 00:37:08.931268 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.238095
I0429 00:37:08.931284 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.35061 (* 0.3 = 1.00518 loss)
I0429 00:37:08.931298 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.933656 (* 0.3 = 0.280097 loss)
I0429 00:37:08.931315 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.90003 (* 0.0272727 = 0.0790918 loss)
I0429 00:37:08.931330 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.64593 (* 0.0272727 = 0.0994344 loss)
I0429 00:37:08.931344 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.16144 (* 0.0272727 = 0.086221 loss)
I0429 00:37:08.931357 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.18879 (* 0.0272727 = 0.086967 loss)
I0429 00:37:08.931371 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 1.96113 (* 0.0272727 = 0.0534852 loss)
I0429 00:37:08.931385 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.27792 (* 0.0272727 = 0.0348523 loss)
I0429 00:37:08.931398 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.707946 (* 0.0272727 = 0.0193076 loss)
I0429 00:37:08.931412 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.515105 (* 0.0272727 = 0.0140483 loss)
I0429 00:37:08.931426 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.427339 (* 0.0272727 = 0.0116547 loss)
I0429 00:37:08.931439 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.477238 (* 0.0272727 = 0.0130156 loss)
I0429 00:37:08.931453 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.398126 (* 0.0272727 = 0.010858 loss)
I0429 00:37:08.931480 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0760661 (* 0.0272727 = 0.00207453 loss)
I0429 00:37:08.931499 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0539063 (* 0.0272727 = 0.00147017 loss)
I0429 00:37:08.931534 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0277285 (* 0.0272727 = 0.000756232 loss)
I0429 00:37:08.931550 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0186182 (* 0.0272727 = 0.000507768 loss)
I0429 00:37:08.931563 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0119215 (* 0.0272727 = 0.000325131 loss)
I0429 00:37:08.931576 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00302734 (* 0.0272727 = 8.25639e-05 loss)
I0429 00:37:08.931591 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00173568 (* 0.0272727 = 4.73367e-05 loss)
I0429 00:37:08.931604 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00112336 (* 0.0272727 = 3.0637e-05 loss)
I0429 00:37:08.931618 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000880345 (* 0.0272727 = 2.40094e-05 loss)
I0429 00:37:08.931632 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000568456 (* 0.0272727 = 1.55033e-05 loss)
I0429 00:37:08.931645 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00042013 (* 0.0272727 = 1.14581e-05 loss)
I0429 00:37:08.931658 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.047619
I0429 00:37:08.931669 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.375
I0429 00:37:08.931681 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:37:08.931694 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:37:08.931704 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.375
I0429 00:37:08.931716 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 00:37:08.931728 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.625
I0429 00:37:08.931740 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.875
I0429 00:37:08.931752 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 00:37:08.931764 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:37:08.931776 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:37:08.931787 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:37:08.931799 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:37:08.931810 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:37:08.931823 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:37:08.931833 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:37:08.931844 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:37:08.931856 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:37:08.931867 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:37:08.931879 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:37:08.931890 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:37:08.931901 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:37:08.931913 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:37:08.931924 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.761364
I0429 00:37:08.931936 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.166667
I0429 00:37:08.931951 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.42136 (* 0.3 = 1.02641 loss)
I0429 00:37:08.931963 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.939554 (* 0.3 = 0.281866 loss)
I0429 00:37:08.931977 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.79032 (* 0.0272727 = 0.0760996 loss)
I0429 00:37:08.931995 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.42879 (* 0.0272727 = 0.0935124 loss)
I0429 00:37:08.932021 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.26218 (* 0.0272727 = 0.0889686 loss)
I0429 00:37:08.932036 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.37691 (* 0.0272727 = 0.0920976 loss)
I0429 00:37:08.932050 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 1.96675 (* 0.0272727 = 0.0536386 loss)
I0429 00:37:08.932063 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.44486 (* 0.0272727 = 0.0394052 loss)
I0429 00:37:08.932076 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.595359 (* 0.0272727 = 0.0162371 loss)
I0429 00:37:08.932091 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.305556 (* 0.0272727 = 0.00833334 loss)
I0429 00:37:08.932104 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.453771 (* 0.0272727 = 0.0123756 loss)
I0429 00:37:08.932118 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.547672 (* 0.0272727 = 0.0149365 loss)
I0429 00:37:08.932132 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.409875 (* 0.0272727 = 0.0111784 loss)
I0429 00:37:08.932145 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.012617 (* 0.0272727 = 0.000344099 loss)
I0429 00:37:08.932159 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00638011 (* 0.0272727 = 0.000174003 loss)
I0429 00:37:08.932169 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00275083 (* 0.0272727 = 7.50228e-05 loss)
I0429 00:37:08.932179 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0019361 (* 0.0272727 = 5.28027e-05 loss)
I0429 00:37:08.932193 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.000676778 (* 0.0272727 = 1.84576e-05 loss)
I0429 00:37:08.932207 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00036638 (* 0.0272727 = 9.99218e-06 loss)
I0429 00:37:08.932220 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000384387 (* 0.0272727 = 1.04833e-05 loss)
I0429 00:37:08.932235 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000168785 (* 0.0272727 = 4.60324e-06 loss)
I0429 00:37:08.932248 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000158128 (* 0.0272727 = 4.31258e-06 loss)
I0429 00:37:08.932261 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000248333 (* 0.0272727 = 6.77272e-06 loss)
I0429 00:37:08.932276 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000188715 (* 0.0272727 = 5.14679e-06 loss)
I0429 00:37:08.932287 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.047619
I0429 00:37:08.932299 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0429 00:37:08.932312 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:37:08.932322 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.375
I0429 00:37:08.932334 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:37:08.932345 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.625
I0429 00:37:08.932358 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.625
I0429 00:37:08.932371 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.875
I0429 00:37:08.932384 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:37:08.932394 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:37:08.932406 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:37:08.932417 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:37:08.932430 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:37:08.932440 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:37:08.932452 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:37:08.932463 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:37:08.932484 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:37:08.932497 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:37:08.932508 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:37:08.932520 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:37:08.932531 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:37:08.932543 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:37:08.932554 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:37:08.932565 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.75
I0429 00:37:08.932577 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.261905
I0429 00:37:08.932591 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.23234 (* 1 = 3.23234 loss)
I0429 00:37:08.932605 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.883514 (* 1 = 0.883514 loss)
I0429 00:37:08.932618 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.32157 (* 0.0909091 = 0.211052 loss)
I0429 00:37:08.932632 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.28839 (* 0.0909091 = 0.298944 loss)
I0429 00:37:08.932646 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.86514 (* 0.0909091 = 0.260468 loss)
I0429 00:37:08.932659 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.92359 (* 0.0909091 = 0.265781 loss)
I0429 00:37:08.932672 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.73366 (* 0.0909091 = 0.157605 loss)
I0429 00:37:08.932685 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.38779 (* 0.0909091 = 0.126163 loss)
I0429 00:37:08.932699 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.681374 (* 0.0909091 = 0.0619431 loss)
I0429 00:37:08.932713 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.462532 (* 0.0909091 = 0.0420484 loss)
I0429 00:37:08.932726 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.544089 (* 0.0909091 = 0.0494626 loss)
I0429 00:37:08.932739 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.603559 (* 0.0909091 = 0.054869 loss)
I0429 00:37:08.932754 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.618279 (* 0.0909091 = 0.0562072 loss)
I0429 00:37:08.932767 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00975051 (* 0.0909091 = 0.00088641 loss)
I0429 00:37:08.932781 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00519105 (* 0.0909091 = 0.000471914 loss)
I0429 00:37:08.932796 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00397967 (* 0.0909091 = 0.000361788 loss)
I0429 00:37:08.932808 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00277649 (* 0.0909091 = 0.000252408 loss)
I0429 00:37:08.932822 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00160245 (* 0.0909091 = 0.000145677 loss)
I0429 00:37:08.932837 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000903265 (* 0.0909091 = 8.2115e-05 loss)
I0429 00:37:08.932850 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000593499 (* 0.0909091 = 5.39545e-05 loss)
I0429 00:37:08.932863 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000468374 (* 0.0909091 = 4.25795e-05 loss)
I0429 00:37:08.932878 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000308778 (* 0.0909091 = 2.80707e-05 loss)
I0429 00:37:08.932891 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000224892 (* 0.0909091 = 2.04447e-05 loss)
I0429 00:37:08.932905 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000143828 (* 0.0909091 = 1.30753e-05 loss)
I0429 00:37:08.932917 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:37:08.932929 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:37:08.932950 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000962013
I0429 00:37:08.932962 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00382823
I0429 00:37:08.932974 6470 sgd_solver.cpp:106] Iteration 16000, lr = 0.01
I0429 00:39:25.607525 6470 solver.cpp:229] Iteration 16500, loss = 9.34188
I0429 00:39:25.607669 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0869565
I0429 00:39:25.607689 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:39:25.607702 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:39:25.607714 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:39:25.607727 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 00:39:25.607738 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:39:25.607750 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:39:25.607763 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 00:39:25.607774 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:39:25.607786 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:39:25.607797 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:39:25.607810 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:39:25.607820 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:39:25.607832 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:39:25.607844 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:39:25.607856 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:39:25.607867 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:39:25.607879 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:39:25.607890 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:39:25.607902 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:39:25.607913 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:39:25.607924 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:39:25.607936 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:39:25.607947 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.744318
I0429 00:39:25.607959 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.26087
I0429 00:39:25.607975 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.14573 (* 0.3 = 0.943718 loss)
I0429 00:39:25.607990 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.941379 (* 0.3 = 0.282414 loss)
I0429 00:39:25.608005 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.63325 (* 0.0272727 = 0.071816 loss)
I0429 00:39:25.608018 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.98281 (* 0.0272727 = 0.108622 loss)
I0429 00:39:25.608032 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.2058 (* 0.0272727 = 0.0874308 loss)
I0429 00:39:25.608047 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.06924 (* 0.0272727 = 0.0837065 loss)
I0429 00:39:25.608060 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.57919 (* 0.0272727 = 0.0703416 loss)
I0429 00:39:25.608073 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.72525 (* 0.0272727 = 0.0743249 loss)
I0429 00:39:25.608086 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.981808 (* 0.0272727 = 0.0267766 loss)
I0429 00:39:25.608100 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.790844 (* 0.0272727 = 0.0215685 loss)
I0429 00:39:25.608114 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.334008 (* 0.0272727 = 0.00910932 loss)
I0429 00:39:25.608129 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.163525 (* 0.0272727 = 0.00445977 loss)
I0429 00:39:25.608142 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0389139 (* 0.0272727 = 0.00106129 loss)
I0429 00:39:25.608156 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.01593 (* 0.0272727 = 0.000434454 loss)
I0429 00:39:25.608170 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0111759 (* 0.0272727 = 0.000304798 loss)
I0429 00:39:25.608202 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00618037 (* 0.0272727 = 0.000168555 loss)
I0429 00:39:25.608217 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0050689 (* 0.0272727 = 0.000138243 loss)
I0429 00:39:25.608232 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00315606 (* 0.0272727 = 8.60743e-05 loss)
I0429 00:39:25.608245 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00310892 (* 0.0272727 = 8.47886e-05 loss)
I0429 00:39:25.608259 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00148327 (* 0.0272727 = 4.04529e-05 loss)
I0429 00:39:25.608273 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000710759 (* 0.0272727 = 1.93843e-05 loss)
I0429 00:39:25.608288 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000617565 (* 0.0272727 = 1.68427e-05 loss)
I0429 00:39:25.608301 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000324784 (* 0.0272727 = 8.85774e-06 loss)
I0429 00:39:25.608317 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000304449 (* 0.0272727 = 8.30316e-06 loss)
I0429 00:39:25.608330 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.173913
I0429 00:39:25.608342 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0429 00:39:25.608355 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:39:25.608366 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:39:25.608377 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:39:25.608389 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:39:25.608397 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 00:39:25.608405 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 00:39:25.608417 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:39:25.608429 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:39:25.608440 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:39:25.608453 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:39:25.608464 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:39:25.608474 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:39:25.608486 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:39:25.608497 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:39:25.608508 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:39:25.608520 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:39:25.608531 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:39:25.608542 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:39:25.608553 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:39:25.608566 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:39:25.608577 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:39:25.608587 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.767045
I0429 00:39:25.608599 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.304348
I0429 00:39:25.608613 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.16129 (* 0.3 = 0.948388 loss)
I0429 00:39:25.608626 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.0077 (* 0.3 = 0.302309 loss)
I0429 00:39:25.608640 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.84314 (* 0.0272727 = 0.0775401 loss)
I0429 00:39:25.608654 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.94461 (* 0.0272727 = 0.10758 loss)
I0429 00:39:25.608678 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.18276 (* 0.0272727 = 0.0868026 loss)
I0429 00:39:25.608697 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.06605 (* 0.0272727 = 0.0836195 loss)
I0429 00:39:25.608711 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.34965 (* 0.0272727 = 0.0640813 loss)
I0429 00:39:25.608724 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.32683 (* 0.0272727 = 0.0634589 loss)
I0429 00:39:25.608737 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.27294 (* 0.0272727 = 0.0347165 loss)
I0429 00:39:25.608752 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.03086 (* 0.0272727 = 0.0281142 loss)
I0429 00:39:25.608764 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.39735 (* 0.0272727 = 0.0108368 loss)
I0429 00:39:25.608778 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.267455 (* 0.0272727 = 0.00729423 loss)
I0429 00:39:25.608793 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.111527 (* 0.0272727 = 0.00304165 loss)
I0429 00:39:25.608806 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0737583 (* 0.0272727 = 0.00201159 loss)
I0429 00:39:25.608819 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0538434 (* 0.0272727 = 0.00146846 loss)
I0429 00:39:25.608834 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0412805 (* 0.0272727 = 0.00112583 loss)
I0429 00:39:25.608847 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.028947 (* 0.0272727 = 0.000789462 loss)
I0429 00:39:25.608860 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0184876 (* 0.0272727 = 0.000504208 loss)
I0429 00:39:25.608875 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00662258 (* 0.0272727 = 0.000180616 loss)
I0429 00:39:25.608888 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00677016 (* 0.0272727 = 0.000184641 loss)
I0429 00:39:25.608901 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0028373 (* 0.0272727 = 7.7381e-05 loss)
I0429 00:39:25.608916 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00228826 (* 0.0272727 = 6.24072e-05 loss)
I0429 00:39:25.608929 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00104516 (* 0.0272727 = 2.85042e-05 loss)
I0429 00:39:25.608943 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000598555 (* 0.0272727 = 1.63242e-05 loss)
I0429 00:39:25.608955 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0869565
I0429 00:39:25.608966 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 00:39:25.608978 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:39:25.608989 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:39:25.609001 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:39:25.609014 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:39:25.609025 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:39:25.609036 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 00:39:25.609047 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:39:25.609060 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:39:25.609071 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:39:25.609082 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:39:25.609093 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:39:25.609104 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:39:25.609115 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:39:25.609127 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:39:25.609138 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:39:25.609159 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:39:25.609172 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:39:25.609184 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:39:25.609195 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:39:25.609206 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:39:25.609217 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:39:25.609230 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.732955
I0429 00:39:25.609241 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.326087
I0429 00:39:25.609254 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 3.12576 (* 1 = 3.12576 loss)
I0429 00:39:25.609268 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.0068 (* 1 = 1.0068 loss)
I0429 00:39:25.609282 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.65509 (* 0.0909091 = 0.241372 loss)
I0429 00:39:25.609297 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.94198 (* 0.0909091 = 0.358362 loss)
I0429 00:39:25.609309 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.02185 (* 0.0909091 = 0.274713 loss)
I0429 00:39:25.609323 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.0601 (* 0.0909091 = 0.278191 loss)
I0429 00:39:25.609336 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.34085 (* 0.0909091 = 0.212804 loss)
I0429 00:39:25.609349 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.58459 (* 0.0909091 = 0.234963 loss)
I0429 00:39:25.609365 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.22047 (* 0.0909091 = 0.110952 loss)
I0429 00:39:25.609380 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.00542 (* 0.0909091 = 0.0914019 loss)
I0429 00:39:25.609395 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.444729 (* 0.0909091 = 0.0404299 loss)
I0429 00:39:25.609408 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.292838 (* 0.0909091 = 0.0266217 loss)
I0429 00:39:25.609421 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.113835 (* 0.0909091 = 0.0103486 loss)
I0429 00:39:25.609436 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0625321 (* 0.0909091 = 0.00568473 loss)
I0429 00:39:25.609448 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.02572 (* 0.0909091 = 0.00233818 loss)
I0429 00:39:25.609462 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0125929 (* 0.0909091 = 0.00114481 loss)
I0429 00:39:25.609475 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00889076 (* 0.0909091 = 0.000808251 loss)
I0429 00:39:25.609489 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00634006 (* 0.0909091 = 0.000576369 loss)
I0429 00:39:25.609503 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00640375 (* 0.0909091 = 0.000582159 loss)
I0429 00:39:25.609518 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0058138 (* 0.0909091 = 0.000528527 loss)
I0429 00:39:25.609531 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00531786 (* 0.0909091 = 0.000483441 loss)
I0429 00:39:25.609544 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00572094 (* 0.0909091 = 0.000520086 loss)
I0429 00:39:25.609558 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00666631 (* 0.0909091 = 0.000606028 loss)
I0429 00:39:25.609572 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.006642 (* 0.0909091 = 0.000603818 loss)
I0429 00:39:25.609585 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:39:25.609596 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:39:25.609607 6470 solver.cpp:245] Train net output #149: total_confidence = 2.2332e-05
I0429 00:39:25.609628 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 3.78149e-05
I0429 00:39:25.609642 6470 sgd_solver.cpp:106] Iteration 16500, lr = 0.01
I0429 00:41:42.282546 6470 solver.cpp:229] Iteration 17000, loss = 9.32978
I0429 00:41:42.282716 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.105263
I0429 00:41:42.282737 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.375
I0429 00:41:42.282752 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:41:42.282763 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:41:42.282775 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 00:41:42.282788 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:41:42.282799 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.75
I0429 00:41:42.282811 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 1
I0429 00:41:42.282824 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 00:41:42.282835 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:41:42.282847 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:41:42.282860 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:41:42.282871 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:41:42.282882 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:41:42.282894 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:41:42.282907 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:41:42.282918 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:41:42.282930 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:41:42.282943 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:41:42.282954 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:41:42.282965 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:41:42.282977 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:41:42.282989 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:41:42.283001 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.795455
I0429 00:41:42.283013 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.315789
I0429 00:41:42.283030 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.90545 (* 0.3 = 0.871636 loss)
I0429 00:41:42.283044 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.698573 (* 0.3 = 0.209572 loss)
I0429 00:41:42.283059 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.43341 (* 0.0272727 = 0.0663658 loss)
I0429 00:41:42.283073 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.71065 (* 0.0272727 = 0.0739269 loss)
I0429 00:41:42.283087 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.71558 (* 0.0272727 = 0.101334 loss)
I0429 00:41:42.283102 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.12289 (* 0.0272727 = 0.057897 loss)
I0429 00:41:42.283115 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.31967 (* 0.0272727 = 0.0632637 loss)
I0429 00:41:42.283129 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.33397 (* 0.0272727 = 0.0363809 loss)
I0429 00:41:42.283143 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.200062 (* 0.0272727 = 0.00545623 loss)
I0429 00:41:42.283159 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.0296292 (* 0.0272727 = 0.000808069 loss)
I0429 00:41:42.283172 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.00112 (* 0.0272727 = 3.05455e-05 loss)
I0429 00:41:42.283187 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.00059083 (* 0.0272727 = 1.61135e-05 loss)
I0429 00:41:42.283201 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.000662927 (* 0.0272727 = 1.80798e-05 loss)
I0429 00:41:42.283215 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.000380962 (* 0.0272727 = 1.03899e-05 loss)
I0429 00:41:42.283249 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.000292236 (* 0.0272727 = 7.97008e-06 loss)
I0429 00:41:42.283265 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.000379714 (* 0.0272727 = 1.03558e-05 loss)
I0429 00:41:42.283279 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.000323198 (* 0.0272727 = 8.81448e-06 loss)
I0429 00:41:42.283293 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00044005 (* 0.0272727 = 1.20014e-05 loss)
I0429 00:41:42.283308 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.000500061 (* 0.0272727 = 1.3638e-05 loss)
I0429 00:41:42.283324 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00063265 (* 0.0272727 = 1.72541e-05 loss)
I0429 00:41:42.283339 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000684896 (* 0.0272727 = 1.8679e-05 loss)
