firenet150x50_len22_3_trainval

I0404 23:42:02.924360 26022 solver.cpp:280] Solving 
I0404 23:42:02.924372 26022 solver.cpp:281] Learning Rate Policy: poly
I0404 23:42:13.433935 26022 solver.cpp:229] Iteration 0, loss = 4.304
I0404 23:42:13.433985 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0
I0404 23:42:13.434006 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0404 23:42:13.434020 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0404 23:42:13.434031 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0404 23:42:13.434043 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0404 23:42:13.434056 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0
I0404 23:42:13.434067 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0
I0404 23:42:13.434079 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0
I0404 23:42:13.434092 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0
I0404 23:42:13.434103 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0
I0404 23:42:13.434115 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 0
I0404 23:42:13.434126 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 0
I0404 23:42:13.434139 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 0
I0404 23:42:13.434150 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 0
I0404 23:42:13.434162 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 0
I0404 23:42:13.434175 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 0
I0404 23:42:13.434186 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 0
I0404 23:42:13.434198 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 0
I0404 23:42:13.434209 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 0
I0404 23:42:13.434221 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 0
I0404 23:42:13.434233 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 0.3125
I0404 23:42:13.434245 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 0
I0404 23:42:13.434262 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.30402 (* 0.0454545 = 0.195637 loss)
I0404 23:42:13.434276 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.30405 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434290 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.30406 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434305 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.30405 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434319 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.30405 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434334 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.304 (* 0.0454545 = 0.195636 loss)
I0404 23:42:13.434350 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 4.30391 (* 0.0454545 = 0.195632 loss)
I0404 23:42:13.434391 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 4.30403 (* 0.0454545 = 0.195638 loss)
I0404 23:42:13.434406 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 4.30395 (* 0.0454545 = 0.195634 loss)
I0404 23:42:13.434420 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 4.30411 (* 0.0454545 = 0.195641 loss)
I0404 23:42:13.434435 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.30406 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434449 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.30401 (* 0.0454545 = 0.195637 loss)
I0404 23:42:13.434463 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.30414 (* 0.0454545 = 0.195643 loss)
I0404 23:42:13.434478 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.30407 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434492 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.30405 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434507 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.30409 (* 0.0454545 = 0.195641 loss)
I0404 23:42:13.434521 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.30373 (* 0.0454545 = 0.195624 loss)
I0404 23:42:13.434536 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.30389 (* 0.0454545 = 0.195632 loss)
I0404 23:42:13.434551 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.30397 (* 0.0454545 = 0.195635 loss)
I0404 23:42:13.434564 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.30389 (* 0.0454545 = 0.195631 loss)
I0404 23:42:13.434578 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.30371 (* 0.0454545 = 0.195623 loss)
I0404 23:42:13.434592 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.30406 (* 0.0454545 = 0.195639 loss)
I0404 23:42:13.434605 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0404 23:42:13.434617 26022 solver.cpp:245]     Train net output #45: total_confidence = 7.61045e-42
I0404 23:42:13.434643 26022 sgd_solver.cpp:106] Iteration 0, lr = 0.04
I0404 23:49:09.543875 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 56.3543 > 30) by scale factor 0.532346
I0404 23:49:20.420500 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 101.693 > 30) by scale factor 0.295007
I0404 23:49:31.279211 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 55.8417 > 30) by scale factor 0.537233
I0404 23:49:42.136785 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 196.495 > 30) by scale factor 0.152675
I0404 23:49:53.043862 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 165.578 > 30) by scale factor 0.181184
I0404 23:50:03.964578 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 150.38 > 30) by scale factor 0.199495
I0404 23:50:14.862637 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 78.1494 > 30) by scale factor 0.38388
I0404 23:50:25.711493 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 39.4703 > 30) by scale factor 0.760065
I0404 23:51:19.566433 26022 solver.cpp:229] Iteration 50, loss = 3.38569
I0404 23:51:19.566547 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0404 23:51:19.566567 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0404 23:51:19.566581 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0404 23:51:19.566594 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0404 23:51:19.566607 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0404 23:51:19.566619 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.03125
I0404 23:51:19.566632 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0404 23:51:19.566645 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0404 23:51:19.566658 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0404 23:51:19.566671 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0404 23:51:19.566684 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0404 23:51:19.566696 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0404 23:51:19.566709 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0404 23:51:19.566720 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0404 23:51:19.566731 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0404 23:51:19.566743 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0404 23:51:19.566756 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0404 23:51:19.566767 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0404 23:51:19.566779 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0404 23:51:19.566794 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0404 23:51:19.566807 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0404 23:51:19.566818 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0404 23:51:19.566834 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.11541 (* 0.0454545 = 0.187064 loss)
I0404 23:51:19.566849 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.20302 (* 0.0454545 = 0.191047 loss)
I0404 23:51:19.566864 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.21355 (* 0.0454545 = 0.191525 loss)
I0404 23:51:19.566879 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.1936 (* 0.0454545 = 0.190618 loss)
I0404 23:51:19.566893 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.25647 (* 0.0454545 = 0.193476 loss)
I0404 23:51:19.566908 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.21738 (* 0.0454545 = 0.191699 loss)
I0404 23:51:19.566922 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.55866 (* 0.0454545 = 0.070848 loss)
I0404 23:51:19.566937 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.40903 (* 0.0454545 = 0.0640468 loss)
I0404 23:51:19.566951 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.491821 (* 0.0454545 = 0.0223555 loss)
I0404 23:51:19.566967 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.277854 (* 0.0454545 = 0.0126297 loss)
I0404 23:51:19.566982 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.00364972 (* 0.0454545 = 0.000165896 loss)
I0404 23:51:19.567003 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.00311343 (* 0.0454545 = 0.00014152 loss)
I0404 23:51:19.567018 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.00431902 (* 0.0454545 = 0.000196319 loss)
I0404 23:51:19.567034 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.00340891 (* 0.0454545 = 0.00015495 loss)
I0404 23:51:19.567049 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.00353399 (* 0.0454545 = 0.000160636 loss)
I0404 23:51:19.567065 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.00344168 (* 0.0454545 = 0.00015644 loss)
I0404 23:51:19.567080 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00272917 (* 0.0454545 = 0.000124053 loss)
I0404 23:51:19.567111 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.00305568 (* 0.0454545 = 0.000138895 loss)
I0404 23:51:19.567127 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.00348132 (* 0.0454545 = 0.000158242 loss)
I0404 23:51:19.567142 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.00342272 (* 0.0454545 = 0.000155578 loss)
I0404 23:51:19.567157 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.00288049 (* 0.0454545 = 0.000130931 loss)
I0404 23:51:19.567173 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.00366961 (* 0.0454545 = 0.0001668 loss)
I0404 23:51:19.567185 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0404 23:51:19.567198 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.23406e-10
I0404 23:51:19.567214 26022 sgd_solver.cpp:106] Iteration 50, lr = 0.039998
I0405 00:00:22.380501 26022 solver.cpp:229] Iteration 100, loss = 1.25884
I0405 00:00:22.380619 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 00:00:22.380645 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 00:00:22.380659 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 00:00:22.380673 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 00:00:22.380686 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 00:00:22.380698 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.03125
I0405 00:00:22.380712 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 00:00:22.380724 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 00:00:22.380736 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 00:00:22.380749 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.90625
I0405 00:00:22.380761 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:00:22.380774 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:00:22.380786 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:00:22.380798 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:00:22.380810 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:00:22.380821 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:00:22.380833 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:00:22.380846 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:00:22.380858 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:00:22.380870 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:00:22.380882 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:00:22.380894 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:00:22.380910 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.97759 (* 0.0454545 = 0.1808 loss)
I0405 00:00:22.380925 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.01867 (* 0.0454545 = 0.182667 loss)
I0405 00:00:22.380939 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.04347 (* 0.0454545 = 0.183794 loss)
I0405 00:00:22.380954 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.93673 (* 0.0454545 = 0.178942 loss)
I0405 00:00:22.380969 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.13129 (* 0.0454545 = 0.187786 loss)
I0405 00:00:22.380983 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.21934 (* 0.0454545 = 0.191788 loss)
I0405 00:00:22.380998 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.89246 (* 0.0454545 = 0.086021 loss)
I0405 00:00:22.381012 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.922598 (* 0.0454545 = 0.0419363 loss)
I0405 00:00:22.381027 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.722389 (* 0.0454545 = 0.0328359 loss)
I0405 00:00:22.381042 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.833156 (* 0.0454545 = 0.0378707 loss)
I0405 00:00:22.381057 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000507901 (* 0.0454545 = 2.30864e-05 loss)
I0405 00:00:22.381072 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000485268 (* 0.0454545 = 2.20576e-05 loss)
I0405 00:00:22.381088 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000547472 (* 0.0454545 = 2.48851e-05 loss)
I0405 00:00:22.381103 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000502838 (* 0.0454545 = 2.28563e-05 loss)
I0405 00:00:22.381117 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000508602 (* 0.0454545 = 2.31183e-05 loss)
I0405 00:00:22.381132 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000503386 (* 0.0454545 = 2.28812e-05 loss)
I0405 00:00:22.381147 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000472033 (* 0.0454545 = 2.14561e-05 loss)
I0405 00:00:22.381181 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000503801 (* 0.0454545 = 2.29001e-05 loss)
I0405 00:00:22.381196 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000510518 (* 0.0454545 = 2.32054e-05 loss)
I0405 00:00:22.381211 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000499319 (* 0.0454545 = 2.26963e-05 loss)
I0405 00:00:22.381227 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000487538 (* 0.0454545 = 2.21608e-05 loss)
I0405 00:00:22.381242 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000516871 (* 0.0454545 = 2.34941e-05 loss)
I0405 00:00:22.381254 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:00:22.381266 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.97162e-09
I0405 00:00:22.381281 26022 sgd_solver.cpp:106] Iteration 100, lr = 0.039996
I0405 00:09:24.807351 26022 solver.cpp:229] Iteration 150, loss = 1.2254
I0405 00:09:24.807485 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 00:09:24.807507 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 00:09:24.807520 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 00:09:24.807533 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 00:09:24.807545 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 00:09:24.807557 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.03125
I0405 00:09:24.807570 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 00:09:24.807582 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 00:09:24.807595 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 00:09:24.807607 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 00:09:24.807620 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:09:24.807631 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:09:24.807643 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:09:24.807656 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:09:24.807667 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:09:24.807678 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:09:24.807693 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:09:24.807704 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:09:24.807716 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:09:24.807729 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:09:24.807739 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:09:24.807751 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:09:24.807766 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.80308 (* 0.0454545 = 0.172867 loss)
I0405 00:09:24.807781 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.79913 (* 0.0454545 = 0.172688 loss)
I0405 00:09:24.807796 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.71088 (* 0.0454545 = 0.168676 loss)
I0405 00:09:24.807811 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.77479 (* 0.0454545 = 0.171581 loss)
I0405 00:09:24.807826 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.05454 (* 0.0454545 = 0.184297 loss)
I0405 00:09:24.807839 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.16416 (* 0.0454545 = 0.18928 loss)
I0405 00:09:24.807853 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.35372 (* 0.0454545 = 0.0615329 loss)
I0405 00:09:24.807868 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.991306 (* 0.0454545 = 0.0450594 loss)
I0405 00:09:24.807883 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.262823 (* 0.0454545 = 0.0119465 loss)
I0405 00:09:24.807900 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0170909 (* 0.0454545 = 0.000776858 loss)
I0405 00:09:24.807916 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000153686 (* 0.0454545 = 6.98572e-06 loss)
I0405 00:09:24.807931 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000147521 (* 0.0454545 = 6.70552e-06 loss)
I0405 00:09:24.807946 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000164077 (* 0.0454545 = 7.45806e-06 loss)
I0405 00:09:24.807961 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000152622 (* 0.0454545 = 6.93737e-06 loss)
I0405 00:09:24.807976 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000155618 (* 0.0454545 = 7.07353e-06 loss)
I0405 00:09:24.807991 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000154165 (* 0.0454545 = 7.00749e-06 loss)
I0405 00:09:24.808007 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000144202 (* 0.0454545 = 6.55462e-06 loss)
I0405 00:09:24.808037 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000153678 (* 0.0454545 = 6.98538e-06 loss)
I0405 00:09:24.808053 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000153842 (* 0.0454545 = 6.99284e-06 loss)
I0405 00:09:24.808089 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000151504 (* 0.0454545 = 6.88656e-06 loss)
I0405 00:09:24.808109 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000148455 (* 0.0454545 = 6.74795e-06 loss)
I0405 00:09:24.808123 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000155815 (* 0.0454545 = 7.08251e-06 loss)
I0405 00:09:24.808136 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:09:24.808148 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.35536e-09
I0405 00:09:24.808163 26022 sgd_solver.cpp:106] Iteration 150, lr = 0.039994
I0405 00:18:27.087574 26022 solver.cpp:229] Iteration 200, loss = 1.21771
I0405 00:18:27.087697 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 00:18:27.087718 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 00:18:27.087731 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 00:18:27.087743 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 00:18:27.087755 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 00:18:27.087769 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.0625
I0405 00:18:27.087781 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 00:18:27.087793 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 00:18:27.087806 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 00:18:27.087818 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 00:18:27.087831 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:18:27.087842 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:18:27.087857 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:18:27.087872 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:18:27.087883 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:18:27.087894 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:18:27.087906 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:18:27.087918 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:18:27.087929 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:18:27.087941 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:18:27.087954 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:18:27.087966 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:18:27.087981 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.65819 (* 0.0454545 = 0.166282 loss)
I0405 00:18:27.087996 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.72011 (* 0.0454545 = 0.169096 loss)
I0405 00:18:27.088011 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.77717 (* 0.0454545 = 0.17169 loss)
I0405 00:18:27.088026 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.91118 (* 0.0454545 = 0.177781 loss)
I0405 00:18:27.088039 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.04271 (* 0.0454545 = 0.18376 loss)
I0405 00:18:27.088053 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.07657 (* 0.0454545 = 0.185299 loss)
I0405 00:18:27.088083 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.32247 (* 0.0454545 = 0.0601121 loss)
I0405 00:18:27.088101 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.716166 (* 0.0454545 = 0.032553 loss)
I0405 00:18:27.088115 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.256884 (* 0.0454545 = 0.0116766 loss)
I0405 00:18:27.088129 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.290546 (* 0.0454545 = 0.0132067 loss)
I0405 00:18:27.088145 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.57313e-05 (* 0.0454545 = 2.98778e-06 loss)
I0405 00:18:27.088160 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.28738e-05 (* 0.0454545 = 2.8579e-06 loss)
I0405 00:18:27.088174 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 7.01647e-05 (* 0.0454545 = 3.1893e-06 loss)
I0405 00:18:27.088191 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 6.47403e-05 (* 0.0454545 = 2.94274e-06 loss)
I0405 00:18:27.088204 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 6.57201e-05 (* 0.0454545 = 2.98728e-06 loss)
I0405 00:18:27.088219 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 6.59771e-05 (* 0.0454545 = 2.99896e-06 loss)
I0405 00:18:27.088234 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.14506e-05 (* 0.0454545 = 2.79321e-06 loss)
I0405 00:18:27.088265 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.57685e-05 (* 0.0454545 = 2.98948e-06 loss)
I0405 00:18:27.088281 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 6.58654e-05 (* 0.0454545 = 2.99388e-06 loss)
I0405 00:18:27.088296 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 6.4215e-05 (* 0.0454545 = 2.91886e-06 loss)
I0405 00:18:27.088311 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 6.23857e-05 (* 0.0454545 = 2.83571e-06 loss)
I0405 00:18:27.088326 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.57573e-05 (* 0.0454545 = 2.98897e-06 loss)
I0405 00:18:27.088338 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:18:27.088351 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.65969e-09
I0405 00:18:27.088366 26022 sgd_solver.cpp:106] Iteration 200, lr = 0.039992
I0405 00:27:29.311956 26022 solver.cpp:229] Iteration 250, loss = 1.21699
I0405 00:27:29.312165 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 00:27:29.312186 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 00:27:29.312199 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 00:27:29.312211 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 00:27:29.312223 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 00:27:29.312235 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.03125
I0405 00:27:29.312248 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.625
I0405 00:27:29.312260 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 00:27:29.312273 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 00:27:29.312283 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 00:27:29.312295 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:27:29.312306 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:27:29.312317 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:27:29.312330 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:27:29.312340 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:27:29.312352 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:27:29.312363 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:27:29.312374 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:27:29.312386 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:27:29.312397 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:27:29.312408 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:27:29.312420 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:27:29.312435 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.62696 (* 0.0454545 = 0.164862 loss)
I0405 00:27:29.312450 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.76131 (* 0.0454545 = 0.170968 loss)
I0405 00:27:29.312464 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.81545 (* 0.0454545 = 0.17343 loss)
I0405 00:27:29.312479 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.91503 (* 0.0454545 = 0.177956 loss)
I0405 00:27:29.312496 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.07815 (* 0.0454545 = 0.185371 loss)
I0405 00:27:29.312510 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.12747 (* 0.0454545 = 0.187612 loss)
I0405 00:27:29.312525 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.00379 (* 0.0454545 = 0.0910813 loss)
I0405 00:27:29.312538 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.423143 (* 0.0454545 = 0.0192338 loss)
I0405 00:27:29.312553 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0391133 (* 0.0454545 = 0.00177788 loss)
I0405 00:27:29.312572 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0166231 (* 0.0454545 = 0.000755594 loss)
I0405 00:27:29.312608 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.48453e-05 (* 0.0454545 = 2.49297e-06 loss)
I0405 00:27:29.312624 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.36904e-05 (* 0.0454545 = 2.44047e-06 loss)
I0405 00:27:29.312639 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.02733e-05 (* 0.0454545 = 2.7397e-06 loss)
I0405 00:27:29.312654 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.55121e-05 (* 0.0454545 = 2.52328e-06 loss)
I0405 00:27:29.312669 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.55457e-05 (* 0.0454545 = 2.5248e-06 loss)
I0405 00:27:29.312683 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.5881e-05 (* 0.0454545 = 2.54004e-06 loss)
I0405 00:27:29.312698 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.27143e-05 (* 0.0454545 = 2.3961e-06 loss)
I0405 00:27:29.312727 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.51955e-05 (* 0.0454545 = 2.50888e-06 loss)
I0405 00:27:29.312746 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.58437e-05 (* 0.0454545 = 2.53835e-06 loss)
I0405 00:27:29.312760 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.45584e-05 (* 0.0454545 = 2.47993e-06 loss)
I0405 00:27:29.312775 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.33476e-05 (* 0.0454545 = 2.42489e-06 loss)
I0405 00:27:29.312790 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.65553e-05 (* 0.0454545 = 2.57069e-06 loss)
I0405 00:27:29.312803 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:27:29.312814 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.14555e-09
I0405 00:27:29.312829 26022 sgd_solver.cpp:106] Iteration 250, lr = 0.03999
I0405 00:36:31.516175 26022 solver.cpp:229] Iteration 300, loss = 1.21257
I0405 00:36:31.516315 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 00:36:31.516346 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 00:36:31.516372 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 00:36:31.516396 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 00:36:31.516418 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 00:36:31.516443 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.03125
I0405 00:36:31.516469 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 00:36:31.516492 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 00:36:31.516515 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 00:36:31.516537 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 00:36:31.516559 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:36:31.516582 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:36:31.516602 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:36:31.516624 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:36:31.516645 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:36:31.516667 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:36:31.516688 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:36:31.516710 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:36:31.516732 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:36:31.516753 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:36:31.516775 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:36:31.516798 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:36:31.516829 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.69181 (* 0.0454545 = 0.16781 loss)
I0405 00:36:31.516860 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.94683 (* 0.0454545 = 0.179401 loss)
I0405 00:36:31.516886 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.99374 (* 0.0454545 = 0.181534 loss)
I0405 00:36:31.516913 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.97806 (* 0.0454545 = 0.180821 loss)
I0405 00:36:31.516942 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.13163 (* 0.0454545 = 0.187801 loss)
I0405 00:36:31.516968 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.09403 (* 0.0454545 = 0.186092 loss)
I0405 00:36:31.516994 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.86977 (* 0.0454545 = 0.0849897 loss)
I0405 00:36:31.517021 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.07859 (* 0.0454545 = 0.0490269 loss)
I0405 00:36:31.517047 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.455684 (* 0.0454545 = 0.0207129 loss)
I0405 00:36:31.517074 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0152058 (* 0.0454545 = 0.000691174 loss)
I0405 00:36:31.517102 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.08366e-05 (* 0.0454545 = 2.31076e-06 loss)
I0405 00:36:31.517134 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.02741e-05 (* 0.0454545 = 2.28519e-06 loss)
I0405 00:36:31.517163 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.56872e-05 (* 0.0454545 = 2.53124e-06 loss)
I0405 00:36:31.517190 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.14513e-05 (* 0.0454545 = 2.3387e-06 loss)
I0405 00:36:31.517217 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.16115e-05 (* 0.0454545 = 2.34598e-06 loss)
I0405 00:36:31.517244 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.19505e-05 (* 0.0454545 = 2.36139e-06 loss)
I0405 00:36:31.517271 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.82623e-05 (* 0.0454545 = 2.19374e-06 loss)
I0405 00:36:31.517323 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.03151e-05 (* 0.0454545 = 2.28705e-06 loss)
I0405 00:36:31.517354 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.07733e-05 (* 0.0454545 = 2.30788e-06 loss)
I0405 00:36:31.517382 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.89366e-05 (* 0.0454545 = 2.22439e-06 loss)
I0405 00:36:31.517410 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.8687e-05 (* 0.0454545 = 2.21305e-06 loss)
I0405 00:36:31.517437 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.0967e-05 (* 0.0454545 = 2.31668e-06 loss)
I0405 00:36:31.517459 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:36:31.517482 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.1362e-09
I0405 00:36:31.517505 26022 sgd_solver.cpp:106] Iteration 300, lr = 0.039988
I0405 00:45:33.696166 26022 solver.cpp:229] Iteration 350, loss = 1.21175
I0405 00:45:33.696311 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 00:45:33.696332 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 00:45:33.696346 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 00:45:33.696357 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 00:45:33.696369 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 00:45:33.696382 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0
I0405 00:45:33.696393 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 00:45:33.696405 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 00:45:33.696419 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 00:45:33.696429 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 00:45:33.696444 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:45:33.696457 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:45:33.696468 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:45:33.696480 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:45:33.696491 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:45:33.696503 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:45:33.696514 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:45:33.696527 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:45:33.696538 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:45:33.696549 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:45:33.696562 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:45:33.696576 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:45:33.696593 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.74156 (* 0.0454545 = 0.170071 loss)
I0405 00:45:33.696607 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.68681 (* 0.0454545 = 0.167582 loss)
I0405 00:45:33.696621 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.85817 (* 0.0454545 = 0.175371 loss)
I0405 00:45:33.696636 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.04357 (* 0.0454545 = 0.183799 loss)
I0405 00:45:33.696650 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.19907 (* 0.0454545 = 0.190867 loss)
I0405 00:45:33.696665 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.98907 (* 0.0454545 = 0.181322 loss)
I0405 00:45:33.696678 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.48096 (* 0.0454545 = 0.0673165 loss)
I0405 00:45:33.696693 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.648692 (* 0.0454545 = 0.029486 loss)
I0405 00:45:33.696707 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.699356 (* 0.0454545 = 0.0317889 loss)
I0405 00:45:33.696722 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.282322 (* 0.0454545 = 0.0128328 loss)
I0405 00:45:33.696737 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.74576e-05 (* 0.0454545 = 2.15716e-06 loss)
I0405 00:45:33.696751 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.62767e-05 (* 0.0454545 = 2.10348e-06 loss)
I0405 00:45:33.696766 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.09968e-05 (* 0.0454545 = 2.31804e-06 loss)
I0405 00:45:33.696780 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.74353e-05 (* 0.0454545 = 2.15615e-06 loss)
I0405 00:45:33.696795 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.78451e-05 (* 0.0454545 = 2.17478e-06 loss)
I0405 00:45:33.696810 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.73757e-05 (* 0.0454545 = 2.15344e-06 loss)
I0405 00:45:33.696825 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.46225e-05 (* 0.0454545 = 2.0283e-06 loss)
I0405 00:45:33.696856 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.70963e-05 (* 0.0454545 = 2.14074e-06 loss)
I0405 00:45:33.696873 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.73533e-05 (* 0.0454545 = 2.15242e-06 loss)
I0405 00:45:33.696887 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.53825e-05 (* 0.0454545 = 2.06284e-06 loss)
I0405 00:45:33.696902 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.4779e-05 (* 0.0454545 = 2.03541e-06 loss)
I0405 00:45:33.696918 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.71931e-05 (* 0.0454545 = 2.14514e-06 loss)
I0405 00:45:33.696929 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:45:33.696941 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.40231e-09
I0405 00:45:33.696956 26022 sgd_solver.cpp:106] Iteration 350, lr = 0.039986
I0405 00:54:36.178771 26022 solver.cpp:229] Iteration 400, loss = 1.2071
I0405 00:54:36.178887 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 00:54:36.178908 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 00:54:36.178921 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 00:54:36.178935 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 00:54:36.178946 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 00:54:36.178959 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0
I0405 00:54:36.178972 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 00:54:36.178983 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 00:54:36.178995 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 00:54:36.179008 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 00:54:36.179020 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 00:54:36.179033 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 00:54:36.179044 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 00:54:36.179055 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 00:54:36.179067 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 00:54:36.179080 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 00:54:36.179091 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 00:54:36.179103 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 00:54:36.179116 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 00:54:36.179127 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 00:54:36.179138 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 00:54:36.179162 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 00:54:36.179179 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.74155 (* 0.0454545 = 0.17007 loss)
I0405 00:54:36.179194 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.01111 (* 0.0454545 = 0.182323 loss)
I0405 00:54:36.179209 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.96256 (* 0.0454545 = 0.180116 loss)
I0405 00:54:36.179224 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.00344 (* 0.0454545 = 0.181975 loss)
I0405 00:54:36.179237 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.11675 (* 0.0454545 = 0.187125 loss)
I0405 00:54:36.179252 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 4.01699 (* 0.0454545 = 0.182591 loss)
I0405 00:54:36.179266 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.37247 (* 0.0454545 = 0.0623849 loss)
I0405 00:54:36.179280 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.280451 (* 0.0454545 = 0.0127478 loss)
I0405 00:54:36.179296 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.245318 (* 0.0454545 = 0.0111508 loss)
I0405 00:54:36.179311 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.017813 (* 0.0454545 = 0.000809682 loss)
I0405 00:54:36.179325 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.23463e-05 (* 0.0454545 = 1.92483e-06 loss)
I0405 00:54:36.179340 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.14373e-05 (* 0.0454545 = 1.88351e-06 loss)
I0405 00:54:36.179358 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.54459e-05 (* 0.0454545 = 2.06572e-06 loss)
I0405 00:54:36.179374 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.24543e-05 (* 0.0454545 = 1.92974e-06 loss)
I0405 00:54:36.179389 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.28939e-05 (* 0.0454545 = 1.94972e-06 loss)
I0405 00:54:36.179404 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.27412e-05 (* 0.0454545 = 1.94278e-06 loss)
I0405 00:54:36.179419 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.88555e-05 (* 0.0454545 = 1.76616e-06 loss)
I0405 00:54:36.179450 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.05059e-05 (* 0.0454545 = 1.84118e-06 loss)
I0405 00:54:36.179466 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.15118e-05 (* 0.0454545 = 1.8869e-06 loss)
I0405 00:54:36.179481 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.95336e-05 (* 0.0454545 = 1.79698e-06 loss)
I0405 00:54:36.179497 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.0111e-05 (* 0.0454545 = 1.82323e-06 loss)
I0405 00:54:36.179512 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.13143e-05 (* 0.0454545 = 1.87792e-06 loss)
I0405 00:54:36.179524 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 00:54:36.179536 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.736e-09
I0405 00:54:36.179553 26022 sgd_solver.cpp:106] Iteration 400, lr = 0.039984
I0405 01:03:38.388528 26022 solver.cpp:229] Iteration 450, loss = 1.20338
I0405 01:03:38.388721 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 01:03:38.388742 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 01:03:38.388756 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 01:03:38.388769 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 01:03:38.388782 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 01:03:38.388795 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 01:03:38.388808 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 01:03:38.388821 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.8125
I0405 01:03:38.388833 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 01:03:38.388846 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 01:03:38.388859 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:03:38.388870 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:03:38.388882 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:03:38.388895 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:03:38.388906 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:03:38.388917 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:03:38.388929 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:03:38.388941 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:03:38.388952 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:03:38.388964 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:03:38.388977 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:03:38.388988 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:03:38.389003 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.63701 (* 0.0454545 = 0.165319 loss)