I0429 00:41:42.283352 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000720667 (* 0.0272727 = 1.96545e-05 loss)
I0429 00:41:42.283366 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000295957 (* 0.0272727 = 8.07155e-06 loss)
I0429 00:41:42.283380 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000496916 (* 0.0272727 = 1.35522e-05 loss)
I0429 00:41:42.283391 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0789474
I0429 00:41:42.283403 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0429 00:41:42.283416 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:41:42.283427 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:41:42.283439 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.375
I0429 00:41:42.283452 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 00:41:42.283463 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.75
I0429 00:41:42.283490 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 1
I0429 00:41:42.283504 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 00:41:42.283515 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:41:42.283529 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:41:42.283541 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:41:42.283552 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:41:42.283565 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:41:42.283576 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:41:42.283587 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:41:42.283599 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:41:42.283610 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:41:42.283622 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:41:42.283633 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:41:42.283645 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:41:42.283656 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:41:42.283668 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:41:42.283679 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.801136
I0429 00:41:42.283691 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.210526
I0429 00:41:42.283706 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.98897 (* 0.3 = 0.896691 loss)
I0429 00:41:42.283720 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.707373 (* 0.3 = 0.212212 loss)
I0429 00:41:42.283738 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.69084 (* 0.0272727 = 0.0733866 loss)
I0429 00:41:42.283752 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.87877 (* 0.0272727 = 0.0785118 loss)
I0429 00:41:42.283779 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.73282 (* 0.0272727 = 0.101804 loss)
I0429 00:41:42.283793 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.15335 (* 0.0272727 = 0.0587277 loss)
I0429 00:41:42.283807 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.11971 (* 0.0272727 = 0.0578104 loss)
I0429 00:41:42.283821 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.23609 (* 0.0272727 = 0.0337116 loss)
I0429 00:41:42.283835 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.15638 (* 0.0272727 = 0.00426492 loss)
I0429 00:41:42.283849 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.0199915 (* 0.0272727 = 0.000545223 loss)
I0429 00:41:42.283864 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.000209485 (* 0.0272727 = 5.71322e-06 loss)
I0429 00:41:42.283877 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.000119188 (* 0.0272727 = 3.25059e-06 loss)
I0429 00:41:42.283890 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 8.73031e-05 (* 0.0272727 = 2.38099e-06 loss)
I0429 00:41:42.283905 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 8.46309e-05 (* 0.0272727 = 2.30811e-06 loss)
I0429 00:41:42.283918 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.000125811 (* 0.0272727 = 3.4312e-06 loss)
I0429 00:41:42.283932 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 4.44014e-05 (* 0.0272727 = 1.21095e-06 loss)
I0429 00:41:42.283946 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 5.76875e-05 (* 0.0272727 = 1.57329e-06 loss)
I0429 00:41:42.283960 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 6.83132e-05 (* 0.0272727 = 1.86309e-06 loss)
I0429 00:41:42.283973 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 7.42557e-05 (* 0.0272727 = 2.02515e-06 loss)
I0429 00:41:42.283988 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000181751 (* 0.0272727 = 4.95684e-06 loss)
I0429 00:41:42.284003 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000133906 (* 0.0272727 = 3.65199e-06 loss)
I0429 00:41:42.284015 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 7.78411e-05 (* 0.0272727 = 2.12294e-06 loss)
I0429 00:41:42.284029 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000151375 (* 0.0272727 = 4.12842e-06 loss)
I0429 00:41:42.284044 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000118597 (* 0.0272727 = 3.23447e-06 loss)
I0429 00:41:42.284056 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.131579
I0429 00:41:42.284067 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.5
I0429 00:41:42.284080 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.25
I0429 00:41:42.284091 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:41:42.284103 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.5
I0429 00:41:42.284116 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0429 00:41:42.284127 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.75
I0429 00:41:42.284138 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 1
I0429 00:41:42.284150 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0429 00:41:42.284162 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:41:42.284173 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:41:42.284184 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:41:42.284196 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:41:42.284207 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:41:42.284219 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:41:42.284230 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:41:42.284251 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:41:42.284265 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:41:42.284276 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:41:42.284287 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:41:42.284299 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:41:42.284310 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:41:42.284322 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:41:42.284334 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.801136
I0429 00:41:42.284346 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.289474
I0429 00:41:42.284359 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.91236 (* 1 = 2.91236 loss)
I0429 00:41:42.284376 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.701799 (* 1 = 0.701799 loss)
I0429 00:41:42.284391 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.11162 (* 0.0909091 = 0.191965 loss)
I0429 00:41:42.284404 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.61014 (* 0.0909091 = 0.237286 loss)
I0429 00:41:42.284418 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.81248 (* 0.0909091 = 0.346589 loss)
I0429 00:41:42.284432 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.00207 (* 0.0909091 = 0.182006 loss)
I0429 00:41:42.284446 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.21046 (* 0.0909091 = 0.200951 loss)
I0429 00:41:42.284459 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.12753 (* 0.0909091 = 0.102503 loss)
I0429 00:41:42.284472 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.156413 (* 0.0909091 = 0.0142194 loss)
I0429 00:41:42.284487 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.0136214 (* 0.0909091 = 0.00123831 loss)
I0429 00:41:42.284500 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.000338688 (* 0.0909091 = 3.07898e-05 loss)
I0429 00:41:42.284514 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 5.73533e-05 (* 0.0909091 = 5.21393e-06 loss)
I0429 00:41:42.284528 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 2.29336e-05 (* 0.0909091 = 2.08487e-06 loss)
I0429 00:41:42.284543 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 1.40968e-05 (* 0.0909091 = 1.28152e-06 loss)
I0429 00:41:42.284557 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 1.08184e-05 (* 0.0909091 = 9.8349e-07 loss)
I0429 00:41:42.284570 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 7.28674e-06 (* 0.0909091 = 6.62431e-07 loss)
I0429 00:41:42.284584 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 7.58478e-06 (* 0.0909091 = 6.89526e-07 loss)
I0429 00:41:42.284598 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 9.98397e-06 (* 0.0909091 = 9.07634e-07 loss)
I0429 00:41:42.284612 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 5.85624e-06 (* 0.0909091 = 5.32385e-07 loss)
I0429 00:41:42.284626 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 6.97387e-06 (* 0.0909091 = 6.33989e-07 loss)
I0429 00:41:42.284641 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 1.00586e-05 (* 0.0909091 = 9.14417e-07 loss)
I0429 00:41:42.284654 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 1.29198e-05 (* 0.0909091 = 1.17453e-06 loss)
I0429 00:41:42.284668 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 2.01476e-05 (* 0.0909091 = 1.8316e-06 loss)
I0429 00:41:42.284682 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 1.97601e-05 (* 0.0909091 = 1.79638e-06 loss)
I0429 00:41:42.284694 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:41:42.284705 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:41:42.284725 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00233839
I0429 00:41:42.284739 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.0130168
I0429 00:41:42.284751 6470 sgd_solver.cpp:106] Iteration 17000, lr = 0.01
I0429 00:43:58.940855 6470 solver.cpp:229] Iteration 17500, loss = 9.21333
I0429 00:43:58.940996 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535
I0429 00:43:58.941025 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.375
I0429 00:43:58.941047 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:43:58.941068 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:43:58.941090 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:43:58.941112 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:43:58.941134 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:43:58.941155 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:43:58.941177 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 00:43:58.941200 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:43:58.941221 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 00:43:58.941241 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:43:58.941263 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:43:58.941285 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:43:58.941310 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:43:58.941332 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:43:58.941355 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:43:58.941377 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:43:58.941400 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:43:58.941421 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:43:58.941444 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:43:58.941471 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:43:58.941494 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:43:58.941516 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.789773
I0429 00:43:58.941537 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.372093
I0429 00:43:58.941566 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.52026 (* 0.3 = 0.756077 loss)
I0429 00:43:58.941597 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.637124 (* 0.3 = 0.191137 loss)
I0429 00:43:58.941624 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.09182 (* 0.0272727 = 0.0570497 loss)
I0429 00:43:58.941651 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.42074 (* 0.0272727 = 0.0660203 loss)
I0429 00:43:58.941678 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.7066 (* 0.0272727 = 0.0738162 loss)
I0429 00:43:58.941705 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.55353 (* 0.0272727 = 0.0696416 loss)
I0429 00:43:58.941731 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.00969 (* 0.0272727 = 0.0548097 loss)
I0429 00:43:58.941758 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.64037 (* 0.0272727 = 0.0447372 loss)
I0429 00:43:58.941787 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.25803 (* 0.0272727 = 0.0343099 loss)
I0429 00:43:58.941813 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.0757812 (* 0.0272727 = 0.00206676 loss)
I0429 00:43:58.941840 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0091898 (* 0.0272727 = 0.000250631 loss)
I0429 00:43:58.941867 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.00374807 (* 0.0272727 = 0.00010222 loss)
I0429 00:43:58.941895 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.0024363 (* 0.0272727 = 6.64446e-05 loss)
I0429 00:43:58.941922 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.00131126 (* 0.0272727 = 3.57616e-05 loss)
I0429 00:43:58.941948 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.00128966 (* 0.0272727 = 3.51725e-05 loss)
I0429 00:43:58.942003 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.000755791 (* 0.0272727 = 2.06125e-05 loss)
I0429 00:43:58.942033 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00123429 (* 0.0272727 = 3.36623e-05 loss)
I0429 00:43:58.942060 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00136703 (* 0.0272727 = 3.72826e-05 loss)
I0429 00:43:58.942090 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.000717206 (* 0.0272727 = 1.95602e-05 loss)
I0429 00:43:58.942117 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000919258 (* 0.0272727 = 2.50707e-05 loss)
I0429 00:43:58.942147 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000412334 (* 0.0272727 = 1.12455e-05 loss)
I0429 00:43:58.942176 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000386545 (* 0.0272727 = 1.05421e-05 loss)
I0429 00:43:58.942204 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000261037 (* 0.0272727 = 7.11919e-06 loss)
I0429 00:43:58.942231 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000713447 (* 0.0272727 = 1.94577e-05 loss)
I0429 00:43:58.942255 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.209302
I0429 00:43:58.942277 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:43:58.942301 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.5
I0429 00:43:58.942323 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:43:58.942347 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:43:58.942369 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:43:58.942392 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:43:58.942415 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:43:58.942437 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 00:43:58.942459 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 00:43:58.942481 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 00:43:58.942503 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:43:58.942531 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:43:58.942554 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:43:58.942576 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:43:58.942598 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:43:58.942622 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:43:58.942646 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:43:58.942668 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:43:58.942690 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:43:58.942713 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:43:58.942734 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:43:58.942756 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:43:58.942778 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.806818
I0429 00:43:58.942800 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.348837
I0429 00:43:58.942828 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.47682 (* 0.3 = 0.743047 loss)
I0429 00:43:58.942854 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.643925 (* 0.3 = 0.193178 loss)
I0429 00:43:58.942881 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.08859 (* 0.0272727 = 0.0569616 loss)
I0429 00:43:58.942909 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.08478 (* 0.0272727 = 0.0568578 loss)
I0429 00:43:58.942951 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.53776 (* 0.0272727 = 0.0692116 loss)
I0429 00:43:58.942978 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.49324 (* 0.0272727 = 0.0679973 loss)
I0429 00:43:58.943004 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.55177 (* 0.0272727 = 0.0695937 loss)
I0429 00:43:58.943032 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.43386 (* 0.0272727 = 0.0391053 loss)
I0429 00:43:58.943058 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.17316 (* 0.0272727 = 0.0319952 loss)
I0429 00:43:58.943084 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.153281 (* 0.0272727 = 0.00418038 loss)
I0429 00:43:58.943110 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.00829867 (* 0.0272727 = 0.000226327 loss)
I0429 00:43:58.943137 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.00378012 (* 0.0272727 = 0.000103094 loss)
I0429 00:43:58.943164 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.00345198 (* 0.0272727 = 9.4145e-05 loss)
I0429 00:43:58.943191 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00425644 (* 0.0272727 = 0.000116085 loss)
I0429 00:43:58.943217 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0017216 (* 0.0272727 = 4.69528e-05 loss)
I0429 00:43:58.943243 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00130495 (* 0.0272727 = 3.55895e-05 loss)
I0429 00:43:58.943270 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.000908621 (* 0.0272727 = 2.47806e-05 loss)
I0429 00:43:58.943295 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.000666811 (* 0.0272727 = 1.81858e-05 loss)
I0429 00:43:58.943321 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.000521949 (* 0.0272727 = 1.4235e-05 loss)
I0429 00:43:58.943349 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00026107 (* 0.0272727 = 7.12008e-06 loss)
I0429 00:43:58.943375 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000270404 (* 0.0272727 = 7.37466e-06 loss)
I0429 00:43:58.943402 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000215035 (* 0.0272727 = 5.86459e-06 loss)
I0429 00:43:58.943428 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 9.8398e-05 (* 0.0272727 = 2.68358e-06 loss)
I0429 00:43:58.943456 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000139628 (* 0.0272727 = 3.80805e-06 loss)
I0429 00:43:58.943495 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.162791
I0429 00:43:58.943518 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0429 00:43:58.943542 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:43:58.943563 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:43:58.943590 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 00:43:58.943614 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 00:43:58.943635 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:43:58.943656 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:43:58.943682 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0429 00:43:58.943706 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:43:58.943727 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 00:43:58.943747 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:43:58.943768 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:43:58.943789 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:43:58.943810 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:43:58.943831 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:43:58.943852 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:43:58.943889 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:43:58.943912 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:43:58.943934 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:43:58.943955 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:43:58.943975 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:43:58.943997 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:43:58.944020 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.795455
I0429 00:43:58.944041 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.488372
I0429 00:43:58.944062 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.38121 (* 1 = 2.38121 loss)
I0429 00:43:58.944087 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.614391 (* 1 = 0.614391 loss)
I0429 00:43:58.944113 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 1.78913 (* 0.0909091 = 0.162648 loss)
I0429 00:43:58.944140 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.25507 (* 0.0909091 = 0.205006 loss)
I0429 00:43:58.944167 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.51031 (* 0.0909091 = 0.22821 loss)
I0429 00:43:58.944193 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.38467 (* 0.0909091 = 0.216788 loss)
I0429 00:43:58.944218 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.03771 (* 0.0909091 = 0.185246 loss)
I0429 00:43:58.944247 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.36632 (* 0.0909091 = 0.124211 loss)
I0429 00:43:58.944272 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.32513 (* 0.0909091 = 0.120466 loss)
I0429 00:43:58.944298 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.0768993 (* 0.0909091 = 0.00699085 loss)
I0429 00:43:58.944325 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.00652558 (* 0.0909091 = 0.000593235 loss)
I0429 00:43:58.944351 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.00308104 (* 0.0909091 = 0.000280095 loss)
I0429 00:43:58.944377 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00258136 (* 0.0909091 = 0.00023467 loss)
I0429 00:43:58.944403 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.00149261 (* 0.0909091 = 0.000135692 loss)
I0429 00:43:58.944429 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.00130918 (* 0.0909091 = 0.000119016 loss)
I0429 00:43:58.944455 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.00103266 (* 0.0909091 = 9.38778e-05 loss)
I0429 00:43:58.944481 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000520652 (* 0.0909091 = 4.7332e-05 loss)
I0429 00:43:58.944509 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00031178 (* 0.0909091 = 2.83436e-05 loss)
I0429 00:43:58.944535 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000149486 (* 0.0909091 = 1.35897e-05 loss)
I0429 00:43:58.944563 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 9.02308e-05 (* 0.0909091 = 8.2028e-06 loss)
I0429 00:43:58.944593 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 7.46774e-05 (* 0.0909091 = 6.78885e-06 loss)
I0429 00:43:58.944615 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 5.66511e-05 (* 0.0909091 = 5.1501e-06 loss)
I0429 00:43:58.944653 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 5.15539e-05 (* 0.0909091 = 4.68672e-06 loss)
I0429 00:43:58.944680 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 3.50786e-05 (* 0.0909091 = 3.18897e-06 loss)
I0429 00:43:58.944702 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:43:58.944726 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:43:58.944766 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00270834
I0429 00:43:58.944790 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00373966
I0429 00:43:58.944813 6470 sgd_solver.cpp:106] Iteration 17500, lr = 0.01
I0429 00:46:15.580801 6470 solver.cpp:229] Iteration 18000, loss = 9.32899
I0429 00:46:15.580977 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.115385
I0429 00:46:15.580998 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:46:15.581012 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.375
I0429 00:46:15.581023 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:46:15.581035 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 00:46:15.581048 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.25
I0429 00:46:15.581059 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:46:15.581071 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0429 00:46:15.581084 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 00:46:15.581095 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:46:15.581107 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:46:15.581120 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:46:15.581130 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:46:15.581142 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:46:15.581154 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:46:15.581166 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:46:15.581178 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:46:15.581190 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:46:15.581202 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:46:15.581213 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:46:15.581225 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:46:15.581238 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:46:15.581249 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:46:15.581260 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.732955
I0429 00:46:15.581272 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.326923
I0429 00:46:15.581289 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.99131 (* 0.3 = 0.897393 loss)
I0429 00:46:15.581303 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.02801 (* 0.3 = 0.308402 loss)
I0429 00:46:15.581321 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.32633 (* 0.0272727 = 0.0634453 loss)
I0429 00:46:15.581334 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.43699 (* 0.0272727 = 0.0664633 loss)
I0429 00:46:15.581348 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.18947 (* 0.0272727 = 0.0869856 loss)
I0429 00:46:15.581363 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.77493 (* 0.0272727 = 0.0756798 loss)
I0429 00:46:15.581377 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.36261 (* 0.0272727 = 0.0644349 loss)
I0429 00:46:15.581392 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.05377 (* 0.0272727 = 0.0560119 loss)
I0429 00:46:15.581405 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.38332 (* 0.0272727 = 0.0649996 loss)
I0429 00:46:15.581419 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.27431 (* 0.0272727 = 0.00748117 loss)
I0429 00:46:15.581434 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.527889 (* 0.0272727 = 0.014397 loss)
I0429 00:46:15.581449 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.393373 (* 0.0272727 = 0.0107284 loss)
I0429 00:46:15.581461 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.450992 (* 0.0272727 = 0.0122998 loss)
I0429 00:46:15.581476 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.328105 (* 0.0272727 = 0.00894832 loss)
I0429 00:46:15.581509 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.322138 (* 0.0272727 = 0.00878559 loss)
I0429 00:46:15.581524 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.176779 (* 0.0272727 = 0.00482125 loss)
I0429 00:46:15.581538 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0714972 (* 0.0272727 = 0.00194992 loss)
I0429 00:46:15.581553 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0319861 (* 0.0272727 = 0.000872347 loss)
I0429 00:46:15.581568 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00697574 (* 0.0272727 = 0.000190247 loss)
I0429 00:46:15.581581 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00218761 (* 0.0272727 = 5.96621e-05 loss)
I0429 00:46:15.581595 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000872953 (* 0.0272727 = 2.38078e-05 loss)
I0429 00:46:15.581609 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000669078 (* 0.0272727 = 1.82476e-05 loss)
I0429 00:46:15.581624 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000530888 (* 0.0272727 = 1.44788e-05 loss)
I0429 00:46:15.581637 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000223708 (* 0.0272727 = 6.10113e-06 loss)
I0429 00:46:15.581650 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.153846
I0429 00:46:15.581661 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:46:15.581673 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.25
I0429 00:46:15.581686 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:46:15.581696 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:46:15.581707 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 00:46:15.581719 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:46:15.581732 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0429 00:46:15.581743 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 00:46:15.581754 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:46:15.581768 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:46:15.581775 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:46:15.581784 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:46:15.581795 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:46:15.581807 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:46:15.581818 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:46:15.581830 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:46:15.581841 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:46:15.581852 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:46:15.581864 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:46:15.581876 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:46:15.581887 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:46:15.581897 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:46:15.581909 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.732955
I0429 00:46:15.581920 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.346154
I0429 00:46:15.581934 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.90361 (* 0.3 = 0.871082 loss)
I0429 00:46:15.581948 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.997686 (* 0.3 = 0.299306 loss)