I0405 01:03:38.389019 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.57354 (* 0.0454545 = 0.162434 loss)
I0405 01:03:38.389034 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.89127 (* 0.0454545 = 0.176876 loss)
I0405 01:03:38.389048 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.11844 (* 0.0454545 = 0.187202 loss)
I0405 01:03:38.389063 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.07295 (* 0.0454545 = 0.185134 loss)
I0405 01:03:38.389077 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.98479 (* 0.0454545 = 0.181127 loss)
I0405 01:03:38.389092 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.79609 (* 0.0454545 = 0.0816405 loss)
I0405 01:03:38.389106 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.14981 (* 0.0454545 = 0.0522642 loss)
I0405 01:03:38.389120 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.68047 (* 0.0454545 = 0.0309305 loss)
I0405 01:03:38.389134 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.540566 (* 0.0454545 = 0.0245712 loss)
I0405 01:03:38.389149 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.16451e-05 (* 0.0454545 = 2.3475e-06 loss)
I0405 01:03:38.389165 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.05423e-05 (* 0.0454545 = 2.29738e-06 loss)
I0405 01:03:38.389180 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.44615e-05 (* 0.0454545 = 2.47552e-06 loss)
I0405 01:03:38.389194 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.15855e-05 (* 0.0454545 = 2.34479e-06 loss)
I0405 01:03:38.389209 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.17568e-05 (* 0.0454545 = 2.35258e-06 loss)
I0405 01:03:38.389225 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.18164e-05 (* 0.0454545 = 2.35529e-06 loss)
I0405 01:03:38.389240 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.78861e-05 (* 0.0454545 = 2.17664e-06 loss)
I0405 01:03:38.389267 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.87653e-05 (* 0.0454545 = 2.2166e-06 loss)
I0405 01:03:38.389283 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.9678e-05 (* 0.0454545 = 2.25809e-06 loss)
I0405 01:03:38.389298 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.81655e-05 (* 0.0454545 = 2.18934e-06 loss)
I0405 01:03:38.389313 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.85492e-05 (* 0.0454545 = 2.20678e-06 loss)
I0405 01:03:38.389328 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.98941e-05 (* 0.0454545 = 2.26791e-06 loss)
I0405 01:03:38.389341 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:03:38.389353 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.98228e-09
I0405 01:03:38.389369 26022 sgd_solver.cpp:106] Iteration 450, lr = 0.039982
I0405 01:12:40.545673 26022 solver.cpp:229] Iteration 500, loss = 1.20326
I0405 01:12:40.545821 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 01:12:40.545850 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 01:12:40.545873 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 01:12:40.545897 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 01:12:40.545918 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 01:12:40.545941 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.0625
I0405 01:12:40.545966 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 01:12:40.545989 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 01:12:40.546011 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 01:12:40.546036 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 01:12:40.546062 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:12:40.546087 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:12:40.546108 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:12:40.546131 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:12:40.546154 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:12:40.546175 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:12:40.546197 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:12:40.546219 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:12:40.546241 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:12:40.546263 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:12:40.546283 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:12:40.546305 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:12:40.546334 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.82883 (* 0.0454545 = 0.174038 loss)
I0405 01:12:40.546362 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.68237 (* 0.0454545 = 0.167381 loss)
I0405 01:12:40.546391 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.76116 (* 0.0454545 = 0.170962 loss)
I0405 01:12:40.546416 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.81041 (* 0.0454545 = 0.1732 loss)
I0405 01:12:40.546443 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.9482 (* 0.0454545 = 0.179464 loss)
I0405 01:12:40.546470 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.90104 (* 0.0454545 = 0.17732 loss)
I0405 01:12:40.546497 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.73042 (* 0.0454545 = 0.0786555 loss)
I0405 01:12:40.546525 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.06146 (* 0.0454545 = 0.0482482 loss)
I0405 01:12:40.546552 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.725997 (* 0.0454545 = 0.0329999 loss)
I0405 01:12:40.546579 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.284488 (* 0.0454545 = 0.0129313 loss)
I0405 01:12:40.546607 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.50444e-05 (* 0.0454545 = 1.59293e-06 loss)
I0405 01:12:40.546638 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.43254e-05 (* 0.0454545 = 1.56025e-06 loss)
I0405 01:12:40.546669 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.74138e-05 (* 0.0454545 = 1.70063e-06 loss)
I0405 01:12:40.546697 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.50295e-05 (* 0.0454545 = 1.59225e-06 loss)
I0405 01:12:40.546725 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.54691e-05 (* 0.0454545 = 1.61223e-06 loss)
I0405 01:12:40.546752 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.54878e-05 (* 0.0454545 = 1.61308e-06 loss)
I0405 01:12:40.546778 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.28837e-05 (* 0.0454545 = 1.49471e-06 loss)
I0405 01:12:40.546833 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.34276e-05 (* 0.0454545 = 1.51944e-06 loss)
I0405 01:12:40.546862 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.39641e-05 (* 0.0454545 = 1.54382e-06 loss)
I0405 01:12:40.546891 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.29433e-05 (* 0.0454545 = 1.49742e-06 loss)
I0405 01:12:40.546919 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.34313e-05 (* 0.0454545 = 1.51961e-06 loss)
I0405 01:12:40.546947 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.43701e-05 (* 0.0454545 = 1.56228e-06 loss)
I0405 01:12:40.546970 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:12:40.546993 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.06655e-09
I0405 01:12:40.547018 26022 sgd_solver.cpp:106] Iteration 500, lr = 0.03998
I0405 01:21:42.751653 26022 solver.cpp:229] Iteration 550, loss = 1.19917
I0405 01:21:42.751802 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 01:21:42.751833 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 01:21:42.751848 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 01:21:42.751863 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 01:21:42.751875 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 01:21:42.751888 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.09375
I0405 01:21:42.751900 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 01:21:42.751912 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.8125
I0405 01:21:42.751925 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 01:21:42.751937 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 01:21:42.751950 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:21:42.751961 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:21:42.751973 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:21:42.751986 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:21:42.751997 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:21:42.752008 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:21:42.752020 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:21:42.752032 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:21:42.752043 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:21:42.752056 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:21:42.752082 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:21:42.752099 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:21:42.752115 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.13842 (* 0.0454545 = 0.18811 loss)
I0405 01:21:42.752130 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.8249 (* 0.0454545 = 0.173859 loss)
I0405 01:21:42.752143 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.91765 (* 0.0454545 = 0.178075 loss)
I0405 01:21:42.752158 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.01818 (* 0.0454545 = 0.182645 loss)
I0405 01:21:42.752172 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.09537 (* 0.0454545 = 0.186153 loss)
I0405 01:21:42.752187 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.89882 (* 0.0454545 = 0.177219 loss)
I0405 01:21:42.752202 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.56936 (* 0.0454545 = 0.0713348 loss)
I0405 01:21:42.752216 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.11834 (* 0.0454545 = 0.0508335 loss)
I0405 01:21:42.752234 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.47592 (* 0.0454545 = 0.0216327 loss)
I0405 01:21:42.752249 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.245379 (* 0.0454545 = 0.0111536 loss)
I0405 01:21:42.752264 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.12547e-05 (* 0.0454545 = 1.87522e-06 loss)
I0405 01:21:42.752280 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.03606e-05 (* 0.0454545 = 1.83457e-06 loss)
I0405 01:21:42.752295 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.37471e-05 (* 0.0454545 = 1.9885e-06 loss)
I0405 01:21:42.752310 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.08002e-05 (* 0.0454545 = 1.85456e-06 loss)
I0405 01:21:42.752324 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.14112e-05 (* 0.0454545 = 1.88233e-06 loss)
I0405 01:21:42.752339 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.18732e-05 (* 0.0454545 = 1.90333e-06 loss)
I0405 01:21:42.752354 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.82073e-05 (* 0.0454545 = 1.7367e-06 loss)
I0405 01:21:42.752387 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.94032e-05 (* 0.0454545 = 1.79105e-06 loss)
I0405 01:21:42.752403 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.99285e-05 (* 0.0454545 = 1.81493e-06 loss)
I0405 01:21:42.752418 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.87438e-05 (* 0.0454545 = 1.76108e-06 loss)
I0405 01:21:42.752431 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.86544e-05 (* 0.0454545 = 1.75702e-06 loss)
I0405 01:21:42.752446 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.007e-05 (* 0.0454545 = 1.82137e-06 loss)
I0405 01:21:42.752460 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:21:42.752471 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.80854e-09
I0405 01:21:42.752486 26022 sgd_solver.cpp:106] Iteration 550, lr = 0.039978
I0405 01:30:45.345886 26022 solver.cpp:229] Iteration 600, loss = 1.19695
I0405 01:30:45.346031 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 01:30:45.346051 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 01:30:45.346065 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 01:30:45.346079 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 01:30:45.346092 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 01:30:45.346104 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 01:30:45.346117 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 01:30:45.346130 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 01:30:45.346143 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 01:30:45.346155 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 01:30:45.346168 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:30:45.346180 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:30:45.346192 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:30:45.346204 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:30:45.346215 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:30:45.346228 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:30:45.346240 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:30:45.346252 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:30:45.346264 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:30:45.346276 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:30:45.346288 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:30:45.346299 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:30:45.346315 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.09366 (* 0.0454545 = 0.186076 loss)
I0405 01:30:45.346330 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.87665 (* 0.0454545 = 0.176211 loss)
I0405 01:30:45.346348 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.94917 (* 0.0454545 = 0.179508 loss)
I0405 01:30:45.346364 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.21312 (* 0.0454545 = 0.191506 loss)
I0405 01:30:45.346379 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.10543 (* 0.0454545 = 0.18661 loss)
I0405 01:30:45.346393 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.92069 (* 0.0454545 = 0.178213 loss)
I0405 01:30:45.346408 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.18667 (* 0.0454545 = 0.0993939 loss)
I0405 01:30:45.346423 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.942238 (* 0.0454545 = 0.042829 loss)
I0405 01:30:45.346438 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.433421 (* 0.0454545 = 0.019701 loss)
I0405 01:30:45.346453 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.293195 (* 0.0454545 = 0.0133271 loss)
I0405 01:30:45.346468 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.56061e-05 (* 0.0454545 = 2.073e-06 loss)
I0405 01:30:45.346483 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.53788e-05 (* 0.0454545 = 2.06267e-06 loss)
I0405 01:30:45.346498 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.87802e-05 (* 0.0454545 = 2.21728e-06 loss)
I0405 01:30:45.346513 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.61835e-05 (* 0.0454545 = 2.09925e-06 loss)
I0405 01:30:45.346529 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.63698e-05 (* 0.0454545 = 2.10772e-06 loss)
I0405 01:30:45.346544 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.66455e-05 (* 0.0454545 = 2.12025e-06 loss)
I0405 01:30:45.346559 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.29126e-05 (* 0.0454545 = 1.95057e-06 loss)
I0405 01:30:45.346592 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.37992e-05 (* 0.0454545 = 1.99087e-06 loss)
I0405 01:30:45.346609 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.41979e-05 (* 0.0454545 = 2.00899e-06 loss)
I0405 01:30:45.346624 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.32553e-05 (* 0.0454545 = 1.96615e-06 loss)
I0405 01:30:45.346639 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.30653e-05 (* 0.0454545 = 1.95751e-06 loss)
I0405 01:30:45.346654 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.50994e-05 (* 0.0454545 = 2.04997e-06 loss)
I0405 01:30:45.346668 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:30:45.346680 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.31689e-09
I0405 01:30:45.346695 26022 sgd_solver.cpp:106] Iteration 600, lr = 0.039976
I0405 01:39:47.418421 26022 solver.cpp:229] Iteration 650, loss = 1.19264
I0405 01:39:47.418647 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 01:39:47.418668 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 01:39:47.418683 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 01:39:47.418695 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 01:39:47.418707 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 01:39:47.418720 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 01:39:47.418732 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 01:39:47.418745 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 01:39:47.418758 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 01:39:47.418771 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 01:39:47.418783 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:39:47.418795 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:39:47.418807 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:39:47.418819 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:39:47.418831 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:39:47.418843 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:39:47.418855 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:39:47.418867 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:39:47.418879 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:39:47.418891 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:39:47.418903 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:39:47.418915 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:39:47.418931 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.08955 (* 0.0454545 = 0.185889 loss)
I0405 01:39:47.418947 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.78136 (* 0.0454545 = 0.17188 loss)
I0405 01:39:47.418962 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.82794 (* 0.0454545 = 0.173997 loss)
I0405 01:39:47.418977 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.96804 (* 0.0454545 = 0.180365 loss)
I0405 01:39:47.418990 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.00449 (* 0.0454545 = 0.182022 loss)
I0405 01:39:47.419008 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.85389 (* 0.0454545 = 0.175177 loss)
I0405 01:39:47.419023 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.84292 (* 0.0454545 = 0.0837689 loss)
I0405 01:39:47.419036 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.892869 (* 0.0454545 = 0.0405849 loss)
I0405 01:39:47.419051 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.728063 (* 0.0454545 = 0.0330938 loss)
I0405 01:39:47.419065 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.516918 (* 0.0454545 = 0.0234963 loss)
I0405 01:39:47.419080 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.20147e-05 (* 0.0454545 = 1.90976e-06 loss)
I0405 01:39:47.419095 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.17838e-05 (* 0.0454545 = 1.89926e-06 loss)
I0405 01:39:47.419111 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.50175e-05 (* 0.0454545 = 2.04625e-06 loss)
I0405 01:39:47.419126 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.21936e-05 (* 0.0454545 = 1.91789e-06 loss)
I0405 01:39:47.419140 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.28492e-05 (* 0.0454545 = 1.94769e-06 loss)
I0405 01:39:47.419155 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.24357e-05 (* 0.0454545 = 1.9289e-06 loss)
I0405 01:39:47.419173 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.91014e-05 (* 0.0454545 = 1.77734e-06 loss)
I0405 01:39:47.419201 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.97832e-05 (* 0.0454545 = 1.80833e-06 loss)
I0405 01:39:47.419219 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.00961e-05 (* 0.0454545 = 1.82255e-06 loss)
I0405 01:39:47.419232 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.91536e-05 (* 0.0454545 = 1.77971e-06 loss)
I0405 01:39:47.419247 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.94293e-05 (* 0.0454545 = 1.79224e-06 loss)
I0405 01:39:47.419261 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.0491e-05 (* 0.0454545 = 1.8405e-06 loss)
I0405 01:39:47.419275 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:39:47.419287 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.97012e-09
I0405 01:39:47.419302 26022 sgd_solver.cpp:106] Iteration 650, lr = 0.039974
I0405 01:48:49.629164 26022 solver.cpp:229] Iteration 700, loss = 1.19176
I0405 01:48:49.629295 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 01:48:49.629317 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 01:48:49.629329 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 01:48:49.629343 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 01:48:49.629355 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 01:48:49.629367 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 01:48:49.629380 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.625
I0405 01:48:49.629392 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 01:48:49.629405 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 01:48:49.629417 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 01:48:49.629431 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:48:49.629441 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:48:49.629453 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:48:49.629465 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:48:49.629477 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:48:49.629488 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:48:49.629500 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:48:49.629511 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:48:49.629523 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:48:49.629534 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:48:49.629546 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:48:49.629559 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:48:49.629573 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.89847 (* 0.0454545 = 0.177203 loss)
I0405 01:48:49.629588 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.73435 (* 0.0454545 = 0.169743 loss)
I0405 01:48:49.629602 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.8765 (* 0.0454545 = 0.176205 loss)
I0405 01:48:49.629617 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.83797 (* 0.0454545 = 0.174453 loss)
I0405 01:48:49.629631 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.99441 (* 0.0454545 = 0.181564 loss)
I0405 01:48:49.629645 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.8418 (* 0.0454545 = 0.174627 loss)
I0405 01:48:49.629660 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.99011 (* 0.0454545 = 0.0904595 loss)
I0405 01:48:49.629675 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.971691 (* 0.0454545 = 0.0441678 loss)
I0405 01:48:49.629689 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.24031 (* 0.0454545 = 0.0109232 loss)
I0405 01:48:49.629704 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0187645 (* 0.0454545 = 0.00085293 loss)
I0405 01:48:49.629719 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.60652e-05 (* 0.0454545 = 1.63933e-06 loss)
I0405 01:48:49.629734 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.54207e-05 (* 0.0454545 = 1.61003e-06 loss)
I0405 01:48:49.629750 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.8498e-05 (* 0.0454545 = 1.74991e-06 loss)
I0405 01:48:49.629765 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.61509e-05 (* 0.0454545 = 1.64322e-06 loss)
I0405 01:48:49.629781 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.63856e-05 (* 0.0454545 = 1.65389e-06 loss)
I0405 01:48:49.629796 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.60988e-05 (* 0.0454545 = 1.64085e-06 loss)
I0405 01:48:49.629811 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.33159e-05 (* 0.0454545 = 1.51436e-06 loss)
I0405 01:48:49.629842 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.40908e-05 (* 0.0454545 = 1.54958e-06 loss)
I0405 01:48:49.629858 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.46347e-05 (* 0.0454545 = 1.5743e-06 loss)
I0405 01:48:49.629873 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.34537e-05 (* 0.0454545 = 1.52062e-06 loss)
I0405 01:48:49.629889 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.35655e-05 (* 0.0454545 = 1.5257e-06 loss)
I0405 01:48:49.629904 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.42919e-05 (* 0.0454545 = 1.55872e-06 loss)
I0405 01:48:49.629916 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:48:49.629928 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.42252e-09
I0405 01:48:49.629942 26022 sgd_solver.cpp:106] Iteration 700, lr = 0.039972
I0405 01:57:51.837373 26022 solver.cpp:229] Iteration 750, loss = 1.19013
I0405 01:57:51.837510 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 01:57:51.837534 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 01:57:51.837549 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 01:57:51.837561 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 01:57:51.837574 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 01:57:51.837586 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 01:57:51.837599 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 01:57:51.837611 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 01:57:51.837623 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 01:57:51.837636 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 01:57:51.837647 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 01:57:51.837659 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 01:57:51.837671 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 01:57:51.837683 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 01:57:51.837694 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 01:57:51.837707 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 01:57:51.837718 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 01:57:51.837730 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 01:57:51.837743 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 01:57:51.837754 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 01:57:51.837765 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 01:57:51.837777 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 01:57:51.837795 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.09375 (* 0.0454545 = 0.18608 loss)
I0405 01:57:51.837811 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.17536 (* 0.0454545 = 0.189789 loss)
I0405 01:57:51.837829 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.17791 (* 0.0454545 = 0.189905 loss)
I0405 01:57:51.837844 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.21709 (* 0.0454545 = 0.191686 loss)
I0405 01:57:51.837859 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.17427 (* 0.0454545 = 0.18974 loss)
I0405 01:57:51.837873 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.7357 (* 0.0454545 = 0.169805 loss)
I0405 01:57:51.837888 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.1876 (* 0.0454545 = 0.0539818 loss)
I0405 01:57:51.837903 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.265084 (* 0.0454545 = 0.0120493 loss)
I0405 01:57:51.837918 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.245304 (* 0.0454545 = 0.0111502 loss)
I0405 01:57:51.837932 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0162712 (* 0.0454545 = 0.000739599 loss)
I0405 01:57:51.837947 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.54695e-05 (* 0.0454545 = 1.61225e-06 loss)
I0405 01:57:51.837962 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.50075e-05 (* 0.0454545 = 1.59125e-06 loss)
I0405 01:57:51.837977 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.77383e-05 (* 0.0454545 = 1.71538e-06 loss)
I0405 01:57:51.837992 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.56446e-05 (* 0.0454545 = 1.62021e-06 loss)
I0405 01:57:51.838007 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.57266e-05 (* 0.0454545 = 1.62393e-06 loss)
I0405 01:57:51.838022 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.61326e-05 (* 0.0454545 = 1.64239e-06 loss)
I0405 01:57:51.838037 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.27126e-05 (* 0.0454545 = 1.48694e-06 loss)
I0405 01:57:51.838068 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.34018e-05 (* 0.0454545 = 1.51827e-06 loss)
I0405 01:57:51.838084 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.40091e-05 (* 0.0454545 = 1.54587e-06 loss)
I0405 01:57:51.838099 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.28207e-05 (* 0.0454545 = 1.49185e-06 loss)
I0405 01:57:51.838114 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.30218e-05 (* 0.0454545 = 1.50099e-06 loss)
I0405 01:57:51.838129 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.41879e-05 (* 0.0454545 = 1.554e-06 loss)
I0405 01:57:51.838141 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 01:57:51.838153 26022 solver.cpp:245]     Train net output #45: total_confidence = 7.64973e-09
I0405 01:57:51.838168 26022 sgd_solver.cpp:106] Iteration 750, lr = 0.03997
I0405 02:06:54.083160 26022 solver.cpp:229] Iteration 800, loss = 1.18456
I0405 02:06:54.083279 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 02:06:54.083299 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 02:06:54.083313 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 02:06:54.083325 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 02:06:54.083338 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 02:06:54.083350 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 02:06:54.083364 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 02:06:54.083376 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 02:06:54.083389 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 02:06:54.083400 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 02:06:54.083412 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:06:54.083425 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:06:54.083436 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:06:54.083448 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:06:54.083459 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:06:54.083472 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:06:54.083483 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:06:54.083495 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:06:54.083508 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:06:54.083519 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:06:54.083531 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:06:54.083544 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:06:54.083559 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.794 (* 0.0454545 = 0.172455 loss)
I0405 02:06:54.083575 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.68009 (* 0.0454545 = 0.167277 loss)
I0405 02:06:54.083588 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.86346 (* 0.0454545 = 0.175612 loss)
I0405 02:06:54.083606 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.917 (* 0.0454545 = 0.178045 loss)
I0405 02:06:54.083621 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.99748 (* 0.0454545 = 0.181704 loss)
I0405 02:06:54.083636 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.78248 (* 0.0454545 = 0.171931 loss)
I0405 02:06:54.083650 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.84701 (* 0.0454545 = 0.083955 loss)
I0405 02:06:54.083664 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.956132 (* 0.0454545 = 0.0434605 loss)
I0405 02:06:54.083679 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.639284 (* 0.0454545 = 0.0290584 loss)
I0405 02:06:54.083696 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.478019 (* 0.0454545 = 0.0217282 loss)
I0405 02:06:54.083714 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.72008e-05 (* 0.0454545 = 3.05458e-06 loss)
I0405 02:06:54.083729 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.60942e-05 (* 0.0454545 = 3.00428e-06 loss)
I0405 02:06:54.083744 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 7.11075e-05 (* 0.0454545 = 3.23216e-06 loss)
I0405 02:06:54.083758 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 6.81939e-05 (* 0.0454545 = 3.09972e-06 loss)
I0405 02:06:54.083773 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 6.8369e-05 (* 0.0454545 = 3.10768e-06 loss)
I0405 02:06:54.083788 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 6.83112e-05 (* 0.0454545 = 3.10506e-06 loss)
I0405 02:06:54.083803 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.2549e-05 (* 0.0454545 = 2.84314e-06 loss)
I0405 02:06:54.083837 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.51498e-05 (* 0.0454545 = 2.96135e-06 loss)
I0405 02:06:54.083854 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 6.47176e-05 (* 0.0454545 = 2.94171e-06 loss)
I0405 02:06:54.083869 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 6.27931e-05 (* 0.0454545 = 2.85423e-06 loss)
I0405 02:06:54.083884 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 6.42034e-05 (* 0.0454545 = 2.91833e-06 loss)
I0405 02:06:54.083899 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.52727e-05 (* 0.0454545 = 2.96694e-06 loss)
I0405 02:06:54.083912 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:06:54.083925 26022 solver.cpp:245]     Train net output #45: total_confidence = 8.20168e-09
I0405 02:06:54.083940 26022 sgd_solver.cpp:106] Iteration 800, lr = 0.039968
I0405 02:15:56.182812 26022 solver.cpp:229] Iteration 850, loss = 1.17905
I0405 02:15:56.183007 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 02:15:56.183028 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 02:15:56.183043 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 02:15:56.183054 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 02:15:56.183068 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 02:15:56.183079 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 02:15:56.183092 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 02:15:56.183104 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 02:15:56.183117 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 02:15:56.183130 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 02:15:56.183141 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:15:56.183153 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:15:56.183164 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:15:56.183176 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:15:56.183188 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:15:56.183200 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:15:56.183212 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:15:56.183224 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:15:56.183236 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:15:56.183248 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:15:56.183259 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:15:56.183270 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:15:56.183285 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.55068 (* 0.0454545 = 0.161394 loss)
I0405 02:15:56.183301 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.79589 (* 0.0454545 = 0.172541 loss)
I0405 02:15:56.183316 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.05486 (* 0.0454545 = 0.184312 loss)
I0405 02:15:56.183331 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.90259 (* 0.0454545 = 0.17739 loss)
I0405 02:15:56.183346 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.03407 (* 0.0454545 = 0.183367 loss)
I0405 02:15:56.183360 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.67407 (* 0.0454545 = 0.167003 loss)
I0405 02:15:56.183375 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.5217 (* 0.0454545 = 0.0691683 loss)
I0405 02:15:56.183389 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.00957 (* 0.0454545 = 0.0458898 loss)
I0405 02:15:56.183404 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.649077 (* 0.0454545 = 0.0295035 loss)
I0405 02:15:56.183420 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0186128 (* 0.0454545 = 0.000846034 loss)
I0405 02:15:56.183454 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.00012097 (* 0.0454545 = 5.49863e-06 loss)
I0405 02:15:56.183471 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000118609 (* 0.0454545 = 5.39132e-06 loss)
I0405 02:15:56.183487 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000126176 (* 0.0454545 = 5.73527e-06 loss)
I0405 02:15:56.183502 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000119694 (* 0.0454545 = 5.44062e-06 loss)
I0405 02:15:56.183517 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000121087 (* 0.0454545 = 5.50396e-06 loss)
I0405 02:15:56.183534 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000119675 (* 0.0454545 = 5.43976e-06 loss)
I0405 02:15:56.183560 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000111752 (* 0.0454545 = 5.07965e-06 loss)
I0405 02:15:56.183591 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000114234 (* 0.0454545 = 5.19247e-06 loss)
I0405 02:15:56.183611 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.00011587 (* 0.0454545 = 5.26682e-06 loss)
I0405 02:15:56.183626 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000111324 (* 0.0454545 = 5.06017e-06 loss)
I0405 02:15:56.183641 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000113571 (* 0.0454545 = 5.16231e-06 loss)
I0405 02:15:56.183656 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000114558 (* 0.0454545 = 5.2072e-06 loss)
I0405 02:15:56.183670 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:15:56.183682 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.62544e-09
I0405 02:15:56.183697 26022 sgd_solver.cpp:106] Iteration 850, lr = 0.039966
I0405 02:24:58.318820 26022 solver.cpp:229] Iteration 900, loss = 1.17929
I0405 02:24:58.318936 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 02:24:58.318956 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 02:24:58.318970 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 02:24:58.318982 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 02:24:58.318995 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 02:24:58.319007 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 02:24:58.319020 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 02:24:58.319032 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 02:24:58.319046 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 02:24:58.319057 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 02:24:58.319069 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:24:58.319082 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:24:58.319092 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:24:58.319104 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:24:58.319116 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:24:58.319128 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:24:58.319139 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:24:58.319151 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:24:58.319162 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:24:58.319175 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:24:58.319185 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:24:58.319197 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:24:58.319212 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.87605 (* 0.0454545 = 0.176184 loss)
I0405 02:24:58.319227 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.07787 (* 0.0454545 = 0.185358 loss)
I0405 02:24:58.319242 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.00669 (* 0.0454545 = 0.182122 loss)
I0405 02:24:58.319257 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.06781 (* 0.0454545 = 0.184901 loss)
I0405 02:24:58.319272 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.09216 (* 0.0454545 = 0.186007 loss)
I0405 02:24:58.319285 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.78551 (* 0.0454545 = 0.172069 loss)
I0405 02:24:58.319300 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.89327 (* 0.0454545 = 0.0860579 loss)
I0405 02:24:58.319314 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.972107 (* 0.0454545 = 0.0441867 loss)
I0405 02:24:58.319329 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.407752 (* 0.0454545 = 0.0185342 loss)
I0405 02:24:58.319344 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0135476 (* 0.0454545 = 0.000615802 loss)
I0405 02:24:58.319358 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 7.89697e-05 (* 0.0454545 = 3.58953e-06 loss)
I0405 02:24:58.319375 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 7.81389e-05 (* 0.0454545 = 3.55177e-06 loss)
I0405 02:24:58.319389 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 8.24464e-05 (* 0.0454545 = 3.74756e-06 loss)
I0405 02:24:58.319403 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 7.96518e-05 (* 0.0454545 = 3.62054e-06 loss)
I0405 02:24:58.319418 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.95325e-05 (* 0.0454545 = 3.61511e-06 loss)
I0405 02:24:58.319432 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.10342e-05 (* 0.0454545 = 3.68337e-06 loss)
I0405 02:24:58.319447 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 7.43306e-05 (* 0.0454545 = 3.37866e-06 loss)
I0405 02:24:58.319476 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 7.59404e-05 (* 0.0454545 = 3.45183e-06 loss)
I0405 02:24:58.319494 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 7.60074e-05 (* 0.0454545 = 3.45488e-06 loss)
I0405 02:24:58.319507 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 7.43642e-05 (* 0.0454545 = 3.38019e-06 loss)
I0405 02:24:58.319522 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 7.50703e-05 (* 0.0454545 = 3.41229e-06 loss)
I0405 02:24:58.319537 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 7.68496e-05 (* 0.0454545 = 3.49316e-06 loss)
I0405 02:24:58.319550 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:24:58.319563 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.46849e-08
I0405 02:24:58.319578 26022 sgd_solver.cpp:106] Iteration 900, lr = 0.039964
I0405 02:34:00.609575 26022 solver.cpp:229] Iteration 950, loss = 1.17626
I0405 02:34:00.609750 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 02:34:00.609782 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 02:34:00.609808 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 02:34:00.609833 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 02:34:00.609859 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 02:34:00.609882 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 02:34:00.609906 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.59375
I0405 02:34:00.609930 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 02:34:00.609946 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 02:34:00.609957 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 02:34:00.609969 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:34:00.609982 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:34:00.609992 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:34:00.610004 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:34:00.610016 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:34:00.610028 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:34:00.610040 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:34:00.610051 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:34:00.610064 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:34:00.610074 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:34:00.610086 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:34:00.610097 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:34:00.610113 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.83618 (* 0.0454545 = 0.174372 loss)
I0405 02:34:00.610128 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.79912 (* 0.0454545 = 0.172687 loss)
I0405 02:34:00.610142 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.74721 (* 0.0454545 = 0.170328 loss)
I0405 02:34:00.610157 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.85507 (* 0.0454545 = 0.175231 loss)
I0405 02:34:00.610172 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.99369 (* 0.0454545 = 0.181531 loss)
I0405 02:34:00.610185 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.77534 (* 0.0454545 = 0.171606 loss)
I0405 02:34:00.610199 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.23163 (* 0.0454545 = 0.101438 loss)
I0405 02:34:00.610213 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.496714 (* 0.0454545 = 0.0225779 loss)
I0405 02:34:00.610229 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0443468 (* 0.0454545 = 0.00201576 loss)
I0405 02:34:00.610244 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0173282 (* 0.0454545 = 0.000787644 loss)
I0405 02:34:00.610258 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000149539 (* 0.0454545 = 6.79724e-06 loss)
I0405 02:34:00.610272 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000147124 (* 0.0454545 = 6.68745e-06 loss)
I0405 02:34:00.610287 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000154465 (* 0.0454545 = 7.02114e-06 loss)
I0405 02:34:00.610302 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000152737 (* 0.0454545 = 6.94261e-06 loss)
I0405 02:34:00.610317 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000150288 (* 0.0454545 = 6.83129e-06 loss)
I0405 02:34:00.610332 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000152029 (* 0.0454545 = 6.91041e-06 loss)
I0405 02:34:00.610347 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000142097 (* 0.0454545 = 6.45897e-06 loss)
I0405 02:34:00.610379 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000143046 (* 0.0454545 = 6.5021e-06 loss)
I0405 02:34:00.610395 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000146541 (* 0.0454545 = 6.66095e-06 loss)
I0405 02:34:00.610410 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000141129 (* 0.0454545 = 6.41494e-06 loss)
I0405 02:34:00.610425 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000142819 (* 0.0454545 = 6.49175e-06 loss)
I0405 02:34:00.610440 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000146446 (* 0.0454545 = 6.65661e-06 loss)
I0405 02:34:00.610452 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:34:00.610465 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.00246e-08
I0405 02:34:00.610479 26022 sgd_solver.cpp:106] Iteration 950, lr = 0.039962
I0405 02:42:52.367350 26022 solver.cpp:338] Iteration 1000, Testing net (#0)
I0405 02:43:09.018882 26022 solver.cpp:393] Test loss: 1.0976
I0405 02:43:09.018930 26022 solver.cpp:406]     Test net output #0: loss/accuracy01 = 0
I0405 02:43:09.018947 26022 solver.cpp:406]     Test net output #1: loss/accuracy02 = 0.066
I0405 02:43:09.018961 26022 solver.cpp:406]     Test net output #2: loss/accuracy03 = 0
I0405 02:43:09.018973 26022 solver.cpp:406]     Test net output #3: loss/accuracy04 = 0.002
I0405 02:43:09.018986 26022 solver.cpp:406]     Test net output #4: loss/accuracy05 = 0.007
I0405 02:43:09.018998 26022 solver.cpp:406]     Test net output #5: loss/accuracy06 = 0.501
I0405 02:43:09.019011 26022 solver.cpp:406]     Test net output #6: loss/accuracy07 = 0.894
I0405 02:43:09.019022 26022 solver.cpp:406]     Test net output #7: loss/accuracy08 = 0.97
I0405 02:43:09.019034 26022 solver.cpp:406]     Test net output #8: loss/accuracy09 = 0.995
I0405 02:43:09.019045 26022 solver.cpp:406]     Test net output #9: loss/accuracy10 = 0.998
I0405 02:43:09.019057 26022 solver.cpp:406]     Test net output #10: loss/accuracy11 = 1
I0405 02:43:09.019068 26022 solver.cpp:406]     Test net output #11: loss/accuracy12 = 1
I0405 02:43:09.019080 26022 solver.cpp:406]     Test net output #12: loss/accuracy13 = 1
I0405 02:43:09.019093 26022 solver.cpp:406]     Test net output #13: loss/accuracy14 = 1
I0405 02:43:09.019104 26022 solver.cpp:406]     Test net output #14: loss/accuracy15 = 1
I0405 02:43:09.019117 26022 solver.cpp:406]     Test net output #15: loss/accuracy16 = 1
I0405 02:43:09.019130 26022 solver.cpp:406]     Test net output #16: loss/accuracy17 = 1
I0405 02:43:09.019141 26022 solver.cpp:406]     Test net output #17: loss/accuracy18 = 1
I0405 02:43:09.019153 26022 solver.cpp:406]     Test net output #18: loss/accuracy19 = 1
I0405 02:43:09.019167 26022 solver.cpp:406]     Test net output #19: loss/accuracy20 = 1
I0405 02:43:09.019179 26022 solver.cpp:406]     Test net output #20: loss/accuracy21 = 1
I0405 02:43:09.019191 26022 solver.cpp:406]     Test net output #21: loss/accuracy22 = 1
I0405 02:43:09.019207 26022 solver.cpp:406]     Test net output #22: loss/loss01 = 3.47848 (* 0.0454545 = 0.158113 loss)
I0405 02:43:09.019222 26022 solver.cpp:406]     Test net output #23: loss/loss02 = 3.72914 (* 0.0454545 = 0.169506 loss)
I0405 02:43:09.019235 26022 solver.cpp:406]     Test net output #24: loss/loss03 = 3.99079 (* 0.0454545 = 0.181399 loss)
I0405 02:43:09.019249 26022 solver.cpp:406]     Test net output #25: loss/loss04 = 3.99051 (* 0.0454545 = 0.181387 loss)
I0405 02:43:09.019263 26022 solver.cpp:406]     Test net output #26: loss/loss05 = 4.01313 (* 0.0454545 = 0.182415 loss)
I0405 02:43:09.019276 26022 solver.cpp:406]     Test net output #27: loss/loss06 = 3.53388 (* 0.0454545 = 0.160631 loss)
I0405 02:43:09.019290 26022 solver.cpp:406]     Test net output #28: loss/loss07 = 0.956657 (* 0.0454545 = 0.0434844 loss)
I0405 02:43:09.019305 26022 solver.cpp:406]     Test net output #29: loss/loss08 = 0.334567 (* 0.0454545 = 0.0152076 loss)
I0405 02:43:09.019320 26022 solver.cpp:406]     Test net output #30: loss/loss09 = 0.0821666 (* 0.0454545 = 0.00373485 loss)
I0405 02:43:09.019335 26022 solver.cpp:406]     Test net output #31: loss/loss10 = 0.0353198 (* 0.0454545 = 0.00160545 loss)
I0405 02:43:09.019348 26022 solver.cpp:406]     Test net output #32: loss/loss11 = 0.000223803 (* 0.0454545 = 1.01729e-05 loss)
I0405 02:43:09.019363 26022 solver.cpp:406]     Test net output #33: loss/loss12 = 0.000219568 (* 0.0454545 = 9.98036e-06 loss)
I0405 02:43:09.019378 26022 solver.cpp:406]     Test net output #34: loss/loss13 = 0.000230953 (* 0.0454545 = 1.04979e-05 loss)
I0405 02:43:09.019393 26022 solver.cpp:406]     Test net output #35: loss/loss14 = 0.000224668 (* 0.0454545 = 1.02122e-05 loss)
I0405 02:43:09.019407 26022 solver.cpp:406]     Test net output #36: loss/loss15 = 0.000224206 (* 0.0454545 = 1.01912e-05 loss)
I0405 02:43:09.019423 26022 solver.cpp:406]     Test net output #37: loss/loss16 = 0.00022409 (* 0.0454545 = 1.01859e-05 loss)
I0405 02:43:09.019436 26022 solver.cpp:406]     Test net output #38: loss/loss17 = 0.000211603 (* 0.0454545 = 9.61834e-06 loss)
I0405 02:43:09.019480 26022 solver.cpp:406]     Test net output #39: loss/loss18 = 0.000213975 (* 0.0454545 = 9.72614e-06 loss)
I0405 02:43:09.019496 26022 solver.cpp:406]     Test net output #40: loss/loss19 = 0.000215922 (* 0.0454545 = 9.81466e-06 loss)
I0405 02:43:09.019511 26022 solver.cpp:406]     Test net output #41: loss/loss20 = 0.000210648 (* 0.0454545 = 9.57489e-06 loss)
I0405 02:43:09.019526 26022 solver.cpp:406]     Test net output #42: loss/loss21 = 0.000212573 (* 0.0454545 = 9.66243e-06 loss)
I0405 02:43:09.019541 26022 solver.cpp:406]     Test net output #43: loss/loss22 = 0.000218494 (* 0.0454545 = 9.93153e-06 loss)
I0405 02:43:09.019552 26022 solver.cpp:406]     Test net output #44: total_accuracy = 0
I0405 02:43:09.019564 26022 solver.cpp:406]     Test net output #45: total_confidence = 6.39859e-09
I0405 02:43:19.335474 26022 solver.cpp:229] Iteration 1000, loss = 1.1725
I0405 02:43:19.335530 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 02:43:19.335549 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 02:43:19.335563 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 02:43:19.335575 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 02:43:19.335588 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 02:43:19.335602 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 02:43:19.335616 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 02:43:19.335629 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 02:43:19.335641 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 02:43:19.335654 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 02:43:19.335665 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:43:19.335677 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:43:19.335690 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:43:19.335701 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:43:19.335713 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:43:19.335726 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:43:19.335737 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:43:19.335748 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:43:19.335760 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:43:19.335772 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:43:19.335783 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:43:19.335795 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:43:19.335810 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.05236 (* 0.0454545 = 0.184198 loss)
I0405 02:43:19.335825 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.144 (* 0.0454545 = 0.188364 loss)
I0405 02:43:19.335840 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.01089 (* 0.0454545 = 0.182313 loss)
I0405 02:43:19.335855 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.97703 (* 0.0454545 = 0.180774 loss)
I0405 02:43:19.335870 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.06477 (* 0.0454545 = 0.184762 loss)
I0405 02:43:19.335885 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.73746 (* 0.0454545 = 0.169884 loss)
I0405 02:43:19.335898 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.70262 (* 0.0454545 = 0.0773917 loss)
I0405 02:43:19.335912 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.582836 (* 0.0454545 = 0.0264926 loss)
I0405 02:43:19.335927 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0427708 (* 0.0454545 = 0.00194413 loss)
I0405 02:43:19.335947 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0184913 (* 0.0454545 = 0.000840514 loss)
I0405 02:43:19.335991 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000237069 (* 0.0454545 = 1.07759e-05 loss)
I0405 02:43:19.336009 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000233027 (* 0.0454545 = 1.05921e-05 loss)
I0405 02:43:19.336024 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000243927 (* 0.0454545 = 1.10876e-05 loss)
I0405 02:43:19.336040 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.0002364 (* 0.0454545 = 1.07454e-05 loss)
I0405 02:43:19.336055 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000238737 (* 0.0454545 = 1.08517e-05 loss)
I0405 02:43:19.336082 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000237071 (* 0.0454545 = 1.0776e-05 loss)
I0405 02:43:19.336100 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000224487 (* 0.0454545 = 1.0204e-05 loss)
I0405 02:43:19.336115 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000226802 (* 0.0454545 = 1.03092e-05 loss)
I0405 02:43:19.336130 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000226482 (* 0.0454545 = 1.02946e-05 loss)
I0405 02:43:19.336145 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000224365 (* 0.0454545 = 1.01984e-05 loss)
I0405 02:43:19.336159 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000226528 (* 0.0454545 = 1.02967e-05 loss)
I0405 02:43:19.336174 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000233715 (* 0.0454545 = 1.06234e-05 loss)
I0405 02:43:19.336187 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:43:19.336199 26022 solver.cpp:245]     Train net output #45: total_confidence = 7.46125e-09
I0405 02:43:19.336215 26022 sgd_solver.cpp:106] Iteration 1000, lr = 0.03996
I0405 02:52:21.431280 26022 solver.cpp:229] Iteration 1050, loss = 1.1733
I0405 02:52:21.431480 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 02:52:21.431505 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 02:52:21.431519 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 02:52:21.431531 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 02:52:21.431545 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 02:52:21.431557 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 02:52:21.431571 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 02:52:21.431582 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.8125
I0405 02:52:21.431594 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.875
I0405 02:52:21.431607 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 02:52:21.431620 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 02:52:21.431632 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 02:52:21.431644 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 02:52:21.431656 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 02:52:21.431668 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 02:52:21.431679 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 02:52:21.431691 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 02:52:21.431704 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 02:52:21.431715 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 02:52:21.431727 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 02:52:21.431740 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 02:52:21.431753 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 02:52:21.431769 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.63954 (* 0.0454545 = 0.165434 loss)
I0405 02:52:21.431784 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.86525 (* 0.0454545 = 0.175693 loss)
I0405 02:52:21.431799 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.79662 (* 0.0454545 = 0.172574 loss)
I0405 02:52:21.431814 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.96958 (* 0.0454545 = 0.180435 loss)
I0405 02:52:21.431828 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.03395 (* 0.0454545 = 0.183361 loss)
I0405 02:52:21.431843 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.61449 (* 0.0454545 = 0.164295 loss)
I0405 02:52:21.431857 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.67549 (* 0.0454545 = 0.0761585 loss)
I0405 02:52:21.431871 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.24515 (* 0.0454545 = 0.0565975 loss)
I0405 02:52:21.431885 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.969847 (* 0.0454545 = 0.044084 loss)
I0405 02:52:21.431900 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.562512 (* 0.0454545 = 0.0255687 loss)
I0405 02:52:21.431915 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000133148 (* 0.0454545 = 6.05217e-06 loss)
I0405 02:52:21.431931 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.00013128 (* 0.0454545 = 5.96728e-06 loss)
I0405 02:52:21.431946 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000137 (* 0.0454545 = 6.22728e-06 loss)
I0405 02:52:21.431964 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000132684 (* 0.0454545 = 6.03107e-06 loss)
I0405 02:52:21.431979 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000134203 (* 0.0454545 = 6.10013e-06 loss)
I0405 02:52:21.431994 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000131878 (* 0.0454545 = 5.99448e-06 loss)
I0405 02:52:21.432009 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00012726 (* 0.0454545 = 5.78456e-06 loss)
I0405 02:52:21.432041 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000127987 (* 0.0454545 = 5.8176e-06 loss)
I0405 02:52:21.432057 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000129294 (* 0.0454545 = 5.87698e-06 loss)
I0405 02:52:21.432088 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.00012535 (* 0.0454545 = 5.69772e-06 loss)
I0405 02:52:21.432106 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000126901 (* 0.0454545 = 5.76822e-06 loss)
I0405 02:52:21.432121 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000129448 (* 0.0454545 = 5.88402e-06 loss)
I0405 02:52:21.432133 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 02:52:21.432145 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.33633e-08
I0405 02:52:21.432160 26022 sgd_solver.cpp:106] Iteration 1050, lr = 0.039958
I0405 03:01:23.483218 26022 solver.cpp:229] Iteration 1100, loss = 1.16672
I0405 03:01:23.483366 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 03:01:23.483389 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.15625
I0405 03:01:23.483404 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 03:01:23.483417 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 03:01:23.483430 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 03:01:23.483443 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 03:01:23.483455 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 03:01:23.483469 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 03:01:23.483480 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 03:01:23.483494 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 03:01:23.483505 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:01:23.483517 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:01:23.483530 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:01:23.483541 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:01:23.483553 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:01:23.483566 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:01:23.483577 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:01:23.483589 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:01:23.483602 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:01:23.483613 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:01:23.483625 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:01:23.483636 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:01:23.483652 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.64008 (* 0.0454545 = 0.165458 loss)
I0405 03:01:23.483669 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.6677 (* 0.0454545 = 0.166714 loss)
I0405 03:01:23.483682 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.78213 (* 0.0454545 = 0.171915 loss)
I0405 03:01:23.483696 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.7499 (* 0.0454545 = 0.17045 loss)
I0405 03:01:23.483711 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.8809 (* 0.0454545 = 0.176405 loss)
I0405 03:01:23.483726 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.71625 (* 0.0454545 = 0.168921 loss)
I0405 03:01:23.483739 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.00426 (* 0.0454545 = 0.0911029 loss)
I0405 03:01:23.483754 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.565093 (* 0.0454545 = 0.025686 loss)
I0405 03:01:23.483768 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.575229 (* 0.0454545 = 0.0261468 loss)
I0405 03:01:23.483783 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.395119 (* 0.0454545 = 0.01796 loss)
I0405 03:01:23.483798 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000361151 (* 0.0454545 = 1.6416e-05 loss)
I0405 03:01:23.483817 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000351951 (* 0.0454545 = 1.59978e-05 loss)
I0405 03:01:23.483832 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000371088 (* 0.0454545 = 1.68676e-05 loss)
I0405 03:01:23.483847 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.0003604 (* 0.0454545 = 1.63818e-05 loss)
I0405 03:01:23.483862 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000361026 (* 0.0454545 = 1.64103e-05 loss)
I0405 03:01:23.483877 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000358865 (* 0.0454545 = 1.63121e-05 loss)
I0405 03:01:23.483892 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000339598 (* 0.0454545 = 1.54363e-05 loss)
I0405 03:01:23.483924 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000342388 (* 0.0454545 = 1.55631e-05 loss)
I0405 03:01:23.483940 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000349614 (* 0.0454545 = 1.58916e-05 loss)
I0405 03:01:23.483955 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000340666 (* 0.0454545 = 1.54848e-05 loss)
I0405 03:01:23.483973 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000343324 (* 0.0454545 = 1.56056e-05 loss)
I0405 03:01:23.483989 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000350932 (* 0.0454545 = 1.59514e-05 loss)
I0405 03:01:23.484001 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:01:23.484014 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.14692e-08
I0405 03:01:23.484030 26022 sgd_solver.cpp:106] Iteration 1100, lr = 0.039956
I0405 03:10:25.616874 26022 solver.cpp:229] Iteration 1150, loss = 1.16862
I0405 03:10:25.617025 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 03:10:25.617048 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 03:10:25.617061 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 03:10:25.617074 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 03:10:25.617085 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0
I0405 03:10:25.617097 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.46875
I0405 03:10:25.617110 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 03:10:25.617122 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 03:10:25.617136 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 03:10:25.617147 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 03:10:25.617159 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:10:25.617172 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:10:25.617182 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:10:25.617194 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:10:25.617207 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:10:25.617218 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:10:25.617229 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:10:25.617241 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:10:25.617252 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:10:25.617264 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:10:25.617276 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:10:25.617288 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:10:25.617305 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 4.13078 (* 0.0454545 = 0.187763 loss)
I0405 03:10:25.617319 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.72054 (* 0.0454545 = 0.169115 loss)
I0405 03:10:25.617334 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 4.0911 (* 0.0454545 = 0.185959 loss)
I0405 03:10:25.617348 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.9875 (* 0.0454545 = 0.18125 loss)
I0405 03:10:25.617363 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.92422 (* 0.0454545 = 0.178374 loss)
I0405 03:10:25.617377 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.49351 (* 0.0454545 = 0.158796 loss)
I0405 03:10:25.617393 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.12282 (* 0.0454545 = 0.0510372 loss)
I0405 03:10:25.617406 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.557153 (* 0.0454545 = 0.0253252 loss)
I0405 03:10:25.617421 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.423012 (* 0.0454545 = 0.0192278 loss)
I0405 03:10:25.617439 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.43682 (* 0.0454545 = 0.0198554 loss)
I0405 03:10:25.617455 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 8.06846e-05 (* 0.0454545 = 3.66748e-06 loss)
I0405 03:10:25.617470 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 7.81224e-05 (* 0.0454545 = 3.55102e-06 loss)
I0405 03:10:25.617485 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 8.08744e-05 (* 0.0454545 = 3.67611e-06 loss)
I0405 03:10:25.617499 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 8.06844e-05 (* 0.0454545 = 3.66747e-06 loss)
I0405 03:10:25.617513 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.92608e-05 (* 0.0454545 = 3.60277e-06 loss)
I0405 03:10:25.617528 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.10889e-05 (* 0.0454545 = 3.68586e-06 loss)
I0405 03:10:25.617543 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 7.51355e-05 (* 0.0454545 = 3.41525e-06 loss)
I0405 03:10:25.617574 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 7.67865e-05 (* 0.0454545 = 3.4903e-06 loss)
I0405 03:10:25.617590 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 7.65888e-05 (* 0.0454545 = 3.48131e-06 loss)
I0405 03:10:25.617605 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 7.63598e-05 (* 0.0454545 = 3.4709e-06 loss)
I0405 03:10:25.617620 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 7.68723e-05 (* 0.0454545 = 3.49419e-06 loss)
I0405 03:10:25.617635 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 7.94232e-05 (* 0.0454545 = 3.61014e-06 loss)
I0405 03:10:25.617647 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:10:25.617660 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.71604e-08
I0405 03:10:25.617674 26022 sgd_solver.cpp:106] Iteration 1150, lr = 0.039954
I0405 03:19:27.962780 26022 solver.cpp:229] Iteration 1200, loss = 1.1609
I0405 03:19:27.962961 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 03:19:27.962982 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 03:19:27.962996 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 03:19:27.963008 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 03:19:27.963021 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 03:19:27.963032 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 03:19:27.963045 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 03:19:27.963057 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 03:19:27.963070 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 03:19:27.963083 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 03:19:27.963094 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:19:27.963106 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:19:27.963119 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:19:27.963129 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:19:27.963141 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:19:27.963153 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:19:27.963165 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:19:27.963176 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:19:27.963187 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:19:27.963199 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:19:27.963212 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:19:27.963224 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:19:27.963240 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.55007 (* 0.0454545 = 0.161367 loss)
I0405 03:19:27.963254 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.75748 (* 0.0454545 = 0.170794 loss)
I0405 03:19:27.963269 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.78453 (* 0.0454545 = 0.172024 loss)
I0405 03:19:27.963284 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.81877 (* 0.0454545 = 0.17358 loss)
I0405 03:19:27.963299 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.77564 (* 0.0454545 = 0.17162 loss)
I0405 03:19:27.963313 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.48725 (* 0.0454545 = 0.158511 loss)
I0405 03:19:27.963327 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.85898 (* 0.0454545 = 0.0844989 loss)
I0405 03:19:27.963342 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.224943 (* 0.0454545 = 0.0102247 loss)
I0405 03:19:27.963356 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0295486 (* 0.0454545 = 0.00134312 loss)
I0405 03:19:27.963372 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0109242 (* 0.0454545 = 0.000496554 loss)
I0405 03:19:27.963388 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000145794 (* 0.0454545 = 6.627e-06 loss)
I0405 03:19:27.963404 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000142488 (* 0.0454545 = 6.47675e-06 loss)
I0405 03:19:27.963433 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000146855 (* 0.0454545 = 6.67523e-06 loss)
I0405 03:19:27.963449 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000144862 (* 0.0454545 = 6.58464e-06 loss)
I0405 03:19:27.963464 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000144791 (* 0.0454545 = 6.5814e-06 loss)
I0405 03:19:27.963481 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000142254 (* 0.0454545 = 6.46609e-06 loss)
I0405 03:19:27.963497 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000138881 (* 0.0454545 = 6.31278e-06 loss)
I0405 03:19:27.963526 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000138891 (* 0.0454545 = 6.31325e-06 loss)
I0405 03:19:27.963541 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000140559 (* 0.0454545 = 6.38905e-06 loss)
I0405 03:19:27.963557 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000135657 (* 0.0454545 = 6.16624e-06 loss)
I0405 03:19:27.963572 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000140024 (* 0.0454545 = 6.36472e-06 loss)
I0405 03:19:27.963587 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000141032 (* 0.0454545 = 6.41056e-06 loss)
I0405 03:19:27.963599 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:19:27.963611 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.6985e-07
I0405 03:19:27.963626 26022 sgd_solver.cpp:106] Iteration 1200, lr = 0.039952
I0405 03:28:30.218416 26022 solver.cpp:229] Iteration 1250, loss = 1.1381
I0405 03:28:30.218570 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 03:28:30.218590 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 03:28:30.218605 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 03:28:30.218616 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 03:28:30.218628 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 03:28:30.218641 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 03:28:30.218653 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 03:28:30.218665 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 03:28:30.218677 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 03:28:30.218689 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 03:28:30.218701 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:28:30.218713 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:28:30.218725 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:28:30.218736 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:28:30.218749 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:28:30.218760 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:28:30.218771 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:28:30.218782 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:28:30.218794 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:28:30.218807 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:28:30.218828 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:28:30.218852 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:28:30.218871 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.62494 (* 0.0454545 = 0.16477 loss)
I0405 03:28:30.218886 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.73217 (* 0.0454545 = 0.169644 loss)
I0405 03:28:30.218900 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.73145 (* 0.0454545 = 0.169611 loss)
I0405 03:28:30.218915 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.23568 (* 0.0454545 = 0.192531 loss)
I0405 03:28:30.218930 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.34047 (* 0.0454545 = 0.15184 loss)
I0405 03:28:30.218945 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.90087 (* 0.0454545 = 0.131858 loss)
I0405 03:28:30.218958 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.90291 (* 0.0454545 = 0.086496 loss)
I0405 03:28:30.218977 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.770584 (* 0.0454545 = 0.0350266 loss)
I0405 03:28:30.218991 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.252814 (* 0.0454545 = 0.0114916 loss)
I0405 03:28:30.219007 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00851682 (* 0.0454545 = 0.000387128 loss)
I0405 03:28:30.219022 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 7.61399e-05 (* 0.0454545 = 3.46091e-06 loss)
I0405 03:28:30.219036 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 7.56815e-05 (* 0.0454545 = 3.44007e-06 loss)
I0405 03:28:30.219051 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 7.88528e-05 (* 0.0454545 = 3.58422e-06 loss)
I0405 03:28:30.219066 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 7.72614e-05 (* 0.0454545 = 3.51188e-06 loss)
I0405 03:28:30.219080 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.6926e-05 (* 0.0454545 = 3.49664e-06 loss)
I0405 03:28:30.219095 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 7.7459e-05 (* 0.0454545 = 3.52086e-06 loss)
I0405 03:28:30.219110 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 7.37063e-05 (* 0.0454545 = 3.35029e-06 loss)
I0405 03:28:30.219143 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 7.43324e-05 (* 0.0454545 = 3.37875e-06 loss)
I0405 03:28:30.219158 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 7.43476e-05 (* 0.0454545 = 3.37943e-06 loss)
I0405 03:28:30.219173 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 7.27897e-05 (* 0.0454545 = 3.30862e-06 loss)
I0405 03:28:30.219187 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 7.36319e-05 (* 0.0454545 = 3.3469e-06 loss)
I0405 03:28:30.219202 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 7.49622e-05 (* 0.0454545 = 3.40738e-06 loss)
I0405 03:28:30.219214 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:28:30.219226 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.60822e-07
I0405 03:28:30.219241 26022 sgd_solver.cpp:106] Iteration 1250, lr = 0.03995
I0405 03:37:32.297591 26022 solver.cpp:229] Iteration 1300, loss = 1.12825
I0405 03:37:32.297755 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 03:37:32.297777 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 03:37:32.297791 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 03:37:32.297804 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 03:37:32.297816 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 03:37:32.297829 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 03:37:32.297842 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 03:37:32.297854 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 03:37:32.297866 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 03:37:32.297878 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 03:37:32.297890 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:37:32.297902 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:37:32.297914 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:37:32.297925 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:37:32.297937 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:37:32.297948 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:37:32.297960 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:37:32.297972 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:37:32.297984 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:37:32.297996 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:37:32.298007 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:37:32.298019 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:37:32.298034 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.89656 (* 0.0454545 = 0.177116 loss)
I0405 03:37:32.298049 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.85652 (* 0.0454545 = 0.175296 loss)
I0405 03:37:32.298063 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.78311 (* 0.0454545 = 0.17196 loss)
I0405 03:37:32.298079 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.87697 (* 0.0454545 = 0.176226 loss)
I0405 03:37:32.298092 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.48238 (* 0.0454545 = 0.15829 loss)
I0405 03:37:32.298107 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.22216 (* 0.0454545 = 0.146462 loss)
I0405 03:37:32.298121 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.56118 (* 0.0454545 = 0.0709628 loss)
I0405 03:37:32.298136 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.41919 (* 0.0454545 = 0.0190541 loss)
I0405 03:37:32.298151 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0428736 (* 0.0454545 = 0.0019488 loss)
I0405 03:37:32.298166 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0220191 (* 0.0454545 = 0.00100087 loss)
I0405 03:37:32.298182 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.00040926 (* 0.0454545 = 1.86027e-05 loss)
I0405 03:37:32.298197 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000413913 (* 0.0454545 = 1.88142e-05 loss)
I0405 03:37:32.298212 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000419584 (* 0.0454545 = 1.9072e-05 loss)
I0405 03:37:32.298226 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000413896 (* 0.0454545 = 1.88135e-05 loss)
I0405 03:37:32.298241 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000417166 (* 0.0454545 = 1.89621e-05 loss)
I0405 03:37:32.298256 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000412759 (* 0.0454545 = 1.87618e-05 loss)
I0405 03:37:32.298271 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000401908 (* 0.0454545 = 1.82685e-05 loss)
I0405 03:37:32.298303 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000410834 (* 0.0454545 = 1.86743e-05 loss)
I0405 03:37:32.298321 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.00039866 (* 0.0454545 = 1.81209e-05 loss)
I0405 03:37:32.298336 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000400482 (* 0.0454545 = 1.82037e-05 loss)
I0405 03:37:32.298351 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000397744 (* 0.0454545 = 1.80793e-05 loss)
I0405 03:37:32.298365 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000404162 (* 0.0454545 = 1.8371e-05 loss)
I0405 03:37:32.298382 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:37:32.298394 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.50374e-07
I0405 03:37:32.298410 26022 sgd_solver.cpp:106] Iteration 1300, lr = 0.039948
I0405 03:46:34.349927 26022 solver.cpp:229] Iteration 1350, loss = 1.12643
I0405 03:46:34.350078 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 03:46:34.350100 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 03:46:34.350114 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 03:46:34.350127 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 03:46:34.350139 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.09375
I0405 03:46:34.350152 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 03:46:34.350164 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 03:46:34.350178 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 03:46:34.350189 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 03:46:34.350201 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 03:46:34.350214 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:46:34.350225 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:46:34.350240 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:46:34.350253 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:46:34.350265 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:46:34.350276 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:46:34.350288 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:46:34.350301 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:46:34.350312 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:46:34.350323 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:46:34.350335 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:46:34.350347 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:46:34.350363 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.75357 (* 0.0454545 = 0.170617 loss)
I0405 03:46:34.350380 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 4.01441 (* 0.0454545 = 0.182473 loss)
I0405 03:46:34.350395 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.85568 (* 0.0454545 = 0.175258 loss)