I0429 00:46:15.581961 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.19796 (* 0.0272727 = 0.0599445 loss)
I0429 00:46:15.581975 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.47779 (* 0.0272727 = 0.0675762 loss)
I0429 00:46:15.582002 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.0799 (* 0.0272727 = 0.0839972 loss)
I0429 00:46:15.582018 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.96105 (* 0.0272727 = 0.0807558 loss)
I0429 00:46:15.582031 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.33175 (* 0.0272727 = 0.0635932 loss)
I0429 00:46:15.582046 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.02983 (* 0.0272727 = 0.0553591 loss)
I0429 00:46:15.582059 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.89352 (* 0.0272727 = 0.0789142 loss)
I0429 00:46:15.582073 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.326582 (* 0.0272727 = 0.00890677 loss)
I0429 00:46:15.582087 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.514992 (* 0.0272727 = 0.0140452 loss)
I0429 00:46:15.582101 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.360869 (* 0.0272727 = 0.00984187 loss)
I0429 00:46:15.582115 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.327291 (* 0.0272727 = 0.00892612 loss)
I0429 00:46:15.582129 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.366608 (* 0.0272727 = 0.0099984 loss)
I0429 00:46:15.582144 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.195636 (* 0.0272727 = 0.00533552 loss)
I0429 00:46:15.582157 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0903993 (* 0.0272727 = 0.00246544 loss)
I0429 00:46:15.582170 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0825505 (* 0.0272727 = 0.00225138 loss)
I0429 00:46:15.582185 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0160525 (* 0.0272727 = 0.000437796 loss)
I0429 00:46:15.582198 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00871146 (* 0.0272727 = 0.000237585 loss)
I0429 00:46:15.582212 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00243233 (* 0.0272727 = 6.63364e-05 loss)
I0429 00:46:15.582226 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000766159 (* 0.0272727 = 2.08953e-05 loss)
I0429 00:46:15.582239 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00043001 (* 0.0272727 = 1.17275e-05 loss)
I0429 00:46:15.582253 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000148978 (* 0.0272727 = 4.06305e-06 loss)
I0429 00:46:15.582267 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 9.1741e-05 (* 0.0272727 = 2.50203e-06 loss)
I0429 00:46:15.582279 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.192308
I0429 00:46:15.582291 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0429 00:46:15.582304 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.25
I0429 00:46:15.582315 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 00:46:15.582327 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 00:46:15.582339 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:46:15.582350 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.625
I0429 00:46:15.582362 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.5
I0429 00:46:15.582376 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0429 00:46:15.582388 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:46:15.582401 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:46:15.582412 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:46:15.582423 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:46:15.582435 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:46:15.582448 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:46:15.582458 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:46:15.582469 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:46:15.582490 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:46:15.582504 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:46:15.582515 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:46:15.582526 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:46:15.582537 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:46:15.582550 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:46:15.582561 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0429 00:46:15.582573 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.480769
I0429 00:46:15.582587 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.78616 (* 1 = 2.78616 loss)
I0429 00:46:15.582600 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.964858 (* 1 = 0.964858 loss)
I0429 00:46:15.582614 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 1.697 (* 0.0909091 = 0.154273 loss)
I0429 00:46:15.582628 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.53284 (* 0.0909091 = 0.230258 loss)
I0429 00:46:15.582643 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.23325 (* 0.0909091 = 0.293932 loss)
I0429 00:46:15.582655 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.08956 (* 0.0909091 = 0.280869 loss)
I0429 00:46:15.582669 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.28803 (* 0.0909091 = 0.208002 loss)
I0429 00:46:15.582682 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.83662 (* 0.0909091 = 0.166966 loss)
I0429 00:46:15.582695 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.92914 (* 0.0909091 = 0.266285 loss)
I0429 00:46:15.582710 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.42969 (* 0.0909091 = 0.0390627 loss)
I0429 00:46:15.582723 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.585267 (* 0.0909091 = 0.0532061 loss)
I0429 00:46:15.582736 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.396021 (* 0.0909091 = 0.0360019 loss)
I0429 00:46:15.582751 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.441858 (* 0.0909091 = 0.040169 loss)
I0429 00:46:15.582764 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.379451 (* 0.0909091 = 0.0344955 loss)
I0429 00:46:15.582777 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.126933 (* 0.0909091 = 0.0115394 loss)
I0429 00:46:15.582792 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.094257 (* 0.0909091 = 0.00856882 loss)
I0429 00:46:15.582805 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0678582 (* 0.0909091 = 0.00616893 loss)
I0429 00:46:15.582818 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0346365 (* 0.0909091 = 0.00314878 loss)
I0429 00:46:15.582833 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0181627 (* 0.0909091 = 0.00165115 loss)
I0429 00:46:15.582846 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0104919 (* 0.0909091 = 0.000953808 loss)
I0429 00:46:15.582860 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0053802 (* 0.0909091 = 0.00048911 loss)
I0429 00:46:15.582875 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.0026434 (* 0.0909091 = 0.000240309 loss)
I0429 00:46:15.582888 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00168321 (* 0.0909091 = 0.000153019 loss)
I0429 00:46:15.582902 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000620515 (* 0.0909091 = 5.64104e-05 loss)
I0429 00:46:15.582914 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:46:15.582926 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:46:15.582937 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00226456
I0429 00:46:15.582957 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00424943
I0429 00:46:15.582973 6470 sgd_solver.cpp:106] Iteration 18000, lr = 0.01
I0429 00:48:32.216272 6470 solver.cpp:229] Iteration 18500, loss = 9.17274
I0429 00:48:32.216439 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.208333
I0429 00:48:32.216467 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:48:32.216490 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.25
I0429 00:48:32.216513 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:48:32.216534 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0
I0429 00:48:32.216557 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 00:48:32.216578 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 00:48:32.216601 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:48:32.216624 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:48:32.216645 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:48:32.216665 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:48:32.216687 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:48:32.216709 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:48:32.216734 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:48:32.216758 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:48:32.216779 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:48:32.216802 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:48:32.216823 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:48:32.216845 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:48:32.216866 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:48:32.216888 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:48:32.216909 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:48:32.216931 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:48:32.216953 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.744318
I0429 00:48:32.216975 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.354167
I0429 00:48:32.217005 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.97562 (* 0.3 = 0.892686 loss)
I0429 00:48:32.217031 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.20573 (* 0.3 = 0.36172 loss)
I0429 00:48:32.217059 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.9692 (* 0.0272727 = 0.108251 loss)
I0429 00:48:32.217085 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.50507 (* 0.0272727 = 0.0955927 loss)
I0429 00:48:32.217113 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.90936 (* 0.0272727 = 0.106619 loss)
I0429 00:48:32.217140 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 4.06083 (* 0.0272727 = 0.11075 loss)
I0429 00:48:32.217166 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 3.19892 (* 0.0272727 = 0.0872432 loss)
I0429 00:48:32.217195 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.38839 (* 0.0272727 = 0.065138 loss)
I0429 00:48:32.217219 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.80934 (* 0.0272727 = 0.0493456 loss)
I0429 00:48:32.217247 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.03562 (* 0.0272727 = 0.0282441 loss)
I0429 00:48:32.217273 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.412484 (* 0.0272727 = 0.0112496 loss)
I0429 00:48:32.217299 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.447324 (* 0.0272727 = 0.0121998 loss)
I0429 00:48:32.217330 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.423437 (* 0.0272727 = 0.0115483 loss)
I0429 00:48:32.217360 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.192315 (* 0.0272727 = 0.00524496 loss)
I0429 00:48:32.217387 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.134033 (* 0.0272727 = 0.00365546 loss)
I0429 00:48:32.217448 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0867568 (* 0.0272727 = 0.00236609 loss)
I0429 00:48:32.217479 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0512126 (* 0.0272727 = 0.00139671 loss)
I0429 00:48:32.217509 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.0289076 (* 0.0272727 = 0.000788389 loss)
I0429 00:48:32.217540 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0176735 (* 0.0272727 = 0.000482004 loss)
I0429 00:48:32.217569 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.0149683 (* 0.0272727 = 0.000408226 loss)
I0429 00:48:32.217597 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00775737 (* 0.0272727 = 0.000211565 loss)
I0429 00:48:32.217624 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.00487629 (* 0.0272727 = 0.00013299 loss)
I0429 00:48:32.217653 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.00156838 (* 0.0272727 = 4.2774e-05 loss)
I0429 00:48:32.217680 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00189167 (* 0.0272727 = 5.15909e-05 loss)
I0429 00:48:32.217705 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.145833
I0429 00:48:32.217726 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:48:32.217748 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:48:32.217770 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:48:32.217793 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:48:32.217815 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.125
I0429 00:48:32.217839 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:48:32.217860 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 00:48:32.217882 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:48:32.217905 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:48:32.217928 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:48:32.217952 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:48:32.217974 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:48:32.217996 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:48:32.218019 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:48:32.218040 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:48:32.218061 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:48:32.218082 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:48:32.218103 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:48:32.218125 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:48:32.218147 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:48:32.218168 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:48:32.218190 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:48:32.218212 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.732955
I0429 00:48:32.218235 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.25
I0429 00:48:32.218262 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.03516 (* 0.3 = 0.910549 loss)
I0429 00:48:32.218289 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.16999 (* 0.3 = 0.350998 loss)
I0429 00:48:32.218315 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 3.94795 (* 0.0272727 = 0.107671 loss)
I0429 00:48:32.218343 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 4.29777 (* 0.0272727 = 0.117212 loss)
I0429 00:48:32.218391 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.75446 (* 0.0272727 = 0.102394 loss)
I0429 00:48:32.218420 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.9947 (* 0.0272727 = 0.108946 loss)
I0429 00:48:32.218446 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.83037 (* 0.0272727 = 0.0771918 loss)
I0429 00:48:32.218472 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.29562 (* 0.0272727 = 0.0626079 loss)
I0429 00:48:32.218507 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.75318 (* 0.0272727 = 0.047814 loss)
I0429 00:48:32.218533 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.988884 (* 0.0272727 = 0.0269696 loss)
I0429 00:48:32.218560 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.395823 (* 0.0272727 = 0.0107952 loss)
I0429 00:48:32.218587 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.394645 (* 0.0272727 = 0.0107631 loss)
I0429 00:48:32.218613 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.456782 (* 0.0272727 = 0.0124577 loss)
I0429 00:48:32.218641 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.119703 (* 0.0272727 = 0.00326462 loss)
I0429 00:48:32.218667 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0569889 (* 0.0272727 = 0.00155424 loss)
I0429 00:48:32.218693 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0413795 (* 0.0272727 = 0.00112853 loss)
I0429 00:48:32.218720 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0238056 (* 0.0272727 = 0.000649244 loss)
I0429 00:48:32.218746 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0176025 (* 0.0272727 = 0.000480069 loss)
I0429 00:48:32.218772 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0242037 (* 0.0272727 = 0.000660101 loss)
I0429 00:48:32.218798 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0122932 (* 0.0272727 = 0.000335269 loss)
I0429 00:48:32.218825 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00409509 (* 0.0272727 = 0.000111684 loss)
I0429 00:48:32.218852 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00363207 (* 0.0272727 = 9.90566e-05 loss)
I0429 00:48:32.218879 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00124053 (* 0.0272727 = 3.38327e-05 loss)
I0429 00:48:32.218906 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000779418 (* 0.0272727 = 2.12569e-05 loss)
I0429 00:48:32.218930 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.166667
I0429 00:48:32.218951 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:48:32.218974 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:48:32.218996 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:48:32.219017 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0
I0429 00:48:32.219038 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 00:48:32.219059 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 00:48:32.219080 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:48:32.219101 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:48:32.219125 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:48:32.219144 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:48:32.219166 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:48:32.219188 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:48:32.219210 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:48:32.219230 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:48:32.219250 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:48:32.219272 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:48:32.219310 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:48:32.219332 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:48:32.219354 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:48:32.219377 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:48:32.219396 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:48:32.219434 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:48:32.219462 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.727273
I0429 00:48:32.219485 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.375
I0429 00:48:32.219511 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.98641 (* 1 = 2.98641 loss)
I0429 00:48:32.219537 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.20519 (* 1 = 1.20519 loss)
I0429 00:48:32.219571 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 3.70449 (* 0.0909091 = 0.336772 loss)
I0429 00:48:32.219599 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 3.75891 (* 0.0909091 = 0.341719 loss)
I0429 00:48:32.219625 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.66323 (* 0.0909091 = 0.333021 loss)
I0429 00:48:32.219651 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 4.01993 (* 0.0909091 = 0.365448 loss)
I0429 00:48:32.219677 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.12597 (* 0.0909091 = 0.284179 loss)
I0429 00:48:32.219705 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.50303 (* 0.0909091 = 0.227548 loss)
I0429 00:48:32.219730 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.77516 (* 0.0909091 = 0.161378 loss)
I0429 00:48:32.219758 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.01145 (* 0.0909091 = 0.0919496 loss)
I0429 00:48:32.219785 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.452877 (* 0.0909091 = 0.0411707 loss)
I0429 00:48:32.219806 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.398111 (* 0.0909091 = 0.0361919 loss)
I0429 00:48:32.219835 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.431017 (* 0.0909091 = 0.0391833 loss)
I0429 00:48:32.219863 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0796118 (* 0.0909091 = 0.00723744 loss)
I0429 00:48:32.219890 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.0329716 (* 0.0909091 = 0.00299742 loss)
I0429 00:48:32.219916 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0114275 (* 0.0909091 = 0.00103886 loss)
I0429 00:48:32.219944 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00652965 (* 0.0909091 = 0.000593604 loss)
I0429 00:48:32.219970 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00469144 (* 0.0909091 = 0.000426495 loss)
I0429 00:48:32.219995 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00348512 (* 0.0909091 = 0.000316829 loss)
I0429 00:48:32.220023 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00250873 (* 0.0909091 = 0.000228067 loss)
I0429 00:48:32.220048 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00184739 (* 0.0909091 = 0.000167945 loss)
I0429 00:48:32.220074 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00136835 (* 0.0909091 = 0.000124395 loss)
I0429 00:48:32.220101 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00118138 (* 0.0909091 = 0.000107398 loss)
I0429 00:48:32.220127 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000891727 (* 0.0909091 = 8.10661e-05 loss)
I0429 00:48:32.220149 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:48:32.220170 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:48:32.220191 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000219542
I0429 00:48:32.220229 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000375935
I0429 00:48:32.220254 6470 sgd_solver.cpp:106] Iteration 18500, lr = 0.01
I0429 00:50:48.816560 6470 solver.cpp:229] Iteration 19000, loss = 9.19409
I0429 00:50:48.816725 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.148936
I0429 00:50:48.816752 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:50:48.816776 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 00:50:48.816797 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.25
I0429 00:50:48.816819 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 00:50:48.816840 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0429 00:50:48.816864 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:50:48.816885 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 00:50:48.816907 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0429 00:50:48.816929 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:50:48.816951 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:50:48.816972 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:50:48.816993 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:50:48.817018 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:50:48.817041 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:50:48.817065 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:50:48.817087 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:50:48.817108 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:50:48.817129 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:50:48.817152 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:50:48.817173 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:50:48.817194 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:50:48.817217 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:50:48.817239 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.761364
I0429 00:50:48.817260 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.297872
I0429 00:50:48.817289 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.0702 (* 0.3 = 0.921059 loss)
I0429 00:50:48.817320 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.966725 (* 0.3 = 0.290018 loss)
I0429 00:50:48.817349 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 3.10817 (* 0.0272727 = 0.0847682 loss)
I0429 00:50:48.817378 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.50119 (* 0.0272727 = 0.095487 loss)
I0429 00:50:48.817404 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.95198 (* 0.0272727 = 0.0805086 loss)
I0429 00:50:48.817430 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.38654 (* 0.0272727 = 0.0650875 loss)
I0429 00:50:48.817456 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.34618 (* 0.0272727 = 0.0639867 loss)
I0429 00:50:48.817483 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.81285 (* 0.0272727 = 0.0494413 loss)
I0429 00:50:48.817510 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.72073 (* 0.0272727 = 0.0469289 loss)
I0429 00:50:48.817536 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.84486 (* 0.0272727 = 0.0503144 loss)
I0429 00:50:48.817564 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.442196 (* 0.0272727 = 0.0120599 loss)
I0429 00:50:48.817590 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.394362 (* 0.0272727 = 0.0107553 loss)
I0429 00:50:48.817616 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.465305 (* 0.0272727 = 0.0126901 loss)
I0429 00:50:48.817644 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.07033 (* 0.0272727 = 0.00191809 loss)
I0429 00:50:48.817672 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0365304 (* 0.0272727 = 0.000996284 loss)
I0429 00:50:48.817724 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0118845 (* 0.0272727 = 0.000324123 loss)
I0429 00:50:48.817759 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0055845 (* 0.0272727 = 0.000152305 loss)
I0429 00:50:48.817788 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00321147 (* 0.0272727 = 8.75857e-05 loss)
I0429 00:50:48.817818 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0024056 (* 0.0272727 = 6.56072e-05 loss)
I0429 00:50:48.817848 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00067309 (* 0.0272727 = 1.8357e-05 loss)
I0429 00:50:48.817878 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000258125 (* 0.0272727 = 7.03977e-06 loss)
I0429 00:50:48.817905 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 8.0979e-05 (* 0.0272727 = 2.20852e-06 loss)
I0429 00:50:48.817932 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 2.93564e-05 (* 0.0272727 = 8.0063e-07 loss)
I0429 00:50:48.817960 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 2.76873e-05 (* 0.0272727 = 7.55109e-07 loss)
I0429 00:50:48.817983 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.148936
I0429 00:50:48.818007 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:50:48.818030 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.25
I0429 00:50:48.818053 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.25
I0429 00:50:48.818075 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.5
I0429 00:50:48.818097 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 00:50:48.818120 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:50:48.818143 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.5
I0429 00:50:48.818166 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 00:50:48.818189 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:50:48.818212 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:50:48.818233 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:50:48.818256 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:50:48.818279 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:50:48.818300 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:50:48.818322 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:50:48.818343 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:50:48.818368 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:50:48.818392 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:50:48.818413 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:50:48.818436 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:50:48.818459 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:50:48.818480 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:50:48.818503 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.755682
I0429 00:50:48.818524 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.319149
I0429 00:50:48.818552 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.91384 (* 0.3 = 0.874152 loss)
I0429 00:50:48.818578 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.968651 (* 0.3 = 0.290595 loss)
I0429 00:50:48.818605 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.90111 (* 0.0272727 = 0.0791212 loss)
I0429 00:50:48.818634 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.01658 (* 0.0272727 = 0.0822703 loss)
I0429 00:50:48.818676 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.76758 (* 0.0272727 = 0.0754796 loss)
I0429 00:50:48.818706 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.50271 (* 0.0272727 = 0.0682558 loss)
I0429 00:50:48.818732 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.14465 (* 0.0272727 = 0.0584904 loss)
I0429 00:50:48.818758 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.02497 (* 0.0272727 = 0.0552263 loss)
I0429 00:50:48.818789 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.95865 (* 0.0272727 = 0.0534178 loss)
I0429 00:50:48.818815 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.96001 (* 0.0272727 = 0.0534547 loss)
I0429 00:50:48.818841 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.441652 (* 0.0272727 = 0.012045 loss)
I0429 00:50:48.818869 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.359293 (* 0.0272727 = 0.00979891 loss)
I0429 00:50:48.818897 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.341213 (* 0.0272727 = 0.0093058 loss)
I0429 00:50:48.818922 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.1094 (* 0.0272727 = 0.00298362 loss)
I0429 00:50:48.818948 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0748807 (* 0.0272727 = 0.0020422 loss)
I0429 00:50:48.818975 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0352624 (* 0.0272727 = 0.000961701 loss)
I0429 00:50:48.819001 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0131982 (* 0.0272727 = 0.000359952 loss)
I0429 00:50:48.819028 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00669038 (* 0.0272727 = 0.000182465 loss)
I0429 00:50:48.819054 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00557964 (* 0.0272727 = 0.000152172 loss)
I0429 00:50:48.819079 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0020125 (* 0.0272727 = 5.48865e-05 loss)
I0429 00:50:48.819106 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000845951 (* 0.0272727 = 2.30714e-05 loss)
I0429 00:50:48.819133 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000585783 (* 0.0272727 = 1.59759e-05 loss)
I0429 00:50:48.819159 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000245865 (* 0.0272727 = 6.7054e-06 loss)
I0429 00:50:48.819185 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000132675 (* 0.0272727 = 3.61842e-06 loss)