I0405 03:46:34.350409 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.99214 (* 0.0454545 = 0.181461 loss)
I0405 03:46:34.350425 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.26844 (* 0.0454545 = 0.19402 loss)
I0405 03:46:34.350438 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.24766 (* 0.0454545 = 0.147621 loss)
I0405 03:46:34.350452 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.73778 (* 0.0454545 = 0.0789898 loss)
I0405 03:46:34.350467 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.437903 (* 0.0454545 = 0.0199047 loss)
I0405 03:46:34.350482 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.487206 (* 0.0454545 = 0.0221457 loss)
I0405 03:46:34.350497 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00725108 (* 0.0454545 = 0.000329595 loss)
I0405 03:46:34.350512 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.41011e-05 (* 0.0454545 = 1.55005e-06 loss)
I0405 03:46:34.350528 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.27503e-05 (* 0.0454545 = 1.48865e-06 loss)
I0405 03:46:34.350543 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.41625e-05 (* 0.0454545 = 1.55284e-06 loss)
I0405 03:46:34.350558 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.35552e-05 (* 0.0454545 = 1.52523e-06 loss)
I0405 03:46:34.350572 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.42091e-05 (* 0.0454545 = 1.55496e-06 loss)
I0405 03:46:34.350587 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.42277e-05 (* 0.0454545 = 1.55581e-06 loss)
I0405 03:46:34.350602 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.17667e-05 (* 0.0454545 = 1.44394e-06 loss)
I0405 03:46:34.350632 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.35217e-05 (* 0.0454545 = 1.52371e-06 loss)
I0405 03:46:34.350649 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.28249e-05 (* 0.0454545 = 1.49204e-06 loss)
I0405 03:46:34.350664 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.27057e-05 (* 0.0454545 = 1.48662e-06 loss)
I0405 03:46:34.350679 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.18412e-05 (* 0.0454545 = 1.44733e-06 loss)
I0405 03:46:34.350693 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.31341e-05 (* 0.0454545 = 1.5061e-06 loss)
I0405 03:46:34.350706 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:46:34.350719 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.00655e-07
I0405 03:46:34.350735 26022 sgd_solver.cpp:106] Iteration 1350, lr = 0.039946
I0405 03:55:36.418175 26022 solver.cpp:229] Iteration 1400, loss = 1.1228
I0405 03:55:36.418367 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 03:55:36.418387 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 03:55:36.418401 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 03:55:36.418414 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 03:55:36.418426 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.09375
I0405 03:55:36.418439 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 03:55:36.418452 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 03:55:36.418463 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 03:55:36.418478 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 03:55:36.418491 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 03:55:36.418506 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 03:55:36.418519 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 03:55:36.418531 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 03:55:36.418543 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 03:55:36.418555 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 03:55:36.418566 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 03:55:36.418578 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 03:55:36.418591 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 03:55:36.418601 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 03:55:36.418613 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 03:55:36.418625 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 03:55:36.418637 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 03:55:36.418653 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.49327 (* 0.0454545 = 0.158785 loss)
I0405 03:55:36.418668 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.73294 (* 0.0454545 = 0.169679 loss)
I0405 03:55:36.418683 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.80687 (* 0.0454545 = 0.173039 loss)
I0405 03:55:36.418696 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.92497 (* 0.0454545 = 0.178408 loss)
I0405 03:55:36.418710 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.04877 (* 0.0454545 = 0.184035 loss)
I0405 03:55:36.418725 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.231 (* 0.0454545 = 0.146864 loss)
I0405 03:55:36.418740 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.84302 (* 0.0454545 = 0.0837737 loss)
I0405 03:55:36.418753 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.696209 (* 0.0454545 = 0.0316459 loss)
I0405 03:55:36.418767 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.22786 (* 0.0454545 = 0.0103573 loss)
I0405 03:55:36.418781 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.271998 (* 0.0454545 = 0.0123635 loss)
I0405 03:55:36.418797 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.45932e-05 (* 0.0454545 = 2.93605e-06 loss)
I0405 03:55:36.418812 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.22548e-05 (* 0.0454545 = 2.82977e-06 loss)
I0405 03:55:36.418826 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.48916e-05 (* 0.0454545 = 2.94962e-06 loss)
I0405 03:55:36.418841 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 6.39859e-05 (* 0.0454545 = 2.90845e-06 loss)
I0405 03:55:36.418856 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 6.46716e-05 (* 0.0454545 = 2.93962e-06 loss)
I0405 03:55:36.418870 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 6.34992e-05 (* 0.0454545 = 2.88633e-06 loss)
I0405 03:55:36.418885 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.13678e-05 (* 0.0454545 = 2.78944e-06 loss)
I0405 03:55:36.418913 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.29033e-05 (* 0.0454545 = 2.85924e-06 loss)
I0405 03:55:36.418929 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 6.18112e-05 (* 0.0454545 = 2.8096e-06 loss)
I0405 03:55:36.418944 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 6.05575e-05 (* 0.0454545 = 2.75261e-06 loss)
I0405 03:55:36.418959 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 6.13807e-05 (* 0.0454545 = 2.79003e-06 loss)
I0405 03:55:36.418974 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.24078e-05 (* 0.0454545 = 2.83672e-06 loss)
I0405 03:55:36.418987 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 03:55:36.418999 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.89047e-07
I0405 03:55:36.419014 26022 sgd_solver.cpp:106] Iteration 1400, lr = 0.039944
I0405 04:04:38.533105 26022 solver.cpp:229] Iteration 1450, loss = 1.12391
I0405 04:04:38.533213 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 04:04:38.533233 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 04:04:38.533246 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 04:04:38.533259 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0
I0405 04:04:38.533272 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 04:04:38.533283 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 04:04:38.533295 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 04:04:38.533308 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 04:04:38.533319 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 04:04:38.533331 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 04:04:38.533344 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:04:38.533355 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:04:38.533367 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:04:38.533380 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:04:38.533390 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:04:38.533402 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:04:38.533414 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:04:38.533427 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:04:38.533437 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:04:38.533449 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:04:38.533460 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:04:38.533471 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:04:38.533486 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.6995 (* 0.0454545 = 0.168159 loss)
I0405 04:04:38.533501 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.94351 (* 0.0454545 = 0.17925 loss)
I0405 04:04:38.533515 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.83284 (* 0.0454545 = 0.17422 loss)
I0405 04:04:38.533530 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 4.04877 (* 0.0454545 = 0.184035 loss)
I0405 04:04:38.533545 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.76955 (* 0.0454545 = 0.171343 loss)
I0405 04:04:38.533560 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.25242 (* 0.0454545 = 0.147837 loss)
I0405 04:04:38.533573 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.75625 (* 0.0454545 = 0.0798294 loss)
I0405 04:04:38.533587 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.952103 (* 0.0454545 = 0.0432774 loss)
I0405 04:04:38.533602 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.404014 (* 0.0454545 = 0.0183643 loss)
I0405 04:04:38.533617 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0188919 (* 0.0454545 = 0.000858721 loss)
I0405 04:04:38.533632 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000189323 (* 0.0454545 = 8.60559e-06 loss)
I0405 04:04:38.533646 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000183443 (* 0.0454545 = 8.3383e-06 loss)
I0405 04:04:38.533661 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000189442 (* 0.0454545 = 8.61101e-06 loss)
I0405 04:04:38.533676 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000185785 (* 0.0454545 = 8.44476e-06 loss)
I0405 04:04:38.533691 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000188017 (* 0.0454545 = 8.54621e-06 loss)
I0405 04:04:38.533706 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000187082 (* 0.0454545 = 8.50374e-06 loss)
I0405 04:04:38.533720 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000182423 (* 0.0454545 = 8.29194e-06 loss)
I0405 04:04:38.533752 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000181944 (* 0.0454545 = 8.27017e-06 loss)
I0405 04:04:38.533768 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000182844 (* 0.0454545 = 8.31107e-06 loss)
I0405 04:04:38.533782 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000179215 (* 0.0454545 = 8.14612e-06 loss)
I0405 04:04:38.533797 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000182276 (* 0.0454545 = 8.28525e-06 loss)
I0405 04:04:38.533812 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000184968 (* 0.0454545 = 8.40763e-06 loss)
I0405 04:04:38.533824 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:04:38.533838 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.20314e-07
I0405 04:04:38.533851 26022 sgd_solver.cpp:106] Iteration 1450, lr = 0.039942
I0405 04:13:40.631392 26022 solver.cpp:229] Iteration 1500, loss = 1.12137
I0405 04:13:40.631564 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0
I0405 04:13:40.631587 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 04:13:40.631600 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 04:13:40.631613 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 04:13:40.631625 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.21875
I0405 04:13:40.631638 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 04:13:40.631649 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.875
I0405 04:13:40.631661 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 04:13:40.631674 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 04:13:40.631685 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 04:13:40.631697 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:13:40.631710 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:13:40.631721 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:13:40.631732 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:13:40.631744 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:13:40.631757 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:13:40.631769 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:13:40.631780 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:13:40.631793 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:13:40.631804 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:13:40.631815 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:13:40.631826 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:13:40.631846 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.65038 (* 0.0454545 = 0.165927 loss)
I0405 04:13:40.631863 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.61049 (* 0.0454545 = 0.164113 loss)
I0405 04:13:40.631877 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.6769 (* 0.0454545 = 0.167132 loss)
I0405 04:13:40.631891 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.9023 (* 0.0454545 = 0.177377 loss)
I0405 04:13:40.631906 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.51507 (* 0.0454545 = 0.159776 loss)
I0405 04:13:40.631921 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.82968 (* 0.0454545 = 0.128622 loss)
I0405 04:13:40.631934 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.04746 (* 0.0454545 = 0.0476118 loss)
I0405 04:13:40.631949 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.258801 (* 0.0454545 = 0.0117637 loss)
I0405 04:13:40.631963 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0294044 (* 0.0454545 = 0.00133656 loss)
I0405 04:13:40.631978 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0130509 (* 0.0454545 = 0.000593224 loss)
I0405 04:13:40.631994 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000107699 (* 0.0454545 = 4.89541e-06 loss)
I0405 04:13:40.632009 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000105751 (* 0.0454545 = 4.80685e-06 loss)
I0405 04:13:40.632024 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000107114 (* 0.0454545 = 4.86883e-06 loss)
I0405 04:13:40.632038 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000106963 (* 0.0454545 = 4.86198e-06 loss)
I0405 04:13:40.632053 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000106561 (* 0.0454545 = 4.84367e-06 loss)
I0405 04:13:40.632083 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000105647 (* 0.0454545 = 4.80213e-06 loss)
I0405 04:13:40.632102 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000105495 (* 0.0454545 = 4.79523e-06 loss)
I0405 04:13:40.632135 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000106619 (* 0.0454545 = 4.8463e-06 loss)
I0405 04:13:40.632153 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000103603 (* 0.0454545 = 4.70923e-06 loss)
I0405 04:13:40.632166 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.00010419 (* 0.0454545 = 4.7359e-06 loss)
I0405 04:13:40.632181 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000102492 (* 0.0454545 = 4.65871e-06 loss)
I0405 04:13:40.632196 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000104346 (* 0.0454545 = 4.74299e-06 loss)
I0405 04:13:40.632208 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:13:40.632220 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.9828e-07
I0405 04:13:40.632236 26022 sgd_solver.cpp:106] Iteration 1500, lr = 0.03994
I0405 04:22:42.707343 26022 solver.cpp:229] Iteration 1550, loss = 1.11948
I0405 04:22:42.707502 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 04:22:42.707525 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 04:22:42.707537 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 04:22:42.707551 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 04:22:42.707563 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 04:22:42.707576 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 04:22:42.707588 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 04:22:42.707600 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 04:22:42.707613 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 04:22:42.707625 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 04:22:42.707638 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:22:42.707649 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:22:42.707661 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:22:42.707674 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:22:42.707684 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:22:42.707695 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:22:42.707708 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:22:42.707720 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:22:42.707731 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:22:42.707743 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:22:42.707756 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:22:42.707767 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:22:42.707782 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.7368 (* 0.0454545 = 0.169854 loss)
I0405 04:22:42.707798 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.65482 (* 0.0454545 = 0.166128 loss)
I0405 04:22:42.707811 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.61159 (* 0.0454545 = 0.164163 loss)
I0405 04:22:42.707826 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.89921 (* 0.0454545 = 0.177237 loss)
I0405 04:22:42.707840 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.36574 (* 0.0454545 = 0.152988 loss)
I0405 04:22:42.707855 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.28189 (* 0.0454545 = 0.149177 loss)
I0405 04:22:42.707870 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.23491 (* 0.0454545 = 0.0561321 loss)
I0405 04:22:42.707885 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.625242 (* 0.0454545 = 0.0284201 loss)
I0405 04:22:42.707898 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.220204 (* 0.0454545 = 0.0100093 loss)
I0405 04:22:42.707912 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.268099 (* 0.0454545 = 0.0121863 loss)
I0405 04:22:42.707927 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000141006 (* 0.0454545 = 6.40937e-06 loss)
I0405 04:22:42.707943 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000136625 (* 0.0454545 = 6.21023e-06 loss)
I0405 04:22:42.707958 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000136766 (* 0.0454545 = 6.21663e-06 loss)
I0405 04:22:42.707973 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000137565 (* 0.0454545 = 6.25293e-06 loss)
I0405 04:22:42.707988 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.00013913 (* 0.0454545 = 6.32407e-06 loss)
I0405 04:22:42.708003 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000138566 (* 0.0454545 = 6.29844e-06 loss)
I0405 04:22:42.708019 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000133657 (* 0.0454545 = 6.0753e-06 loss)
I0405 04:22:42.708051 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000136642 (* 0.0454545 = 6.21099e-06 loss)
I0405 04:22:42.708086 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000136234 (* 0.0454545 = 6.19247e-06 loss)
I0405 04:22:42.708106 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000129774 (* 0.0454545 = 5.8988e-06 loss)
I0405 04:22:42.708120 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000133112 (* 0.0454545 = 6.05056e-06 loss)
I0405 04:22:42.708135 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000134918 (* 0.0454545 = 6.13263e-06 loss)
I0405 04:22:42.708148 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:22:42.708160 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.10238e-07
I0405 04:22:42.708175 26022 sgd_solver.cpp:106] Iteration 1550, lr = 0.039938
I0405 04:31:44.849534 26022 solver.cpp:229] Iteration 1600, loss = 1.12035
I0405 04:31:44.849730 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 04:31:44.849750 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 04:31:44.849764 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 04:31:44.849777 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 04:31:44.849789 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 04:31:44.849802 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.25
I0405 04:31:44.849814 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.59375
I0405 04:31:44.849825 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.75
I0405 04:31:44.849838 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 04:31:44.849850 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 04:31:44.849863 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:31:44.849874 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:31:44.849886 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:31:44.849898 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:31:44.849910 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:31:44.849921 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:31:44.849933 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:31:44.849944 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:31:44.849956 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:31:44.849967 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:31:44.849979 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:31:44.849990 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:31:44.850006 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.92378 (* 0.0454545 = 0.178354 loss)
I0405 04:31:44.850021 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.86668 (* 0.0454545 = 0.175758 loss)
I0405 04:31:44.850035 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.74211 (* 0.0454545 = 0.170096 loss)
I0405 04:31:44.850049 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.79458 (* 0.0454545 = 0.172481 loss)
I0405 04:31:44.850064 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.77885 (* 0.0454545 = 0.171766 loss)
I0405 04:31:44.850078 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.66865 (* 0.0454545 = 0.166757 loss)
I0405 04:31:44.850093 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.1614 (* 0.0454545 = 0.0982453 loss)
I0405 04:31:44.850107 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.32715 (* 0.0454545 = 0.0603249 loss)
I0405 04:31:44.850122 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.605397 (* 0.0454545 = 0.0275181 loss)
I0405 04:31:44.850137 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.205891 (* 0.0454545 = 0.00935868 loss)
I0405 04:31:44.850152 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000295737 (* 0.0454545 = 1.34426e-05 loss)
I0405 04:31:44.850167 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000301722 (* 0.0454545 = 1.37146e-05 loss)
I0405 04:31:44.850181 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000307488 (* 0.0454545 = 1.39767e-05 loss)
I0405 04:31:44.850196 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000303907 (* 0.0454545 = 1.3814e-05 loss)
I0405 04:31:44.850211 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000309246 (* 0.0454545 = 1.40566e-05 loss)
I0405 04:31:44.850226 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000300088 (* 0.0454545 = 1.36404e-05 loss)
I0405 04:31:44.850244 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00028915 (* 0.0454545 = 1.31432e-05 loss)
I0405 04:31:44.850273 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000298417 (* 0.0454545 = 1.35644e-05 loss)
I0405 04:31:44.850289 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000302208 (* 0.0454545 = 1.37367e-05 loss)
I0405 04:31:44.850306 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000286137 (* 0.0454545 = 1.30062e-05 loss)
I0405 04:31:44.850322 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000295425 (* 0.0454545 = 1.34284e-05 loss)
I0405 04:31:44.850337 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.00029914 (* 0.0454545 = 1.35973e-05 loss)
I0405 04:31:44.850350 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:31:44.850363 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.50095e-07
I0405 04:31:44.850378 26022 sgd_solver.cpp:106] Iteration 1600, lr = 0.039936
I0405 04:40:46.862489 26022 solver.cpp:229] Iteration 1650, loss = 1.10699
I0405 04:40:46.862624 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 04:40:46.862646 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 04:40:46.862660 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 04:40:46.862673 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 04:40:46.862686 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 04:40:46.862699 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 04:40:46.862711 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 04:40:46.862723 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 04:40:46.862735 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 04:40:46.862747 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 04:40:46.862759 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:40:46.862771 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:40:46.862785 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:40:46.862797 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:40:46.862809 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:40:46.862821 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:40:46.862833 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:40:46.862844 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:40:46.862856 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:40:46.862870 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:40:46.862884 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:40:46.862895 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:40:46.862910 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.77215 (* 0.0454545 = 0.171461 loss)
I0405 04:40:46.862925 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.75691 (* 0.0454545 = 0.170769 loss)
I0405 04:40:46.862939 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.63627 (* 0.0454545 = 0.165285 loss)
I0405 04:40:46.862953 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.8985 (* 0.0454545 = 0.177205 loss)
I0405 04:40:46.862968 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.74775 (* 0.0454545 = 0.170352 loss)
I0405 04:40:46.862982 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.93792 (* 0.0454545 = 0.133542 loss)
I0405 04:40:46.862996 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.45836 (* 0.0454545 = 0.0662892 loss)
I0405 04:40:46.863010 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.456041 (* 0.0454545 = 0.0207292 loss)
I0405 04:40:46.863025 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.279212 (* 0.0454545 = 0.0126914 loss)
I0405 04:40:46.863040 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0154636 (* 0.0454545 = 0.000702889 loss)
I0405 04:40:46.863055 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000301855 (* 0.0454545 = 1.37207e-05 loss)
I0405 04:40:46.863070 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000352835 (* 0.0454545 = 1.60379e-05 loss)
I0405 04:40:46.863085 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000348977 (* 0.0454545 = 1.58626e-05 loss)
I0405 04:40:46.863100 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000336347 (* 0.0454545 = 1.52885e-05 loss)
I0405 04:40:46.863113 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000351044 (* 0.0454545 = 1.59565e-05 loss)
I0405 04:40:46.863128 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000321046 (* 0.0454545 = 1.4593e-05 loss)
I0405 04:40:46.863143 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000323894 (* 0.0454545 = 1.47224e-05 loss)
I0405 04:40:46.863175 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000339334 (* 0.0454545 = 1.54243e-05 loss)
I0405 04:40:46.863191 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000329075 (* 0.0454545 = 1.4958e-05 loss)
I0405 04:40:46.863206 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000325912 (* 0.0454545 = 1.48142e-05 loss)
I0405 04:40:46.863220 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000329801 (* 0.0454545 = 1.4991e-05 loss)
I0405 04:40:46.863235 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000346268 (* 0.0454545 = 1.57395e-05 loss)
I0405 04:40:46.863247 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:40:46.863260 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.28413e-06
I0405 04:40:46.863276 26022 sgd_solver.cpp:106] Iteration 1650, lr = 0.039934
I0405 04:49:48.875283 26022 solver.cpp:229] Iteration 1700, loss = 1.09533
I0405 04:49:48.875432 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 04:49:48.875452 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 04:49:48.875465 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 04:49:48.875479 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 04:49:48.875491 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.0625
I0405 04:49:48.875504 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 04:49:48.875516 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 04:49:48.875529 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 04:49:48.875540 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 04:49:48.875552 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 04:49:48.875565 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:49:48.875576 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:49:48.875587 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:49:48.875599 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:49:48.875612 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:49:48.875622 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:49:48.875634 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:49:48.875645 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:49:48.875658 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:49:48.875669 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:49:48.875679 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:49:48.875691 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:49:48.875706 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.61083 (* 0.0454545 = 0.164128 loss)
I0405 04:49:48.875722 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.96377 (* 0.0454545 = 0.180171 loss)
I0405 04:49:48.875736 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.90553 (* 0.0454545 = 0.177524 loss)
I0405 04:49:48.875751 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.87717 (* 0.0454545 = 0.176235 loss)
I0405 04:49:48.875764 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 4.08364 (* 0.0454545 = 0.18562 loss)
I0405 04:49:48.875779 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.14117 (* 0.0454545 = 0.14278 loss)
I0405 04:49:48.875793 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.41758 (* 0.0454545 = 0.0644356 loss)
I0405 04:49:48.875808 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.670318 (* 0.0454545 = 0.030469 loss)
I0405 04:49:48.875821 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.218472 (* 0.0454545 = 0.00993053 loss)
I0405 04:49:48.875836 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.032196 (* 0.0454545 = 0.00146345 loss)
I0405 04:49:48.875851 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000342488 (* 0.0454545 = 1.55676e-05 loss)
I0405 04:49:48.875867 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000432211 (* 0.0454545 = 1.9646e-05 loss)
I0405 04:49:48.875882 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000430321 (* 0.0454545 = 1.956e-05 loss)
I0405 04:49:48.875897 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000405939 (* 0.0454545 = 1.84518e-05 loss)
I0405 04:49:48.875912 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.00043219 (* 0.0454545 = 1.9645e-05 loss)
I0405 04:49:48.875926 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000372476 (* 0.0454545 = 1.69307e-05 loss)
I0405 04:49:48.875941 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000391327 (* 0.0454545 = 1.77876e-05 loss)
I0405 04:49:48.875972 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000412334 (* 0.0454545 = 1.87424e-05 loss)
I0405 04:49:48.875988 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000392499 (* 0.0454545 = 1.78409e-05 loss)
I0405 04:49:48.876003 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000391067 (* 0.0454545 = 1.77758e-05 loss)
I0405 04:49:48.876019 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000397539 (* 0.0454545 = 1.807e-05 loss)
I0405 04:49:48.876032 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.00042274 (* 0.0454545 = 1.92154e-05 loss)
I0405 04:49:48.876045 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:49:48.876057 26022 solver.cpp:245]     Train net output #45: total_confidence = 7.44678e-07
I0405 04:49:48.876093 26022 sgd_solver.cpp:106] Iteration 1700, lr = 0.039932
I0405 04:58:50.914773 26022 solver.cpp:229] Iteration 1750, loss = 1.08695
I0405 04:58:50.914924 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 04:58:50.914945 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 04:58:50.914958 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 04:58:50.914973 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.21875
I0405 04:58:50.914985 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 04:58:50.914997 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 04:58:50.915009 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 04:58:50.915021 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 04:58:50.915035 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 04:58:50.915046 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 04:58:50.915058 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 04:58:50.915071 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 04:58:50.915081 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 04:58:50.915093 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 04:58:50.915105 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 04:58:50.915117 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 04:58:50.915128 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 04:58:50.915140 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 04:58:50.915153 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 04:58:50.915164 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 04:58:50.915176 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 04:58:50.915187 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 04:58:50.915204 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.608 (* 0.0454545 = 0.164 loss)
I0405 04:58:50.915220 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.74516 (* 0.0454545 = 0.170235 loss)
I0405 04:58:50.915233 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.9471 (* 0.0454545 = 0.179414 loss)
I0405 04:58:50.915247 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.33867 (* 0.0454545 = 0.151758 loss)
I0405 04:58:50.915262 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.28752 (* 0.0454545 = 0.149433 loss)
I0405 04:58:50.915277 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.72643 (* 0.0454545 = 0.123929 loss)
I0405 04:58:50.915290 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.0919 (* 0.0454545 = 0.0496316 loss)
I0405 04:58:50.915304 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.63007 (* 0.0454545 = 0.0286396 loss)
I0405 04:58:50.915319 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.243341 (* 0.0454545 = 0.011061 loss)
I0405 04:58:50.915334 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0223476 (* 0.0454545 = 0.0010158 loss)
I0405 04:58:50.915349 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000125371 (* 0.0454545 = 5.69868e-06 loss)
I0405 04:58:50.915364 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000170279 (* 0.0454545 = 7.73995e-06 loss)
I0405 04:58:50.915379 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000167372 (* 0.0454545 = 7.60781e-06 loss)
I0405 04:58:50.915393 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000156062 (* 0.0454545 = 7.09372e-06 loss)
I0405 04:58:50.915408 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000170134 (* 0.0454545 = 7.73336e-06 loss)
I0405 04:58:50.915423 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000137632 (* 0.0454545 = 6.25602e-06 loss)
I0405 04:58:50.915437 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000151974 (* 0.0454545 = 6.9079e-06 loss)
I0405 04:58:50.915469 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000158972 (* 0.0454545 = 7.22599e-06 loss)
I0405 04:58:50.915485 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000150366 (* 0.0454545 = 6.83483e-06 loss)
I0405 04:58:50.915500 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000151645 (* 0.0454545 = 6.89295e-06 loss)
I0405 04:58:50.915514 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000153619 (* 0.0454545 = 6.98269e-06 loss)
I0405 04:58:50.915529 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000165581 (* 0.0454545 = 7.5264e-06 loss)
I0405 04:58:50.915541 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 04:58:50.915554 26022 solver.cpp:245]     Train net output #45: total_confidence = 9.40163e-07
I0405 04:58:50.915568 26022 sgd_solver.cpp:106] Iteration 1750, lr = 0.03993
I0405 05:07:52.987673 26022 solver.cpp:229] Iteration 1800, loss = 1.08421
I0405 05:07:52.987887 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 05:07:52.987920 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 05:07:52.987942 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 05:07:52.987967 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 05:07:52.987990 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 05:07:52.988013 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 05:07:52.988034 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 05:07:52.988056 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 05:07:52.988108 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 05:07:52.988136 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 05:07:52.988158 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:07:52.988183 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:07:52.988204 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:07:52.988225 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:07:52.988246 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:07:52.988266 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:07:52.988288 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:07:52.988312 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:07:52.988333 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:07:52.988354 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:07:52.988390 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:07:52.988415 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:07:52.988441 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.40658 (* 0.0454545 = 0.154845 loss)
I0405 05:07:52.988468 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.56145 (* 0.0454545 = 0.161884 loss)
I0405 05:07:52.988494 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.63802 (* 0.0454545 = 0.165364 loss)
I0405 05:07:52.988523 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.59114 (* 0.0454545 = 0.163234 loss)
I0405 05:07:52.988550 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.52469 (* 0.0454545 = 0.160213 loss)
I0405 05:07:52.988577 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.42737 (* 0.0454545 = 0.15579 loss)
I0405 05:07:52.988602 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.51929 (* 0.0454545 = 0.0690587 loss)
I0405 05:07:52.988628 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.974019 (* 0.0454545 = 0.0442736 loss)
I0405 05:07:52.988654 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.447813 (* 0.0454545 = 0.0203551 loss)
I0405 05:07:52.988682 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.473413 (* 0.0454545 = 0.0215188 loss)
I0405 05:07:52.988708 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000102096 (* 0.0454545 = 4.64075e-06 loss)
I0405 05:07:52.988734 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000138019 (* 0.0454545 = 6.27359e-06 loss)
I0405 05:07:52.988762 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000135999 (* 0.0454545 = 6.18176e-06 loss)
I0405 05:07:52.988788 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.00012364 (* 0.0454545 = 5.61999e-06 loss)
I0405 05:07:52.988814 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000134096 (* 0.0454545 = 6.09529e-06 loss)
I0405 05:07:52.988840 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000112692 (* 0.0454545 = 5.12237e-06 loss)
I0405 05:07:52.988867 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000122226 (* 0.0454545 = 5.55573e-06 loss)
I0405 05:07:52.988915 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000129187 (* 0.0454545 = 5.87215e-06 loss)
I0405 05:07:52.988945 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000120222 (* 0.0454545 = 5.46462e-06 loss)
I0405 05:07:52.988971 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000123209 (* 0.0454545 = 5.6004e-06 loss)
I0405 05:07:52.988997 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000121235 (* 0.0454545 = 5.51067e-06 loss)
I0405 05:07:52.989024 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.00013379 (* 0.0454545 = 6.08138e-06 loss)
I0405 05:07:52.989048 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:07:52.989070 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.08386e-06
I0405 05:07:52.989094 26022 sgd_solver.cpp:106] Iteration 1800, lr = 0.039928
I0405 05:16:55.073053 26022 solver.cpp:229] Iteration 1850, loss = 1.07775
I0405 05:16:55.073222 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 05:16:55.073242 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 05:16:55.073256 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 05:16:55.073269 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 05:16:55.073282 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 05:16:55.073294 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 05:16:55.073307 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.84375