I0429 00:50:48.819208 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.191489
I0429 00:50:48.819229 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:50:48.819252 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.25
I0429 00:50:48.819274 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.375
I0429 00:50:48.819295 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.5
I0429 00:50:48.819316 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 00:50:48.819339 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:50:48.819360 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 00:50:48.819381 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0429 00:50:48.819402 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:50:48.819447 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:50:48.819473 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:50:48.819496 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:50:48.819517 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:50:48.819540 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:50:48.819561 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:50:48.819581 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:50:48.819620 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:50:48.819644 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:50:48.819665 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:50:48.819686 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:50:48.819707 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:50:48.819728 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:50:48.819749 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.761364
I0429 00:50:48.819772 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.319149
I0429 00:50:48.819798 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.79452 (* 1 = 2.79452 loss)
I0429 00:50:48.819825 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.93696 (* 1 = 0.93696 loss)
I0429 00:50:48.819856 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.57405 (* 0.0909091 = 0.234005 loss)
I0429 00:50:48.819885 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.83849 (* 0.0909091 = 0.258044 loss)
I0429 00:50:48.819911 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.54844 (* 0.0909091 = 0.231677 loss)
I0429 00:50:48.819936 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.26169 (* 0.0909091 = 0.205608 loss)
I0429 00:50:48.819963 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.22818 (* 0.0909091 = 0.202562 loss)
I0429 00:50:48.819990 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.87059 (* 0.0909091 = 0.170053 loss)
I0429 00:50:48.820015 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.64589 (* 0.0909091 = 0.149626 loss)
I0429 00:50:48.820041 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.57041 (* 0.0909091 = 0.142764 loss)
I0429 00:50:48.820067 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.380847 (* 0.0909091 = 0.0346225 loss)
I0429 00:50:48.820092 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.303091 (* 0.0909091 = 0.0275537 loss)
I0429 00:50:48.820119 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.443908 (* 0.0909091 = 0.0403552 loss)
I0429 00:50:48.820145 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.197937 (* 0.0909091 = 0.0179943 loss)
I0429 00:50:48.820171 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.153166 (* 0.0909091 = 0.0139241 loss)
I0429 00:50:48.820200 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.094629 (* 0.0909091 = 0.00860264 loss)
I0429 00:50:48.820222 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0508558 (* 0.0909091 = 0.00462326 loss)
I0429 00:50:48.820251 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0268552 (* 0.0909091 = 0.00244138 loss)
I0429 00:50:48.820279 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0245315 (* 0.0909091 = 0.00223013 loss)
I0429 00:50:48.820307 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.0111193 (* 0.0909091 = 0.00101085 loss)
I0429 00:50:48.820333 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00461082 (* 0.0909091 = 0.000419166 loss)
I0429 00:50:48.820358 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00132567 (* 0.0909091 = 0.000120515 loss)
I0429 00:50:48.820384 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000326666 (* 0.0909091 = 2.96969e-05 loss)
I0429 00:50:48.820410 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 5.39783e-05 (* 0.0909091 = 4.90712e-06 loss)
I0429 00:50:48.820433 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:50:48.820456 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:50:48.820480 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000792324
I0429 00:50:48.820518 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00716923
I0429 00:50:48.820543 6470 sgd_solver.cpp:106] Iteration 19000, lr = 0.01
I0429 00:53:05.389253 6470 solver.cpp:229] Iteration 19500, loss = 9.17665
I0429 00:53:05.389371 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.173913
I0429 00:53:05.389401 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 00:53:05.389426 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:53:05.389448 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 00:53:05.389470 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 00:53:05.389492 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:53:05.389519 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:53:05.389544 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 00:53:05.389569 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 00:53:05.389591 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:53:05.389614 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:53:05.389637 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 00:53:05.389659 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:53:05.389681 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:53:05.389703 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 00:53:05.389722 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 00:53:05.389744 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 00:53:05.389765 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:53:05.389787 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:53:05.389809 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:53:05.389830 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:53:05.389853 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:53:05.389874 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:53:05.389895 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.778409
I0429 00:53:05.389919 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.369565
I0429 00:53:05.389946 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.85683 (* 0.3 = 0.857048 loss)
I0429 00:53:05.389973 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.842991 (* 0.3 = 0.252897 loss)
I0429 00:53:05.390002 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.89658 (* 0.0272727 = 0.0789977 loss)
I0429 00:53:05.390027 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.17821 (* 0.0272727 = 0.0866784 loss)
I0429 00:53:05.390053 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.81045 (* 0.0272727 = 0.0766487 loss)
I0429 00:53:05.390081 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.08745 (* 0.0272727 = 0.0569305 loss)
I0429 00:53:05.390110 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.07449 (* 0.0272727 = 0.056577 loss)
I0429 00:53:05.390143 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.91195 (* 0.0272727 = 0.052144 loss)
I0429 00:53:05.390172 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.96429 (* 0.0272727 = 0.0262988 loss)
I0429 00:53:05.390198 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.835716 (* 0.0272727 = 0.0227923 loss)
I0429 00:53:05.390225 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.649627 (* 0.0272727 = 0.0177171 loss)
I0429 00:53:05.390252 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.382738 (* 0.0272727 = 0.0104383 loss)
I0429 00:53:05.390280 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.133376 (* 0.0272727 = 0.00363753 loss)
I0429 00:53:05.390306 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.0608618 (* 0.0272727 = 0.00165987 loss)
I0429 00:53:05.390334 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0390093 (* 0.0272727 = 0.00106389 loss)
I0429 00:53:05.390383 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.041852 (* 0.0272727 = 0.00114142 loss)
I0429 00:53:05.390413 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0115184 (* 0.0272727 = 0.000314137 loss)
I0429 00:53:05.390441 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00974325 (* 0.0272727 = 0.000265725 loss)
I0429 00:53:05.390470 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00623598 (* 0.0272727 = 0.000170072 loss)
I0429 00:53:05.390496 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00122394 (* 0.0272727 = 3.338e-05 loss)
I0429 00:53:05.390523 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000504628 (* 0.0272727 = 1.37626e-05 loss)
I0429 00:53:05.390552 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000425421 (* 0.0272727 = 1.16024e-05 loss)
I0429 00:53:05.390583 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000103934 (* 0.0272727 = 2.83456e-06 loss)
I0429 00:53:05.390611 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 5.05412e-05 (* 0.0272727 = 1.3784e-06 loss)
I0429 00:53:05.390635 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0652174
I0429 00:53:05.390657 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0
I0429 00:53:05.390679 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:53:05.390700 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 00:53:05.390723 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:53:05.390744 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 00:53:05.390766 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.625
I0429 00:53:05.390789 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 00:53:05.390810 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 00:53:05.390831 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:53:05.390854 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:53:05.390875 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 00:53:05.390897 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 00:53:05.390919 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 00:53:05.390940 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 00:53:05.390962 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 00:53:05.390985 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 00:53:05.391005 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:53:05.391026 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:53:05.391047 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:53:05.391069 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:53:05.391090 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:53:05.391111 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:53:05.391134 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.75
I0429 00:53:05.391155 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.326087
I0429 00:53:05.391185 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.87147 (* 0.3 = 0.861442 loss)
I0429 00:53:05.391212 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.826065 (* 0.3 = 0.24782 loss)
I0429 00:53:05.391240 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.78504 (* 0.0272727 = 0.0759555 loss)
I0429 00:53:05.391266 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.01641 (* 0.0272727 = 0.0822658 loss)
I0429 00:53:05.391309 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.80797 (* 0.0272727 = 0.0765811 loss)
I0429 00:53:05.391335 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.19881 (* 0.0272727 = 0.0599676 loss)
I0429 00:53:05.391362 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.2762 (* 0.0272727 = 0.0620782 loss)
I0429 00:53:05.391389 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.17421 (* 0.0272727 = 0.0592966 loss)
I0429 00:53:05.391412 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.938202 (* 0.0272727 = 0.0255873 loss)
I0429 00:53:05.391435 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.849781 (* 0.0272727 = 0.0231758 loss)
I0429 00:53:05.391461 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.561933 (* 0.0272727 = 0.0153255 loss)
I0429 00:53:05.391507 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.340798 (* 0.0272727 = 0.00929448 loss)
I0429 00:53:05.391535 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0415226 (* 0.0272727 = 0.00113243 loss)
I0429 00:53:05.391561 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.0172242 (* 0.0272727 = 0.000469751 loss)
I0429 00:53:05.391587 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0110628 (* 0.0272727 = 0.000301713 loss)
I0429 00:53:05.391618 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.0050024 (* 0.0272727 = 0.000136429 loss)
I0429 00:53:05.391645 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00370338 (* 0.0272727 = 0.000101001 loss)
I0429 00:53:05.391671 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00255874 (* 0.0272727 = 6.97838e-05 loss)
I0429 00:53:05.391698 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00178492 (* 0.0272727 = 4.86795e-05 loss)
I0429 00:53:05.391726 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000345667 (* 0.0272727 = 9.42729e-06 loss)
I0429 00:53:05.391753 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000164639 (* 0.0272727 = 4.49016e-06 loss)
I0429 00:53:05.391783 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000106387 (* 0.0272727 = 2.90146e-06 loss)
I0429 00:53:05.391815 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000228543 (* 0.0272727 = 6.233e-06 loss)
I0429 00:53:05.391844 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 8.11099e-05 (* 0.0272727 = 2.21209e-06 loss)
I0429 00:53:05.391868 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.108696
I0429 00:53:05.391891 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:53:05.391912 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 00:53:05.391934 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.25
I0429 00:53:05.391957 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.375
I0429 00:53:05.391978 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 00:53:05.391999 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:53:05.392020 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 00:53:05.392041 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 00:53:05.392063 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:53:05.392084 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:53:05.392107 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:53:05.392127 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 00:53:05.392148 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 00:53:05.392168 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:53:05.392189 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 00:53:05.392210 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 00:53:05.392256 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:53:05.392279 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:53:05.392300 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:53:05.392321 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:53:05.392343 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:53:05.392364 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:53:05.392385 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.761364
I0429 00:53:05.392405 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.391304
I0429 00:53:05.392432 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.74972 (* 1 = 2.74972 loss)
I0429 00:53:05.392459 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.806262 (* 1 = 0.806262 loss)
I0429 00:53:05.392485 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.4724 (* 0.0909091 = 0.224764 loss)
I0429 00:53:05.392513 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.76101 (* 0.0909091 = 0.251001 loss)
I0429 00:53:05.392539 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.79542 (* 0.0909091 = 0.254129 loss)
I0429 00:53:05.392565 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.1167 (* 0.0909091 = 0.192427 loss)
I0429 00:53:05.392590 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.99482 (* 0.0909091 = 0.181347 loss)
I0429 00:53:05.392617 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.91628 (* 0.0909091 = 0.174208 loss)
I0429 00:53:05.392643 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.976461 (* 0.0909091 = 0.0887692 loss)
I0429 00:53:05.392673 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.661219 (* 0.0909091 = 0.0601108 loss)
I0429 00:53:05.392699 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.608262 (* 0.0909091 = 0.0552965 loss)
I0429 00:53:05.392725 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.398838 (* 0.0909091 = 0.036258 loss)
I0429 00:53:05.392752 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.0364382 (* 0.0909091 = 0.00331257 loss)
I0429 00:53:05.392778 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.0206359 (* 0.0909091 = 0.00187599 loss)
I0429 00:53:05.392805 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.0140454 (* 0.0909091 = 0.00127686 loss)
I0429 00:53:05.392830 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0111681 (* 0.0909091 = 0.00101528 loss)
I0429 00:53:05.392858 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.00860693 (* 0.0909091 = 0.000782449 loss)
I0429 00:53:05.392884 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00728474 (* 0.0909091 = 0.000662249 loss)
I0429 00:53:05.392910 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00274119 (* 0.0909091 = 0.000249199 loss)
I0429 00:53:05.392936 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00151718 (* 0.0909091 = 0.000137926 loss)
I0429 00:53:05.392964 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000947812 (* 0.0909091 = 8.61647e-05 loss)
I0429 00:53:05.392990 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000660782 (* 0.0909091 = 6.00711e-05 loss)
I0429 00:53:05.393016 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.0003235 (* 0.0909091 = 2.94091e-05 loss)
I0429 00:53:05.393043 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000181078 (* 0.0909091 = 1.64616e-05 loss)
I0429 00:53:05.393065 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:53:05.393086 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:53:05.393122 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000549055
I0429 00:53:05.393146 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00145488
I0429 00:53:05.393167 6470 sgd_solver.cpp:106] Iteration 19500, lr = 0.01
I0429 00:55:21.797332 6470 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm14_bn_iter_20000.caffemodel
I0429 00:55:23.854998 6470 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm14_bn_iter_20000.solverstate
I0429 00:55:24.942581 6470 solver.cpp:338] Iteration 20000, Testing net (#0)
I0429 00:56:06.068272 6470 solver.cpp:393] Test loss: 8.17889
I0429 00:56:06.068418 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.117295
I0429 00:56:06.068446 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.213
I0429 00:56:06.068466 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.133
I0429 00:56:06.068487 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.117
I0429 00:56:06.068509 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.208
I0429 00:56:06.068531 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.356
I0429 00:56:06.068552 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.505
I0429 00:56:06.068572 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.745
I0429 00:56:06.068594 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.913
I0429 00:56:06.068616 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.993
I0429 00:56:06.068637 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.999
I0429 00:56:06.068658 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 1
I0429 00:56:06.068680 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 1
I0429 00:56:06.068702 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 1
I0429 00:56:06.068725 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 1
I0429 00:56:06.068747 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 1
I0429 00:56:06.068768 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 1
I0429 00:56:06.068790 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 1
I0429 00:56:06.068810 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 1
I0429 00:56:06.068831 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:56:06.068852 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:56:06.068873 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:56:06.068893 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:56:06.068915 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.775138
I0429 00:56:06.068936 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.311728
I0429 00:56:06.068964 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 2.9352 (* 0.3 = 0.88056 loss)
I0429 00:56:06.068991 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.793053 (* 0.3 = 0.237916 loss)
I0429 00:56:06.069017 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.56346 (* 0.0272727 = 0.0699127 loss)
I0429 00:56:06.069044 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 2.84316 (* 0.0272727 = 0.0775407 loss)
I0429 00:56:06.069069 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 2.94764 (* 0.0272727 = 0.0803902 loss)
I0429 00:56:06.069097 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.72534 (* 0.0272727 = 0.0743275 loss)
I0429 00:56:06.069121 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.29264 (* 0.0272727 = 0.0625265 loss)
I0429 00:56:06.069146 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 1.84895 (* 0.0272727 = 0.050426 loss)
I0429 00:56:06.069172 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.04294 (* 0.0272727 = 0.0284437 loss)
I0429 00:56:06.069197 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.434398 (* 0.0272727 = 0.0118472 loss)
I0429 00:56:06.069224 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.0550587 (* 0.0272727 = 0.0015016 loss)
I0429 00:56:06.069250 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.026236 (* 0.0272727 = 0.000715528 loss)
I0429 00:56:06.069278 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.0110731 (* 0.0272727 = 0.000301994 loss)
I0429 00:56:06.069305 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.00780819 (* 0.0272727 = 0.000212951 loss)
I0429 00:56:06.069335 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.00525272 (* 0.0272727 = 0.000143256 loss)
I0429 00:56:06.069386 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.00392049 (* 0.0272727 = 0.000106923 loss)
I0429 00:56:06.069416 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.00303249 (* 0.0272727 = 8.27044e-05 loss)
I0429 00:56:06.069447 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.00210541 (* 0.0272727 = 5.74202e-05 loss)
I0429 00:56:06.069474 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.00136753 (* 0.0272727 = 3.72962e-05 loss)
I0429 00:56:06.069504 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.00105927 (* 0.0272727 = 2.88893e-05 loss)
I0429 00:56:06.069532 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.000815787 (* 0.0272727 = 2.22487e-05 loss)
I0429 00:56:06.069560 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.000696873 (* 0.0272727 = 1.90056e-05 loss)
I0429 00:56:06.069586 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.000508274 (* 0.0272727 = 1.3862e-05 loss)
I0429 00:56:06.069613 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.000454098 (* 0.0272727 = 1.23845e-05 loss)
I0429 00:56:06.069635 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.10388
I0429 00:56:06.069658 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.186
I0429 00:56:06.069680 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.123
I0429 00:56:06.069702 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.088
I0429 00:56:06.069725 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.199
I0429 00:56:06.069746 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.347
I0429 00:56:06.069767 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.494
I0429 00:56:06.069789 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.739
I0429 00:56:06.069811 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.913
I0429 00:56:06.069831 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.993
I0429 00:56:06.069854 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.998
I0429 00:56:06.069875 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 1
I0429 00:56:06.069895 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 1
I0429 00:56:06.069916 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 1
I0429 00:56:06.069937 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 1
I0429 00:56:06.069958 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 1
I0429 00:56:06.069980 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 1
I0429 00:56:06.070000 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 1
I0429 00:56:06.070021 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 1
I0429 00:56:06.070042 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:56:06.070062 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:56:06.070083 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:56:06.070104 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:56:06.070125 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.771501
I0429 00:56:06.070147 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.289914
I0429 00:56:06.070173 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 2.97057 (* 0.3 = 0.89117 loss)
I0429 00:56:06.070199 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.798807 (* 0.3 = 0.239642 loss)
I0429 00:56:06.070226 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.58202 (* 0.0272727 = 0.0704188 loss)
I0429 00:56:06.070252 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 2.87638 (* 0.0272727 = 0.0784466 loss)
I0429 00:56:06.070277 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 2.98955 (* 0.0272727 = 0.0815331 loss)
I0429 00:56:06.070320 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.76293 (* 0.0272727 = 0.0753526 loss)
I0429 00:56:06.070348 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.32298 (* 0.0272727 = 0.0633539 loss)
I0429 00:56:06.070379 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 1.86973 (* 0.0272727 = 0.0509927 loss)
I0429 00:56:06.070404 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.03888 (* 0.0272727 = 0.028333 loss)
I0429 00:56:06.070430 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.431189 (* 0.0272727 = 0.0117597 loss)
I0429 00:56:06.070456 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.0514118 (* 0.0272727 = 0.00140214 loss)
I0429 00:56:06.070485 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.0228123 (* 0.0272727 = 0.000622155 loss)
I0429 00:56:06.070511 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.00962134 (* 0.0272727 = 0.0002624 loss)
I0429 00:56:06.070539 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.00670043 (* 0.0272727 = 0.000182739 loss)
I0429 00:56:06.070564 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.00422109 (* 0.0272727 = 0.000115121 loss)
I0429 00:56:06.070590 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.00294475 (* 0.0272727 = 8.03114e-05 loss)
I0429 00:56:06.070616 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.00188498 (* 0.0272727 = 5.14084e-05 loss)
I0429 00:56:06.070641 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.00132083 (* 0.0272727 = 3.60225e-05 loss)
I0429 00:56:06.070667 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.000783825 (* 0.0272727 = 2.1377e-05 loss)
I0429 00:56:06.070693 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.000563467 (* 0.0272727 = 1.53673e-05 loss)
I0429 00:56:06.070718 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.000371435 (* 0.0272727 = 1.013e-05 loss)
I0429 00:56:06.070744 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.000279782 (* 0.0272727 = 7.63043e-06 loss)
I0429 00:56:06.070771 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.000220002 (* 0.0272727 = 6.00006e-06 loss)
I0429 00:56:06.070796 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000183135 (* 0.0272727 = 4.99458e-06 loss)
I0429 00:56:06.070817 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.155085
I0429 00:56:06.070839 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.335
I0429 00:56:06.070860 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.153
I0429 00:56:06.070880 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.154
I0429 00:56:06.070901 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.225
I0429 00:56:06.070922 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.359
I0429 00:56:06.070943 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.505
I0429 00:56:06.070965 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.735
I0429 00:56:06.070986 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.91
I0429 00:56:06.071007 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.991
I0429 00:56:06.071027 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.995
I0429 00:56:06.071048 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.999
I0429 00:56:06.071069 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 1
I0429 00:56:06.071089 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 1
I0429 00:56:06.071108 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 1
I0429 00:56:06.071130 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 1
I0429 00:56:06.071151 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 1
I0429 00:56:06.071185 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 1
I0429 00:56:06.071208 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 1
I0429 00:56:06.071228 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:56:06.071249 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:56:06.071269 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:56:06.071290 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:56:06.071310 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.780183
I0429 00:56:06.071331 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.375196
I0429 00:56:06.071355 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.76891 (* 1 = 2.76891 loss)
I0429 00:56:06.071380 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.766397 (* 1 = 0.766397 loss)
I0429 00:56:06.071408 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.20382 (* 0.0909091 = 0.200347 loss)
I0429 00:56:06.071452 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 2.77974 (* 0.0909091 = 0.252704 loss)
I0429 00:56:06.071480 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 2.88432 (* 0.0909091 = 0.262211 loss)
I0429 00:56:06.071507 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.64723 (* 0.0909091 = 0.240657 loss)
I0429 00:56:06.071537 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.28115 (* 0.0909091 = 0.207377 loss)
I0429 00:56:06.071564 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 1.83381 (* 0.0909091 = 0.16671 loss)
I0429 00:56:06.071591 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.02298 (* 0.0909091 = 0.0929983 loss)
I0429 00:56:06.071617 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.435164 (* 0.0909091 = 0.0395604 loss)
I0429 00:56:06.071642 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.0539593 (* 0.0909091 = 0.00490539 loss)
I0429 00:56:06.071669 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.026763 (* 0.0909091 = 0.002433 loss)
I0429 00:56:06.071694 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.0103355 (* 0.0909091 = 0.000939591 loss)
I0429 00:56:06.071719 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.00689664 (* 0.0909091 = 0.000626968 loss)
I0429 00:56:06.071746 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.00469162 (* 0.0909091 = 0.000426511 loss)
I0429 00:56:06.071771 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.0032573 (* 0.0909091 = 0.000296119 loss)
I0429 00:56:06.071796 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.00190815 (* 0.0909091 = 0.000173468 loss)
I0429 00:56:06.071825 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.0011421 (* 0.0909091 = 0.000103827 loss)
I0429 00:56:06.071846 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.000598089 (* 0.0909091 = 5.43717e-05 loss)
I0429 00:56:06.071874 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.000322669 (* 0.0909091 = 2.93335e-05 loss)
I0429 00:56:06.071902 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.000229542 (* 0.0909091 = 2.08675e-05 loss)
I0429 00:56:06.071928 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.000173747 (* 0.0909091 = 1.57952e-05 loss)
I0429 00:56:06.071954 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.000159126 (* 0.0909091 = 1.4466e-05 loss)
I0429 00:56:06.071979 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.000128679 (* 0.0909091 = 1.16981e-05 loss)
I0429 00:56:06.072001 6470 solver.cpp:406] Test net output #147: total_accuracy = 0.003
I0429 00:56:06.072022 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.012
I0429 00:56:06.072042 6470 solver.cpp:406] Test net output #149: total_confidence = 0.00264012
I0429 00:56:06.072082 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.00799304
I0429 00:56:06.072105 6470 solver.cpp:338] Iteration 20000, Testing net (#1)
I0429 00:56:46.903954 6470 solver.cpp:393] Test loss: 8.72884
I0429 00:56:46.904091 6470 solver.cpp:406] Test net output #0: loss1/accuracy = 0.11938
I0429 00:56:46.904122 6470 solver.cpp:406] Test net output #1: loss1/accuracy01 = 0.23
I0429 00:56:46.904147 6470 solver.cpp:406] Test net output #2: loss1/accuracy02 = 0.133
I0429 00:56:46.904170 6470 solver.cpp:406] Test net output #3: loss1/accuracy03 = 0.106
I0429 00:56:46.904193 6470 solver.cpp:406] Test net output #4: loss1/accuracy04 = 0.199
I0429 00:56:46.904219 6470 solver.cpp:406] Test net output #5: loss1/accuracy05 = 0.342
I0429 00:56:46.904242 6470 solver.cpp:406] Test net output #6: loss1/accuracy06 = 0.446
I0429 00:56:46.904265 6470 solver.cpp:406] Test net output #7: loss1/accuracy07 = 0.667
I0429 00:56:46.904287 6470 solver.cpp:406] Test net output #8: loss1/accuracy08 = 0.813
I0429 00:56:46.904309 6470 solver.cpp:406] Test net output #9: loss1/accuracy09 = 0.897
I0429 00:56:46.904335 6470 solver.cpp:406] Test net output #10: loss1/accuracy10 = 0.912
I0429 00:56:46.904357 6470 solver.cpp:406] Test net output #11: loss1/accuracy11 = 0.923
I0429 00:56:46.904379 6470 solver.cpp:406] Test net output #12: loss1/accuracy12 = 0.942
I0429 00:56:46.904400 6470 solver.cpp:406] Test net output #13: loss1/accuracy13 = 0.961
I0429 00:56:46.904423 6470 solver.cpp:406] Test net output #14: loss1/accuracy14 = 0.966
I0429 00:56:46.904444 6470 solver.cpp:406] Test net output #15: loss1/accuracy15 = 0.975
I0429 00:56:46.904465 6470 solver.cpp:406] Test net output #16: loss1/accuracy16 = 0.989
I0429 00:56:46.904487 6470 solver.cpp:406] Test net output #17: loss1/accuracy17 = 0.996
I0429 00:56:46.904510 6470 solver.cpp:406] Test net output #18: loss1/accuracy18 = 1
I0429 00:56:46.904533 6470 solver.cpp:406] Test net output #19: loss1/accuracy19 = 1
I0429 00:56:46.904556 6470 solver.cpp:406] Test net output #20: loss1/accuracy20 = 1
I0429 00:56:46.904577 6470 solver.cpp:406] Test net output #21: loss1/accuracy21 = 1
I0429 00:56:46.904597 6470 solver.cpp:406] Test net output #22: loss1/accuracy22 = 1
I0429 00:56:46.904618 6470 solver.cpp:406] Test net output #23: loss1/accuracy_incl_empty = 0.742682
I0429 00:56:46.904640 6470 solver.cpp:406] Test net output #24: loss1/accuracy_top3 = 0.313825
I0429 00:56:46.904669 6470 solver.cpp:406] Test net output #25: loss1/cross_entropy_loss = 2.89942 (* 0.3 = 0.869825 loss)
I0429 00:56:46.904695 6470 solver.cpp:406] Test net output #26: loss1/cross_entropy_loss_incl_empty = 0.914754 (* 0.3 = 0.274426 loss)
I0429 00:56:46.904723 6470 solver.cpp:406] Test net output #27: loss1/loss01 = 2.55798 (* 0.0272727 = 0.0697632 loss)
I0429 00:56:46.904749 6470 solver.cpp:406] Test net output #28: loss1/loss02 = 2.85594 (* 0.0272727 = 0.0778892 loss)
I0429 00:56:46.904777 6470 solver.cpp:406] Test net output #29: loss1/loss03 = 2.92354 (* 0.0272727 = 0.079733 loss)
I0429 00:56:46.904801 6470 solver.cpp:406] Test net output #30: loss1/loss04 = 2.73765 (* 0.0272727 = 0.0746633 loss)
I0429 00:56:46.904827 6470 solver.cpp:406] Test net output #31: loss1/loss05 = 2.29583 (* 0.0272727 = 0.0626134 loss)
I0429 00:56:46.904855 6470 solver.cpp:406] Test net output #32: loss1/loss06 = 2.04546 (* 0.0272727 = 0.0557853 loss)
I0429 00:56:46.904878 6470 solver.cpp:406] Test net output #33: loss1/loss07 = 1.34892 (* 0.0272727 = 0.0367888 loss)
I0429 00:56:46.904904 6470 solver.cpp:406] Test net output #34: loss1/loss08 = 0.756353 (* 0.0272727 = 0.0206278 loss)
I0429 00:56:46.904930 6470 solver.cpp:406] Test net output #35: loss1/loss09 = 0.452748 (* 0.0272727 = 0.0123477 loss)
I0429 00:56:46.904955 6470 solver.cpp:406] Test net output #36: loss1/loss10 = 0.364759 (* 0.0272727 = 0.00994796 loss)
I0429 00:56:46.904983 6470 solver.cpp:406] Test net output #37: loss1/loss11 = 0.311667 (* 0.0272727 = 0.00850002 loss)
I0429 00:56:46.905007 6470 solver.cpp:406] Test net output #38: loss1/loss12 = 0.242779 (* 0.0272727 = 0.00662124 loss)
I0429 00:56:46.905035 6470 solver.cpp:406] Test net output #39: loss1/loss13 = 0.181217 (* 0.0272727 = 0.00494228 loss)
I0429 00:56:46.905084 6470 solver.cpp:406] Test net output #40: loss1/loss14 = 0.159536 (* 0.0272727 = 0.00435098 loss)
I0429 00:56:46.905117 6470 solver.cpp:406] Test net output #41: loss1/loss15 = 0.126648 (* 0.0272727 = 0.00345403 loss)
I0429 00:56:46.905144 6470 solver.cpp:406] Test net output #42: loss1/loss16 = 0.0690548 (* 0.0272727 = 0.00188331 loss)
I0429 00:56:46.905172 6470 solver.cpp:406] Test net output #43: loss1/loss17 = 0.0318243 (* 0.0272727 = 0.000867935 loss)
I0429 00:56:46.905199 6470 solver.cpp:406] Test net output #44: loss1/loss18 = 0.00374857 (* 0.0272727 = 0.000102234 loss)
I0429 00:56:46.905226 6470 solver.cpp:406] Test net output #45: loss1/loss19 = 0.00212579 (* 0.0272727 = 5.79762e-05 loss)
I0429 00:56:46.905253 6470 solver.cpp:406] Test net output #46: loss1/loss20 = 0.00128149 (* 0.0272727 = 3.49497e-05 loss)
I0429 00:56:46.905280 6470 solver.cpp:406] Test net output #47: loss1/loss21 = 0.000670786 (* 0.0272727 = 1.82942e-05 loss)
I0429 00:56:46.905306 6470 solver.cpp:406] Test net output #48: loss1/loss22 = 0.000476108 (* 0.0272727 = 1.29848e-05 loss)
I0429 00:56:46.905328 6470 solver.cpp:406] Test net output #49: loss2/accuracy = 0.110366
I0429 00:56:46.905350 6470 solver.cpp:406] Test net output #50: loss2/accuracy01 = 0.205
I0429 00:56:46.905375 6470 solver.cpp:406] Test net output #51: loss2/accuracy02 = 0.131
I0429 00:56:46.905398 6470 solver.cpp:406] Test net output #52: loss2/accuracy03 = 0.1
I0429 00:56:46.905419 6470 solver.cpp:406] Test net output #53: loss2/accuracy04 = 0.191
I0429 00:56:46.905442 6470 solver.cpp:406] Test net output #54: loss2/accuracy05 = 0.344
I0429 00:56:46.905465 6470 solver.cpp:406] Test net output #55: loss2/accuracy06 = 0.442
I0429 00:56:46.905489 6470 solver.cpp:406] Test net output #56: loss2/accuracy07 = 0.664
I0429 00:56:46.905514 6470 solver.cpp:406] Test net output #57: loss2/accuracy08 = 0.813
I0429 00:56:46.905535 6470 solver.cpp:406] Test net output #58: loss2/accuracy09 = 0.895
I0429 00:56:46.905560 6470 solver.cpp:406] Test net output #59: loss2/accuracy10 = 0.917
I0429 00:56:46.905581 6470 solver.cpp:406] Test net output #60: loss2/accuracy11 = 0.923
I0429 00:56:46.905602 6470 solver.cpp:406] Test net output #61: loss2/accuracy12 = 0.941
I0429 00:56:46.905625 6470 solver.cpp:406] Test net output #62: loss2/accuracy13 = 0.958
I0429 00:56:46.905645 6470 solver.cpp:406] Test net output #63: loss2/accuracy14 = 0.965
I0429 00:56:46.905668 6470 solver.cpp:406] Test net output #64: loss2/accuracy15 = 0.977
I0429 00:56:46.905689 6470 solver.cpp:406] Test net output #65: loss2/accuracy16 = 0.989
I0429 00:56:46.905711 6470 solver.cpp:406] Test net output #66: loss2/accuracy17 = 0.996
I0429 00:56:46.905733 6470 solver.cpp:406] Test net output #67: loss2/accuracy18 = 1
I0429 00:56:46.905755 6470 solver.cpp:406] Test net output #68: loss2/accuracy19 = 1
I0429 00:56:46.905776 6470 solver.cpp:406] Test net output #69: loss2/accuracy20 = 1
I0429 00:56:46.905797 6470 solver.cpp:406] Test net output #70: loss2/accuracy21 = 1
I0429 00:56:46.905818 6470 solver.cpp:406] Test net output #71: loss2/accuracy22 = 1
I0429 00:56:46.905840 6470 solver.cpp:406] Test net output #72: loss2/accuracy_incl_empty = 0.741546
I0429 00:56:46.905863 6470 solver.cpp:406] Test net output #73: loss2/accuracy_top3 = 0.301551
I0429 00:56:46.905887 6470 solver.cpp:406] Test net output #74: loss2/cross_entropy_loss = 2.93549 (* 0.3 = 0.880646 loss)
I0429 00:56:46.905913 6470 solver.cpp:406] Test net output #75: loss2/cross_entropy_loss_incl_empty = 0.916055 (* 0.3 = 0.274817 loss)
I0429 00:56:46.905939 6470 solver.cpp:406] Test net output #76: loss2/loss01 = 2.57546 (* 0.0272727 = 0.0702398 loss)
I0429 00:56:46.905966 6470 solver.cpp:406] Test net output #77: loss2/loss02 = 2.88801 (* 0.0272727 = 0.0787638 loss)
I0429 00:56:46.906008 6470 solver.cpp:406] Test net output #78: loss2/loss03 = 2.96839 (* 0.0272727 = 0.080956 loss)
I0429 00:56:46.906036 6470 solver.cpp:406] Test net output #79: loss2/loss04 = 2.7572 (* 0.0272727 = 0.0751963 loss)
I0429 00:56:46.906061 6470 solver.cpp:406] Test net output #80: loss2/loss05 = 2.31252 (* 0.0272727 = 0.0630686 loss)
I0429 00:56:46.906087 6470 solver.cpp:406] Test net output #81: loss2/loss06 = 2.07008 (* 0.0272727 = 0.0564567 loss)
I0429 00:56:46.906112 6470 solver.cpp:406] Test net output #82: loss2/loss07 = 1.34106 (* 0.0272727 = 0.0365743 loss)
I0429 00:56:46.906137 6470 solver.cpp:406] Test net output #83: loss2/loss08 = 0.761378 (* 0.0272727 = 0.0207649 loss)
I0429 00:56:46.906168 6470 solver.cpp:406] Test net output #84: loss2/loss09 = 0.461832 (* 0.0272727 = 0.0125954 loss)
I0429 00:56:46.906195 6470 solver.cpp:406] Test net output #85: loss2/loss10 = 0.365991 (* 0.0272727 = 0.00998157 loss)
I0429 00:56:46.906221 6470 solver.cpp:406] Test net output #86: loss2/loss11 = 0.321844 (* 0.0272727 = 0.00877756 loss)
I0429 00:56:46.906249 6470 solver.cpp:406] Test net output #87: loss2/loss12 = 0.250999 (* 0.0272727 = 0.00684542 loss)
I0429 00:56:46.906273 6470 solver.cpp:406] Test net output #88: loss2/loss13 = 0.18438 (* 0.0272727 = 0.00502855 loss)
I0429 00:56:46.906299 6470 solver.cpp:406] Test net output #89: loss2/loss14 = 0.158972 (* 0.0272727 = 0.00433559 loss)
I0429 00:56:46.906325 6470 solver.cpp:406] Test net output #90: loss2/loss15 = 0.125606 (* 0.0272727 = 0.00342561 loss)
I0429 00:56:46.906352 6470 solver.cpp:406] Test net output #91: loss2/loss16 = 0.0689262 (* 0.0272727 = 0.0018798 loss)
I0429 00:56:46.906376 6470 solver.cpp:406] Test net output #92: loss2/loss17 = 0.0336303 (* 0.0272727 = 0.000917189 loss)
I0429 00:56:46.906404 6470 solver.cpp:406] Test net output #93: loss2/loss18 = 0.00285057 (* 0.0272727 = 7.77427e-05 loss)
I0429 00:56:46.906432 6470 solver.cpp:406] Test net output #94: loss2/loss19 = 0.001524 (* 0.0272727 = 4.15637e-05 loss)
I0429 00:56:46.906460 6470 solver.cpp:406] Test net output #95: loss2/loss20 = 0.00100312 (* 0.0272727 = 2.73578e-05 loss)
I0429 00:56:46.906487 6470 solver.cpp:406] Test net output #96: loss2/loss21 = 0.000639726 (* 0.0272727 = 1.74471e-05 loss)
I0429 00:56:46.906512 6470 solver.cpp:406] Test net output #97: loss2/loss22 = 0.000502327 (* 0.0272727 = 1.36998e-05 loss)
I0429 00:56:46.906533 6470 solver.cpp:406] Test net output #98: loss3/accuracy = 0.161837
I0429 00:56:46.906554 6470 solver.cpp:406] Test net output #99: loss3/accuracy01 = 0.327
I0429 00:56:46.906576 6470 solver.cpp:406] Test net output #100: loss3/accuracy02 = 0.151
I0429 00:56:46.906596 6470 solver.cpp:406] Test net output #101: loss3/accuracy03 = 0.139
I0429 00:56:46.906617 6470 solver.cpp:406] Test net output #102: loss3/accuracy04 = 0.23
I0429 00:56:46.906638 6470 solver.cpp:406] Test net output #103: loss3/accuracy05 = 0.357
I0429 00:56:46.906659 6470 solver.cpp:406] Test net output #104: loss3/accuracy06 = 0.453
I0429 00:56:46.906679 6470 solver.cpp:406] Test net output #105: loss3/accuracy07 = 0.66
I0429 00:56:46.906700 6470 solver.cpp:406] Test net output #106: loss3/accuracy08 = 0.814
I0429 00:56:46.906723 6470 solver.cpp:406] Test net output #107: loss3/accuracy09 = 0.898
I0429 00:56:46.906743 6470 solver.cpp:406] Test net output #108: loss3/accuracy10 = 0.914
I0429 00:56:46.906764 6470 solver.cpp:406] Test net output #109: loss3/accuracy11 = 0.926
I0429 00:56:46.906785 6470 solver.cpp:406] Test net output #110: loss3/accuracy12 = 0.941
I0429 00:56:46.906805 6470 solver.cpp:406] Test net output #111: loss3/accuracy13 = 0.956
I0429 00:56:46.906826 6470 solver.cpp:406] Test net output #112: loss3/accuracy14 = 0.963
I0429 00:56:46.906847 6470 solver.cpp:406] Test net output #113: loss3/accuracy15 = 0.973
I0429 00:56:46.906868 6470 solver.cpp:406] Test net output #114: loss3/accuracy16 = 0.989
I0429 00:56:46.906904 6470 solver.cpp:406] Test net output #115: loss3/accuracy17 = 0.996
I0429 00:56:46.906927 6470 solver.cpp:406] Test net output #116: loss3/accuracy18 = 1
I0429 00:56:46.906949 6470 solver.cpp:406] Test net output #117: loss3/accuracy19 = 1
I0429 00:56:46.906970 6470 solver.cpp:406] Test net output #118: loss3/accuracy20 = 1
I0429 00:56:46.906990 6470 solver.cpp:406] Test net output #119: loss3/accuracy21 = 1
I0429 00:56:46.907011 6470 solver.cpp:406] Test net output #120: loss3/accuracy22 = 1
I0429 00:56:46.907030 6470 solver.cpp:406] Test net output #121: loss3/accuracy_incl_empty = 0.751455
I0429 00:56:46.907052 6470 solver.cpp:406] Test net output #122: loss3/accuracy_top3 = 0.362341
I0429 00:56:46.907078 6470 solver.cpp:406] Test net output #123: loss3/cross_entropy_loss = 2.7707 (* 1 = 2.7707 loss)
I0429 00:56:46.907102 6470 solver.cpp:406] Test net output #124: loss3/cross_entropy_loss_incl_empty = 0.87793 (* 1 = 0.87793 loss)
I0429 00:56:46.907129 6470 solver.cpp:406] Test net output #125: loss3/loss01 = 2.2777 (* 0.0909091 = 0.207064 loss)
I0429 00:56:46.907153 6470 solver.cpp:406] Test net output #126: loss3/loss02 = 2.77667 (* 0.0909091 = 0.252424 loss)
I0429 00:56:46.907178 6470 solver.cpp:406] Test net output #127: loss3/loss03 = 2.86329 (* 0.0909091 = 0.260299 loss)
I0429 00:56:46.907205 6470 solver.cpp:406] Test net output #128: loss3/loss04 = 2.67253 (* 0.0909091 = 0.242957 loss)
I0429 00:56:46.907234 6470 solver.cpp:406] Test net output #129: loss3/loss05 = 2.24783 (* 0.0909091 = 0.204348 loss)
I0429 00:56:46.907261 6470 solver.cpp:406] Test net output #130: loss3/loss06 = 2.0292 (* 0.0909091 = 0.184472 loss)
I0429 00:56:46.907286 6470 solver.cpp:406] Test net output #131: loss3/loss07 = 1.32652 (* 0.0909091 = 0.120593 loss)
I0429 00:56:46.907312 6470 solver.cpp:406] Test net output #132: loss3/loss08 = 0.765646 (* 0.0909091 = 0.0696041 loss)
I0429 00:56:46.907337 6470 solver.cpp:406] Test net output #133: loss3/loss09 = 0.443813 (* 0.0909091 = 0.0403466 loss)
I0429 00:56:46.907363 6470 solver.cpp:406] Test net output #134: loss3/loss10 = 0.353742 (* 0.0909091 = 0.0321583 loss)
I0429 00:56:46.907389 6470 solver.cpp:406] Test net output #135: loss3/loss11 = 0.302997 (* 0.0909091 = 0.0275452 loss)
I0429 00:56:46.907414 6470 solver.cpp:406] Test net output #136: loss3/loss12 = 0.237406 (* 0.0909091 = 0.0215824 loss)
I0429 00:56:46.907439 6470 solver.cpp:406] Test net output #137: loss3/loss13 = 0.173217 (* 0.0909091 = 0.015747 loss)
I0429 00:56:46.907482 6470 solver.cpp:406] Test net output #138: loss3/loss14 = 0.151052 (* 0.0909091 = 0.013732 loss)
I0429 00:56:46.907513 6470 solver.cpp:406] Test net output #139: loss3/loss15 = 0.122634 (* 0.0909091 = 0.0111486 loss)
I0429 00:56:46.907541 6470 solver.cpp:406] Test net output #140: loss3/loss16 = 0.0691888 (* 0.0909091 = 0.00628989 loss)
I0429 00:56:46.907567 6470 solver.cpp:406] Test net output #141: loss3/loss17 = 0.030412 (* 0.0909091 = 0.00276473 loss)
I0429 00:56:46.907593 6470 solver.cpp:406] Test net output #142: loss3/loss18 = 0.00243725 (* 0.0909091 = 0.000221568 loss)
I0429 00:56:46.907618 6470 solver.cpp:406] Test net output #143: loss3/loss19 = 0.00126923 (* 0.0909091 = 0.000115385 loss)
I0429 00:56:46.907644 6470 solver.cpp:406] Test net output #144: loss3/loss20 = 0.000543767 (* 0.0909091 = 4.94334e-05 loss)
I0429 00:56:46.907670 6470 solver.cpp:406] Test net output #145: loss3/loss21 = 0.000285835 (* 0.0909091 = 2.5985e-05 loss)
I0429 00:56:46.907694 6470 solver.cpp:406] Test net output #146: loss3/loss22 = 0.000136084 (* 0.0909091 = 1.23713e-05 loss)
I0429 00:56:46.907716 6470 solver.cpp:406] Test net output #147: total_accuracy = 0.005
I0429 00:56:46.907738 6470 solver.cpp:406] Test net output #148: total_accuracy_not_rec = 0.005
I0429 00:56:46.907759 6470 solver.cpp:406] Test net output #149: total_confidence = 0.00265252
I0429 00:56:46.907798 6470 solver.cpp:406] Test net output #150: total_confidence_not_rec = 0.00730615
I0429 00:56:47.086171 6470 solver.cpp:229] Iteration 20000, loss = 9.05395
I0429 00:56:47.086231 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.153846
I0429 00:56:47.086266 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.375
I0429 00:56:47.086293 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:56:47.086318 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:56:47.086341 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 00:56:47.086366 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0429 00:56:47.086392 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 00:56:47.086416 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 00:56:47.086439 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 00:56:47.086462 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 00:56:47.086483 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:56:47.086505 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:56:47.086529 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 00:56:47.086555 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 00:56:47.086578 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:56:47.086601 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:56:47.086623 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0429 00:56:47.086645 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 0.875
I0429 00:56:47.086669 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:56:47.086691 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:56:47.086714 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:56:47.086736 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:56:47.086760 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:56:47.086786 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.744318
I0429 00:56:47.086810 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.307692
I0429 00:56:47.086838 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 3.06311 (* 0.3 = 0.918932 loss)
I0429 00:56:47.086866 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.968517 (* 0.3 = 0.290555 loss)
I0429 00:56:47.086894 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.6304 (* 0.0272727 = 0.0717383 loss)
I0429 00:56:47.086922 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.76564 (* 0.0272727 = 0.0754264 loss)
I0429 00:56:47.086948 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.66806 (* 0.0272727 = 0.0727652 loss)
I0429 00:56:47.086977 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.98618 (* 0.0272727 = 0.0814412 loss)
I0429 00:56:47.087003 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.28081 (* 0.0272727 = 0.0622039 loss)
I0429 00:56:47.087030 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.7763 (* 0.0272727 = 0.0484444 loss)
I0429 00:56:47.087056 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.1487 (* 0.0272727 = 0.0313281 loss)
I0429 00:56:47.087083 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.2448 (* 0.0272727 = 0.00667637 loss)
I0429 00:56:47.087111 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.447338 (* 0.0272727 = 0.0122001 loss)
I0429 00:56:47.087138 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.397202 (* 0.0272727 = 0.0108328 loss)
I0429 00:56:47.087167 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.338244 (* 0.0272727 = 0.00922485 loss)
I0429 00:56:47.087232 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.634904 (* 0.0272727 = 0.0173156 loss)
I0429 00:56:47.087261 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.663235 (* 0.0272727 = 0.0180882 loss)
I0429 00:56:47.087290 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.491062 (* 0.0272727 = 0.0133926 loss)
I0429 00:56:47.087319 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.488561 (* 0.0272727 = 0.0133244 loss)
I0429 00:56:47.087348 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.620445 (* 0.0272727 = 0.0169212 loss)
I0429 00:56:47.087376 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.655464 (* 0.0272727 = 0.0178763 loss)
I0429 00:56:47.087405 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00244533 (* 0.0272727 = 6.66909e-05 loss)
I0429 00:56:47.087432 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000289927 (* 0.0272727 = 7.9071e-06 loss)
I0429 00:56:47.087460 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 6.94645e-05 (* 0.0272727 = 1.89449e-06 loss)
I0429 00:56:47.087507 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 4.16302e-05 (* 0.0272727 = 1.13537e-06 loss)
I0429 00:56:47.087537 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 2.81647e-05 (* 0.0272727 = 7.68127e-07 loss)
I0429 00:56:47.087559 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0769231
I0429 00:56:47.087582 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:56:47.087609 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 00:56:47.087631 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:56:47.087653 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 00:56:47.087676 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 00:56:47.087698 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 00:56:47.087720 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 00:56:47.087743 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 00:56:47.087764 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:56:47.087786 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:56:47.087808 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:56:47.087831 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:56:47.087852 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:56:47.087874 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:56:47.087901 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:56:47.087925 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0429 00:56:47.087947 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 0.875
I0429 00:56:47.087970 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:56:47.087992 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:56:47.088014 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:56:47.088035 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:56:47.088057 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:56:47.088079 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.721591
I0429 00:56:47.088101 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.211538
I0429 00:56:47.088127 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 3.07363 (* 0.3 = 0.92209 loss)
I0429 00:56:47.088155 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.967722 (* 0.3 = 0.290317 loss)
I0429 00:56:47.088199 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.68787 (* 0.0272727 = 0.0733055 loss)
I0429 00:56:47.088228 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 3.11563 (* 0.0272727 = 0.0849718 loss)
I0429 00:56:47.088254 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.10514 (* 0.0272727 = 0.0846857 loss)
I0429 00:56:47.088279 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.62339 (* 0.0272727 = 0.0715469 loss)
I0429 00:56:47.088306 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.06304 (* 0.0272727 = 0.0562648 loss)
I0429 00:56:47.088333 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.90215 (* 0.0272727 = 0.0518767 loss)
I0429 00:56:47.088359 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.33504 (* 0.0272727 = 0.0364103 loss)
I0429 00:56:47.088392 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.262355 (* 0.0272727 = 0.00715515 loss)
I0429 00:56:47.088420 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.594861 (* 0.0272727 = 0.0162235 loss)
I0429 00:56:47.088449 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.366704 (* 0.0272727 = 0.010001 loss)
I0429 00:56:47.088476 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.445341 (* 0.0272727 = 0.0121457 loss)
I0429 00:56:47.088503 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.891565 (* 0.0272727 = 0.0243154 loss)
I0429 00:56:47.088531 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.599572 (* 0.0272727 = 0.016352 loss)
I0429 00:56:47.088557 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.539423 (* 0.0272727 = 0.0147115 loss)
I0429 00:56:47.088583 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.865979 (* 0.0272727 = 0.0236176 loss)
I0429 00:56:47.088609 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.94573 (* 0.0272727 = 0.0257926 loss)
I0429 00:56:47.088635 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.927524 (* 0.0272727 = 0.0252961 loss)
I0429 00:56:47.088667 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00133251 (* 0.0272727 = 3.6341e-05 loss)
I0429 00:56:47.088695 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000488059 (* 0.0272727 = 1.33107e-05 loss)
I0429 00:56:47.088721 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000278638 (* 0.0272727 = 7.59922e-06 loss)
I0429 00:56:47.088747 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00045519 (* 0.0272727 = 1.24143e-05 loss)
I0429 00:56:47.088773 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000224851 (* 0.0272727 = 6.1323e-06 loss)
I0429 00:56:47.088795 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.134615
I0429 00:56:47.088817 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:56:47.088838 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:56:47.088860 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:56:47.088882 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 00:56:47.088904 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 00:56:47.088927 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 00:56:47.088948 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 00:56:47.088969 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 00:56:47.088990 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 00:56:47.089011 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:56:47.089033 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 00:56:47.089056 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:56:47.089076 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:56:47.089112 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 0.875
I0429 00:56:47.089138 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:56:47.089157 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0429 00:56:47.089179 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 0.875
I0429 00:56:47.089201 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:56:47.089224 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:56:47.089246 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:56:47.089267 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:56:47.089289 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:56:47.089311 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0429 00:56:47.089334 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.288462
I0429 00:56:47.089359 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.93467 (* 1 = 2.93467 loss)
I0429 00:56:47.089386 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.90921 (* 1 = 0.90921 loss)