I0405 05:16:55.073318 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 05:16:55.073330 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 05:16:55.073343 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 05:16:55.073354 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:16:55.073366 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:16:55.073379 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:16:55.073390 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:16:55.073401 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:16:55.073412 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:16:55.073424 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:16:55.073436 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:16:55.073448 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:16:55.073460 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:16:55.073472 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:16:55.073483 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:16:55.073499 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.22815 (* 0.0454545 = 0.146734 loss)
I0405 05:16:55.073514 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.63088 (* 0.0454545 = 0.16504 loss)
I0405 05:16:55.073529 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.66731 (* 0.0454545 = 0.166696 loss)
I0405 05:16:55.073544 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.60324 (* 0.0454545 = 0.163784 loss)
I0405 05:16:55.073557 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.22995 (* 0.0454545 = 0.146816 loss)
I0405 05:16:55.073571 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.12951 (* 0.0454545 = 0.14225 loss)
I0405 05:16:55.073585 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.02389 (* 0.0454545 = 0.0465406 loss)
I0405 05:16:55.073599 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.294799 (* 0.0454545 = 0.0133999 loss)
I0405 05:16:55.073616 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0350087 (* 0.0454545 = 0.00159131 loss)
I0405 05:16:55.073631 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0125076 (* 0.0454545 = 0.000568527 loss)
I0405 05:16:55.073645 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.67043e-05 (* 0.0454545 = 2.12292e-06 loss)
I0405 05:16:55.073660 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.51602e-05 (* 0.0454545 = 2.96183e-06 loss)
I0405 05:16:55.073675 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.18653e-05 (* 0.0454545 = 2.81206e-06 loss)
I0405 05:16:55.073693 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.67219e-05 (* 0.0454545 = 2.57827e-06 loss)
I0405 05:16:55.073709 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 6.29047e-05 (* 0.0454545 = 2.8593e-06 loss)
I0405 05:16:55.073724 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.04438e-05 (* 0.0454545 = 2.2929e-06 loss)
I0405 05:16:55.073740 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.78126e-05 (* 0.0454545 = 2.62785e-06 loss)
I0405 05:16:55.073771 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.77132e-05 (* 0.0454545 = 2.62333e-06 loss)
I0405 05:16:55.073787 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.47151e-05 (* 0.0454545 = 2.48705e-06 loss)
I0405 05:16:55.073802 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.53669e-05 (* 0.0454545 = 2.51668e-06 loss)
I0405 05:16:55.073817 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.67034e-05 (* 0.0454545 = 2.57743e-06 loss)
I0405 05:16:55.073832 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.14962e-05 (* 0.0454545 = 2.79528e-06 loss)
I0405 05:16:55.073844 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:16:55.073856 26022 solver.cpp:245]     Train net output #45: total_confidence = 7.55418e-07
I0405 05:16:55.073871 26022 sgd_solver.cpp:106] Iteration 1850, lr = 0.039926
I0405 05:25:57.123227 26022 solver.cpp:229] Iteration 1900, loss = 1.07608
I0405 05:25:57.123370 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 05:25:57.123390 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 05:25:57.123404 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 05:25:57.123417 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 05:25:57.123430 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.21875
I0405 05:25:57.123441 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 05:25:57.123455 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 05:25:57.123466 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 05:25:57.123478 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 05:25:57.123491 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 05:25:57.123502 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:25:57.123514 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:25:57.123525 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:25:57.123538 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:25:57.123550 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:25:57.123563 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:25:57.123574 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:25:57.123585 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:25:57.123596 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:25:57.123608 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:25:57.123620 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:25:57.123631 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:25:57.123647 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.16138 (* 0.0454545 = 0.143699 loss)
I0405 05:25:57.123662 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.54021 (* 0.0454545 = 0.160919 loss)
I0405 05:25:57.123677 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.44852 (* 0.0454545 = 0.156751 loss)
I0405 05:25:57.123692 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.46473 (* 0.0454545 = 0.157488 loss)
I0405 05:25:57.123705 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.45744 (* 0.0454545 = 0.157156 loss)
I0405 05:25:57.123719 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.43915 (* 0.0454545 = 0.156325 loss)
I0405 05:25:57.123733 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.75418 (* 0.0454545 = 0.0797356 loss)
I0405 05:25:57.123747 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.798965 (* 0.0454545 = 0.0363166 loss)
I0405 05:25:57.123762 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.370926 (* 0.0454545 = 0.0168603 loss)
I0405 05:25:57.123776 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.463308 (* 0.0454545 = 0.0210594 loss)
I0405 05:25:57.123792 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000139444 (* 0.0454545 = 6.33835e-06 loss)
I0405 05:25:57.123806 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000192836 (* 0.0454545 = 8.76528e-06 loss)
I0405 05:25:57.123821 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000189002 (* 0.0454545 = 8.591e-06 loss)
I0405 05:25:57.123837 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000165056 (* 0.0454545 = 7.50257e-06 loss)
I0405 05:25:57.123852 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.00018078 (* 0.0454545 = 8.21727e-06 loss)
I0405 05:25:57.123867 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000150456 (* 0.0454545 = 6.83891e-06 loss)
I0405 05:25:57.123881 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000173001 (* 0.0454545 = 7.86368e-06 loss)
I0405 05:25:57.123913 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.00017825 (* 0.0454545 = 8.10226e-06 loss)
I0405 05:25:57.123929 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000161776 (* 0.0454545 = 7.35346e-06 loss)
I0405 05:25:57.123944 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000170194 (* 0.0454545 = 7.73608e-06 loss)
I0405 05:25:57.123958 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000165563 (* 0.0454545 = 7.5256e-06 loss)
I0405 05:25:57.123972 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000186544 (* 0.0454545 = 8.47928e-06 loss)
I0405 05:25:57.123986 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:25:57.123998 26022 solver.cpp:245]     Train net output #45: total_confidence = 1.79919e-06
I0405 05:25:57.124012 26022 sgd_solver.cpp:106] Iteration 1900, lr = 0.039924
I0405 05:34:59.136031 26022 solver.cpp:229] Iteration 1950, loss = 1.06555
I0405 05:34:59.136237 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 05:34:59.136260 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 05:34:59.136273 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 05:34:59.136286 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 05:34:59.136299 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 05:34:59.136312 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 05:34:59.136323 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.875
I0405 05:34:59.136337 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 05:34:59.136349 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 05:34:59.136361 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 05:34:59.136373 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:34:59.136384 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:34:59.136395 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:34:59.136407 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:34:59.136420 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:34:59.136430 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:34:59.136442 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:34:59.136454 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:34:59.136466 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:34:59.136477 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:34:59.136489 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:34:59.136500 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:34:59.136517 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.38565 (* 0.0454545 = 0.153893 loss)
I0405 05:34:59.136531 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.64951 (* 0.0454545 = 0.165887 loss)
I0405 05:34:59.136545 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.7439 (* 0.0454545 = 0.170177 loss)
I0405 05:34:59.136560 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.64981 (* 0.0454545 = 0.1659 loss)
I0405 05:34:59.136574 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.23874 (* 0.0454545 = 0.147215 loss)
I0405 05:34:59.136590 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.88875 (* 0.0454545 = 0.131307 loss)
I0405 05:34:59.136603 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.787059 (* 0.0454545 = 0.0357754 loss)
I0405 05:34:59.136617 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.300835 (* 0.0454545 = 0.0136743 loss)
I0405 05:34:59.136632 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0409181 (* 0.0454545 = 0.00185991 loss)
I0405 05:34:59.136647 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0120105 (* 0.0454545 = 0.00054593 loss)
I0405 05:34:59.136662 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 7.38063e-05 (* 0.0454545 = 3.35483e-06 loss)
I0405 05:34:59.136677 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000107621 (* 0.0454545 = 4.89188e-06 loss)
I0405 05:34:59.136693 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000105466 (* 0.0454545 = 4.79389e-06 loss)
I0405 05:34:59.136708 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 8.71873e-05 (* 0.0454545 = 3.96306e-06 loss)
I0405 05:34:59.136721 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 9.15702e-05 (* 0.0454545 = 4.16228e-06 loss)
I0405 05:34:59.136736 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 7.98792e-05 (* 0.0454545 = 3.63087e-06 loss)
I0405 05:34:59.136751 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 9.56017e-05 (* 0.0454545 = 4.34553e-06 loss)
I0405 05:34:59.136783 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000100443 (* 0.0454545 = 4.56561e-06 loss)
I0405 05:34:59.136800 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 9.02716e-05 (* 0.0454545 = 4.10326e-06 loss)
I0405 05:34:59.136814 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 8.8643e-05 (* 0.0454545 = 4.02923e-06 loss)
I0405 05:34:59.136831 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 8.78489e-05 (* 0.0454545 = 3.99313e-06 loss)
I0405 05:34:59.136847 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000101777 (* 0.0454545 = 4.62624e-06 loss)
I0405 05:34:59.136859 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:34:59.136872 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.71124e-06
I0405 05:34:59.136886 26022 sgd_solver.cpp:106] Iteration 1950, lr = 0.039922
I0405 05:38:58.151005 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.4473 > 30) by scale factor 0.953978
I0405 05:39:30.672099 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 38.8336 > 30) by scale factor 0.772526
I0405 05:43:50.842366 26022 solver.cpp:338] Iteration 2000, Testing net (#0)
I0405 05:44:04.500149 26022 solver.cpp:393] Test loss: 0.976496
I0405 05:44:04.500206 26022 solver.cpp:406]     Test net output #0: loss/accuracy01 = 0.259
I0405 05:44:04.500223 26022 solver.cpp:406]     Test net output #1: loss/accuracy02 = 0.067
I0405 05:44:04.500236 26022 solver.cpp:406]     Test net output #2: loss/accuracy03 = 0.083
I0405 05:44:04.500248 26022 solver.cpp:406]     Test net output #3: loss/accuracy04 = 0.088
I0405 05:44:04.500260 26022 solver.cpp:406]     Test net output #4: loss/accuracy05 = 0.213
I0405 05:44:04.500272 26022 solver.cpp:406]     Test net output #5: loss/accuracy06 = 0.502
I0405 05:44:04.500284 26022 solver.cpp:406]     Test net output #6: loss/accuracy07 = 0.894
I0405 05:44:04.500296 26022 solver.cpp:406]     Test net output #7: loss/accuracy08 = 0.97
I0405 05:44:04.500308 26022 solver.cpp:406]     Test net output #8: loss/accuracy09 = 0.995
I0405 05:44:04.500319 26022 solver.cpp:406]     Test net output #9: loss/accuracy10 = 0.998
I0405 05:44:04.500330 26022 solver.cpp:406]     Test net output #10: loss/accuracy11 = 1
I0405 05:44:04.500342 26022 solver.cpp:406]     Test net output #11: loss/accuracy12 = 1
I0405 05:44:04.500355 26022 solver.cpp:406]     Test net output #12: loss/accuracy13 = 1
I0405 05:44:04.500366 26022 solver.cpp:406]     Test net output #13: loss/accuracy14 = 1
I0405 05:44:04.500377 26022 solver.cpp:406]     Test net output #14: loss/accuracy15 = 1
I0405 05:44:04.500390 26022 solver.cpp:406]     Test net output #15: loss/accuracy16 = 1
I0405 05:44:04.500401 26022 solver.cpp:406]     Test net output #16: loss/accuracy17 = 1
I0405 05:44:04.500411 26022 solver.cpp:406]     Test net output #17: loss/accuracy18 = 1
I0405 05:44:04.500422 26022 solver.cpp:406]     Test net output #18: loss/accuracy19 = 1
I0405 05:44:04.500434 26022 solver.cpp:406]     Test net output #19: loss/accuracy20 = 1
I0405 05:44:04.500445 26022 solver.cpp:406]     Test net output #20: loss/accuracy21 = 1
I0405 05:44:04.500457 26022 solver.cpp:406]     Test net output #21: loss/accuracy22 = 1
I0405 05:44:04.500473 26022 solver.cpp:406]     Test net output #22: loss/loss01 = 3.30976 (* 0.0454545 = 0.150444 loss)
I0405 05:44:04.500488 26022 solver.cpp:406]     Test net output #23: loss/loss02 = 3.61106 (* 0.0454545 = 0.164139 loss)
I0405 05:44:04.500501 26022 solver.cpp:406]     Test net output #24: loss/loss03 = 3.52363 (* 0.0454545 = 0.160165 loss)
I0405 05:44:04.500515 26022 solver.cpp:406]     Test net output #25: loss/loss04 = 3.52764 (* 0.0454545 = 0.160347 loss)
I0405 05:44:04.500530 26022 solver.cpp:406]     Test net output #26: loss/loss05 = 3.61129 (* 0.0454545 = 0.16415 loss)
I0405 05:44:04.500543 26022 solver.cpp:406]     Test net output #27: loss/loss06 = 2.63173 (* 0.0454545 = 0.119624 loss)
I0405 05:44:04.500557 26022 solver.cpp:406]     Test net output #28: loss/loss07 = 0.895689 (* 0.0454545 = 0.0407132 loss)
I0405 05:44:04.500571 26022 solver.cpp:406]     Test net output #29: loss/loss08 = 0.278613 (* 0.0454545 = 0.0126642 loss)
I0405 05:44:04.500586 26022 solver.cpp:406]     Test net output #30: loss/loss09 = 0.0661727 (* 0.0454545 = 0.00300785 loss)
I0405 05:44:04.500599 26022 solver.cpp:406]     Test net output #31: loss/loss10 = 0.0269832 (* 0.0454545 = 0.00122651 loss)
I0405 05:44:04.500614 26022 solver.cpp:406]     Test net output #32: loss/loss11 = 2.29458e-05 (* 0.0454545 = 1.04299e-06 loss)
I0405 05:44:04.500629 26022 solver.cpp:406]     Test net output #33: loss/loss12 = 3.2726e-05 (* 0.0454545 = 1.48755e-06 loss)
I0405 05:44:04.500644 26022 solver.cpp:406]     Test net output #34: loss/loss13 = 3.15673e-05 (* 0.0454545 = 1.43488e-06 loss)
I0405 05:44:04.500659 26022 solver.cpp:406]     Test net output #35: loss/loss14 = 2.6249e-05 (* 0.0454545 = 1.19314e-06 loss)
I0405 05:44:04.500674 26022 solver.cpp:406]     Test net output #36: loss/loss15 = 2.66096e-05 (* 0.0454545 = 1.20953e-06 loss)
I0405 05:44:04.500689 26022 solver.cpp:406]     Test net output #37: loss/loss16 = 2.44362e-05 (* 0.0454545 = 1.11074e-06 loss)
I0405 05:44:04.500704 26022 solver.cpp:406]     Test net output #38: loss/loss17 = 2.90888e-05 (* 0.0454545 = 1.32222e-06 loss)
I0405 05:44:04.500751 26022 solver.cpp:406]     Test net output #39: loss/loss18 = 3.05173e-05 (* 0.0454545 = 1.38715e-06 loss)
I0405 05:44:04.500767 26022 solver.cpp:406]     Test net output #40: loss/loss19 = 2.91623e-05 (* 0.0454545 = 1.32556e-06 loss)
I0405 05:44:04.500782 26022 solver.cpp:406]     Test net output #41: loss/loss20 = 2.64048e-05 (* 0.0454545 = 1.20022e-06 loss)
I0405 05:44:04.500797 26022 solver.cpp:406]     Test net output #42: loss/loss21 = 2.61203e-05 (* 0.0454545 = 1.18729e-06 loss)
I0405 05:44:04.500810 26022 solver.cpp:406]     Test net output #43: loss/loss22 = 3.10265e-05 (* 0.0454545 = 1.41029e-06 loss)
I0405 05:44:04.500823 26022 solver.cpp:406]     Test net output #44: total_accuracy = 0.001
I0405 05:44:04.500835 26022 solver.cpp:406]     Test net output #45: total_confidence = 4.78457e-05
I0405 05:44:14.812119 26022 solver.cpp:229] Iteration 2000, loss = 1.06042
I0405 05:44:14.812167 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 05:44:14.812186 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 05:44:14.812199 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.15625
I0405 05:44:14.812212 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 05:44:14.812224 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.3125
I0405 05:44:14.812237 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.46875
I0405 05:44:14.812249 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 05:44:14.812261 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 05:44:14.812273 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 05:44:14.812286 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 05:44:14.812299 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:44:14.812310 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:44:14.812322 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:44:14.812333 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:44:14.812345 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:44:14.812357 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:44:14.812368 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:44:14.812381 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:44:14.812392 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:44:14.812403 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:44:14.812415 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:44:14.812427 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:44:14.812443 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.16283 (* 0.0454545 = 0.143765 loss)
I0405 05:44:14.812458 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.39338 (* 0.0454545 = 0.154244 loss)
I0405 05:44:14.812471 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.39031 (* 0.0454545 = 0.154105 loss)
I0405 05:44:14.812486 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.42935 (* 0.0454545 = 0.15588 loss)
I0405 05:44:14.812500 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.11024 (* 0.0454545 = 0.141375 loss)
I0405 05:44:14.812515 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.48534 (* 0.0454545 = 0.11297 loss)
I0405 05:44:14.812530 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.4104 (* 0.0454545 = 0.0641089 loss)
I0405 05:44:14.812543 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.526895 (* 0.0454545 = 0.0239498 loss)
I0405 05:44:14.812558 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.205056 (* 0.0454545 = 0.00932073 loss)
I0405 05:44:14.812602 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00657359 (* 0.0454545 = 0.000298799 loss)
I0405 05:44:14.812619 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.14763e-05 (* 0.0454545 = 9.76197e-07 loss)
I0405 05:44:14.812634 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.02028e-05 (* 0.0454545 = 1.37285e-06 loss)
I0405 05:44:14.812649 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.97483e-05 (* 0.0454545 = 1.3522e-06 loss)
I0405 05:44:14.812664 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.41981e-05 (* 0.0454545 = 1.09991e-06 loss)
I0405 05:44:14.812680 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.51222e-05 (* 0.0454545 = 1.14192e-06 loss)
I0405 05:44:14.812695 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.30953e-05 (* 0.0454545 = 1.04979e-06 loss)
I0405 05:44:14.812708 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.7112e-05 (* 0.0454545 = 1.23236e-06 loss)
I0405 05:44:14.812723 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.92229e-05 (* 0.0454545 = 1.32831e-06 loss)
I0405 05:44:14.812738 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.77977e-05 (* 0.0454545 = 1.26353e-06 loss)
I0405 05:44:14.812753 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.44627e-05 (* 0.0454545 = 1.11194e-06 loss)
I0405 05:44:14.812772 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.48167e-05 (* 0.0454545 = 1.12803e-06 loss)
I0405 05:44:14.812786 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.90086e-05 (* 0.0454545 = 1.31857e-06 loss)
I0405 05:44:14.812799 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:44:14.812811 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000117295
I0405 05:44:14.812826 26022 sgd_solver.cpp:106] Iteration 2000, lr = 0.03992
I0405 05:53:16.875116 26022 solver.cpp:229] Iteration 2050, loss = 1.04963
I0405 05:53:16.875254 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 05:53:16.875273 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 05:53:16.875288 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 05:53:16.875300 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 05:53:16.875313 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 05:53:16.875325 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 05:53:16.875339 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 05:53:16.875350 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 05:53:16.875363 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 05:53:16.875375 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 05:53:16.875387 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 05:53:16.875399 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 05:53:16.875411 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 05:53:16.875422 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 05:53:16.875434 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 05:53:16.875447 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 05:53:16.875458 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 05:53:16.875469 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 05:53:16.875481 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 05:53:16.875494 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 05:53:16.875504 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 05:53:16.875516 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 05:53:16.875531 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.10413 (* 0.0454545 = 0.141097 loss)
I0405 05:53:16.875546 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.63586 (* 0.0454545 = 0.165266 loss)
I0405 05:53:16.875560 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.59294 (* 0.0454545 = 0.163315 loss)
I0405 05:53:16.875576 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.5076 (* 0.0454545 = 0.159437 loss)
I0405 05:53:16.875591 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.25507 (* 0.0454545 = 0.147958 loss)
I0405 05:53:16.875604 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.9397 (* 0.0454545 = 0.133623 loss)
I0405 05:53:16.875618 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.73763 (* 0.0454545 = 0.0789834 loss)
I0405 05:53:16.875633 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.823319 (* 0.0454545 = 0.0374236 loss)
I0405 05:53:16.875648 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.427028 (* 0.0454545 = 0.0194103 loss)
I0405 05:53:16.875663 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.206037 (* 0.0454545 = 0.00936532 loss)
I0405 05:53:16.875677 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.35394e-05 (* 0.0454545 = 1.97907e-06 loss)
I0405 05:53:16.875692 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.12159e-05 (* 0.0454545 = 2.78254e-06 loss)
I0405 05:53:16.875707 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.09273e-05 (* 0.0454545 = 2.76942e-06 loss)
I0405 05:53:16.875722 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.76398e-05 (* 0.0454545 = 2.16544e-06 loss)
I0405 05:53:16.875737 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.75466e-05 (* 0.0454545 = 2.16121e-06 loss)
I0405 05:53:16.875751 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.62352e-05 (* 0.0454545 = 2.1016e-06 loss)
I0405 05:53:16.875766 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.44511e-05 (* 0.0454545 = 2.47505e-06 loss)
I0405 05:53:16.875798 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.78978e-05 (* 0.0454545 = 2.63172e-06 loss)
I0405 05:53:16.875814 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.71322e-05 (* 0.0454545 = 2.59692e-06 loss)
I0405 05:53:16.875829 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.84036e-05 (* 0.0454545 = 2.20017e-06 loss)
I0405 05:53:16.875847 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.97283e-05 (* 0.0454545 = 2.26038e-06 loss)
I0405 05:53:16.875864 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.74058e-05 (* 0.0454545 = 2.60935e-06 loss)
I0405 05:53:16.875875 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 05:53:16.875887 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000105631
I0405 05:53:16.875902 26022 sgd_solver.cpp:106] Iteration 2050, lr = 0.039918
I0405 06:02:18.943953 26022 solver.cpp:229] Iteration 2100, loss = 1.05269
I0405 06:02:18.944134 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 06:02:18.944155 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 06:02:18.944169 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 06:02:18.944181 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 06:02:18.944195 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.09375
I0405 06:02:18.944207 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 06:02:18.944221 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.625
I0405 06:02:18.944232 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 06:02:18.944244 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 06:02:18.944255 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 06:02:18.944268 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:02:18.944280 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:02:18.944291 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:02:18.944303 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:02:18.944315 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:02:18.944326 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:02:18.944339 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:02:18.944350 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:02:18.944361 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:02:18.944373 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:02:18.944385 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:02:18.944396 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:02:18.944411 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.15477 (* 0.0454545 = 0.143399 loss)
I0405 06:02:18.944427 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.7016 (* 0.0454545 = 0.168255 loss)
I0405 06:02:18.944442 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.54766 (* 0.0454545 = 0.161257 loss)
I0405 06:02:18.944456 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.59132 (* 0.0454545 = 0.163242 loss)
I0405 06:02:18.944471 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.74816 (* 0.0454545 = 0.170371 loss)
I0405 06:02:18.944485 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.17241 (* 0.0454545 = 0.1442 loss)
I0405 06:02:18.944500 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.0926 (* 0.0454545 = 0.0951182 loss)
I0405 06:02:18.944515 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.275549 (* 0.0454545 = 0.0125249 loss)
I0405 06:02:18.944532 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0378832 (* 0.0454545 = 0.00172196 loss)
I0405 06:02:18.944548 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0164502 (* 0.0454545 = 0.000747738 loss)
I0405 06:02:18.944563 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.02812e-05 (* 0.0454545 = 2.28551e-06 loss)
I0405 06:02:18.944578 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.95598e-05 (* 0.0454545 = 3.16181e-06 loss)
I0405 06:02:18.944592 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.68455e-05 (* 0.0454545 = 3.03843e-06 loss)
I0405 06:02:18.944608 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.3476e-05 (* 0.0454545 = 2.43073e-06 loss)
I0405 06:02:18.944623 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.26041e-05 (* 0.0454545 = 2.3911e-06 loss)
I0405 06:02:18.944638 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.21012e-05 (* 0.0454545 = 2.36824e-06 loss)
I0405 06:02:18.944656 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.11948e-05 (* 0.0454545 = 2.78158e-06 loss)
I0405 06:02:18.944689 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.70299e-05 (* 0.0454545 = 3.04681e-06 loss)
I0405 06:02:18.944705 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 6.51837e-05 (* 0.0454545 = 2.96289e-06 loss)
I0405 06:02:18.944720 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.56894e-05 (* 0.0454545 = 2.53134e-06 loss)
I0405 06:02:18.944735 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.61701e-05 (* 0.0454545 = 2.55319e-06 loss)
I0405 06:02:18.944749 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.5968e-05 (* 0.0454545 = 2.99855e-06 loss)
I0405 06:02:18.944762 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:02:18.944774 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.37273e-05
I0405 06:02:18.944789 26022 sgd_solver.cpp:106] Iteration 2100, lr = 0.039916
I0405 06:11:21.063184 26022 solver.cpp:229] Iteration 2150, loss = 1.04902
I0405 06:11:21.063406 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 06:11:21.063428 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 06:11:21.063442 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 06:11:21.063455 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 06:11:21.063468 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 06:11:21.063480 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 06:11:21.063493 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 06:11:21.063504 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 06:11:21.063516 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 06:11:21.063529 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 06:11:21.063540 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:11:21.063555 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:11:21.063567 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:11:21.063580 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:11:21.063591 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:11:21.063603 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:11:21.063614 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:11:21.063627 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:11:21.063638 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:11:21.063649 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:11:21.063662 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:11:21.063673 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:11:21.063688 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.17069 (* 0.0454545 = 0.144122 loss)
I0405 06:11:21.063704 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.44571 (* 0.0454545 = 0.156623 loss)
I0405 06:11:21.063719 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.49802 (* 0.0454545 = 0.159001 loss)
I0405 06:11:21.063735 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.38808 (* 0.0454545 = 0.154004 loss)
I0405 06:11:21.063750 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.25272 (* 0.0454545 = 0.147851 loss)
I0405 06:11:21.063765 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.90225 (* 0.0454545 = 0.131921 loss)
I0405 06:11:21.063779 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.76221 (* 0.0454545 = 0.0801004 loss)
I0405 06:11:21.063793 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.738473 (* 0.0454545 = 0.033567 loss)
I0405 06:11:21.063807 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.221624 (* 0.0454545 = 0.0100738 loss)
I0405 06:11:21.063823 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0102922 (* 0.0454545 = 0.000467827 loss)
I0405 06:11:21.063838 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.4005e-05 (* 0.0454545 = 1.54568e-06 loss)
I0405 06:11:21.063859 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.75975e-05 (* 0.0454545 = 2.16352e-06 loss)
I0405 06:11:21.063889 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.51692e-05 (* 0.0454545 = 2.05314e-06 loss)
I0405 06:11:21.063920 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.55176e-05 (* 0.0454545 = 1.61444e-06 loss)
I0405 06:11:21.063963 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.48582e-05 (* 0.0454545 = 1.58446e-06 loss)
I0405 06:11:21.063984 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.61721e-05 (* 0.0454545 = 1.64419e-06 loss)
I0405 06:11:21.063999 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.01357e-05 (* 0.0454545 = 1.82435e-06 loss)
I0405 06:11:21.064029 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.32775e-05 (* 0.0454545 = 1.96716e-06 loss)
I0405 06:11:21.064045 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.38369e-05 (* 0.0454545 = 1.99259e-06 loss)
I0405 06:11:21.064060 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.69806e-05 (* 0.0454545 = 1.68094e-06 loss)
I0405 06:11:21.064095 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.76031e-05 (* 0.0454545 = 1.70923e-06 loss)
I0405 06:11:21.064111 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.41086e-05 (* 0.0454545 = 2.00493e-06 loss)
I0405 06:11:21.064123 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:11:21.064136 26022 solver.cpp:245]     Train net output #45: total_confidence = 8.99676e-05
I0405 06:11:21.064151 26022 sgd_solver.cpp:106] Iteration 2150, lr = 0.039914
I0405 06:20:23.211572 26022 solver.cpp:229] Iteration 2200, loss = 1.04157
I0405 06:20:23.211741 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 06:20:23.211762 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 06:20:23.211776 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 06:20:23.211788 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 06:20:23.211802 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.3125
I0405 06:20:23.211813 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 06:20:23.211827 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 06:20:23.211838 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 06:20:23.211850 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 06:20:23.211863 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 06:20:23.211875 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:20:23.211887 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:20:23.211899 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:20:23.211911 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:20:23.211922 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:20:23.211935 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:20:23.211946 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:20:23.211957 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:20:23.211969 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:20:23.211980 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:20:23.211992 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:20:23.212004 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:20:23.212020 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.1739 (* 0.0454545 = 0.144268 loss)
I0405 06:20:23.212035 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.56057 (* 0.0454545 = 0.161844 loss)
I0405 06:20:23.212049 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.55262 (* 0.0454545 = 0.161483 loss)
I0405 06:20:23.212064 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.60216 (* 0.0454545 = 0.163735 loss)
I0405 06:20:23.212100 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.03437 (* 0.0454545 = 0.137926 loss)
I0405 06:20:23.212117 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.28901 (* 0.0454545 = 0.104046 loss)
I0405 06:20:23.212131 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.51273 (* 0.0454545 = 0.0687606 loss)
I0405 06:20:23.212146 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.442683 (* 0.0454545 = 0.0201219 loss)
I0405 06:20:23.212160 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.21846 (* 0.0454545 = 0.00993002 loss)
I0405 06:20:23.212175 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.200569 (* 0.0454545 = 0.00911678 loss)
I0405 06:20:23.212190 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.81832e-05 (* 0.0454545 = 1.7356e-06 loss)
I0405 06:20:23.212205 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.22439e-05 (* 0.0454545 = 2.37472e-06 loss)
I0405 06:20:23.212220 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.20132e-05 (* 0.0454545 = 2.36424e-06 loss)
I0405 06:20:23.212235 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.02997e-05 (* 0.0454545 = 1.8318e-06 loss)
I0405 06:20:23.212255 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.8235e-05 (* 0.0454545 = 1.73796e-06 loss)
I0405 06:20:23.212270 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.99367e-05 (* 0.0454545 = 1.81531e-06 loss)
I0405 06:20:23.212285 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.5356e-05 (* 0.0454545 = 2.06164e-06 loss)
I0405 06:20:23.212316 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.91352e-05 (* 0.0454545 = 2.23342e-06 loss)
I0405 06:20:23.212332 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.01571e-05 (* 0.0454545 = 2.27987e-06 loss)
I0405 06:20:23.212347 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.05662e-05 (* 0.0454545 = 1.84392e-06 loss)
I0405 06:20:23.212363 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.09056e-05 (* 0.0454545 = 1.85935e-06 loss)
I0405 06:20:23.212378 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.89187e-05 (* 0.0454545 = 2.22358e-06 loss)
I0405 06:20:23.212390 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:20:23.212404 26022 solver.cpp:245]     Train net output #45: total_confidence = 8.03094e-05
I0405 06:20:23.212421 26022 sgd_solver.cpp:106] Iteration 2200, lr = 0.039912
I0405 06:23:17.099337 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 47.8066 > 30) by scale factor 0.627528
I0405 06:29:25.259403 26022 solver.cpp:229] Iteration 2250, loss = 1.0418
I0405 06:29:25.259558 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 06:29:25.259579 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 06:29:25.259593 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 06:29:25.259605 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 06:29:25.259618 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 06:29:25.259630 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 06:29:25.259644 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.84375
I0405 06:29:25.259655 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 06:29:25.259667 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 06:29:25.259680 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 06:29:25.259695 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:29:25.259706 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:29:25.259718 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:29:25.259730 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:29:25.259742 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:29:25.259753 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:29:25.259764 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:29:25.259776 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:29:25.259788 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:29:25.259799 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:29:25.259810 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:29:25.259822 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:29:25.259837 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.56038 (* 0.0454545 = 0.161835 loss)
I0405 06:29:25.259852 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.76496 (* 0.0454545 = 0.171135 loss)
I0405 06:29:25.259866 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.74323 (* 0.0454545 = 0.170147 loss)
I0405 06:29:25.259881 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.7351 (* 0.0454545 = 0.169777 loss)