I0429 00:56:47.089413 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.2679 (* 0.0909091 = 0.206173 loss)
I0429 00:56:47.089444 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.8177 (* 0.0909091 = 0.256155 loss)
I0429 00:56:47.089473 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.65491 (* 0.0909091 = 0.241356 loss)
I0429 00:56:47.089498 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.51197 (* 0.0909091 = 0.22836 loss)
I0429 00:56:47.089525 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.93178 (* 0.0909091 = 0.175617 loss)
I0429 00:56:47.089552 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.67628 (* 0.0909091 = 0.152389 loss)
I0429 00:56:47.089578 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.16605 (* 0.0909091 = 0.106004 loss)
I0429 00:56:47.089604 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.284088 (* 0.0909091 = 0.0258262 loss)
I0429 00:56:47.089630 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.564064 (* 0.0909091 = 0.0512786 loss)
I0429 00:56:47.089656 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.276617 (* 0.0909091 = 0.025147 loss)
I0429 00:56:47.089684 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.427718 (* 0.0909091 = 0.0388835 loss)
I0429 00:56:47.089714 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.761176 (* 0.0909091 = 0.0691978 loss)
I0429 00:56:47.089740 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.671402 (* 0.0909091 = 0.0610366 loss)
I0429 00:56:47.089766 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.335131 (* 0.0909091 = 0.0304665 loss)
I0429 00:56:47.089792 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.416894 (* 0.0909091 = 0.0378994 loss)
I0429 00:56:47.089818 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.511056 (* 0.0909091 = 0.0464596 loss)
I0429 00:56:47.089844 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.509684 (* 0.0909091 = 0.0463349 loss)
I0429 00:56:47.089871 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00547203 (* 0.0909091 = 0.000497457 loss)
I0429 00:56:47.089900 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00173107 (* 0.0909091 = 0.00015737 loss)
I0429 00:56:47.089926 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00034095 (* 0.0909091 = 3.09955e-05 loss)
I0429 00:56:47.089952 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000116406 (* 0.0909091 = 1.05824e-05 loss)
I0429 00:56:47.089980 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 9.00609e-05 (* 0.0909091 = 8.18735e-06 loss)
I0429 00:56:47.090018 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:56:47.090040 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:56:47.090064 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00495914
I0429 00:56:47.090085 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.0204836
I0429 00:56:47.090107 6470 sgd_solver.cpp:106] Iteration 20000, lr = 0.01
I0429 00:59:03.673919 6470 solver.cpp:229] Iteration 20500, loss = 9.1005
I0429 00:59:03.674101 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.078125
I0429 00:59:03.674123 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 00:59:03.674137 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 00:59:03.674150 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 00:59:03.674162 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 00:59:03.674175 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 00:59:03.674186 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.125
I0429 00:59:03.674198 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.375
I0429 00:59:03.674211 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.375
I0429 00:59:03.674222 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 00:59:03.674234 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 00:59:03.674247 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 00:59:03.674257 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 00:59:03.674269 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 00:59:03.674281 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 0.875
I0429 00:59:03.674293 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 0.875
I0429 00:59:03.674304 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 0.875
I0429 00:59:03.674319 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 00:59:03.674331 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 00:59:03.674343 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 00:59:03.674355 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 00:59:03.674366 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 00:59:03.674378 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 00:59:03.674389 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.659091
I0429 00:59:03.674401 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.359375
I0429 00:59:03.674417 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.70687 (* 0.3 = 0.812061 loss)
I0429 00:59:03.674432 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 1.08899 (* 0.3 = 0.326697 loss)
I0429 00:59:03.674446 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.67874 (* 0.0272727 = 0.0730565 loss)
I0429 00:59:03.674460 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.62065 (* 0.0272727 = 0.0714724 loss)
I0429 00:59:03.674474 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.84419 (* 0.0272727 = 0.0775689 loss)
I0429 00:59:03.674487 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.38636 (* 0.0272727 = 0.0650824 loss)
I0429 00:59:03.674501 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 1.89689 (* 0.0272727 = 0.0517335 loss)
I0429 00:59:03.674515 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 2.63603 (* 0.0272727 = 0.0718918 loss)
I0429 00:59:03.674528 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 2.417 (* 0.0272727 = 0.0659182 loss)
I0429 00:59:03.674542 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 2.49939 (* 0.0272727 = 0.0681652 loss)
I0429 00:59:03.674556 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.265729 (* 0.0272727 = 0.00724715 loss)
I0429 00:59:03.674571 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.608382 (* 0.0272727 = 0.0165922 loss)
I0429 00:59:03.674584 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.381313 (* 0.0272727 = 0.0103994 loss)
I0429 00:59:03.674598 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.336916 (* 0.0272727 = 0.00918863 loss)
I0429 00:59:03.674634 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.306636 (* 0.0272727 = 0.00836279 loss)
I0429 00:59:03.674650 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.268481 (* 0.0272727 = 0.0073222 loss)
I0429 00:59:03.674664 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.450191 (* 0.0272727 = 0.0122779 loss)
I0429 00:59:03.674679 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.479972 (* 0.0272727 = 0.0130902 loss)
I0429 00:59:03.674693 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.0116759 (* 0.0272727 = 0.000318434 loss)
I0429 00:59:03.674707 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00470748 (* 0.0272727 = 0.000128386 loss)
I0429 00:59:03.674721 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00240306 (* 0.0272727 = 6.55381e-05 loss)
I0429 00:59:03.674736 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.001296 (* 0.0272727 = 3.53454e-05 loss)
I0429 00:59:03.674749 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000365497 (* 0.0272727 = 9.96809e-06 loss)
I0429 00:59:03.674763 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00110736 (* 0.0272727 = 3.02006e-05 loss)
I0429 00:59:03.674777 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.109375
I0429 00:59:03.674788 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 00:59:03.674799 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 00:59:03.674811 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 00:59:03.674823 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 00:59:03.674835 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 00:59:03.674849 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.125
I0429 00:59:03.674856 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.375
I0429 00:59:03.674865 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.375
I0429 00:59:03.674877 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 00:59:03.674890 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 00:59:03.674901 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 00:59:03.674914 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 00:59:03.674934 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 00:59:03.674954 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 0.875
I0429 00:59:03.674974 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 0.875
I0429 00:59:03.674995 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 0.875
I0429 00:59:03.675014 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 00:59:03.675034 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 00:59:03.675055 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 00:59:03.675076 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 00:59:03.675102 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 00:59:03.675130 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 00:59:03.675156 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.659091
I0429 00:59:03.675176 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.265625
I0429 00:59:03.675192 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.89492 (* 0.3 = 0.868475 loss)
I0429 00:59:03.675206 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 1.18359 (* 0.3 = 0.355076 loss)
I0429 00:59:03.675220 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.75231 (* 0.0272727 = 0.0750629 loss)
I0429 00:59:03.675235 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.83698 (* 0.0272727 = 0.0773722 loss)
I0429 00:59:03.675262 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.83842 (* 0.0272727 = 0.0774116 loss)
I0429 00:59:03.675277 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.62304 (* 0.0272727 = 0.0715374 loss)
I0429 00:59:03.675290 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.36046 (* 0.0272727 = 0.0643761 loss)
I0429 00:59:03.675304 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 2.91526 (* 0.0272727 = 0.0795072 loss)
I0429 00:59:03.675318 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 2.55338 (* 0.0272727 = 0.0696375 loss)
I0429 00:59:03.675331 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 2.73989 (* 0.0272727 = 0.0747241 loss)
I0429 00:59:03.675345 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.46397 (* 0.0272727 = 0.0126537 loss)
I0429 00:59:03.675359 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.543702 (* 0.0272727 = 0.0148282 loss)
I0429 00:59:03.675377 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.364059 (* 0.0272727 = 0.00992889 loss)
I0429 00:59:03.675391 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.408999 (* 0.0272727 = 0.0111545 loss)
I0429 00:59:03.675405 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.37105 (* 0.0272727 = 0.0101195 loss)
I0429 00:59:03.675420 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.371313 (* 0.0272727 = 0.0101267 loss)
I0429 00:59:03.675432 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.395362 (* 0.0272727 = 0.0107826 loss)
I0429 00:59:03.675446 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.45349 (* 0.0272727 = 0.0123679 loss)
I0429 00:59:03.675460 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0398622 (* 0.0272727 = 0.00108715 loss)
I0429 00:59:03.675490 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.0143908 (* 0.0272727 = 0.000392477 loss)
I0429 00:59:03.675505 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00657121 (* 0.0272727 = 0.000179215 loss)
I0429 00:59:03.675519 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00376642 (* 0.0272727 = 0.000102721 loss)
I0429 00:59:03.675534 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00200325 (* 0.0272727 = 5.46342e-05 loss)
I0429 00:59:03.675547 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00169366 (* 0.0272727 = 4.61907e-05 loss)
I0429 00:59:03.675559 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.109375
I0429 00:59:03.675571 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 00:59:03.675583 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 00:59:03.675595 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 00:59:03.675607 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.375
I0429 00:59:03.675618 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.625
I0429 00:59:03.675631 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.125
I0429 00:59:03.675642 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.375
I0429 00:59:03.675653 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.375
I0429 00:59:03.675665 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 00:59:03.675676 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 00:59:03.675688 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 00:59:03.675699 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 00:59:03.675711 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 00:59:03.675724 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 00:59:03.675734 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 0.875
I0429 00:59:03.675757 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 0.875
I0429 00:59:03.675770 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 00:59:03.675782 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 00:59:03.675794 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 00:59:03.675806 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 00:59:03.675817 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 00:59:03.675828 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 00:59:03.675840 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.664773
I0429 00:59:03.675853 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.3125
I0429 00:59:03.675866 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.71033 (* 1 = 2.71033 loss)
I0429 00:59:03.675880 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 1.0667 (* 1 = 1.0667 loss)
I0429 00:59:03.675894 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.13097 (* 0.0909091 = 0.193725 loss)
I0429 00:59:03.675907 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.65534 (* 0.0909091 = 0.241395 loss)
I0429 00:59:03.675921 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.82977 (* 0.0909091 = 0.257252 loss)
I0429 00:59:03.675935 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.1954 (* 0.0909091 = 0.199582 loss)
I0429 00:59:03.675948 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.99033 (* 0.0909091 = 0.180939 loss)
I0429 00:59:03.675962 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 2.39142 (* 0.0909091 = 0.217402 loss)
I0429 00:59:03.675976 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 2.31965 (* 0.0909091 = 0.210877 loss)
I0429 00:59:03.675990 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 2.2213 (* 0.0909091 = 0.201936 loss)
I0429 00:59:03.676003 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.338598 (* 0.0909091 = 0.0307817 loss)
I0429 00:59:03.676017 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.559429 (* 0.0909091 = 0.0508572 loss)
I0429 00:59:03.676031 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.373929 (* 0.0909091 = 0.0339935 loss)
I0429 00:59:03.676045 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.423764 (* 0.0909091 = 0.038524 loss)
I0429 00:59:03.676059 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.347228 (* 0.0909091 = 0.0315662 loss)
I0429 00:59:03.676072 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.27666 (* 0.0909091 = 0.0251509 loss)
I0429 00:59:03.676086 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.3734 (* 0.0909091 = 0.0339455 loss)
I0429 00:59:03.676100 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.393445 (* 0.0909091 = 0.0357677 loss)
I0429 00:59:03.676113 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0204225 (* 0.0909091 = 0.00185659 loss)
I0429 00:59:03.676127 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.006387 (* 0.0909091 = 0.000580636 loss)
I0429 00:59:03.676141 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00284895 (* 0.0909091 = 0.000258996 loss)
I0429 00:59:03.676159 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00182418 (* 0.0909091 = 0.000165835 loss)
I0429 00:59:03.676174 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00126208 (* 0.0909091 = 0.000114735 loss)
I0429 00:59:03.676188 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.00108731 (* 0.0909091 = 9.88466e-05 loss)
I0429 00:59:03.676200 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 00:59:03.676213 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 00:59:03.676223 6470 solver.cpp:245] Train net output #149: total_confidence = 0.000409159
I0429 00:59:03.676244 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000359521
I0429 00:59:03.676259 6470 sgd_solver.cpp:106] Iteration 20500, lr = 0.01
I0429 01:01:20.215961 6470 solver.cpp:229] Iteration 21000, loss = 9.04722
I0429 01:01:20.216152 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.166667
I0429 01:01:20.216174 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 01:01:20.216188 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.5
I0429 01:01:20.216200 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 01:01:20.216212 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 01:01:20.216224 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 01:01:20.216236 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 01:01:20.216248 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 01:01:20.216260 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 01:01:20.216272 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 01:01:20.216284 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 01:01:20.216295 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.875
I0429 01:01:20.216307 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 01:01:20.216322 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 01:01:20.216334 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 01:01:20.216346 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 01:01:20.216358 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 01:01:20.216370 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 01:01:20.216382 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 01:01:20.216393 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 01:01:20.216404 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 01:01:20.216416 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 01:01:20.216428 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 01:01:20.216439 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.767045
I0429 01:01:20.216450 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.354167
I0429 01:01:20.216466 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.79027 (* 0.3 = 0.837081 loss)
I0429 01:01:20.216481 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.838192 (* 0.3 = 0.251458 loss)
I0429 01:01:20.216495 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.83284 (* 0.0272727 = 0.0772593 loss)
I0429 01:01:20.216509 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.05026 (* 0.0272727 = 0.0559163 loss)
I0429 01:01:20.216522 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.87215 (* 0.0272727 = 0.0783315 loss)
I0429 01:01:20.216536 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 1.83444 (* 0.0272727 = 0.0500301 loss)
I0429 01:01:20.216550 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 1.96048 (* 0.0272727 = 0.0534676 loss)
I0429 01:01:20.216564 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.9283 (* 0.0272727 = 0.0525899 loss)
I0429 01:01:20.216578 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.19886 (* 0.0272727 = 0.0326962 loss)
I0429 01:01:20.216591 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.827977 (* 0.0272727 = 0.0225812 loss)
I0429 01:01:20.216604 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.502866 (* 0.0272727 = 0.0137145 loss)
I0429 01:01:20.216619 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.334895 (* 0.0272727 = 0.00913349 loss)
I0429 01:01:20.216634 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.507957 (* 0.0272727 = 0.0138534 loss)
I0429 01:01:20.216647 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.584575 (* 0.0272727 = 0.0159429 loss)
I0429 01:01:20.216686 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0151233 (* 0.0272727 = 0.000412453 loss)
I0429 01:01:20.216701 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0109593 (* 0.0272727 = 0.000298889 loss)
I0429 01:01:20.216717 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00305538 (* 0.0272727 = 8.33285e-05 loss)
I0429 01:01:20.216730 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00288785 (* 0.0272727 = 7.87596e-05 loss)
I0429 01:01:20.216744 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00275708 (* 0.0272727 = 7.5193e-05 loss)
I0429 01:01:20.216758 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00201985 (* 0.0272727 = 5.50867e-05 loss)
I0429 01:01:20.216771 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.0016056 (* 0.0272727 = 4.37892e-05 loss)
I0429 01:01:20.216785 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000692213 (* 0.0272727 = 1.88785e-05 loss)
I0429 01:01:20.216799 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000923461 (* 0.0272727 = 2.51853e-05 loss)
I0429 01:01:20.216814 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.00114076 (* 0.0272727 = 3.11115e-05 loss)
I0429 01:01:20.216826 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.145833
I0429 01:01:20.216838 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 01:01:20.216846 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.375
I0429 01:01:20.216855 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 01:01:20.216866 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.375
I0429 01:01:20.216877 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 01:01:20.216889 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 01:01:20.216900 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 01:01:20.216912 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 01:01:20.216924 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 01:01:20.216935 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 01:01:20.216948 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.875
I0429 01:01:20.216958 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 01:01:20.216970 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 01:01:20.216981 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 01:01:20.216994 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 01:01:20.217005 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 01:01:20.217016 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 01:01:20.217027 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 01:01:20.217038 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 01:01:20.217049 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 01:01:20.217061 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 01:01:20.217072 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 01:01:20.217084 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.755682
I0429 01:01:20.217095 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.333333
I0429 01:01:20.217109 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.84304 (* 0.3 = 0.852912 loss)
I0429 01:01:20.217123 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.889321 (* 0.3 = 0.266796 loss)
I0429 01:01:20.217136 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.74138 (* 0.0272727 = 0.0747649 loss)
I0429 01:01:20.217150 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.20834 (* 0.0272727 = 0.0602275 loss)
I0429 01:01:20.217180 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.91122 (* 0.0272727 = 0.0793969 loss)
I0429 01:01:20.217195 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.13265 (* 0.0272727 = 0.0581633 loss)
I0429 01:01:20.217208 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.03914 (* 0.0272727 = 0.0556128 loss)
I0429 01:01:20.217222 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.91337 (* 0.0272727 = 0.0521829 loss)
I0429 01:01:20.217236 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.18566 (* 0.0272727 = 0.0323361 loss)
I0429 01:01:20.217248 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.897779 (* 0.0272727 = 0.0244849 loss)
I0429 01:01:20.217262 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.543617 (* 0.0272727 = 0.0148259 loss)
I0429 01:01:20.217275 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.436364 (* 0.0272727 = 0.0119008 loss)
I0429 01:01:20.217289 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.554508 (* 0.0272727 = 0.0151229 loss)
I0429 01:01:20.217303 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.632368 (* 0.0272727 = 0.0172464 loss)
I0429 01:01:20.217317 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.0198178 (* 0.0272727 = 0.000540485 loss)
I0429 01:01:20.217331 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00953825 (* 0.0272727 = 0.000260134 loss)
I0429 01:01:20.217345 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.00570277 (* 0.0272727 = 0.00015553 loss)
I0429 01:01:20.217360 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00263674 (* 0.0272727 = 7.19111e-05 loss)
I0429 01:01:20.217376 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00368842 (* 0.0272727 = 0.000100593 loss)
I0429 01:01:20.217391 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00124487 (* 0.0272727 = 3.3951e-05 loss)
I0429 01:01:20.217406 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.001657 (* 0.0272727 = 4.51908e-05 loss)
I0429 01:01:20.217419 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000971773 (* 0.0272727 = 2.65029e-05 loss)
I0429 01:01:20.217432 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000397266 (* 0.0272727 = 1.08345e-05 loss)
I0429 01:01:20.217447 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000317627 (* 0.0272727 = 8.66256e-06 loss)
I0429 01:01:20.217458 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.125
I0429 01:01:20.217470 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 01:01:20.217483 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.375
I0429 01:01:20.217494 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 01:01:20.217505 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 01:01:20.217517 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 01:01:20.217530 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 01:01:20.217541 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 01:01:20.217553 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 01:01:20.217564 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 01:01:20.217576 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 01:01:20.217587 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.875
I0429 01:01:20.217599 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 01:01:20.217612 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 01:01:20.217623 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 01:01:20.217634 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 01:01:20.217655 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 01:01:20.217669 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 01:01:20.217680 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 01:01:20.217691 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 01:01:20.217703 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 01:01:20.217715 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 01:01:20.217726 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 01:01:20.217737 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0429 01:01:20.217749 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.416667
I0429 01:01:20.217763 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.69157 (* 1 = 2.69157 loss)
I0429 01:01:20.217777 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.839548 (* 1 = 0.839548 loss)
I0429 01:01:20.217792 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.37037 (* 0.0909091 = 0.215488 loss)
I0429 01:01:20.217805 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.28049 (* 0.0909091 = 0.207318 loss)
I0429 01:01:20.217818 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.79654 (* 0.0909091 = 0.254231 loss)
I0429 01:01:20.217831 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.17009 (* 0.0909091 = 0.197281 loss)
I0429 01:01:20.217845 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.10637 (* 0.0909091 = 0.191488 loss)
I0429 01:01:20.217859 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.88809 (* 0.0909091 = 0.171645 loss)
I0429 01:01:20.217872 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.19294 (* 0.0909091 = 0.108449 loss)
I0429 01:01:20.217885 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.83416 (* 0.0909091 = 0.0758327 loss)
I0429 01:01:20.217898 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.563324 (* 0.0909091 = 0.0512113 loss)
I0429 01:01:20.217912 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.398886 (* 0.0909091 = 0.0362623 loss)
I0429 01:01:20.217926 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.476817 (* 0.0909091 = 0.043347 loss)
I0429 01:01:20.217939 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.510322 (* 0.0909091 = 0.0463929 loss)
I0429 01:01:20.217953 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.035129 (* 0.0909091 = 0.00319354 loss)
I0429 01:01:20.217967 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0288793 (* 0.0909091 = 0.00262539 loss)
I0429 01:01:20.217980 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0171069 (* 0.0909091 = 0.00155517 loss)
I0429 01:01:20.217994 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.00649193 (* 0.0909091 = 0.000590175 loss)
I0429 01:01:20.218008 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.00427874 (* 0.0909091 = 0.000388977 loss)
I0429 01:01:20.218021 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00246776 (* 0.0909091 = 0.000224341 loss)
I0429 01:01:20.218035 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.0016208 (* 0.0909091 = 0.000147345 loss)
I0429 01:01:20.218049 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000910318 (* 0.0909091 = 8.27562e-05 loss)
I0429 01:01:20.218062 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00044614 (* 0.0909091 = 4.05582e-05 loss)
I0429 01:01:20.218076 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000205476 (* 0.0909091 = 1.86796e-05 loss)
I0429 01:01:20.218088 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 01:01:20.218101 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 01:01:20.218121 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00816474
I0429 01:01:20.218133 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.0390538
I0429 01:01:20.218147 6470 sgd_solver.cpp:106] Iteration 21000, lr = 0.01
I0429 01:01:40.758960 6470 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.7127 > 30) by scale factor 0.976795
I0429 01:03:36.730485 6470 solver.cpp:229] Iteration 21500, loss = 9.07358
I0429 01:03:36.730655 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.025
I0429 01:03:36.730677 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 01:03:36.730690 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 01:03:36.730702 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 01:03:36.730715 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 01:03:36.730726 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.5
I0429 01:03:36.730738 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.625
I0429 01:03:36.730751 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 01:03:36.730762 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 01:03:36.730774 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 01:03:36.730785 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 01:03:36.730798 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 01:03:36.730808 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 01:03:36.730820 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 01:03:36.730831 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 01:03:36.730844 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 01:03:36.730854 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 01:03:36.730866 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 01:03:36.730878 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 01:03:36.730890 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 01:03:36.730901 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 01:03:36.730913 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 01:03:36.730926 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 01:03:36.730937 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.767045
I0429 01:03:36.730948 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.275
I0429 01:03:36.730965 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.91821 (* 0.3 = 0.875462 loss)
I0429 01:03:36.730979 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.732744 (* 0.3 = 0.219823 loss)
I0429 01:03:36.730996 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.62925 (* 0.0272727 = 0.0717068 loss)
I0429 01:03:36.731021 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.26698 (* 0.0272727 = 0.0890994 loss)
I0429 01:03:36.731050 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.82887 (* 0.0272727 = 0.0771511 loss)
I0429 01:03:36.731072 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.66796 (* 0.0272727 = 0.0727624 loss)
I0429 01:03:36.731096 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 1.70092 (* 0.0272727 = 0.0463888 loss)
I0429 01:03:36.731122 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.674 (* 0.0272727 = 0.0456545 loss)
I0429 01:03:36.731148 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.952251 (* 0.0272727 = 0.0259705 loss)
I0429 01:03:36.731165 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.0512814 (* 0.0272727 = 0.00139858 loss)
I0429 01:03:36.731180 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.00951145 (* 0.0272727 = 0.000259403 loss)
I0429 01:03:36.731194 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.00728885 (* 0.0272727 = 0.000198787 loss)
I0429 01:03:36.731209 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.00604771 (* 0.0272727 = 0.000164938 loss)
I0429 01:03:36.731222 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.00330781 (* 0.0272727 = 9.02129e-05 loss)
I0429 01:03:36.731236 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.0030747 (* 0.0272727 = 8.38553e-05 loss)
I0429 01:03:36.731271 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00223291 (* 0.0272727 = 6.08975e-05 loss)
I0429 01:03:36.731287 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00125726 (* 0.0272727 = 3.4289e-05 loss)
I0429 01:03:36.731302 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00101932 (* 0.0272727 = 2.77996e-05 loss)
I0429 01:03:36.731319 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.000942092 (* 0.0272727 = 2.56934e-05 loss)
I0429 01:03:36.731333 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000810864 (* 0.0272727 = 2.21145e-05 loss)
I0429 01:03:36.731348 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.00102735 (* 0.0272727 = 2.80186e-05 loss)
I0429 01:03:36.731361 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000675195 (* 0.0272727 = 1.84144e-05 loss)
I0429 01:03:36.731375 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000659412 (* 0.0272727 = 1.7984e-05 loss)
I0429 01:03:36.731389 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 0.000481237 (* 0.0272727 = 1.31247e-05 loss)
I0429 01:03:36.731400 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.1
I0429 01:03:36.731413 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.25
I0429 01:03:36.731425 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 01:03:36.731436 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 01:03:36.731448 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 01:03:36.731459 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 01:03:36.731487 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 01:03:36.731500 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 01:03:36.731513 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 1
I0429 01:03:36.731523 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 01:03:36.731535 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 01:03:36.731547 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 01:03:36.731559 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 01:03:36.731570 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 01:03:36.731581 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 01:03:36.731593 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 01:03:36.731605 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 01:03:36.731612 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 01:03:36.731621 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 01:03:36.731632 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 01:03:36.731644 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 01:03:36.731668 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 01:03:36.731695 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 01:03:36.731722 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.789773
I0429 01:03:36.731750 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.325
I0429 01:03:36.731767 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.85717 (* 0.3 = 0.85715 loss)
I0429 01:03:36.731781 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.725937 (* 0.3 = 0.217781 loss)
I0429 01:03:36.731796 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.5335 (* 0.0272727 = 0.0690954 loss)
I0429 01:03:36.731809 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.97939 (* 0.0272727 = 0.081256 loss)
I0429 01:03:36.731837 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.22998 (* 0.0272727 = 0.0880905 loss)
I0429 01:03:36.731853 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.50571 (* 0.0272727 = 0.0683375 loss)
I0429 01:03:36.731866 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 1.78141 (* 0.0272727 = 0.0485838 loss)
I0429 01:03:36.731879 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.47495 (* 0.0272727 = 0.040226 loss)
I0429 01:03:36.731894 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.736397 (* 0.0272727 = 0.0200836 loss)
I0429 01:03:36.731907 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.112275 (* 0.0272727 = 0.00306205 loss)
I0429 01:03:36.731921 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.00462244 (* 0.0272727 = 0.000126066 loss)
I0429 01:03:36.731935 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.00274383 (* 0.0272727 = 7.48316e-05 loss)
I0429 01:03:36.731948 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0026079 (* 0.0272727 = 7.11246e-05 loss)
I0429 01:03:36.731962 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00173156 (* 0.0272727 = 4.72244e-05 loss)
I0429 01:03:36.731976 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00163583 (* 0.0272727 = 4.46135e-05 loss)
I0429 01:03:36.731989 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.000936078 (* 0.0272727 = 2.55294e-05 loss)
I0429 01:03:36.732003 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.000584258 (* 0.0272727 = 1.59343e-05 loss)
I0429 01:03:36.732017 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.000869484 (* 0.0272727 = 2.37132e-05 loss)
I0429 01:03:36.732030 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.000614354 (* 0.0272727 = 1.67551e-05 loss)
I0429 01:03:36.732044 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.000381637 (* 0.0272727 = 1.04083e-05 loss)
I0429 01:03:36.732059 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000233842 (* 0.0272727 = 6.3775e-06 loss)
I0429 01:03:36.732072 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000255975 (* 0.0272727 = 6.98114e-06 loss)
I0429 01:03:36.732085 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.000270119 (* 0.0272727 = 7.36688e-06 loss)
I0429 01:03:36.732100 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 9.51487e-05 (* 0.0272727 = 2.59497e-06 loss)
I0429 01:03:36.732111 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.15
I0429 01:03:36.732123 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.5
I0429 01:03:36.732136 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 01:03:36.732146 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 01:03:36.732158 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 01:03:36.732169 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.5
I0429 01:03:36.732182 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.625
I0429 01:03:36.732193 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 01:03:36.732204 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 1
I0429 01:03:36.732216 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 01:03:36.732228 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 01:03:36.732239 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 01:03:36.732250 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 01:03:36.732261 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 01:03:36.732273 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 01:03:36.732285 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 01:03:36.732296 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 01:03:36.732316 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 01:03:36.732329 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 01:03:36.732342 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 01:03:36.732353 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 01:03:36.732367 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 01:03:36.732380 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 01:03:36.732393 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.795455
I0429 01:03:36.732404 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.4
I0429 01:03:36.732419 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.57967 (* 1 = 2.57967 loss)
I0429 01:03:36.732431 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.642329 (* 1 = 0.642329 loss)
I0429 01:03:36.732445 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 1.61142 (* 0.0909091 = 0.146492 loss)
I0429 01:03:36.732460 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.80315 (* 0.0909091 = 0.254832 loss)
I0429 01:03:36.732472 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 3.00723 (* 0.0909091 = 0.273384 loss)
I0429 01:03:36.732486 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.433 (* 0.0909091 = 0.221182 loss)
I0429 01:03:36.732499 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.58047 (* 0.0909091 = 0.143679 loss)
I0429 01:03:36.732513 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.54418 (* 0.0909091 = 0.14038 loss)
I0429 01:03:36.732527 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.626121 (* 0.0909091 = 0.0569201 loss)
I0429 01:03:36.732540 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.06386 (* 0.0909091 = 0.00580545 loss)
I0429 01:03:36.732554 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.00103302 (* 0.0909091 = 9.39111e-05 loss)
I0429 01:03:36.732568 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.000679974 (* 0.0909091 = 6.18158e-05 loss)
I0429 01:03:36.732581 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.000776265 (* 0.0909091 = 7.05696e-05 loss)
I0429 01:03:36.732595 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.000491765 (* 0.0909091 = 4.47059e-05 loss)
I0429 01:03:36.732609 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.000402875 (* 0.0909091 = 3.6625e-05 loss)
I0429 01:03:36.732622 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.000229453 (* 0.0909091 = 2.08593e-05 loss)
I0429 01:03:36.732636 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000186232 (* 0.0909091 = 1.69302e-05 loss)
I0429 01:03:36.732650 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000141929 (* 0.0909091 = 1.29026e-05 loss)
I0429 01:03:36.732663 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000119934 (* 0.0909091 = 1.09031e-05 loss)
I0429 01:03:36.732677 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 9.12039e-05 (* 0.0909091 = 8.29126e-06 loss)
I0429 01:03:36.732692 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 7.0696e-05 (* 0.0909091 = 6.42691e-06 loss)
I0429 01:03:36.732705 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 6.00901e-05 (* 0.0909091 = 5.46274e-06 loss)
I0429 01:03:36.732723 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 5.27042e-05 (* 0.0909091 = 4.79129e-06 loss)
I0429 01:03:36.732738 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 4.66837e-05 (* 0.0909091 = 4.24397e-06 loss)
I0429 01:03:36.732749 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 01:03:36.732760 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 01:03:36.732771 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00389824
I0429 01:03:36.732794 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.0280572
I0429 01:03:36.732808 6470 sgd_solver.cpp:106] Iteration 21500, lr = 0.01
I0429 01:05:53.413432 6470 solver.cpp:229] Iteration 22000, loss = 9.05501
I0429 01:05:53.413647 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.14
I0429 01:05:53.413671 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.375
I0429 01:05:53.413686 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 01:05:53.413697 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 01:05:53.413709 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 01:05:53.413722 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 01:05:53.413734 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.375
I0429 01:05:53.413745 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.625
I0429 01:05:53.413758 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.625
I0429 01:05:53.413770 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.875
I0429 01:05:53.413781 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.875
I0429 01:05:53.413794 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 01:05:53.413805 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 01:05:53.413816 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 01:05:53.413828 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 01:05:53.413841 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 01:05:53.413851 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 01:05:53.413863 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 01:05:53.413875 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 01:05:53.413887 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 01:05:53.413898 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 01:05:53.413909 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 01:05:53.413921 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 01:05:53.413933 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.744318
I0429 01:05:53.413945 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.42
I0429 01:05:53.413961 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.70695 (* 0.3 = 0.812086 loss)
I0429 01:05:53.413975 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.851525 (* 0.3 = 0.255458 loss)
I0429 01:05:53.413990 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.63348 (* 0.0272727 = 0.0718222 loss)
I0429 01:05:53.414005 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 3.05312 (* 0.0272727 = 0.0832668 loss)
I0429 01:05:53.414018 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.96661 (* 0.0272727 = 0.0809077 loss)
I0429 01:05:53.414032 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.39363 (* 0.0272727 = 0.0652808 loss)
I0429 01:05:53.414046 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.19363 (* 0.0272727 = 0.0598263 loss)
I0429 01:05:53.414059 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.9151 (* 0.0272727 = 0.05223 loss)
I0429 01:05:53.414073 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.23969 (* 0.0272727 = 0.0338098 loss)
I0429 01:05:53.414088 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 1.16904 (* 0.0272727 = 0.0318828 loss)
I0429 01:05:53.414100 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.746702 (* 0.0272727 = 0.0203646 loss)
I0429 01:05:53.414114 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.75843 (* 0.0272727 = 0.0206845 loss)
I0429 01:05:53.414129 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.000839257 (* 0.0272727 = 2.28888e-05 loss)
I0429 01:05:53.414144 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.000200715 (* 0.0272727 = 5.47405e-06 loss)
I0429 01:05:53.414158 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.000173359 (* 0.0272727 = 4.72796e-06 loss)
I0429 01:05:53.414197 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 5.04877e-05 (* 0.0272727 = 1.37694e-06 loss)
I0429 01:05:53.414213 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 3.78207e-05 (* 0.0272727 = 1.03147e-06 loss)
I0429 01:05:53.414227 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 2.1622e-05 (* 0.0272727 = 5.89691e-07 loss)
I0429 01:05:53.414242 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 1.89844e-05 (* 0.0272727 = 5.17757e-07 loss)
I0429 01:05:53.414255 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 2.56605e-05 (* 0.0272727 = 6.99832e-07 loss)
I0429 01:05:53.414269 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 1.60342e-05 (* 0.0272727 = 4.37295e-07 loss)
I0429 01:05:53.414283 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 1.74349e-05 (* 0.0272727 = 4.75496e-07 loss)
I0429 01:05:53.414296 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 6.33306e-06 (* 0.0272727 = 1.7272e-07 loss)
I0429 01:05:53.414310 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 7.28676e-06 (* 0.0272727 = 1.9873e-07 loss)
I0429 01:05:53.414326 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.14
I0429 01:05:53.414340 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 01:05:53.414351 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0
I0429 01:05:53.414362 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0.125
I0429 01:05:53.414374 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.25
I0429 01:05:53.414386 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.5
I0429 01:05:53.414398 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.375
I0429 01:05:53.414410 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 01:05:53.414422 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.625
I0429 01:05:53.414433 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.875
I0429 01:05:53.414445 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.875
I0429 01:05:53.414456 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 01:05:53.414469 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 01:05:53.414480 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 01:05:53.414491 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 01:05:53.414502 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 01:05:53.414515 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 01:05:53.414525 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 01:05:53.414536 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 01:05:53.414547 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 01:05:53.414559 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 01:05:53.414571 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 01:05:53.414582 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 01:05:53.414593 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.744318
I0429 01:05:53.414605 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.38
I0429 01:05:53.414619 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.80655 (* 0.3 = 0.841966 loss)
I0429 01:05:53.414633 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.858461 (* 0.3 = 0.257538 loss)
I0429 01:05:53.414650 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.48562 (* 0.0272727 = 0.0677897 loss)
I0429 01:05:53.414664 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.96151 (* 0.0272727 = 0.0807685 loss)
I0429 01:05:53.414690 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.83709 (* 0.0272727 = 0.0773752 loss)
I0429 01:05:53.414705 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.41825 (* 0.0272727 = 0.0659523 loss)
I0429 01:05:53.414718 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.61955 (* 0.0272727 = 0.0714424 loss)
I0429 01:05:53.414732 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.88538 (* 0.0272727 = 0.0514194 loss)
I0429 01:05:53.414746 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.17079 (* 0.0272727 = 0.0319307 loss)
I0429 01:05:53.414759 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 1.12888 (* 0.0272727 = 0.0307876 loss)
I0429 01:05:53.414773 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.657865 (* 0.0272727 = 0.0179418 loss)
I0429 01:05:53.414788 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.711066 (* 0.0272727 = 0.0193927 loss)
I0429 01:05:53.414801 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.0023614 (* 0.0272727 = 6.44019e-05 loss)
I0429 01:05:53.414815 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00137811 (* 0.0272727 = 3.75848e-05 loss)
I0429 01:05:53.414829 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00110476 (* 0.0272727 = 3.01297e-05 loss)
I0429 01:05:53.414844 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.000637665 (* 0.0272727 = 1.73909e-05 loss)
I0429 01:05:53.414857 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.000555392 (* 0.0272727 = 1.51471e-05 loss)
I0429 01:05:53.414871 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.00117689 (* 0.0272727 = 3.20969e-05 loss)
I0429 01:05:53.414885 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00116604 (* 0.0272727 = 3.18011e-05 loss)
I0429 01:05:53.414899 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00133013 (* 0.0272727 = 3.62762e-05 loss)
I0429 01:05:53.414913 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.000723306 (* 0.0272727 = 1.97265e-05 loss)
I0429 01:05:53.414927 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00137968 (* 0.0272727 = 3.76277e-05 loss)
I0429 01:05:53.414942 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00106388 (* 0.0272727 = 2.90148e-05 loss)
I0429 01:05:53.414954 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.000462708 (* 0.0272727 = 1.26193e-05 loss)
I0429 01:05:53.414966 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.14
I0429 01:05:53.414978 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.375
I0429 01:05:53.414990 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0
I0429 01:05:53.415002 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.125
I0429 01:05:53.415014 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.25
I0429 01:05:53.415025 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.25
I0429 01:05:53.415037 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.375
I0429 01:05:53.415048 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 01:05:53.415060 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.625
I0429 01:05:53.415071 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.875
I0429 01:05:53.415083 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.875
I0429 01:05:53.415094 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 01:05:53.415107 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 01:05:53.415117 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 01:05:53.415129 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 01:05:53.415140 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 01:05:53.415151 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 01:05:53.415172 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 01:05:53.415185 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 01:05:53.415197 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 01:05:53.415208 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 01:05:53.415220 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 01:05:53.415231 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 01:05:53.415243 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.755682
I0429 01:05:53.415254 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.36
I0429 01:05:53.415268 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.74304 (* 1 = 2.74304 loss)
I0429 01:05:53.415282 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.812312 (* 1 = 0.812312 loss)
I0429 01:05:53.415297 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.26887 (* 0.0909091 = 0.206261 loss)
I0429 01:05:53.415310 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.79588 (* 0.0909091 = 0.254171 loss)
I0429 01:05:53.415324 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.74331 (* 0.0909091 = 0.249392 loss)
I0429 01:05:53.415338 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 2.65975 (* 0.0909091 = 0.241796 loss)
I0429 01:05:53.415348 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 2.32531 (* 0.0909091 = 0.211392 loss)
I0429 01:05:53.415357 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.96306 (* 0.0909091 = 0.17846 loss)
I0429 01:05:53.415374 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.11419 (* 0.0909091 = 0.10129 loss)
I0429 01:05:53.415388 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 1.25433 (* 0.0909091 = 0.11403 loss)
I0429 01:05:53.415401 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.825594 (* 0.0909091 = 0.075054 loss)
I0429 01:05:53.415416 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.823568 (* 0.0909091 = 0.0748698 loss)
I0429 01:05:53.415429 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00126173 (* 0.0909091 = 0.000114703 loss)
I0429 01:05:53.415443 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.000781905 (* 0.0909091 = 7.10822e-05 loss)
I0429 01:05:53.415457 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.000726821 (* 0.0909091 = 6.60747e-05 loss)
I0429 01:05:53.415490 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.000554673 (* 0.0909091 = 5.04249e-05 loss)
I0429 01:05:53.415506 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000456976 (* 0.0909091 = 4.15433e-05 loss)
I0429 01:05:53.415519 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000332536 (* 0.0909091 = 3.02305e-05 loss)
I0429 01:05:53.415534 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000334021 (* 0.0909091 = 3.03655e-05 loss)
I0429 01:05:53.415547 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.000312583 (* 0.0909091 = 2.84167e-05 loss)
I0429 01:05:53.415560 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00035679 (* 0.0909091 = 3.24355e-05 loss)
I0429 01:05:53.415575 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000374464 (* 0.0909091 = 3.40421e-05 loss)
I0429 01:05:53.415587 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000416939 (* 0.0909091 = 3.79035e-05 loss)
I0429 01:05:53.415601 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000329772 (* 0.0909091 = 2.99793e-05 loss)
I0429 01:05:53.415613 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 01:05:53.415624 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 01:05:53.415647 6470 solver.cpp:245] Train net output #149: total_confidence = 0.00142536
I0429 01:05:53.415660 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.00146244
I0429 01:05:53.415673 6470 sgd_solver.cpp:106] Iteration 22000, lr = 0.01
I0429 01:08:10.037513 6470 solver.cpp:229] Iteration 22500, loss = 9.01242
I0429 01:08:10.037660 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.108696
I0429 01:08:10.037680 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 01:08:10.037694 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 01:08:10.037706 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 01:08:10.037719 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.125
I0429 01:08:10.037730 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 01:08:10.037742 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.5
I0429 01:08:10.037755 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 01:08:10.037765 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.875
I0429 01:08:10.037777 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 1
I0429 01:08:10.037789 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 1