I0405 06:29:25.259896 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.32177 (* 0.0454545 = 0.150989 loss)
I0405 06:29:25.259909 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.50765 (* 0.0454545 = 0.113984 loss)
I0405 06:29:25.259923 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.973847 (* 0.0454545 = 0.0442658 loss)
I0405 06:29:25.259938 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.424503 (* 0.0454545 = 0.0192956 loss)
I0405 06:29:25.259953 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0435743 (* 0.0454545 = 0.00198065 loss)
I0405 06:29:25.259966 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0188889 (* 0.0454545 = 0.000858585 loss)
I0405 06:29:25.259981 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.82887e-05 (* 0.0454545 = 3.10403e-06 loss)
I0405 06:29:25.259996 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 9.41924e-05 (* 0.0454545 = 4.28147e-06 loss)
I0405 06:29:25.260011 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 9.22585e-05 (* 0.0454545 = 4.19357e-06 loss)
I0405 06:29:25.260025 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 7.40565e-05 (* 0.0454545 = 3.3662e-06 loss)
I0405 06:29:25.260040 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.10813e-05 (* 0.0454545 = 3.23097e-06 loss)
I0405 06:29:25.260054 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 7.27711e-05 (* 0.0454545 = 3.30778e-06 loss)
I0405 06:29:25.260087 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 8.18552e-05 (* 0.0454545 = 3.72069e-06 loss)
I0405 06:29:25.260121 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 8.949e-05 (* 0.0454545 = 4.06773e-06 loss)
I0405 06:29:25.260138 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 8.96559e-05 (* 0.0454545 = 4.07527e-06 loss)
I0405 06:29:25.260155 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 7.57053e-05 (* 0.0454545 = 3.44115e-06 loss)
I0405 06:29:25.260170 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 7.59661e-05 (* 0.0454545 = 3.45301e-06 loss)
I0405 06:29:25.260185 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 9.04998e-05 (* 0.0454545 = 4.11363e-06 loss)
I0405 06:29:25.260197 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:29:25.260210 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.39823e-05
I0405 06:29:25.260224 26022 sgd_solver.cpp:106] Iteration 2250, lr = 0.03991
I0405 06:34:51.221547 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 37.1166 > 30) by scale factor 0.808264
I0405 06:37:01.356776 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 51.7218 > 30) by scale factor 0.580026
I0405 06:38:27.603896 26022 solver.cpp:229] Iteration 2300, loss = 1.04061
I0405 06:38:27.603999 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 06:38:27.604020 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 06:38:27.604034 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 06:38:27.604046 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.25
I0405 06:38:27.604058 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.3125
I0405 06:38:27.604095 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 06:38:27.604120 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 06:38:27.604135 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 06:38:27.604146 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 06:38:27.604161 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 06:38:27.604174 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:38:27.604187 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:38:27.604199 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:38:27.604210 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:38:27.604224 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:38:27.604236 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:38:27.604249 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:38:27.604259 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:38:27.604271 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:38:27.604284 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:38:27.604295 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:38:27.604307 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:38:27.604322 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.28621 (* 0.0454545 = 0.149373 loss)
I0405 06:38:27.604337 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.88135 (* 0.0454545 = 0.176425 loss)
I0405 06:38:27.604351 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.55859 (* 0.0454545 = 0.161754 loss)
I0405 06:38:27.604365 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.35195 (* 0.0454545 = 0.152361 loss)
I0405 06:38:27.604380 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.18797 (* 0.0454545 = 0.144908 loss)
I0405 06:38:27.604394 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.87953 (* 0.0454545 = 0.130888 loss)
I0405 06:38:27.604408 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.40215 (* 0.0454545 = 0.0637339 loss)
I0405 06:38:27.604423 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.564786 (* 0.0454545 = 0.0256721 loss)
I0405 06:38:27.604437 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.446556 (* 0.0454545 = 0.020298 loss)
I0405 06:38:27.604451 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.21338 (* 0.0454545 = 0.00969911 loss)
I0405 06:38:27.604466 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.33235e-05 (* 0.0454545 = 2.4238e-06 loss)
I0405 06:38:27.604481 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 7.05476e-05 (* 0.0454545 = 3.20671e-06 loss)
I0405 06:38:27.604496 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.91986e-05 (* 0.0454545 = 3.14539e-06 loss)
I0405 06:38:27.604511 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.59148e-05 (* 0.0454545 = 2.54158e-06 loss)
I0405 06:38:27.604524 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.18609e-05 (* 0.0454545 = 2.35731e-06 loss)
I0405 06:38:27.604538 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.61385e-05 (* 0.0454545 = 2.55175e-06 loss)
I0405 06:38:27.604553 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.09098e-05 (* 0.0454545 = 2.76863e-06 loss)
I0405 06:38:27.604588 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.73766e-05 (* 0.0454545 = 3.06257e-06 loss)
I0405 06:38:27.604604 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 6.93069e-05 (* 0.0454545 = 3.15031e-06 loss)
I0405 06:38:27.604619 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.76065e-05 (* 0.0454545 = 2.61848e-06 loss)
I0405 06:38:27.604632 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.85306e-05 (* 0.0454545 = 2.66048e-06 loss)
I0405 06:38:27.604647 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.71901e-05 (* 0.0454545 = 3.0541e-06 loss)
I0405 06:38:27.604660 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:38:27.604672 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000166972
I0405 06:38:27.604686 26022 sgd_solver.cpp:106] Iteration 2300, lr = 0.039908
I0405 06:41:43.230902 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 41.3741 > 30) by scale factor 0.725091
I0405 06:44:47.528802 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.4197 > 30) by scale factor 0.954816
I0405 06:47:29.735996 26022 solver.cpp:229] Iteration 2350, loss = 1.04194
I0405 06:47:29.736186 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 06:47:29.736217 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 06:47:29.736239 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 06:47:29.736263 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 06:47:29.736286 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 06:47:29.736309 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 06:47:29.736330 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 06:47:29.736351 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 06:47:29.736374 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 06:47:29.736397 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 06:47:29.736420 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:47:29.736441 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:47:29.736462 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:47:29.736484 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:47:29.736516 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:47:29.736551 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:47:29.736572 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:47:29.736594 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:47:29.736615 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:47:29.736637 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:47:29.736659 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:47:29.736678 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:47:29.736706 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.65373 (* 0.0454545 = 0.166078 loss)
I0405 06:47:29.736735 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.77806 (* 0.0454545 = 0.17173 loss)
I0405 06:47:29.736763 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.71551 (* 0.0454545 = 0.168887 loss)
I0405 06:47:29.736788 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.74906 (* 0.0454545 = 0.170412 loss)
I0405 06:47:29.736814 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.54179 (* 0.0454545 = 0.16099 loss)
I0405 06:47:29.736845 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.08887 (* 0.0454545 = 0.140403 loss)
I0405 06:47:29.736871 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.70116 (* 0.0454545 = 0.0773253 loss)
I0405 06:47:29.736896 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.977217 (* 0.0454545 = 0.044419 loss)
I0405 06:47:29.736922 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.60226 (* 0.0454545 = 0.0273755 loss)
I0405 06:47:29.736949 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.207637 (* 0.0454545 = 0.00943806 loss)
I0405 06:47:29.736976 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.54259e-05 (* 0.0454545 = 1.61027e-06 loss)
I0405 06:47:29.737006 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.58247e-05 (* 0.0454545 = 2.08294e-06 loss)
I0405 06:47:29.737032 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.52118e-05 (* 0.0454545 = 2.05508e-06 loss)
I0405 06:47:29.737059 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.74302e-05 (* 0.0454545 = 1.70137e-06 loss)
I0405 06:47:29.737084 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.41609e-05 (* 0.0454545 = 1.55277e-06 loss)
I0405 06:47:29.737112 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.74714e-05 (* 0.0454545 = 1.70324e-06 loss)
I0405 06:47:29.737136 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.95542e-05 (* 0.0454545 = 1.79792e-06 loss)
I0405 06:47:29.737185 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.26988e-05 (* 0.0454545 = 1.94085e-06 loss)
I0405 06:47:29.737212 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.43084e-05 (* 0.0454545 = 2.01402e-06 loss)
I0405 06:47:29.737239 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.75943e-05 (* 0.0454545 = 1.70883e-06 loss)
I0405 06:47:29.737267 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.77657e-05 (* 0.0454545 = 1.71662e-06 loss)
I0405 06:47:29.737293 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.35295e-05 (* 0.0454545 = 1.97861e-06 loss)
I0405 06:47:29.737315 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:47:29.737339 26022 solver.cpp:245]     Train net output #45: total_confidence = 8.79651e-05
I0405 06:47:29.737362 26022 sgd_solver.cpp:106] Iteration 2350, lr = 0.039906
I0405 06:51:17.853572 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.1101 > 30) by scale factor 0.879505
I0405 06:51:50.367398 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 46.2169 > 30) by scale factor 0.649113
I0405 06:56:31.788909 26022 solver.cpp:229] Iteration 2400, loss = 1.04916
I0405 06:56:31.789065 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.03125
I0405 06:56:31.789096 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 06:56:31.789121 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 06:56:31.789146 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 06:56:31.789168 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 06:56:31.789191 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 06:56:31.789216 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 06:56:31.789239 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.78125
I0405 06:56:31.789261 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.875
I0405 06:56:31.789283 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 06:56:31.789305 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 06:56:31.789330 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 06:56:31.789351 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 06:56:31.789372 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 06:56:31.789393 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 06:56:31.789417 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 06:56:31.789440 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 06:56:31.789463 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 06:56:31.789485 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 06:56:31.789507 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 06:56:31.789530 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 06:56:31.789551 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 06:56:31.789577 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.5361 (* 0.0454545 = 0.160732 loss)
I0405 06:56:31.789604 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.59461 (* 0.0454545 = 0.163391 loss)
I0405 06:56:31.789631 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.67854 (* 0.0454545 = 0.167206 loss)
I0405 06:56:31.789659 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.57854 (* 0.0454545 = 0.162661 loss)
I0405 06:56:31.789686 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.58126 (* 0.0454545 = 0.162784 loss)
I0405 06:56:31.789712 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.80024 (* 0.0454545 = 0.127284 loss)
I0405 06:56:31.789738 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.68087 (* 0.0454545 = 0.0764031 loss)
I0405 06:56:31.789764 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.18533 (* 0.0454545 = 0.0538786 loss)
I0405 06:56:31.789790 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.75491 (* 0.0454545 = 0.0343141 loss)
I0405 06:56:31.789816 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.439005 (* 0.0454545 = 0.0199548 loss)
I0405 06:56:31.789842 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.12527e-05 (* 0.0454545 = 1.87512e-06 loss)
I0405 06:56:31.789868 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.41292e-05 (* 0.0454545 = 2.46042e-06 loss)
I0405 06:56:31.789896 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.40492e-05 (* 0.0454545 = 2.45678e-06 loss)
I0405 06:56:31.789924 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.35197e-05 (* 0.0454545 = 1.97817e-06 loss)
I0405 06:56:31.789950 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.14593e-05 (* 0.0454545 = 1.88452e-06 loss)
I0405 06:56:31.789976 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.34397e-05 (* 0.0454545 = 1.97453e-06 loss)
I0405 06:56:31.790002 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.74469e-05 (* 0.0454545 = 2.15668e-06 loss)
I0405 06:56:31.790055 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.11634e-05 (* 0.0454545 = 2.32561e-06 loss)
I0405 06:56:31.790082 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.20075e-05 (* 0.0454545 = 2.36398e-06 loss)
I0405 06:56:31.790108 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.54702e-05 (* 0.0454545 = 2.06683e-06 loss)
I0405 06:56:31.790135 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.60627e-05 (* 0.0454545 = 2.09376e-06 loss)
I0405 06:56:31.790163 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.13719e-05 (* 0.0454545 = 2.33509e-06 loss)
I0405 06:56:31.790184 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 06:56:31.790205 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.50714e-05
I0405 06:56:31.790231 26022 sgd_solver.cpp:106] Iteration 2400, lr = 0.039904
I0405 07:00:52.514952 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.5976 > 30) by scale factor 0.949439
I0405 07:05:12.669852 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.6915 > 30) by scale factor 0.864764
I0405 07:05:33.865030 26022 solver.cpp:229] Iteration 2450, loss = 1.03176
I0405 07:05:33.865094 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 07:05:33.865113 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 07:05:33.865126 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 07:05:33.865139 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 07:05:33.865151 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 07:05:33.865164 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 07:05:33.865176 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 07:05:33.865188 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 07:05:33.865200 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 07:05:33.865212 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 07:05:33.865227 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:05:33.865241 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:05:33.865252 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:05:33.865263 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:05:33.865275 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:05:33.865288 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:05:33.865298 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:05:33.865310 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:05:33.865321 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:05:33.865334 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:05:33.865345 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:05:33.865356 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:05:33.865371 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.12395 (* 0.0454545 = 0.141998 loss)
I0405 07:05:33.865386 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.58241 (* 0.0454545 = 0.162837 loss)
I0405 07:05:33.865401 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.54781 (* 0.0454545 = 0.161264 loss)
I0405 07:05:33.865416 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.31614 (* 0.0454545 = 0.150734 loss)
I0405 07:05:33.865429 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.48788 (* 0.0454545 = 0.15854 loss)
I0405 07:05:33.865444 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.67842 (* 0.0454545 = 0.121746 loss)
I0405 07:05:33.865458 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.64442 (* 0.0454545 = 0.0747465 loss)
I0405 07:05:33.865473 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.563585 (* 0.0454545 = 0.0256175 loss)
I0405 07:05:33.865490 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.240202 (* 0.0454545 = 0.0109183 loss)
I0405 07:05:33.865505 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.193557 (* 0.0454545 = 0.00879802 loss)
I0405 07:05:33.865520 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 8.40053e-05 (* 0.0454545 = 3.81842e-06 loss)
I0405 07:05:33.865535 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000104934 (* 0.0454545 = 4.76971e-06 loss)
I0405 07:05:33.865550 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000105681 (* 0.0454545 = 4.80368e-06 loss)
I0405 07:05:33.865564 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 8.44336e-05 (* 0.0454545 = 3.83789e-06 loss)
I0405 07:05:33.865579 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.82222e-05 (* 0.0454545 = 3.55555e-06 loss)
I0405 07:05:33.865594 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.7109e-05 (* 0.0454545 = 3.9595e-06 loss)
I0405 07:05:33.865638 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 9.20333e-05 (* 0.0454545 = 4.18333e-06 loss)
I0405 07:05:33.865654 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000103825 (* 0.0454545 = 4.71933e-06 loss)
I0405 07:05:33.865669 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.00010964 (* 0.0454545 = 4.98365e-06 loss)
I0405 07:05:33.865684 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 8.72468e-05 (* 0.0454545 = 3.96576e-06 loss)
I0405 07:05:33.865699 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 8.93672e-05 (* 0.0454545 = 4.06215e-06 loss)
I0405 07:05:33.865712 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000103385 (* 0.0454545 = 4.69932e-06 loss)
I0405 07:05:33.865725 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:05:33.865737 26022 solver.cpp:245]     Train net output #45: total_confidence = 6.49686e-05
I0405 07:05:33.865751 26022 sgd_solver.cpp:106] Iteration 2450, lr = 0.039902
I0405 07:07:55.322444 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.2216 > 30) by scale factor 0.992668
I0405 07:08:27.856645 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 44.5959 > 30) by scale factor 0.672707
I0405 07:14:36.046927 26022 solver.cpp:229] Iteration 2500, loss = 1.03
I0405 07:14:36.047045 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 07:14:36.047065 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 07:14:36.047080 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 07:14:36.047092 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 07:14:36.047104 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 07:14:36.047117 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 07:14:36.047130 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 07:14:36.047142 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 07:14:36.047155 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 07:14:36.047168 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:14:36.047179 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:14:36.047191 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:14:36.047204 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:14:36.047215 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:14:36.047227 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:14:36.047240 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:14:36.047252 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:14:36.047265 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:14:36.047276 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:14:36.047288 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:14:36.047299 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:14:36.047312 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:14:36.047327 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.40011 (* 0.0454545 = 0.15455 loss)
I0405 07:14:36.047343 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.75881 (* 0.0454545 = 0.170855 loss)
I0405 07:14:36.047358 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.74831 (* 0.0454545 = 0.170378 loss)
I0405 07:14:36.047371 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.74678 (* 0.0454545 = 0.170308 loss)
I0405 07:14:36.047386 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.57872 (* 0.0454545 = 0.162669 loss)
I0405 07:14:36.047400 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.09007 (* 0.0454545 = 0.140458 loss)
I0405 07:14:36.047415 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.62409 (* 0.0454545 = 0.0738222 loss)
I0405 07:14:36.047430 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.718967 (* 0.0454545 = 0.0326803 loss)
I0405 07:14:36.047444 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.417162 (* 0.0454545 = 0.0189619 loss)
I0405 07:14:36.047459 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0207939 (* 0.0454545 = 0.000945177 loss)
I0405 07:14:36.047474 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.28359e-05 (* 0.0454545 = 1.49254e-06 loss)
I0405 07:14:36.047489 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.07025e-05 (* 0.0454545 = 1.85012e-06 loss)
I0405 07:14:36.047504 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.1824e-05 (* 0.0454545 = 1.90109e-06 loss)
I0405 07:14:36.047519 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.47024e-05 (* 0.0454545 = 1.57738e-06 loss)
I0405 07:14:36.047534 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.20908e-05 (* 0.0454545 = 1.45867e-06 loss)
I0405 07:14:36.047549 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.49912e-05 (* 0.0454545 = 1.59051e-06 loss)
I0405 07:14:36.047564 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.51812e-05 (* 0.0454545 = 1.59915e-06 loss)
I0405 07:14:36.047596 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.99239e-05 (* 0.0454545 = 1.81472e-06 loss)
I0405 07:14:36.047612 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.20774e-05 (* 0.0454545 = 1.91261e-06 loss)
I0405 07:14:36.047627 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.51029e-05 (* 0.0454545 = 1.59559e-06 loss)
I0405 07:14:36.047642 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.54028e-05 (* 0.0454545 = 1.60922e-06 loss)
I0405 07:14:36.047657 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.09708e-05 (* 0.0454545 = 1.86231e-06 loss)
I0405 07:14:36.047670 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:14:36.047683 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000111447
I0405 07:14:36.047698 26022 sgd_solver.cpp:106] Iteration 2500, lr = 0.0399
I0405 07:19:29.258702 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 60.9616 > 30) by scale factor 0.492113
I0405 07:20:12.605080 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.7894 > 30) by scale factor 0.862331
I0405 07:23:38.102818 26022 solver.cpp:229] Iteration 2550, loss = 1.02848
I0405 07:23:38.102924 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 07:23:38.102944 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 07:23:38.102958 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.1875
I0405 07:23:38.102972 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 07:23:38.102984 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 07:23:38.102996 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 07:23:38.103009 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 07:23:38.103021 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 07:23:38.103034 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 07:23:38.103049 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:23:38.103062 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:23:38.103075 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:23:38.103086 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:23:38.103098 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:23:38.103111 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:23:38.103121 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:23:38.103133 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:23:38.103145 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:23:38.103157 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:23:38.103168 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:23:38.103180 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:23:38.103193 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:23:38.103209 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.22879 (* 0.0454545 = 0.146763 loss)
I0405 07:23:38.103222 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.77785 (* 0.0454545 = 0.17172 loss)
I0405 07:23:38.103237 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.50215 (* 0.0454545 = 0.159189 loss)
I0405 07:23:38.103251 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.46668 (* 0.0454545 = 0.157576 loss)
I0405 07:23:38.103266 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.55308 (* 0.0454545 = 0.161504 loss)
I0405 07:23:38.103281 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.62177 (* 0.0454545 = 0.119171 loss)
I0405 07:23:38.103294 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.37211 (* 0.0454545 = 0.0623689 loss)
I0405 07:23:38.103308 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.477724 (* 0.0454545 = 0.0217147 loss)
I0405 07:23:38.103322 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.552396 (* 0.0454545 = 0.0251089 loss)
I0405 07:23:38.103338 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00720419 (* 0.0454545 = 0.000327463 loss)
I0405 07:23:38.103353 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.56242e-05 (* 0.0454545 = 1.61928e-06 loss)
I0405 07:23:38.103368 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.37934e-05 (* 0.0454545 = 1.99061e-06 loss)
I0405 07:23:38.103382 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.43972e-05 (* 0.0454545 = 2.01805e-06 loss)
I0405 07:23:38.103397 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.91298e-05 (* 0.0454545 = 1.77863e-06 loss)
I0405 07:23:38.103411 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.67752e-05 (* 0.0454545 = 1.6716e-06 loss)
I0405 07:23:38.103426 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.71555e-05 (* 0.0454545 = 1.68889e-06 loss)
I0405 07:23:38.103441 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.98047e-05 (* 0.0454545 = 1.80931e-06 loss)
I0405 07:23:38.103474 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.39947e-05 (* 0.0454545 = 1.99976e-06 loss)
I0405 07:23:38.103492 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.41737e-05 (* 0.0454545 = 2.0079e-06 loss)
I0405 07:23:38.103507 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.85265e-05 (* 0.0454545 = 1.7512e-06 loss)
I0405 07:23:38.103521 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.89849e-05 (* 0.0454545 = 1.77204e-06 loss)
I0405 07:23:38.103536 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 4.46893e-05 (* 0.0454545 = 2.03133e-06 loss)
I0405 07:23:38.103549 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:23:38.103560 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000579929
I0405 07:23:38.103575 26022 sgd_solver.cpp:106] Iteration 2550, lr = 0.039898
I0405 07:32:29.971864 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 34.4484 > 30) by scale factor 0.870868
I0405 07:32:40.353359 26022 solver.cpp:229] Iteration 2600, loss = 1.01697
I0405 07:32:40.353407 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 07:32:40.353425 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 07:32:40.353438 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 07:32:40.353451 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 07:32:40.353464 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 07:32:40.353477 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 07:32:40.353489 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 07:32:40.353502 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 07:32:40.353516 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 07:32:40.353530 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:32:40.353545 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:32:40.353557 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:32:40.353569 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:32:40.353581 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:32:40.353593 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:32:40.353605 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:32:40.353617 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:32:40.353628 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:32:40.353641 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:32:40.353652 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:32:40.353663 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:32:40.353675 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:32:40.353691 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.22794 (* 0.0454545 = 0.146725 loss)
I0405 07:32:40.353708 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.40361 (* 0.0454545 = 0.15471 loss)
I0405 07:32:40.353723 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.53299 (* 0.0454545 = 0.160591 loss)
I0405 07:32:40.353737 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.35214 (* 0.0454545 = 0.15237 loss)
I0405 07:32:40.353751 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.02552 (* 0.0454545 = 0.137524 loss)
I0405 07:32:40.353766 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.56857 (* 0.0454545 = 0.116753 loss)
I0405 07:32:40.353780 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.49965 (* 0.0454545 = 0.0681659 loss)
I0405 07:32:40.353795 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.424517 (* 0.0454545 = 0.0192962 loss)
I0405 07:32:40.353809 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.233856 (* 0.0454545 = 0.0106298 loss)
I0405 07:32:40.353823 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0250415 (* 0.0454545 = 0.00113825 loss)
I0405 07:32:40.353839 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.66886e-05 (* 0.0454545 = 1.21312e-06 loss)
I0405 07:32:40.353854 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.33088e-05 (* 0.0454545 = 1.51404e-06 loss)
I0405 07:32:40.353869 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.46984e-05 (* 0.0454545 = 1.5772e-06 loss)
I0405 07:32:40.353884 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.9818e-05 (* 0.0454545 = 1.35536e-06 loss)
I0405 07:32:40.353899 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.70425e-05 (* 0.0454545 = 1.22921e-06 loss)
I0405 07:32:40.353914 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.95982e-05 (* 0.0454545 = 1.34537e-06 loss)
I0405 07:32:40.353929 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.00639e-05 (* 0.0454545 = 1.36654e-06 loss)
I0405 07:32:40.353974 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.37261e-05 (* 0.0454545 = 1.533e-06 loss)
I0405 07:32:40.353991 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.59166e-05 (* 0.0454545 = 1.63257e-06 loss)
I0405 07:32:40.354007 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.91325e-05 (* 0.0454545 = 1.32421e-06 loss)
I0405 07:32:40.354022 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.9818e-05 (* 0.0454545 = 1.35536e-06 loss)
I0405 07:32:40.354037 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.48474e-05 (* 0.0454545 = 1.58397e-06 loss)
I0405 07:32:40.354048 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:32:40.354060 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000275161
I0405 07:32:40.354075 26022 sgd_solver.cpp:106] Iteration 2600, lr = 0.039896
I0405 07:39:54.510187 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 38.5796 > 30) by scale factor 0.777614
I0405 07:41:42.439062 26022 solver.cpp:229] Iteration 2650, loss = 1.02116
I0405 07:41:42.439160 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 07:41:42.439179 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 07:41:42.439193 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 07:41:42.439206 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 07:41:42.439219 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 07:41:42.439232 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 07:41:42.439245 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 07:41:42.439259 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 07:41:42.439270 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 07:41:42.439285 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:41:42.439296 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:41:42.439308 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:41:42.439321 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:41:42.439332 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:41:42.439344 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:41:42.439355 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:41:42.439368 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:41:42.439379 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:41:42.439393 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:41:42.439404 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:41:42.439416 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:41:42.439429 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:41:42.439443 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.02099 (* 0.0454545 = 0.137318 loss)
I0405 07:41:42.439460 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.5508 (* 0.0454545 = 0.1614 loss)
I0405 07:41:42.439474 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.60096 (* 0.0454545 = 0.16368 loss)
I0405 07:41:42.439488 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.4535 (* 0.0454545 = 0.156977 loss)
I0405 07:41:42.439507 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.50477 (* 0.0454545 = 0.159308 loss)
I0405 07:41:42.439522 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.45924 (* 0.0454545 = 0.111784 loss)
I0405 07:41:42.439537 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.6309 (* 0.0454545 = 0.0741316 loss)
I0405 07:41:42.439550 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.514668 (* 0.0454545 = 0.023394 loss)
I0405 07:41:42.439565 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.301843 (* 0.0454545 = 0.0137201 loss)
I0405 07:41:42.439580 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0298712 (* 0.0454545 = 0.00135778 loss)
I0405 07:41:42.439595 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.0716e-05 (* 0.0454545 = 1.85073e-06 loss)
I0405 07:41:42.439611 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.93299e-05 (* 0.0454545 = 2.24227e-06 loss)
I0405 07:41:42.439626 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.19008e-05 (* 0.0454545 = 2.35913e-06 loss)
I0405 07:41:42.439641 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.57309e-05 (* 0.0454545 = 2.07868e-06 loss)
I0405 07:41:42.439656 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.13085e-05 (* 0.0454545 = 1.87766e-06 loss)
I0405 07:41:42.439671 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.33389e-05 (* 0.0454545 = 1.96995e-06 loss)
I0405 07:41:42.439685 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.63047e-05 (* 0.0454545 = 2.10476e-06 loss)
I0405 07:41:42.439719 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.05335e-05 (* 0.0454545 = 2.29698e-06 loss)
I0405 07:41:42.439735 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.44529e-05 (* 0.0454545 = 2.47513e-06 loss)
I0405 07:41:42.439752 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.33836e-05 (* 0.0454545 = 1.97198e-06 loss)
I0405 07:41:42.439769 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.54142e-05 (* 0.0454545 = 2.06428e-06 loss)
I0405 07:41:42.439784 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.23554e-05 (* 0.0454545 = 2.37979e-06 loss)
I0405 07:41:42.439796 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:41:42.439808 26022 solver.cpp:245]     Train net output #45: total_confidence = 9.90312e-05
I0405 07:41:42.439823 26022 sgd_solver.cpp:106] Iteration 2650, lr = 0.039894
I0405 07:50:44.602888 26022 solver.cpp:229] Iteration 2700, loss = 1.01635
I0405 07:50:44.603058 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 07:50:44.603080 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 07:50:44.603092 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 07:50:44.603104 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 07:50:44.603117 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 07:50:44.603130 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.46875
I0405 07:50:44.603142 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 07:50:44.603153 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 07:50:44.603165 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 07:50:44.603178 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:50:44.603189 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:50:44.603201 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:50:44.603212 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:50:44.603224 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:50:44.603236 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:50:44.603247 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:50:44.603260 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:50:44.603271 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:50:44.603282 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:50:44.603298 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:50:44.603310 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:50:44.603322 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:50:44.603338 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.4866 (* 0.0454545 = 0.158482 loss)
I0405 07:50:44.603351 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.8075 (* 0.0454545 = 0.173068 loss)
I0405 07:50:44.603366 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.76698 (* 0.0454545 = 0.171226 loss)
I0405 07:50:44.603380 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.62488 (* 0.0454545 = 0.164767 loss)
I0405 07:50:44.603394 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.22734 (* 0.0454545 = 0.146697 loss)
I0405 07:50:44.603409 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.56155 (* 0.0454545 = 0.116434 loss)
I0405 07:50:44.603423 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.48513 (* 0.0454545 = 0.067506 loss)
I0405 07:50:44.603437 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.412215 (* 0.0454545 = 0.0187371 loss)
I0405 07:50:44.603451 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0238762 (* 0.0454545 = 0.00108528 loss)
I0405 07:50:44.603466 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.013874 (* 0.0454545 = 0.000630635 loss)
I0405 07:50:44.603482 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000116789 (* 0.0454545 = 5.3086e-06 loss)
I0405 07:50:44.603497 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000143373 (* 0.0454545 = 6.51697e-06 loss)
I0405 07:50:44.603512 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000144423 (* 0.0454545 = 6.5647e-06 loss)
I0405 07:50:44.603526 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000159214 (* 0.0454545 = 7.23701e-06 loss)
I0405 07:50:44.603540 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000146414 (* 0.0454545 = 6.65516e-06 loss)
I0405 07:50:44.603555 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000131189 (* 0.0454545 = 5.96315e-06 loss)
I0405 07:50:44.603570 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00013745 (* 0.0454545 = 6.24775e-06 loss)
I0405 07:50:44.603602 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000141829 (* 0.0454545 = 6.44678e-06 loss)
I0405 07:50:44.603618 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000142325 (* 0.0454545 = 6.46932e-06 loss)
I0405 07:50:44.603633 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000140897 (* 0.0454545 = 6.40443e-06 loss)
I0405 07:50:44.603648 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000141105 (* 0.0454545 = 6.41388e-06 loss)
I0405 07:50:44.603662 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000160749 (* 0.0454545 = 7.30676e-06 loss)
I0405 07:50:44.603675 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:50:44.603688 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000171319
I0405 07:50:44.603701 26022 sgd_solver.cpp:106] Iteration 2700, lr = 0.039892
I0405 07:59:46.679805 26022 solver.cpp:229] Iteration 2750, loss = 1.00627
I0405 07:59:46.679966 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 07:59:46.679987 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 07:59:46.680001 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 07:59:46.680014 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 07:59:46.680027 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 07:59:46.680039 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 07:59:46.680052 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.625
I0405 07:59:46.680063 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 07:59:46.680096 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 07:59:46.680111 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 07:59:46.680124 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 07:59:46.680135 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 07:59:46.680147 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 07:59:46.680160 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 07:59:46.680171 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 07:59:46.680182 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 07:59:46.680193 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 07:59:46.680207 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 07:59:46.680217 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 07:59:46.680229 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 07:59:46.680240 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 07:59:46.680253 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 07:59:46.680268 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.19843 (* 0.0454545 = 0.145383 loss)
I0405 07:59:46.680282 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.49072 (* 0.0454545 = 0.158669 loss)
I0405 07:59:46.680296 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.56795 (* 0.0454545 = 0.16218 loss)
I0405 07:59:46.680311 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.56764 (* 0.0454545 = 0.162165 loss)
I0405 07:59:46.680325 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.19156 (* 0.0454545 = 0.145071 loss)
I0405 07:59:46.680340 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.81714 (* 0.0454545 = 0.128052 loss)
I0405 07:59:46.680353 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.65766 (* 0.0454545 = 0.0753483 loss)
I0405 07:59:46.680371 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.738623 (* 0.0454545 = 0.0335738 loss)
I0405 07:59:46.680387 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.404443 (* 0.0454545 = 0.0183838 loss)
I0405 07:59:46.680402 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.022895 (* 0.0454545 = 0.00104068 loss)
I0405 07:59:46.680416 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 9.2357e-05 (* 0.0454545 = 4.19804e-06 loss)
I0405 07:59:46.680434 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000111608 (* 0.0454545 = 5.0731e-06 loss)
I0405 07:59:46.680449 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000109847 (* 0.0454545 = 4.99304e-06 loss)
I0405 07:59:46.680464 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.00012808 (* 0.0454545 = 5.82181e-06 loss)
I0405 07:59:46.680480 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000118705 (* 0.0454545 = 5.3957e-06 loss)
I0405 07:59:46.680493 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 9.97206e-05 (* 0.0454545 = 4.53275e-06 loss)
I0405 07:59:46.680508 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00010762 (* 0.0454545 = 4.89183e-06 loss)
I0405 07:59:46.680541 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.00011135 (* 0.0454545 = 5.06137e-06 loss)
I0405 07:59:46.680557 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000113639 (* 0.0454545 = 5.16543e-06 loss)
I0405 07:59:46.680572 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000108691 (* 0.0454545 = 4.94048e-06 loss)
I0405 07:59:46.680586 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000110392 (* 0.0454545 = 5.01783e-06 loss)
I0405 07:59:46.680600 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000130658 (* 0.0454545 = 5.939e-06 loss)
I0405 07:59:46.680614 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 07:59:46.680626 26022 solver.cpp:245]     Train net output #45: total_confidence = 5.90404e-05
I0405 07:59:46.680641 26022 sgd_solver.cpp:106] Iteration 2750, lr = 0.03989
I0405 08:02:51.460438 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.4721 > 30) by scale factor 0.984507
I0405 08:03:13.136863 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 43.6104 > 30) by scale factor 0.68791
I0405 08:08:48.736587 26022 solver.cpp:229] Iteration 2800, loss = 1.01006
I0405 08:08:48.736773 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 08:08:48.736802 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 08:08:48.736827 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 08:08:48.736851 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 08:08:48.736874 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 08:08:48.736896 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 08:08:48.736918 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 08:08:48.736939 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 08:08:48.736963 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.875
I0405 08:08:48.736986 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.90625
I0405 08:08:48.737009 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:08:48.737030 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:08:48.737051 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:08:48.737071 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:08:48.737092 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:08:48.737112 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:08:48.737135 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:08:48.737159 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:08:48.737179 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:08:48.737205 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:08:48.737226 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:08:48.737246 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:08:48.737275 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.0103 (* 0.0454545 = 0.136832 loss)
I0405 08:08:48.737303 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.46945 (* 0.0454545 = 0.157702 loss)
I0405 08:08:48.737329 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.38179 (* 0.0454545 = 0.153718 loss)
I0405 08:08:48.737355 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.41237 (* 0.0454545 = 0.155108 loss)
I0405 08:08:48.737383 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.01667 (* 0.0454545 = 0.137121 loss)
I0405 08:08:48.737411 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.36758 (* 0.0454545 = 0.107617 loss)
I0405 08:08:48.737437 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.8134 (* 0.0454545 = 0.0824273 loss)
I0405 08:08:48.737463 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.743737 (* 0.0454545 = 0.0338062 loss)
I0405 08:08:48.737489 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.622293 (* 0.0454545 = 0.028286 loss)
I0405 08:08:48.737516 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.563441 (* 0.0454545 = 0.0256109 loss)
I0405 08:08:48.737546 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000123897 (* 0.0454545 = 5.63166e-06 loss)
I0405 08:08:48.737573 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.00015342 (* 0.0454545 = 6.97364e-06 loss)
I0405 08:08:48.737601 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.00015174 (* 0.0454545 = 6.89727e-06 loss)
I0405 08:08:48.737627 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000177286 (* 0.0454545 = 8.05847e-06 loss)
I0405 08:08:48.737655 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000167621 (* 0.0454545 = 7.61913e-06 loss)
I0405 08:08:48.737682 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.00014175 (* 0.0454545 = 6.4432e-06 loss)
I0405 08:08:48.737709 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000153989 (* 0.0454545 = 6.99951e-06 loss)
I0405 08:08:48.737756 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000157589 (* 0.0454545 = 7.16313e-06 loss)
I0405 08:08:48.737784 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000157715 (* 0.0454545 = 7.16886e-06 loss)
I0405 08:08:48.737810 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000155644 (* 0.0454545 = 7.07471e-06 loss)
I0405 08:08:48.737838 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.0001574 (* 0.0454545 = 7.15455e-06 loss)
I0405 08:08:48.737864 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000181505 (* 0.0454545 = 8.25023e-06 loss)
I0405 08:08:48.737886 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:08:48.737908 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000219318
I0405 08:08:48.737932 26022 sgd_solver.cpp:106] Iteration 2800, lr = 0.039888
I0405 08:17:50.866094 26022 solver.cpp:229] Iteration 2850, loss = 0.999861
I0405 08:17:50.866274 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 08:17:50.866296 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 08:17:50.866308 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 08:17:50.866322 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 08:17:50.866335 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 08:17:50.866348 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5
I0405 08:17:50.866361 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 08:17:50.866374 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 08:17:50.866385 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 08:17:50.866397 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 08:17:50.866410 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:17:50.866421 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:17:50.866432 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:17:50.866444 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:17:50.866456 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:17:50.866467 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:17:50.866478 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:17:50.866490 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:17:50.866502 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:17:50.866513 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:17:50.866523 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:17:50.866534 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:17:50.866550 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.9859 (* 0.0454545 = 0.135723 loss)
I0405 08:17:50.866565 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.33505 (* 0.0454545 = 0.151593 loss)
I0405 08:17:50.866580 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.29851 (* 0.0454545 = 0.149932 loss)
I0405 08:17:50.866592 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.14018 (* 0.0454545 = 0.142735 loss)
I0405 08:17:50.866606 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.09943 (* 0.0454545 = 0.140883 loss)
I0405 08:17:50.866621 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.42721 (* 0.0454545 = 0.110328 loss)
I0405 08:17:50.866636 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.21982 (* 0.0454545 = 0.0554464 loss)
I0405 08:17:50.866649 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.650299 (* 0.0454545 = 0.0295591 loss)
I0405 08:17:50.866663 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.221143 (* 0.0454545 = 0.010052 loss)
I0405 08:17:50.866678 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0141841 (* 0.0454545 = 0.000644734 loss)
I0405 08:17:50.866693 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.8817e-05 (* 0.0454545 = 2.21896e-06 loss)
I0405 08:17:50.866708 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.9227e-05 (* 0.0454545 = 2.69214e-06 loss)
I0405 08:17:50.866722 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.59884e-05 (* 0.0454545 = 2.54493e-06 loss)
I0405 08:17:50.866736 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 6.89509e-05 (* 0.0454545 = 3.13413e-06 loss)
I0405 08:17:50.866750 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 6.47773e-05 (* 0.0454545 = 2.94442e-06 loss)
I0405 08:17:50.866765 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.24497e-05 (* 0.0454545 = 2.38408e-06 loss)
I0405 08:17:50.866780 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.64644e-05 (* 0.0454545 = 2.56656e-06 loss)
I0405 08:17:50.866812 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.7343e-05 (* 0.0454545 = 2.6065e-06 loss)
I0405 08:17:50.866828 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.8494e-05 (* 0.0454545 = 2.65882e-06 loss)
I0405 08:17:50.866842 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.59152e-05 (* 0.0454545 = 2.5416e-06 loss)
I0405 08:17:50.866860 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.74581e-05 (* 0.0454545 = 2.61173e-06 loss)
I0405 08:17:50.866875 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.99565e-05 (* 0.0454545 = 3.17984e-06 loss)
I0405 08:17:50.866888 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:17:50.866899 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000591066
I0405 08:17:50.866914 26022 sgd_solver.cpp:106] Iteration 2850, lr = 0.039886
I0405 08:26:52.900837 26022 solver.cpp:229] Iteration 2900, loss = 0.989039
I0405 08:26:52.901012 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 08:26:52.901033 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 08:26:52.901046 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 08:26:52.901059 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 08:26:52.901072 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.34375
I0405 08:26:52.901083 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 08:26:52.901096 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 08:26:52.901108 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 08:26:52.901120 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 08:26:52.901132 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 08:26:52.901144 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:26:52.901156 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:26:52.901168 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:26:52.901180 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:26:52.901191 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:26:52.901203 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:26:52.901216 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:26:52.901227 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:26:52.901238 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:26:52.901250 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:26:52.901262 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:26:52.901273 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:26:52.901289 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.95186 (* 0.0454545 = 0.134176 loss)
I0405 08:26:52.901304 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.1987 (* 0.0454545 = 0.145395 loss)
I0405 08:26:52.901319 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.23369 (* 0.0454545 = 0.146986 loss)
I0405 08:26:52.901334 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.12794 (* 0.0454545 = 0.142179 loss)
I0405 08:26:52.901350 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.60461 (* 0.0454545 = 0.118391 loss)
I0405 08:26:52.901365 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.38423 (* 0.0454545 = 0.108374 loss)
I0405 08:26:52.901381 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.13068 (* 0.0454545 = 0.0513948 loss)
I0405 08:26:52.901394 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.436172 (* 0.0454545 = 0.019826 loss)
I0405 08:26:52.901409 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.215405 (* 0.0454545 = 0.00979113 loss)
I0405 08:26:52.901424 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.176911 (* 0.0454545 = 0.0080414 loss)
I0405 08:26:52.901438 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000109099 (* 0.0454545 = 4.95903e-06 loss)
I0405 08:26:52.901453 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000131678 (* 0.0454545 = 5.98536e-06 loss)
I0405 08:26:52.901468 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000127641 (* 0.0454545 = 5.80188e-06 loss)
I0405 08:26:52.901485 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000154166 (* 0.0454545 = 7.00754e-06 loss)
I0405 08:26:52.901500 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000151433 (* 0.0454545 = 6.8833e-06 loss)
I0405 08:26:52.901515 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000124601 (* 0.0454545 = 5.66369e-06 loss)
I0405 08:26:52.901530 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000129564 (* 0.0454545 = 5.88929e-06 loss)
I0405 08:26:52.901561 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000137466 (* 0.0454545 = 6.24845e-06 loss)
I0405 08:26:52.901577 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000138558 (* 0.0454545 = 6.29807e-06 loss)
I0405 08:26:52.901592 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000132428 (* 0.0454545 = 6.01944e-06 loss)
I0405 08:26:52.901607 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.00013685 (* 0.0454545 = 6.22044e-06 loss)
I0405 08:26:52.901621 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000164147 (* 0.0454545 = 7.46123e-06 loss)
I0405 08:26:52.901633 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:26:52.901645 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000193541
I0405 08:26:52.901660 26022 sgd_solver.cpp:106] Iteration 2900, lr = 0.039884
I0405 08:29:14.283166 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.5152 > 30) by scale factor 0.951922
I0405 08:35:54.988898 26022 solver.cpp:229] Iteration 2950, loss = 0.988215
I0405 08:35:54.989084 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 08:35:54.989106 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 08:35:54.989120 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 08:35:54.989132 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 08:35:54.989145 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.21875
I0405 08:35:54.989157 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 08:35:54.989169 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 08:35:54.989181 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 08:35:54.989193 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 08:35:54.989205 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 08:35:54.989217 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:35:54.989229 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:35:54.989240 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:35:54.989256 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:35:54.989269 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:35:54.989279 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:35:54.989291 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:35:54.989305 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:35:54.989318 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:35:54.989329 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:35:54.989341 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:35:54.989352 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:35:54.989368 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.02343 (* 0.0454545 = 0.137429 loss)
I0405 08:35:54.989387 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.32558 (* 0.0454545 = 0.151163 loss)
I0405 08:35:54.989401 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.30755 (* 0.0454545 = 0.150343 loss)
I0405 08:35:54.989415 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.41307 (* 0.0454545 = 0.15514 loss)
I0405 08:35:54.989429 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.98157 (* 0.0454545 = 0.135526 loss)
I0405 08:35:54.989444 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.83393 (* 0.0454545 = 0.128815 loss)
I0405 08:35:54.989459 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.941952 (* 0.0454545 = 0.042816 loss)
I0405 08:35:54.989473 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.33843 (* 0.0454545 = 0.0153832 loss)
I0405 08:35:54.989487 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.162089 (* 0.0454545 = 0.00736768 loss)
I0405 08:35:54.989502 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00987351 (* 0.0454545 = 0.000448796 loss)
I0405 08:35:54.989518 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.72058e-05 (* 0.0454545 = 7.82083e-07 loss)
I0405 08:35:54.989533 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.19534e-05 (* 0.0454545 = 9.97883e-07 loss)
I0405 08:35:54.989548 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.00249e-05 (* 0.0454545 = 9.10224e-07 loss)
I0405 08:35:54.989562 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.55962e-05 (* 0.0454545 = 1.16347e-06 loss)
I0405 08:35:54.989578 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.67799e-05 (* 0.0454545 = 1.21727e-06 loss)
I0405 08:35:54.989593 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 1.88696e-05 (* 0.0454545 = 8.57709e-07 loss)
I0405 08:35:54.989608 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.0595e-05 (* 0.0454545 = 9.36135e-07 loss)
I0405 08:35:54.989641 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.13907e-05 (* 0.0454545 = 9.72302e-07 loss)
I0405 08:35:54.989657 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.09023e-05 (* 0.0454545 = 9.50106e-07 loss)
I0405 08:35:54.989671 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.09659e-05 (* 0.0454545 = 9.52998e-07 loss)
I0405 08:35:54.989686 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.23336e-05 (* 0.0454545 = 1.01516e-06 loss)
I0405 08:35:54.989701 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.74153e-05 (* 0.0454545 = 1.24615e-06 loss)
I0405 08:35:54.989713 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:35:54.989725 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000176443
I0405 08:35:54.989740 26022 sgd_solver.cpp:106] Iteration 2950, lr = 0.039882
I0405 08:37:54.875483 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.3554 > 30) by scale factor 0.956773
I0405 08:38:59.938989 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.6858 > 30) by scale factor 0.977651
I0405 08:44:46.853643 26022 solver.cpp:338] Iteration 3000, Testing net (#0)
I0405 08:45:00.555656 26022 solver.cpp:393] Test loss: 0.871374
I0405 08:45:00.555708 26022 solver.cpp:406]     Test net output #0: loss/accuracy01 = 0.298
I0405 08:45:00.555726 26022 solver.cpp:406]     Test net output #1: loss/accuracy02 = 0.087
I0405 08:45:00.555738 26022 solver.cpp:406]     Test net output #2: loss/accuracy03 = 0.07
I0405 08:45:00.555752 26022 solver.cpp:406]     Test net output #3: loss/accuracy04 = 0.089
I0405 08:45:00.555766 26022 solver.cpp:406]     Test net output #4: loss/accuracy05 = 0.211
I0405 08:45:00.555778 26022 solver.cpp:406]     Test net output #5: loss/accuracy06 = 0.501
I0405 08:45:00.555790 26022 solver.cpp:406]     Test net output #6: loss/accuracy07 = 0.894
I0405 08:45:00.555802 26022 solver.cpp:406]     Test net output #7: loss/accuracy08 = 0.97
I0405 08:45:00.555814 26022 solver.cpp:406]     Test net output #8: loss/accuracy09 = 0.995
I0405 08:45:00.555825 26022 solver.cpp:406]     Test net output #9: loss/accuracy10 = 0.998
I0405 08:45:00.555837 26022 solver.cpp:406]     Test net output #10: loss/accuracy11 = 1
I0405 08:45:00.555848 26022 solver.cpp:406]     Test net output #11: loss/accuracy12 = 1
I0405 08:45:00.555860 26022 solver.cpp:406]     Test net output #12: loss/accuracy13 = 1
I0405 08:45:00.555873 26022 solver.cpp:406]     Test net output #13: loss/accuracy14 = 1
I0405 08:45:00.555884 26022 solver.cpp:406]     Test net output #14: loss/accuracy15 = 1
I0405 08:45:00.555896 26022 solver.cpp:406]     Test net output #15: loss/accuracy16 = 1
I0405 08:45:00.555908 26022 solver.cpp:406]     Test net output #16: loss/accuracy17 = 1
I0405 08:45:00.555919 26022 solver.cpp:406]     Test net output #17: loss/accuracy18 = 1
I0405 08:45:00.555930 26022 solver.cpp:406]     Test net output #18: loss/accuracy19 = 1
I0405 08:45:00.555943 26022 solver.cpp:406]     Test net output #19: loss/accuracy20 = 1
I0405 08:45:00.555953 26022 solver.cpp:406]     Test net output #20: loss/accuracy21 = 1
I0405 08:45:00.555964 26022 solver.cpp:406]     Test net output #21: loss/accuracy22 = 1
I0405 08:45:00.555980 26022 solver.cpp:406]     Test net output #22: loss/loss01 = 2.98547 (* 0.0454545 = 0.135703 loss)
I0405 08:45:00.555994 26022 solver.cpp:406]     Test net output #23: loss/loss02 = 3.13533 (* 0.0454545 = 0.142515 loss)
I0405 08:45:00.556008 26022 solver.cpp:406]     Test net output #24: loss/loss03 = 3.25033 (* 0.0454545 = 0.147742 loss)
I0405 08:45:00.556022 26022 solver.cpp:406]     Test net output #25: loss/loss04 = 3.22198 (* 0.0454545 = 0.146453 loss)
I0405 08:45:00.556036 26022 solver.cpp:406]     Test net output #26: loss/loss05 = 3.1364 (* 0.0454545 = 0.142563 loss)
I0405 08:45:00.556051 26022 solver.cpp:406]     Test net output #27: loss/loss06 = 2.37802 (* 0.0454545 = 0.108092 loss)
I0405 08:45:00.556064 26022 solver.cpp:406]     Test net output #28: loss/loss07 = 0.746467 (* 0.0454545 = 0.0339303 loss)
I0405 08:45:00.556097 26022 solver.cpp:406]     Test net output #29: loss/loss08 = 0.237323 (* 0.0454545 = 0.0107874 loss)
I0405 08:45:00.556113 26022 solver.cpp:406]     Test net output #30: loss/loss09 = 0.0515998 (* 0.0454545 = 0.00234545 loss)
I0405 08:45:00.556128 26022 solver.cpp:406]     Test net output #31: loss/loss10 = 0.0226643 (* 0.0454545 = 0.00103019 loss)
I0405 08:45:00.556143 26022 solver.cpp:406]     Test net output #32: loss/loss11 = 0.000309187 (* 0.0454545 = 1.4054e-05 loss)
I0405 08:45:00.556157 26022 solver.cpp:406]     Test net output #33: loss/loss12 = 0.000384221 (* 0.0454545 = 1.74646e-05 loss)
I0405 08:45:00.556172 26022 solver.cpp:406]     Test net output #34: loss/loss13 = 0.00035597 (* 0.0454545 = 1.61804e-05 loss)
I0405 08:45:00.556186 26022 solver.cpp:406]     Test net output #35: loss/loss14 = 0.000436772 (* 0.0454545 = 1.98533e-05 loss)
I0405 08:45:00.556201 26022 solver.cpp:406]     Test net output #36: loss/loss15 = 0.000453498 (* 0.0454545 = 2.06135e-05 loss)
I0405 08:45:00.556216 26022 solver.cpp:406]     Test net output #37: loss/loss16 = 0.000366789 (* 0.0454545 = 1.66722e-05 loss)
I0405 08:45:00.556231 26022 solver.cpp:406]     Test net output #38: loss/loss17 = 0.000372945 (* 0.0454545 = 1.6952e-05 loss)
I0405 08:45:00.556277 26022 solver.cpp:406]     Test net output #39: loss/loss18 = 0.00036605 (* 0.0454545 = 1.66386e-05 loss)
I0405 08:45:00.556293 26022 solver.cpp:406]     Test net output #40: loss/loss19 = 0.000363002 (* 0.0454545 = 1.65001e-05 loss)
I0405 08:45:00.556308 26022 solver.cpp:406]     Test net output #41: loss/loss20 = 0.000381216 (* 0.0454545 = 1.7328e-05 loss)
I0405 08:45:00.556324 26022 solver.cpp:406]     Test net output #42: loss/loss21 = 0.000386985 (* 0.0454545 = 1.75902e-05 loss)
I0405 08:45:00.556337 26022 solver.cpp:406]     Test net output #43: loss/loss22 = 0.000458737 (* 0.0454545 = 2.08517e-05 loss)
I0405 08:45:00.556349 26022 solver.cpp:406]     Test net output #44: total_accuracy = 0
I0405 08:45:00.556362 26022 solver.cpp:406]     Test net output #45: total_confidence = 0.000243673
I0405 08:45:10.869858 26022 solver.cpp:229] Iteration 3000, loss = 0.986703
I0405 08:45:10.869904 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 08:45:10.869921 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 08:45:10.869935 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 08:45:10.869947 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.21875
I0405 08:45:10.869959 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 08:45:10.869972 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 08:45:10.869984 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 08:45:10.869997 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 08:45:10.870008 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 08:45:10.870020 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 08:45:10.870033 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:45:10.870043 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:45:10.870055 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:45:10.870067 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:45:10.870079 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:45:10.870090 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:45:10.870102 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:45:10.870113 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:45:10.870126 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:45:10.870136 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:45:10.870151 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:45:10.870163 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:45:10.870178 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.98562 (* 0.0454545 = 0.13571 loss)
I0405 08:45:10.870193 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.31053 (* 0.0454545 = 0.150479 loss)
I0405 08:45:10.870208 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.12625 (* 0.0454545 = 0.142102 loss)
I0405 08:45:10.870221 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.1603 (* 0.0454545 = 0.14365 loss)
I0405 08:45:10.870235 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.90014 (* 0.0454545 = 0.131825 loss)
I0405 08:45:10.870249 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.24806 (* 0.0454545 = 0.102184 loss)
I0405 08:45:10.870263 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.1097 (* 0.0454545 = 0.0504408 loss)
I0405 08:45:10.870278 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.21263 (* 0.0454545 = 0.00966502 loss)
I0405 08:45:10.870292 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.194163 (* 0.0454545 = 0.0088256 loss)
I0405 08:45:10.870307 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0070138 (* 0.0454545 = 0.000318809 loss)
I0405 08:45:10.870352 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000198522 (* 0.0454545 = 9.02371e-06 loss)
I0405 08:45:10.870371 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000255639 (* 0.0454545 = 1.162e-05 loss)
I0405 08:45:10.870388 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000227167 (* 0.0454545 = 1.03258e-05 loss)
I0405 08:45:10.870403 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000287001 (* 0.0454545 = 1.30455e-05 loss)
I0405 08:45:10.870416 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000294629 (* 0.0454545 = 1.33922e-05 loss)
I0405 08:45:10.870430 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000243 (* 0.0454545 = 1.10455e-05 loss)
I0405 08:45:10.870445 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000242795 (* 0.0454545 = 1.10361e-05 loss)
I0405 08:45:10.870460 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.00024611 (* 0.0454545 = 1.11868e-05 loss)
I0405 08:45:10.870474 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000234736 (* 0.0454545 = 1.06698e-05 loss)
I0405 08:45:10.870489 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000248723 (* 0.0454545 = 1.13056e-05 loss)
I0405 08:45:10.870504 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000253455 (* 0.0454545 = 1.15207e-05 loss)
I0405 08:45:10.870519 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000303101 (* 0.0454545 = 1.37773e-05 loss)
I0405 08:45:10.870532 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:45:10.870544 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000779544
I0405 08:45:10.870558 26022 sgd_solver.cpp:106] Iteration 3000, lr = 0.03988
I0405 08:54:13.055459 26022 solver.cpp:229] Iteration 3050, loss = 0.984372
I0405 08:54:13.055620 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 08:54:13.055640 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 08:54:13.055655 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 08:54:13.055667 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 08:54:13.055680 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 08:54:13.055692 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 08:54:13.055706 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 08:54:13.055718 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 08:54:13.055730 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 08:54:13.055742 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 08:54:13.055754 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 08:54:13.055768 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 08:54:13.055779 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 08:54:13.055791 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 08:54:13.055804 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 08:54:13.055815 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 08:54:13.055827 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 08:54:13.055838 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 08:54:13.055850 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 08:54:13.055861 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 08:54:13.055873 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 08:54:13.055886 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 08:54:13.055903 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.84871 (* 0.0454545 = 0.129487 loss)
I0405 08:54:13.055919 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.19996 (* 0.0454545 = 0.145453 loss)
I0405 08:54:13.055933 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.4137 (* 0.0454545 = 0.155168 loss)
I0405 08:54:13.055948 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.22837 (* 0.0454545 = 0.146744 loss)
I0405 08:54:13.055963 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.09561 (* 0.0454545 = 0.14071 loss)
I0405 08:54:13.055976 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.73 (* 0.0454545 = 0.124091 loss)
I0405 08:54:13.055991 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.48504 (* 0.0454545 = 0.0675019 loss)
I0405 08:54:13.056006 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.471762 (* 0.0454545 = 0.0214437 loss)
I0405 08:54:13.056021 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.205822 (* 0.0454545 = 0.00935555 loss)
I0405 08:54:13.056036 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.024952 (* 0.0454545 = 0.00113418 loss)
I0405 08:54:13.056051 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000278946 (* 0.0454545 = 1.26794e-05 loss)
I0405 08:54:13.056079 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000360127 (* 0.0454545 = 1.63694e-05 loss)
I0405 08:54:13.056099 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.0003368 (* 0.0454545 = 1.53091e-05 loss)
I0405 08:54:13.056114 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000381416 (* 0.0454545 = 1.73371e-05 loss)
I0405 08:54:13.056129 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000397707 (* 0.0454545 = 1.80776e-05 loss)
I0405 08:54:13.056144 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000335743 (* 0.0454545 = 1.5261e-05 loss)
I0405 08:54:13.056159 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000350513 (* 0.0454545 = 1.59324e-05 loss)
I0405 08:54:13.056191 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000349383 (* 0.0454545 = 1.5881e-05 loss)
I0405 08:54:13.056207 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000341423 (* 0.0454545 = 1.55192e-05 loss)
I0405 08:54:13.056222 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000345654 (* 0.0454545 = 1.57115e-05 loss)
I0405 08:54:13.056237 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000371899 (* 0.0454545 = 1.69045e-05 loss)
I0405 08:54:13.056252 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000421919 (* 0.0454545 = 1.91781e-05 loss)
I0405 08:54:13.056265 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 08:54:13.056278 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.93339e-05
I0405 08:54:13.056293 26022 sgd_solver.cpp:106] Iteration 3050, lr = 0.039878
I0405 09:02:10.643656 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.7631 > 30) by scale factor 0.975194
I0405 09:02:21.500195 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 32.0588 > 30) by scale factor 0.935781
I0405 09:03:15.220595 26022 solver.cpp:229] Iteration 3100, loss = 0.985082
I0405 09:03:15.220718 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 09:03:15.220738 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 09:03:15.220752 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 09:03:15.220767 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 09:03:15.220780 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.3125
I0405 09:03:15.220793 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 09:03:15.220805 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 09:03:15.220818 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 09:03:15.220830 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 09:03:15.220842 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 09:03:15.220854 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:03:15.220866 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:03:15.220877 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:03:15.220890 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:03:15.220901 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:03:15.220912 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:03:15.220924 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:03:15.220935 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:03:15.220947 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:03:15.220959 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:03:15.220970 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:03:15.220983 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:03:15.220998 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.30195 (* 0.0454545 = 0.150089 loss)
I0405 09:03:15.221012 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.59886 (* 0.0454545 = 0.163584 loss)
I0405 09:03:15.221026 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.43358 (* 0.0454545 = 0.156072 loss)
I0405 09:03:15.221041 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.34051 (* 0.0454545 = 0.151841 loss)
I0405 09:03:15.221055 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.86087 (* 0.0454545 = 0.130039 loss)
I0405 09:03:15.221070 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.82439 (* 0.0454545 = 0.128381 loss)
I0405 09:03:15.221084 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.28216 (* 0.0454545 = 0.0582799 loss)
I0405 09:03:15.221098 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.678106 (* 0.0454545 = 0.030823 loss)
I0405 09:03:15.221113 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.153552 (* 0.0454545 = 0.00697962 loss)
I0405 09:03:15.221128 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0113403 (* 0.0454545 = 0.000515469 loss)
I0405 09:03:15.221143 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.96821e-05 (* 0.0454545 = 1.34919e-06 loss)
I0405 09:03:15.221158 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 3.16514e-05 (* 0.0454545 = 1.4387e-06 loss)
I0405 09:03:15.221174 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.11634e-05 (* 0.0454545 = 1.41652e-06 loss)
I0405 09:03:15.221187 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.57538e-05 (* 0.0454545 = 1.62517e-06 loss)
I0405 09:03:15.221202 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.40714e-05 (* 0.0454545 = 1.5487e-06 loss)
I0405 09:03:15.221217 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.48243e-05 (* 0.0454545 = 1.58292e-06 loss)
I0405 09:03:15.221231 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 3.21172e-05 (* 0.0454545 = 1.45987e-06 loss)
I0405 09:03:15.221263 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.41649e-05 (* 0.0454545 = 1.55295e-06 loss)
I0405 09:03:15.221279 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.34623e-05 (* 0.0454545 = 1.52101e-06 loss)
I0405 09:03:15.221294 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.10292e-05 (* 0.0454545 = 1.41042e-06 loss)
I0405 09:03:15.221308 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.12341e-05 (* 0.0454545 = 1.41973e-06 loss)
I0405 09:03:15.221323 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.65122e-05 (* 0.0454545 = 1.65964e-06 loss)
I0405 09:03:15.221335 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:03:15.221348 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000899288
I0405 09:03:15.221364 26022 sgd_solver.cpp:106] Iteration 3100, lr = 0.039876
I0405 09:03:15.700597 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.4188 > 30) by scale factor 0.954842
I0405 09:11:23.591235 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.1826 > 30) by scale factor 0.962074
I0405 09:12:17.307250 26022 solver.cpp:229] Iteration 3150, loss = 0.981546
I0405 09:12:17.307353 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 09:12:17.307373 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 09:12:17.307386 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 09:12:17.307399 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 09:12:17.307411 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 09:12:17.307423 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 09:12:17.307435 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 09:12:17.307448 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 09:12:17.307461 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 09:12:17.307472 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 09:12:17.307484 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:12:17.307497 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:12:17.307507 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:12:17.307519 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:12:17.307531 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:12:17.307543 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:12:17.307554 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:12:17.307565 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:12:17.307577 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:12:17.307590 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:12:17.307600 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:12:17.307612 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:12:17.307627 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.23649 (* 0.0454545 = 0.147113 loss)
I0405 09:12:17.307642 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.50091 (* 0.0454545 = 0.159132 loss)
I0405 09:12:17.307657 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.51866 (* 0.0454545 = 0.159939 loss)
I0405 09:12:17.307670 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.42208 (* 0.0454545 = 0.155549 loss)
I0405 09:12:17.307685 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.47634 (* 0.0454545 = 0.158016 loss)
I0405 09:12:17.307699 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.71815 (* 0.0454545 = 0.123552 loss)
I0405 09:12:17.307713 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.21243 (* 0.0454545 = 0.0551105 loss)
I0405 09:12:17.307728 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.528366 (* 0.0454545 = 0.0240166 loss)
I0405 09:12:17.307741 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.32387 (* 0.0454545 = 0.0147214 loss)
I0405 09:12:17.307756 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.363193 (* 0.0454545 = 0.0165088 loss)
I0405 09:12:17.307771 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000290013 (* 0.0454545 = 1.31824e-05 loss)
I0405 09:12:17.307785 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000370774 (* 0.0454545 = 1.68533e-05 loss)
I0405 09:12:17.307801 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.00033943 (* 0.0454545 = 1.54286e-05 loss)
I0405 09:12:17.307816 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.00038756 (* 0.0454545 = 1.76164e-05 loss)
I0405 09:12:17.307831 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000411054 (* 0.0454545 = 1.86843e-05 loss)
I0405 09:12:17.307844 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000355065 (* 0.0454545 = 1.61393e-05 loss)
I0405 09:12:17.307859 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000355835 (* 0.0454545 = 1.61743e-05 loss)
I0405 09:12:17.307891 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000341589 (* 0.0454545 = 1.55268e-05 loss)
I0405 09:12:17.307907 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000328214 (* 0.0454545 = 1.49188e-05 loss)
I0405 09:12:17.307924 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000359869 (* 0.0454545 = 1.63577e-05 loss)
I0405 09:12:17.307940 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000353721 (* 0.0454545 = 1.60782e-05 loss)
I0405 09:12:17.307955 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000418953 (* 0.0454545 = 1.90433e-05 loss)
I0405 09:12:17.307966 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:12:17.307978 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.17224e-05
I0405 09:12:17.307993 26022 sgd_solver.cpp:106] Iteration 3150, lr = 0.039874
I0405 09:21:19.348706 26022 solver.cpp:229] Iteration 3200, loss = 0.978408
I0405 09:21:19.348887 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 09:21:19.348907 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.21875
I0405 09:21:19.348922 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 09:21:19.348934 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 09:21:19.348948 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 09:21:19.348960 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 09:21:19.348974 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 09:21:19.348987 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 09:21:19.349000 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 09:21:19.349012 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 09:21:19.349025 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:21:19.349038 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:21:19.349050 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:21:19.349061 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:21:19.349073 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:21:19.349084 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:21:19.349097 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:21:19.349107 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:21:19.349119 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:21:19.349130 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:21:19.349143 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:21:19.349154 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:21:19.349170 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.07475 (* 0.0454545 = 0.139761 loss)
I0405 09:21:19.349185 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.34893 (* 0.0454545 = 0.152224 loss)
I0405 09:21:19.349200 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.55113 (* 0.0454545 = 0.161415 loss)
I0405 09:21:19.349215 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.46061 (* 0.0454545 = 0.1573 loss)
I0405 09:21:19.349228 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.33657 (* 0.0454545 = 0.151662 loss)
I0405 09:21:19.349242 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.90089 (* 0.0454545 = 0.131858 loss)
I0405 09:21:19.349261 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.22445 (* 0.0454545 = 0.0556568 loss)
I0405 09:21:19.349275 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.351622 (* 0.0454545 = 0.0159828 loss)
I0405 09:21:19.349290 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.213094 (* 0.0454545 = 0.0096861 loss)
I0405 09:21:19.349305 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0254252 (* 0.0454545 = 0.00115569 loss)
I0405 09:21:19.349320 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000736469 (* 0.0454545 = 3.34759e-05 loss)
I0405 09:21:19.349335 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000932625 (* 0.0454545 = 4.23921e-05 loss)
I0405 09:21:19.349349 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000868579 (* 0.0454545 = 3.94809e-05 loss)
I0405 09:21:19.349364 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000959955 (* 0.0454545 = 4.36343e-05 loss)
I0405 09:21:19.349380 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000958705 (* 0.0454545 = 4.35775e-05 loss)
I0405 09:21:19.349395 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000961819 (* 0.0454545 = 4.3719e-05 loss)
I0405 09:21:19.349409 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000918232 (* 0.0454545 = 4.17378e-05 loss)
I0405 09:21:19.349442 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000964681 (* 0.0454545 = 4.38491e-05 loss)
I0405 09:21:19.349457 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000864468 (* 0.0454545 = 3.9294e-05 loss)
I0405 09:21:19.349475 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000911892 (* 0.0454545 = 4.14496e-05 loss)
I0405 09:21:19.349491 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000889877 (* 0.0454545 = 4.0449e-05 loss)
I0405 09:21:19.349505 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.00109956 (* 0.0454545 = 4.99802e-05 loss)
I0405 09:21:19.349519 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:21:19.349530 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000145647
I0405 09:21:19.349545 26022 sgd_solver.cpp:106] Iteration 3200, lr = 0.039872
I0405 09:30:21.481858 26022 solver.cpp:229] Iteration 3250, loss = 0.978292
I0405 09:30:21.482017 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 09:30:21.482038 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 09:30:21.482051 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 09:30:21.482064 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 09:30:21.482076 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 09:30:21.482089 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.25
I0405 09:30:21.482100 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 09:30:21.482113 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 09:30:21.482125 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 09:30:21.482137 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 09:30:21.482149 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:30:21.482161 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:30:21.482172 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:30:21.482184 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:30:21.482199 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:30:21.482211 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:30:21.482223 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:30:21.482234 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:30:21.482246 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:30:21.482259 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:30:21.482270 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:30:21.482281 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:30:21.482296 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.23139 (* 0.0454545 = 0.146881 loss)
I0405 09:30:21.482311 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.48278 (* 0.0454545 = 0.158308 loss)
I0405 09:30:21.482326 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.38683 (* 0.0454545 = 0.153947 loss)
I0405 09:30:21.482339 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.39529 (* 0.0454545 = 0.154331 loss)
I0405 09:30:21.482353 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.26988 (* 0.0454545 = 0.148631 loss)
I0405 09:30:21.482367 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 3.05633 (* 0.0454545 = 0.138924 loss)
I0405 09:30:21.482381 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.39495 (* 0.0454545 = 0.0634067 loss)
I0405 09:30:21.482395 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.715182 (* 0.0454545 = 0.0325083 loss)
I0405 09:30:21.482409 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.410011 (* 0.0454545 = 0.0186369 loss)
I0405 09:30:21.482424 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.171701 (* 0.0454545 = 0.00780461 loss)
I0405 09:30:21.482439 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000161249 (* 0.0454545 = 7.32951e-06 loss)
I0405 09:30:21.482453 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000191182 (* 0.0454545 = 8.69008e-06 loss)
I0405 09:30:21.482468 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000173372 (* 0.0454545 = 7.88054e-06 loss)
I0405 09:30:21.482483 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000202963 (* 0.0454545 = 9.22561e-06 loss)
I0405 09:30:21.482498 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000205958 (* 0.0454545 = 9.36174e-06 loss)
I0405 09:30:21.482512 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000203809 (* 0.0454545 = 9.26404e-06 loss)
I0405 09:30:21.482527 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000190012 (* 0.0454545 = 8.63689e-06 loss)
I0405 09:30:21.482558 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000197348 (* 0.0454545 = 8.97037e-06 loss)
I0405 09:30:21.482573 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000177954 (* 0.0454545 = 8.08881e-06 loss)
I0405 09:30:21.482589 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000184052 (* 0.0454545 = 8.36599e-06 loss)
I0405 09:30:21.482604 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000187379 (* 0.0454545 = 8.51724e-06 loss)
I0405 09:30:21.482621 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000237846 (* 0.0454545 = 1.08112e-05 loss)
I0405 09:30:21.482635 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:30:21.482646 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000312623
I0405 09:30:21.482661 26022 sgd_solver.cpp:106] Iteration 3250, lr = 0.03987
I0405 09:36:52.285923 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 45.2247 > 30) by scale factor 0.663354
I0405 09:39:23.638155 26022 solver.cpp:229] Iteration 3300, loss = 0.972041
I0405 09:39:23.638258 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.21875
I0405 09:39:23.638279 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 09:39:23.638293 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 09:39:23.638305 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 09:39:23.638319 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 09:39:23.638330 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 09:39:23.638344 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 09:39:23.638355 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 09:39:23.638367 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 09:39:23.638380 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 09:39:23.638391 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:39:23.638402 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:39:23.638416 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:39:23.638427 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:39:23.638438 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:39:23.638449 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:39:23.638461 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:39:23.638473 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:39:23.638486 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:39:23.638499 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:39:23.638510 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:39:23.638522 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:39:23.638537 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.96345 (* 0.0454545 = 0.134702 loss)
I0405 09:39:23.638552 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.33228 (* 0.0454545 = 0.151467 loss)
I0405 09:39:23.638566 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.558 (* 0.0454545 = 0.161727 loss)
I0405 09:39:23.638581 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.34071 (* 0.0454545 = 0.15185 loss)
I0405 09:39:23.638594 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.92313 (* 0.0454545 = 0.13287 loss)
I0405 09:39:23.638608 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.67501 (* 0.0454545 = 0.121592 loss)
I0405 09:39:23.638622 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.40884 (* 0.0454545 = 0.0640383 loss)
I0405 09:39:23.638636 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.567185 (* 0.0454545 = 0.0257811 loss)
I0405 09:39:23.638650 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.180387 (* 0.0454545 = 0.0081994 loss)
I0405 09:39:23.638665 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0128076 (* 0.0454545 = 0.000582164 loss)
I0405 09:39:23.638680 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000328412 (* 0.0454545 = 1.49278e-05 loss)
I0405 09:39:23.638696 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.00041396 (* 0.0454545 = 1.88163e-05 loss)
I0405 09:39:23.638711 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000384415 (* 0.0454545 = 1.74734e-05 loss)
I0405 09:39:23.638725 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000407411 (* 0.0454545 = 1.85187e-05 loss)
I0405 09:39:23.638741 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000443831 (* 0.0454545 = 2.01741e-05 loss)
I0405 09:39:23.638754 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000460907 (* 0.0454545 = 2.09503e-05 loss)
I0405 09:39:23.638769 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000387089 (* 0.0454545 = 1.7595e-05 loss)
I0405 09:39:23.638804 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000430807 (* 0.0454545 = 1.95821e-05 loss)
I0405 09:39:23.638821 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000374216 (* 0.0454545 = 1.70098e-05 loss)
I0405 09:39:23.638835 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000399846 (* 0.0454545 = 1.81748e-05 loss)
I0405 09:39:23.638849 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000379526 (* 0.0454545 = 1.72512e-05 loss)
I0405 09:39:23.638864 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000512007 (* 0.0454545 = 2.3273e-05 loss)
I0405 09:39:23.638876 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:39:23.638888 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000235594
I0405 09:39:23.638903 26022 sgd_solver.cpp:106] Iteration 3300, lr = 0.039868
I0405 09:42:28.402530 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 42.2207 > 30) by scale factor 0.710552
I0405 09:48:25.733757 26022 solver.cpp:229] Iteration 3350, loss = 0.97706
I0405 09:48:25.733902 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 09:48:25.733922 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 09:48:25.733935 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 09:48:25.733948 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 09:48:25.733960 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.09375
I0405 09:48:25.733973 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 09:48:25.733984 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 09:48:25.733997 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 09:48:25.734009 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 09:48:25.734021 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 09:48:25.734033 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:48:25.734045 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:48:25.734056 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:48:25.734068 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:48:25.734081 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:48:25.734092 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:48:25.734104 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:48:25.734117 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:48:25.734127 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:48:25.734139 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:48:25.734151 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:48:25.734163 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:48:25.734179 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.28152 (* 0.0454545 = 0.14916 loss)
I0405 09:48:25.734194 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.56566 (* 0.0454545 = 0.162076 loss)
I0405 09:48:25.734208 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.68787 (* 0.0454545 = 0.167631 loss)
I0405 09:48:25.734222 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.57901 (* 0.0454545 = 0.162682 loss)
I0405 09:48:25.734237 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.53068 (* 0.0454545 = 0.160485 loss)
I0405 09:48:25.734251 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.76644 (* 0.0454545 = 0.125747 loss)
I0405 09:48:25.734266 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.56951 (* 0.0454545 = 0.0713412 loss)
I0405 09:48:25.734279 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.620139 (* 0.0454545 = 0.0281881 loss)
I0405 09:48:25.734294 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0410177 (* 0.0454545 = 0.00186444 loss)
I0405 09:48:25.734309 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0154256 (* 0.0454545 = 0.000701163 loss)
I0405 09:48:25.734323 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000245895 (* 0.0454545 = 1.1177e-05 loss)
I0405 09:48:25.734338 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000287044 (* 0.0454545 = 1.30474e-05 loss)
I0405 09:48:25.734354 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000256177 (* 0.0454545 = 1.16444e-05 loss)
I0405 09:48:25.734369 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000286516 (* 0.0454545 = 1.30234e-05 loss)
I0405 09:48:25.734383 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000282452 (* 0.0454545 = 1.28387e-05 loss)
I0405 09:48:25.734398 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000299689 (* 0.0454545 = 1.36222e-05 loss)
I0405 09:48:25.734412 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000278652 (* 0.0454545 = 1.2666e-05 loss)
I0405 09:48:25.734443 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000279352 (* 0.0454545 = 1.26978e-05 loss)
I0405 09:48:25.734460 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000257678 (* 0.0454545 = 1.17127e-05 loss)
I0405 09:48:25.734474 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000262799 (* 0.0454545 = 1.19454e-05 loss)
I0405 09:48:25.734489 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000263471 (* 0.0454545 = 1.1976e-05 loss)
I0405 09:48:25.734503 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000303065 (* 0.0454545 = 1.37757e-05 loss)
I0405 09:48:25.734516 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:48:25.734529 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000124314
I0405 09:48:25.734544 26022 sgd_solver.cpp:106] Iteration 3350, lr = 0.039866
I0405 09:57:27.810667 26022 solver.cpp:229] Iteration 3400, loss = 0.962915
I0405 09:57:27.810847 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 09:57:27.810868 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 09:57:27.810881 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 09:57:27.810894 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 09:57:27.810906 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 09:57:27.810919 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 09:57:27.810931 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.84375
I0405 09:57:27.810945 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 09:57:27.810956 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 09:57:27.810968 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 09:57:27.810981 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 09:57:27.810992 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 09:57:27.811004 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 09:57:27.811015 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 09:57:27.811028 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 09:57:27.811039 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 09:57:27.811051 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 09:57:27.811063 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 09:57:27.811074 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 09:57:27.811085 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 09:57:27.811097 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 09:57:27.811108 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 09:57:27.811125 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.2856 (* 0.0454545 = 0.149346 loss)
I0405 09:57:27.811139 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.4416 (* 0.0454545 = 0.156436 loss)
I0405 09:57:27.811156 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.49471 (* 0.0454545 = 0.158851 loss)
I0405 09:57:27.811170 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.31009 (* 0.0454545 = 0.150459 loss)
I0405 09:57:27.811185 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.97538 (* 0.0454545 = 0.135245 loss)
I0405 09:57:27.811199 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.52638 (* 0.0454545 = 0.114835 loss)
I0405 09:57:27.811213 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.798847 (* 0.0454545 = 0.0363112 loss)
I0405 09:57:27.811228 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.429839 (* 0.0454545 = 0.0195382 loss)
I0405 09:57:27.811242 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.28851 (* 0.0454545 = 0.0131141 loss)
I0405 09:57:27.811256 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.310038 (* 0.0454545 = 0.0140926 loss)
I0405 09:57:27.811271 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000200715 (* 0.0454545 = 9.12341e-06 loss)
I0405 09:57:27.811290 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000222354 (* 0.0454545 = 1.0107e-05 loss)
I0405 09:57:27.811305 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000211346 (* 0.0454545 = 9.60665e-06 loss)
I0405 09:57:27.811321 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000226846 (* 0.0454545 = 1.03112e-05 loss)
I0405 09:57:27.811334 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000219472 (* 0.0454545 = 9.97601e-06 loss)
I0405 09:57:27.811350 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000245371 (* 0.0454545 = 1.11532e-05 loss)
I0405 09:57:27.811365 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00022398 (* 0.0454545 = 1.01809e-05 loss)
I0405 09:57:27.811398 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000244454 (* 0.0454545 = 1.11115e-05 loss)
I0405 09:57:27.811414 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000202651 (* 0.0454545 = 9.21139e-06 loss)
I0405 09:57:27.811427 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000209962 (* 0.0454545 = 9.54373e-06 loss)
I0405 09:57:27.811442 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000204096 (* 0.0454545 = 9.27707e-06 loss)
I0405 09:57:27.811457 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000233759 (* 0.0454545 = 1.06254e-05 loss)
I0405 09:57:27.811470 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 09:57:27.811482 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000290527
I0405 09:57:27.811496 26022 sgd_solver.cpp:106] Iteration 3400, lr = 0.039864
I0405 09:58:33.325523 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.1422 > 30) by scale factor 0.963324
I0405 10:06:29.848544 26022 solver.cpp:229] Iteration 3450, loss = 0.964977
I0405 10:06:29.848727 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 10:06:29.848748 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 10:06:29.848762 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 10:06:29.848775 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 10:06:29.848788 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 10:06:29.848799 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 10:06:29.848812 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.53125
I0405 10:06:29.848824 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 10:06:29.848836 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 10:06:29.848848 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 10:06:29.848860 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:06:29.848872 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:06:29.848883 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:06:29.848896 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:06:29.848907 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:06:29.848918 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:06:29.848932 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:06:29.848944 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:06:29.848956 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:06:29.848968 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:06:29.848979 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:06:29.848991 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:06:29.849007 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.04289 (* 0.0454545 = 0.138313 loss)
I0405 10:06:29.849022 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.38953 (* 0.0454545 = 0.154069 loss)
I0405 10:06:29.849037 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.46353 (* 0.0454545 = 0.157433 loss)
I0405 10:06:29.849051 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.3634 (* 0.0454545 = 0.152882 loss)
I0405 10:06:29.849066 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.32611 (* 0.0454545 = 0.151187 loss)
I0405 10:06:29.849079 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.83281 (* 0.0454545 = 0.128764 loss)
I0405 10:06:29.849094 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 2.26083 (* 0.0454545 = 0.102765 loss)
I0405 10:06:29.849107 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.542907 (* 0.0454545 = 0.0246776 loss)
I0405 10:06:29.849123 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.18267 (* 0.0454545 = 0.00830319 loss)
I0405 10:06:29.849138 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0118268 (* 0.0454545 = 0.000537582 loss)
I0405 10:06:29.849153 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 7.74375e-05 (* 0.0454545 = 3.51989e-06 loss)
I0405 10:06:29.849167 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 8.95121e-05 (* 0.0454545 = 4.06873e-06 loss)
I0405 10:06:29.849182 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 8.54101e-05 (* 0.0454545 = 3.88228e-06 loss)
I0405 10:06:29.849196 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 9.20401e-05 (* 0.0454545 = 4.18364e-06 loss)
I0405 10:06:29.849211 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 8.56335e-05 (* 0.0454545 = 3.89243e-06 loss)
I0405 10:06:29.849225 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000106226 (* 0.0454545 = 4.82847e-06 loss)
I0405 10:06:29.849239 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 9.14082e-05 (* 0.0454545 = 4.15492e-06 loss)
I0405 10:06:29.849272 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000103344 (* 0.0454545 = 4.69746e-06 loss)
I0405 10:06:29.849288 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 7.60455e-05 (* 0.0454545 = 3.45661e-06 loss)
I0405 10:06:29.849303 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 8.67713e-05 (* 0.0454545 = 3.94415e-06 loss)
I0405 10:06:29.849318 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 7.9785e-05 (* 0.0454545 = 3.62659e-06 loss)
I0405 10:06:29.849331 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 9.88945e-05 (* 0.0454545 = 4.49521e-06 loss)
I0405 10:06:29.849344 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:06:29.849356 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000250568
I0405 10:06:29.849370 26022 sgd_solver.cpp:106] Iteration 3450, lr = 0.039862
I0405 10:08:18.756674 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.8216 > 30) by scale factor 0.942756
I0405 10:15:32.225406 26022 solver.cpp:229] Iteration 3500, loss = 0.960145
I0405 10:15:32.225582 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 10:15:32.225603 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 10:15:32.225617 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 10:15:32.225630 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 10:15:32.225642 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 10:15:32.225654 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 10:15:32.225667 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 10:15:32.225679 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.8125
I0405 10:15:32.225692 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 10:15:32.225703 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 10:15:32.225716 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:15:32.225728 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:15:32.225739 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:15:32.225751 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:15:32.225762 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:15:32.225775 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:15:32.225785 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:15:32.225797 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:15:32.225812 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:15:32.225824 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:15:32.225836 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:15:32.225847 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:15:32.225863 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.26918 (* 0.0454545 = 0.148599 loss)
I0405 10:15:32.225878 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.56498 (* 0.0454545 = 0.162045 loss)
I0405 10:15:32.225891 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.64576 (* 0.0454545 = 0.165716 loss)
I0405 10:15:32.225906 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.64105 (* 0.0454545 = 0.165502 loss)
I0405 10:15:32.225920 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.61317 (* 0.0454545 = 0.164235 loss)
I0405 10:15:32.225934 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.80155 (* 0.0454545 = 0.127343 loss)
I0405 10:15:32.225949 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.78027 (* 0.0454545 = 0.0809216 loss)
I0405 10:15:32.225963 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 1.16882 (* 0.0454545 = 0.0531283 loss)
I0405 10:15:32.225977 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.539189 (* 0.0454545 = 0.0245086 loss)
I0405 10:15:32.225992 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0051598 (* 0.0454545 = 0.000234536 loss)
I0405 10:15:32.226008 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.35188e-05 (* 0.0454545 = 1.97813e-06 loss)
I0405 10:15:32.226027 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.02274e-05 (* 0.0454545 = 2.28306e-06 loss)
I0405 10:15:32.226042 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.23005e-05 (* 0.0454545 = 1.92275e-06 loss)
I0405 10:15:32.226058 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 4.95791e-05 (* 0.0454545 = 2.25359e-06 loss)
I0405 10:15:32.226071 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.87868e-05 (* 0.0454545 = 2.21758e-06 loss)
I0405 10:15:32.226086 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.5929e-05 (* 0.0454545 = 2.54223e-06 loss)
I0405 10:15:32.226101 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.75629e-05 (* 0.0454545 = 2.16195e-06 loss)
I0405 10:15:32.226133 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.08531e-05 (* 0.0454545 = 2.31151e-06 loss)
I0405 10:15:32.226150 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.98599e-05 (* 0.0454545 = 1.81181e-06 loss)
I0405 10:15:32.226164 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 4.41862e-05 (* 0.0454545 = 2.00846e-06 loss)
I0405 10:15:32.226179 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.28833e-05 (* 0.0454545 = 1.94924e-06 loss)
I0405 10:15:32.226194 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.39407e-05 (* 0.0454545 = 2.45185e-06 loss)
I0405 10:15:32.226207 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:15:32.226218 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000391016
I0405 10:15:32.226233 26022 sgd_solver.cpp:106] Iteration 3500, lr = 0.03986
I0405 10:24:34.391579 26022 solver.cpp:229] Iteration 3550, loss = 0.956689
I0405 10:24:34.391685 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 10:24:34.391707 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 10:24:34.391721 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 10:24:34.391734 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 10:24:34.391746 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.21875
I0405 10:24:34.391758 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 10:24:34.391772 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.59375
I0405 10:24:34.391783 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 10:24:34.391795 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 10:24:34.391808 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 10:24:34.391819 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:24:34.391830 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:24:34.391842 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:24:34.391855 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:24:34.391865 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:24:34.391876 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:24:34.391888 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:24:34.391899 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:24:34.391911 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:24:34.391923 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:24:34.391934 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:24:34.391947 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:24:34.391962 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.9535 (* 0.0454545 = 0.13425 loss)
I0405 10:24:34.391976 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.22622 (* 0.0454545 = 0.146646 loss)
I0405 10:24:34.391990 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.10281 (* 0.0454545 = 0.141037 loss)
I0405 10:24:34.392005 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.01671 (* 0.0454545 = 0.137123 loss)
I0405 10:24:34.392019 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.89914 (* 0.0454545 = 0.131779 loss)
I0405 10:24:34.392033 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.28026 (* 0.0454545 = 0.103648 loss)
I0405 10:24:34.392048 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.70556 (* 0.0454545 = 0.0775256 loss)
I0405 10:24:34.392062 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.732195 (* 0.0454545 = 0.0332816 loss)
I0405 10:24:34.392099 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.416373 (* 0.0454545 = 0.018926 loss)
I0405 10:24:34.392117 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0308643 (* 0.0454545 = 0.00140292 loss)
I0405 10:24:34.392132 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000121772 (* 0.0454545 = 5.5351e-06 loss)
I0405 10:24:34.392146 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000154533 (* 0.0454545 = 7.02424e-06 loss)
I0405 10:24:34.392161 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000130354 (* 0.0454545 = 5.92516e-06 loss)
I0405 10:24:34.392176 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000141099 (* 0.0454545 = 6.41361e-06 loss)
I0405 10:24:34.392191 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000140398 (* 0.0454545 = 6.38175e-06 loss)
I0405 10:24:34.392206 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.0001816 (* 0.0454545 = 8.25457e-06 loss)
I0405 10:24:34.392220 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000140684 (* 0.0454545 = 6.39475e-06 loss)
I0405 10:24:34.392253 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.00015918 (* 0.0454545 = 7.23545e-06 loss)
I0405 10:24:34.392269 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000101717 (* 0.0454545 = 4.62352e-06 loss)
I0405 10:24:34.392283 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.0001279 (* 0.0454545 = 5.81364e-06 loss)
I0405 10:24:34.392298 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.00012026 (* 0.0454545 = 5.46638e-06 loss)
I0405 10:24:34.392313 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000163878 (* 0.0454545 = 7.44901e-06 loss)
I0405 10:24:34.392325 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:24:34.392338 26022 solver.cpp:245]     Train net output #45: total_confidence = 2.8416e-05
I0405 10:24:34.392351 26022 sgd_solver.cpp:106] Iteration 3550, lr = 0.039858
I0405 10:33:36.593175 26022 solver.cpp:229] Iteration 3600, loss = 0.956311
I0405 10:33:36.593323 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 10:33:36.593345 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 10:33:36.593359 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 10:33:36.593372 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 10:33:36.593384 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 10:33:36.593396 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 10:33:36.593410 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 10:33:36.593421 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 10:33:36.593433 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 10:33:36.593446 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 10:33:36.593457 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:33:36.593471 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:33:36.593482 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:33:36.593493 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:33:36.593505 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:33:36.593518 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:33:36.593529 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:33:36.593540 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:33:36.593552 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:33:36.593564 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:33:36.593575 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:33:36.593587 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:33:36.593602 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.97521 (* 0.0454545 = 0.135237 loss)
I0405 10:33:36.593617 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.35736 (* 0.0454545 = 0.152607 loss)
I0405 10:33:36.593631 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.44287 (* 0.0454545 = 0.156494 loss)
I0405 10:33:36.593648 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.36845 (* 0.0454545 = 0.153111 loss)
I0405 10:33:36.593664 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.96784 (* 0.0454545 = 0.134902 loss)
I0405 