I0429 01:08:10.037801 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 1
I0429 01:08:10.037812 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 1
I0429 01:08:10.037823 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 1
I0429 01:08:10.037835 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 01:08:10.037847 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 01:08:10.037859 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 01:08:10.037870 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 01:08:10.037883 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 01:08:10.037894 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 01:08:10.037904 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 01:08:10.037916 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 01:08:10.037927 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 01:08:10.037940 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.75
I0429 01:08:10.037951 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.217391
I0429 01:08:10.037967 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.9148 (* 0.3 = 0.874441 loss)
I0429 01:08:10.037982 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.897343 (* 0.3 = 0.269203 loss)
I0429 01:08:10.037997 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.75437 (* 0.0272727 = 0.0751192 loss)
I0429 01:08:10.038010 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.74165 (* 0.0272727 = 0.0747722 loss)
I0429 01:08:10.038023 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 3.60768 (* 0.0272727 = 0.0983912 loss)
I0429 01:08:10.038038 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 3.01071 (* 0.0272727 = 0.0821103 loss)
I0429 01:08:10.038051 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.99551 (* 0.0272727 = 0.0816958 loss)
I0429 01:08:10.038064 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.53538 (* 0.0272727 = 0.041874 loss)
I0429 01:08:10.038077 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 1.0155 (* 0.0272727 = 0.0276955 loss)
I0429 01:08:10.038091 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.928754 (* 0.0272727 = 0.0253296 loss)
I0429 01:08:10.038105 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.0134325 (* 0.0272727 = 0.000366341 loss)
I0429 01:08:10.038120 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.00534996 (* 0.0272727 = 0.000145908 loss)
I0429 01:08:10.038133 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.00178435 (* 0.0272727 = 4.8664e-05 loss)
I0429 01:08:10.038147 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.00195205 (* 0.0272727 = 5.32377e-05 loss)
I0429 01:08:10.038161 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 0.00212147 (* 0.0272727 = 5.78583e-05 loss)
I0429 01:08:10.038194 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.00125697 (* 0.0272727 = 3.42809e-05 loss)
I0429 01:08:10.038210 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.00131747 (* 0.0272727 = 3.59309e-05 loss)
I0429 01:08:10.038223 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00203396 (* 0.0272727 = 5.54717e-05 loss)
I0429 01:08:10.038238 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.000455189 (* 0.0272727 = 1.24142e-05 loss)
I0429 01:08:10.038251 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.000285648 (* 0.0272727 = 7.79041e-06 loss)
I0429 01:08:10.038264 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000201998 (* 0.0272727 = 5.50903e-06 loss)
I0429 01:08:10.038278 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000173546 (* 0.0272727 = 4.73307e-06 loss)
I0429 01:08:10.038291 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 0.000106565 (* 0.0272727 = 2.90631e-06 loss)
I0429 01:08:10.038305 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 5.77832e-05 (* 0.0272727 = 1.5759e-06 loss)
I0429 01:08:10.038321 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.0652174
I0429 01:08:10.038333 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.125
I0429 01:08:10.038346 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 01:08:10.038357 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 01:08:10.038368 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.125
I0429 01:08:10.038378 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.25
I0429 01:08:10.038384 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.5
I0429 01:08:10.038393 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.625
I0429 01:08:10.038404 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.875
I0429 01:08:10.038416 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 1
I0429 01:08:10.038427 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 1
I0429 01:08:10.038439 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 1
I0429 01:08:10.038450 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 1
I0429 01:08:10.038461 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 1
I0429 01:08:10.038472 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 01:08:10.038485 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 01:08:10.038496 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 01:08:10.038506 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 01:08:10.038517 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 01:08:10.038529 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 01:08:10.038540 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 01:08:10.038552 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 01:08:10.038563 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 01:08:10.038574 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.744318
I0429 01:08:10.038586 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.195652
I0429 01:08:10.038600 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.8862 (* 0.3 = 0.865861 loss)
I0429 01:08:10.038614 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.87191 (* 0.3 = 0.261573 loss)
I0429 01:08:10.038627 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.57683 (* 0.0272727 = 0.0702772 loss)
I0429 01:08:10.038641 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.94169 (* 0.0272727 = 0.0802278 loss)
I0429 01:08:10.038666 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 3.0327 (* 0.0272727 = 0.0827101 loss)
I0429 01:08:10.038686 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 3.06643 (* 0.0272727 = 0.0836299 loss)
I0429 01:08:10.038698 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.83244 (* 0.0272727 = 0.0772484 loss)
I0429 01:08:10.038712 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.72155 (* 0.0272727 = 0.0469515 loss)
I0429 01:08:10.038725 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 1.18046 (* 0.0272727 = 0.0321943 loss)
I0429 01:08:10.038739 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.65285 (* 0.0272727 = 0.017805 loss)
I0429 01:08:10.038753 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.0185566 (* 0.0272727 = 0.00050609 loss)
I0429 01:08:10.038768 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.00740087 (* 0.0272727 = 0.000201842 loss)
I0429 01:08:10.038781 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.00542838 (* 0.0272727 = 0.000148047 loss)
I0429 01:08:10.038795 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.00372431 (* 0.0272727 = 0.000101572 loss)
I0429 01:08:10.038808 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.00286475 (* 0.0272727 = 7.81297e-05 loss)
I0429 01:08:10.038822 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.00208832 (* 0.0272727 = 5.69541e-05 loss)
I0429 01:08:10.038836 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0019374 (* 0.0272727 = 5.28381e-05 loss)
I0429 01:08:10.038849 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0024294 (* 0.0272727 = 6.62563e-05 loss)
I0429 01:08:10.038862 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.00376174 (* 0.0272727 = 0.000102593 loss)
I0429 01:08:10.038877 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00144113 (* 0.0272727 = 3.93037e-05 loss)
I0429 01:08:10.038890 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.0018143 (* 0.0272727 = 4.94809e-05 loss)
I0429 01:08:10.038903 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.00269215 (* 0.0272727 = 7.34224e-05 loss)
I0429 01:08:10.038918 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 0.00138576 (* 0.0272727 = 3.77935e-05 loss)
I0429 01:08:10.038930 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 0.00201425 (* 0.0272727 = 5.4934e-05 loss)
I0429 01:08:10.038943 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.0652174
I0429 01:08:10.038954 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.125
I0429 01:08:10.038966 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 01:08:10.038978 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0
I0429 01:08:10.038990 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.125
I0429 01:08:10.039001 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.125
I0429 01:08:10.039013 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.5
I0429 01:08:10.039026 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.625
I0429 01:08:10.039036 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.875
I0429 01:08:10.039048 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 1
I0429 01:08:10.039059 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 1
I0429 01:08:10.039070 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 1
I0429 01:08:10.039083 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 1
I0429 01:08:10.039093 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 1
I0429 01:08:10.039104 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 01:08:10.039116 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 01:08:10.039127 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 01:08:10.039149 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 01:08:10.039161 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 01:08:10.039172 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 01:08:10.039183 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 01:08:10.039196 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 01:08:10.039206 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 01:08:10.039217 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.75
I0429 01:08:10.039229 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.347826
I0429 01:08:10.039243 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.73325 (* 1 = 2.73325 loss)
I0429 01:08:10.039258 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.809171 (* 1 = 0.809171 loss)
I0429 01:08:10.039271 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.41149 (* 0.0909091 = 0.219226 loss)
I0429 01:08:10.039284 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.31248 (* 0.0909091 = 0.210225 loss)
I0429 01:08:10.039299 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.99843 (* 0.0909091 = 0.272584 loss)
I0429 01:08:10.039311 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 3.03874 (* 0.0909091 = 0.276249 loss)
I0429 01:08:10.039325 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 3.09013 (* 0.0909091 = 0.280921 loss)
I0429 01:08:10.039338 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.69502 (* 0.0909091 = 0.154093 loss)
I0429 01:08:10.039352 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 1.02941 (* 0.0909091 = 0.093583 loss)
I0429 01:08:10.039368 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.526136 (* 0.0909091 = 0.0478306 loss)
I0429 01:08:10.039383 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.00605349 (* 0.0909091 = 0.000550318 loss)
I0429 01:08:10.039397 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.00286375 (* 0.0909091 = 0.000260341 loss)
I0429 01:08:10.039412 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.00094343 (* 0.0909091 = 8.57664e-05 loss)
I0429 01:08:10.039425 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.000687723 (* 0.0909091 = 6.25202e-05 loss)
I0429 01:08:10.039439 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.000764178 (* 0.0909091 = 6.94707e-05 loss)
I0429 01:08:10.039453 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.000780156 (* 0.0909091 = 7.09233e-05 loss)
I0429 01:08:10.039481 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.000888501 (* 0.0909091 = 8.07729e-05 loss)
I0429 01:08:10.039500 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.000895814 (* 0.0909091 = 8.14376e-05 loss)
I0429 01:08:10.039515 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.000590268 (* 0.0909091 = 5.36607e-05 loss)
I0429 01:08:10.039527 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00048068 (* 0.0909091 = 4.36982e-05 loss)
I0429 01:08:10.039541 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.000402566 (* 0.0909091 = 3.65969e-05 loss)
I0429 01:08:10.039556 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.000454243 (* 0.0909091 = 4.12949e-05 loss)
I0429 01:08:10.039569 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.000408877 (* 0.0909091 = 3.71706e-05 loss)
I0429 01:08:10.039582 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000347868 (* 0.0909091 = 3.16244e-05 loss)
I0429 01:08:10.039594 6470 solver.cpp:245] Train net output #147: total_accuracy = 0
I0429 01:08:10.039607 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0
I0429 01:08:10.039628 6470 solver.cpp:245] Train net output #149: total_confidence = 8.18418e-05
I0429 01:08:10.039643 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.000315324
I0429 01:08:10.039654 6470 sgd_solver.cpp:106] Iteration 22500, lr = 0.01
I0429 01:10:26.694229 6470 solver.cpp:229] Iteration 23000, loss = 8.90115
I0429 01:10:26.694401 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12
I0429 01:10:26.694422 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.125
I0429 01:10:26.694435 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0.125
I0429 01:10:26.694448 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0.125
I0429 01:10:26.694460 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.375
I0429 01:10:26.694473 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.375
I0429 01:10:26.694484 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.75
I0429 01:10:26.694496 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.75
I0429 01:10:26.694509 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 0.75
I0429 01:10:26.694520 6470 solver.cpp:245] Train net output #9: loss1/accuracy09 = 0.75
I0429 01:10:26.694532 6470 solver.cpp:245] Train net output #10: loss1/accuracy10 = 0.75
I0429 01:10:26.694543 6470 solver.cpp:245] Train net output #11: loss1/accuracy11 = 0.75
I0429 01:10:26.694555 6470 solver.cpp:245] Train net output #12: loss1/accuracy12 = 0.875
I0429 01:10:26.694567 6470 solver.cpp:245] Train net output #13: loss1/accuracy13 = 0.875
I0429 01:10:26.694579 6470 solver.cpp:245] Train net output #14: loss1/accuracy14 = 1
I0429 01:10:26.694591 6470 solver.cpp:245] Train net output #15: loss1/accuracy15 = 1
I0429 01:10:26.694604 6470 solver.cpp:245] Train net output #16: loss1/accuracy16 = 1
I0429 01:10:26.694615 6470 solver.cpp:245] Train net output #17: loss1/accuracy17 = 1
I0429 01:10:26.694628 6470 solver.cpp:245] Train net output #18: loss1/accuracy18 = 1
I0429 01:10:26.694639 6470 solver.cpp:245] Train net output #19: loss1/accuracy19 = 1
I0429 01:10:26.694650 6470 solver.cpp:245] Train net output #20: loss1/accuracy20 = 1
I0429 01:10:26.694663 6470 solver.cpp:245] Train net output #21: loss1/accuracy21 = 1
I0429 01:10:26.694674 6470 solver.cpp:245] Train net output #22: loss1/accuracy22 = 1
I0429 01:10:26.694685 6470 solver.cpp:245] Train net output #23: loss1/accuracy_incl_empty = 0.75
I0429 01:10:26.694697 6470 solver.cpp:245] Train net output #24: loss1/accuracy_top3 = 0.3
I0429 01:10:26.694713 6470 solver.cpp:245] Train net output #25: loss1/cross_entropy_loss = 2.89351 (* 0.3 = 0.868053 loss)
I0429 01:10:26.694728 6470 solver.cpp:245] Train net output #26: loss1/cross_entropy_loss_incl_empty = 0.89097 (* 0.3 = 0.267291 loss)
I0429 01:10:26.694742 6470 solver.cpp:245] Train net output #27: loss1/loss01 = 2.72157 (* 0.0272727 = 0.0742246 loss)
I0429 01:10:26.694756 6470 solver.cpp:245] Train net output #28: loss1/loss02 = 2.49819 (* 0.0272727 = 0.0681324 loss)
I0429 01:10:26.694771 6470 solver.cpp:245] Train net output #29: loss1/loss03 = 2.8242 (* 0.0272727 = 0.0770236 loss)
I0429 01:10:26.694784 6470 solver.cpp:245] Train net output #30: loss1/loss04 = 2.5548 (* 0.0272727 = 0.0696763 loss)
I0429 01:10:26.694798 6470 solver.cpp:245] Train net output #31: loss1/loss05 = 2.19992 (* 0.0272727 = 0.0599979 loss)
I0429 01:10:26.694813 6470 solver.cpp:245] Train net output #32: loss1/loss06 = 1.03815 (* 0.0272727 = 0.0283131 loss)
I0429 01:10:26.694826 6470 solver.cpp:245] Train net output #33: loss1/loss07 = 0.961625 (* 0.0272727 = 0.0262261 loss)
I0429 01:10:26.694840 6470 solver.cpp:245] Train net output #34: loss1/loss08 = 0.73721 (* 0.0272727 = 0.0201057 loss)
I0429 01:10:26.694854 6470 solver.cpp:245] Train net output #35: loss1/loss09 = 0.998165 (* 0.0272727 = 0.0272227 loss)
I0429 01:10:26.694869 6470 solver.cpp:245] Train net output #36: loss1/loss10 = 0.618379 (* 0.0272727 = 0.0168649 loss)
I0429 01:10:26.694882 6470 solver.cpp:245] Train net output #37: loss1/loss11 = 0.88145 (* 0.0272727 = 0.0240396 loss)
I0429 01:10:26.694895 6470 solver.cpp:245] Train net output #38: loss1/loss12 = 0.935452 (* 0.0272727 = 0.0255123 loss)
I0429 01:10:26.694910 6470 solver.cpp:245] Train net output #39: loss1/loss13 = 1.03095 (* 0.0272727 = 0.0281167 loss)
I0429 01:10:26.694943 6470 solver.cpp:245] Train net output #40: loss1/loss14 = 0.0395578 (* 0.0272727 = 0.00107885 loss)
I0429 01:10:26.694959 6470 solver.cpp:245] Train net output #41: loss1/loss15 = 0.0200606 (* 0.0272727 = 0.000547107 loss)
I0429 01:10:26.694973 6470 solver.cpp:245] Train net output #42: loss1/loss16 = 0.00612035 (* 0.0272727 = 0.000166919 loss)
I0429 01:10:26.694988 6470 solver.cpp:245] Train net output #43: loss1/loss17 = 0.00707436 (* 0.0272727 = 0.000192937 loss)
I0429 01:10:26.695001 6470 solver.cpp:245] Train net output #44: loss1/loss18 = 0.00116438 (* 0.0272727 = 3.17557e-05 loss)
I0429 01:10:26.695015 6470 solver.cpp:245] Train net output #45: loss1/loss19 = 0.000418975 (* 0.0272727 = 1.14266e-05 loss)
I0429 01:10:26.695029 6470 solver.cpp:245] Train net output #46: loss1/loss20 = 0.000207192 (* 0.0272727 = 5.65068e-06 loss)
I0429 01:10:26.695044 6470 solver.cpp:245] Train net output #47: loss1/loss21 = 6.59936e-05 (* 0.0272727 = 1.79982e-06 loss)
I0429 01:10:26.695057 6470 solver.cpp:245] Train net output #48: loss1/loss22 = 4.34533e-05 (* 0.0272727 = 1.18509e-06 loss)
I0429 01:10:26.695070 6470 solver.cpp:245] Train net output #49: loss2/accuracy = 0.1
I0429 01:10:26.695081 6470 solver.cpp:245] Train net output #50: loss2/accuracy01 = 0.375
I0429 01:10:26.695092 6470 solver.cpp:245] Train net output #51: loss2/accuracy02 = 0.125
I0429 01:10:26.695104 6470 solver.cpp:245] Train net output #52: loss2/accuracy03 = 0
I0429 01:10:26.695116 6470 solver.cpp:245] Train net output #53: loss2/accuracy04 = 0.5
I0429 01:10:26.695128 6470 solver.cpp:245] Train net output #54: loss2/accuracy05 = 0.375
I0429 01:10:26.695140 6470 solver.cpp:245] Train net output #55: loss2/accuracy06 = 0.75
I0429 01:10:26.695152 6470 solver.cpp:245] Train net output #56: loss2/accuracy07 = 0.75
I0429 01:10:26.695163 6470 solver.cpp:245] Train net output #57: loss2/accuracy08 = 0.75
I0429 01:10:26.695175 6470 solver.cpp:245] Train net output #58: loss2/accuracy09 = 0.75
I0429 01:10:26.695188 6470 solver.cpp:245] Train net output #59: loss2/accuracy10 = 0.75
I0429 01:10:26.695200 6470 solver.cpp:245] Train net output #60: loss2/accuracy11 = 0.75
I0429 01:10:26.695212 6470 solver.cpp:245] Train net output #61: loss2/accuracy12 = 0.875
I0429 01:10:26.695225 6470 solver.cpp:245] Train net output #62: loss2/accuracy13 = 0.875
I0429 01:10:26.695236 6470 solver.cpp:245] Train net output #63: loss2/accuracy14 = 1
I0429 01:10:26.695248 6470 solver.cpp:245] Train net output #64: loss2/accuracy15 = 1
I0429 01:10:26.695261 6470 solver.cpp:245] Train net output #65: loss2/accuracy16 = 1
I0429 01:10:26.695271 6470 solver.cpp:245] Train net output #66: loss2/accuracy17 = 1
I0429 01:10:26.695282 6470 solver.cpp:245] Train net output #67: loss2/accuracy18 = 1
I0429 01:10:26.695294 6470 solver.cpp:245] Train net output #68: loss2/accuracy19 = 1
I0429 01:10:26.695305 6470 solver.cpp:245] Train net output #69: loss2/accuracy20 = 1
I0429 01:10:26.695320 6470 solver.cpp:245] Train net output #70: loss2/accuracy21 = 1
I0429 01:10:26.695333 6470 solver.cpp:245] Train net output #71: loss2/accuracy22 = 1
I0429 01:10:26.695344 6470 solver.cpp:245] Train net output #72: loss2/accuracy_incl_empty = 0.727273
I0429 01:10:26.695356 6470 solver.cpp:245] Train net output #73: loss2/accuracy_top3 = 0.28
I0429 01:10:26.695369 6470 solver.cpp:245] Train net output #74: loss2/cross_entropy_loss = 2.91333 (* 0.3 = 0.874 loss)
I0429 01:10:26.695384 6470 solver.cpp:245] Train net output #75: loss2/cross_entropy_loss_incl_empty = 0.926153 (* 0.3 = 0.277846 loss)
I0429 01:10:26.695397 6470 solver.cpp:245] Train net output #76: loss2/loss01 = 2.56262 (* 0.0272727 = 0.0698897 loss)
I0429 01:10:26.695415 6470 solver.cpp:245] Train net output #77: loss2/loss02 = 2.70214 (* 0.0272727 = 0.0736948 loss)
I0429 01:10:26.695441 6470 solver.cpp:245] Train net output #78: loss2/loss03 = 2.99893 (* 0.0272727 = 0.0817889 loss)
I0429 01:10:26.695456 6470 solver.cpp:245] Train net output #79: loss2/loss04 = 2.0432 (* 0.0272727 = 0.0557237 loss)
I0429 01:10:26.695483 6470 solver.cpp:245] Train net output #80: loss2/loss05 = 2.07524 (* 0.0272727 = 0.0565975 loss)
I0429 01:10:26.695499 6470 solver.cpp:245] Train net output #81: loss2/loss06 = 1.1779 (* 0.0272727 = 0.0321247 loss)
I0429 01:10:26.695513 6470 solver.cpp:245] Train net output #82: loss2/loss07 = 0.893823 (* 0.0272727 = 0.024377 loss)
I0429 01:10:26.695528 6470 solver.cpp:245] Train net output #83: loss2/loss08 = 0.76126 (* 0.0272727 = 0.0207616 loss)
I0429 01:10:26.695540 6470 solver.cpp:245] Train net output #84: loss2/loss09 = 0.934416 (* 0.0272727 = 0.0254841 loss)
I0429 01:10:26.695554 6470 solver.cpp:245] Train net output #85: loss2/loss10 = 0.742063 (* 0.0272727 = 0.0202381 loss)
I0429 01:10:26.695569 6470 solver.cpp:245] Train net output #86: loss2/loss11 = 0.725959 (* 0.0272727 = 0.0197989 loss)
I0429 01:10:26.695581 6470 solver.cpp:245] Train net output #87: loss2/loss12 = 0.705982 (* 0.0272727 = 0.019254 loss)
I0429 01:10:26.695595 6470 solver.cpp:245] Train net output #88: loss2/loss13 = 0.873828 (* 0.0272727 = 0.0238317 loss)
I0429 01:10:26.695610 6470 solver.cpp:245] Train net output #89: loss2/loss14 = 0.122529 (* 0.0272727 = 0.00334171 loss)
I0429 01:10:26.695623 6470 solver.cpp:245] Train net output #90: loss2/loss15 = 0.0612142 (* 0.0272727 = 0.00166948 loss)
I0429 01:10:26.695636 6470 solver.cpp:245] Train net output #91: loss2/loss16 = 0.0253537 (* 0.0272727 = 0.000691464 loss)
I0429 01:10:26.695650 6470 solver.cpp:245] Train net output #92: loss2/loss17 = 0.0208371 (* 0.0272727 = 0.000568284 loss)
I0429 01:10:26.695664 6470 solver.cpp:245] Train net output #93: loss2/loss18 = 0.00513101 (* 0.0272727 = 0.000139937 loss)
I0429 01:10:26.695678 6470 solver.cpp:245] Train net output #94: loss2/loss19 = 0.00186684 (* 0.0272727 = 5.09138e-05 loss)
I0429 01:10:26.695693 6470 solver.cpp:245] Train net output #95: loss2/loss20 = 0.000560929 (* 0.0272727 = 1.52981e-05 loss)
I0429 01:10:26.695705 6470 solver.cpp:245] Train net output #96: loss2/loss21 = 9.72365e-05 (* 0.0272727 = 2.6519e-06 loss)
I0429 01:10:26.695719 6470 solver.cpp:245] Train net output #97: loss2/loss22 = 7.06555e-05 (* 0.0272727 = 1.92697e-06 loss)
I0429 01:10:26.695731 6470 solver.cpp:245] Train net output #98: loss3/accuracy = 0.16
I0429 01:10:26.695744 6470 solver.cpp:245] Train net output #99: loss3/accuracy01 = 0.25
I0429 01:10:26.695755 6470 solver.cpp:245] Train net output #100: loss3/accuracy02 = 0.125
I0429 01:10:26.695766 6470 solver.cpp:245] Train net output #101: loss3/accuracy03 = 0.25
I0429 01:10:26.695778 6470 solver.cpp:245] Train net output #102: loss3/accuracy04 = 0.625
I0429 01:10:26.695791 6470 solver.cpp:245] Train net output #103: loss3/accuracy05 = 0.375
I0429 01:10:26.695802 6470 solver.cpp:245] Train net output #104: loss3/accuracy06 = 0.75
I0429 01:10:26.695813 6470 solver.cpp:245] Train net output #105: loss3/accuracy07 = 0.75
I0429 01:10:26.695825 6470 solver.cpp:245] Train net output #106: loss3/accuracy08 = 0.75
I0429 01:10:26.695837 6470 solver.cpp:245] Train net output #107: loss3/accuracy09 = 0.75
I0429 01:10:26.695849 6470 solver.cpp:245] Train net output #108: loss3/accuracy10 = 0.75
I0429 01:10:26.695860 6470 solver.cpp:245] Train net output #109: loss3/accuracy11 = 0.75
I0429 01:10:26.695873 6470 solver.cpp:245] Train net output #110: loss3/accuracy12 = 0.875
I0429 01:10:26.695883 6470 solver.cpp:245] Train net output #111: loss3/accuracy13 = 0.875
I0429 01:10:26.695895 6470 solver.cpp:245] Train net output #112: loss3/accuracy14 = 1
I0429 01:10:26.695906 6470 solver.cpp:245] Train net output #113: loss3/accuracy15 = 1
I0429 01:10:26.695919 6470 solver.cpp:245] Train net output #114: loss3/accuracy16 = 1
I0429 01:10:26.695940 6470 solver.cpp:245] Train net output #115: loss3/accuracy17 = 1
I0429 01:10:26.695955 6470 solver.cpp:245] Train net output #116: loss3/accuracy18 = 1
I0429 01:10:26.695966 6470 solver.cpp:245] Train net output #117: loss3/accuracy19 = 1
I0429 01:10:26.695977 6470 solver.cpp:245] Train net output #118: loss3/accuracy20 = 1
I0429 01:10:26.695989 6470 solver.cpp:245] Train net output #119: loss3/accuracy21 = 1
I0429 01:10:26.696001 6470 solver.cpp:245] Train net output #120: loss3/accuracy22 = 1
I0429 01:10:26.696012 6470 solver.cpp:245] Train net output #121: loss3/accuracy_incl_empty = 0.744318
I0429 01:10:26.696024 6470 solver.cpp:245] Train net output #122: loss3/accuracy_top3 = 0.34
I0429 01:10:26.696038 6470 solver.cpp:245] Train net output #123: loss3/cross_entropy_loss = 2.73598 (* 1 = 2.73598 loss)
I0429 01:10:26.696053 6470 solver.cpp:245] Train net output #124: loss3/cross_entropy_loss_incl_empty = 0.874735 (* 1 = 0.874735 loss)
I0429 01:10:26.696066 6470 solver.cpp:245] Train net output #125: loss3/loss01 = 2.29131 (* 0.0909091 = 0.208301 loss)
I0429 01:10:26.696080 6470 solver.cpp:245] Train net output #126: loss3/loss02 = 2.60031 (* 0.0909091 = 0.236392 loss)
I0429 01:10:26.696094 6470 solver.cpp:245] Train net output #127: loss3/loss03 = 2.51206 (* 0.0909091 = 0.228369 loss)
I0429 01:10:26.696107 6470 solver.cpp:245] Train net output #128: loss3/loss04 = 1.90293 (* 0.0909091 = 0.172994 loss)
I0429 01:10:26.696121 6470 solver.cpp:245] Train net output #129: loss3/loss05 = 1.91467 (* 0.0909091 = 0.174061 loss)
I0429 01:10:26.696135 6470 solver.cpp:245] Train net output #130: loss3/loss06 = 1.11288 (* 0.0909091 = 0.101171 loss)
I0429 01:10:26.696149 6470 solver.cpp:245] Train net output #131: loss3/loss07 = 0.983952 (* 0.0909091 = 0.0894502 loss)
I0429 01:10:26.696163 6470 solver.cpp:245] Train net output #132: loss3/loss08 = 0.721243 (* 0.0909091 = 0.0655676 loss)
I0429 01:10:26.696177 6470 solver.cpp:245] Train net output #133: loss3/loss09 = 0.956196 (* 0.0909091 = 0.0869269 loss)
I0429 01:10:26.696192 6470 solver.cpp:245] Train net output #134: loss3/loss10 = 0.639525 (* 0.0909091 = 0.0581386 loss)
I0429 01:10:26.696204 6470 solver.cpp:245] Train net output #135: loss3/loss11 = 0.867171 (* 0.0909091 = 0.0788337 loss)
I0429 01:10:26.696218 6470 solver.cpp:245] Train net output #136: loss3/loss12 = 0.710468 (* 0.0909091 = 0.064588 loss)
I0429 01:10:26.696233 6470 solver.cpp:245] Train net output #137: loss3/loss13 = 0.789756 (* 0.0909091 = 0.071796 loss)
I0429 01:10:26.696246 6470 solver.cpp:245] Train net output #138: loss3/loss14 = 0.0705774 (* 0.0909091 = 0.00641613 loss)
I0429 01:10:26.696260 6470 solver.cpp:245] Train net output #139: loss3/loss15 = 0.0398996 (* 0.0909091 = 0.00362724 loss)
I0429 01:10:26.696274 6470 solver.cpp:245] Train net output #140: loss3/loss16 = 0.0210394 (* 0.0909091 = 0.00191267 loss)
I0429 01:10:26.696287 6470 solver.cpp:245] Train net output #141: loss3/loss17 = 0.0173165 (* 0.0909091 = 0.00157422 loss)
I0429 01:10:26.696301 6470 solver.cpp:245] Train net output #142: loss3/loss18 = 0.00984091 (* 0.0909091 = 0.000894629 loss)
I0429 01:10:26.696316 6470 solver.cpp:245] Train net output #143: loss3/loss19 = 0.00855665 (* 0.0909091 = 0.000777877 loss)
I0429 01:10:26.696329 6470 solver.cpp:245] Train net output #144: loss3/loss20 = 0.00553591 (* 0.0909091 = 0.000503264 loss)
I0429 01:10:26.696343 6470 solver.cpp:245] Train net output #145: loss3/loss21 = 0.00207667 (* 0.0909091 = 0.000188788 loss)
I0429 01:10:26.696357 6470 solver.cpp:245] Train net output #146: loss3/loss22 = 0.000459373 (* 0.0909091 = 4.17612e-05 loss)
I0429 01:10:26.696372 6470 solver.cpp:245] Train net output #147: total_accuracy = 0.125
I0429 01:10:26.696383 6470 solver.cpp:245] Train net output #148: total_accuracy_not_rec = 0.125
I0429 01:10:26.696395 6470 solver.cpp:245] Train net output #149: total_confidence = 0.0139327
I0429 01:10:26.696416 6470 solver.cpp:245] Train net output #150: total_confidence_not_rec = 0.0220241
I0429 01:10:26.696431 6470 sgd_solver.cpp:106] Iteration 23000, lr = 0.01
I0429 01:12:43.371711 6470 solver.cpp:229] Iteration 23500, loss = 8.87916
I0429 01:12:43.371872 6470 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0612245
I0429 01:12:43.371893 6470 solver.cpp:245] Train net output #1: loss1/accuracy01 = 0.25
I0429 01:12:43.371906 6470 solver.cpp:245] Train net output #2: loss1/accuracy02 = 0
I0429 01:12:43.371918 6470 solver.cpp:245] Train net output #3: loss1/accuracy03 = 0
I0429 01:12:43.371930 6470 solver.cpp:245] Train net output #4: loss1/accuracy04 = 0.25
I0429 01:12:43.371942 6470 solver.cpp:245] Train net output #5: loss1/accuracy05 = 0.125
I0429 01:12:43.371954 6470 solver.cpp:245] Train net output #6: loss1/accuracy06 = 0.25
I0429 01:12:43.371965 6470 solver.cpp:245] Train net output #7: loss1/accuracy07 = 0.5
I0429 01:12:43.371978 6470 solver.cpp:245] Train net output #8: loss1/accuracy08 = 1
I0429 01:12:43.371989
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment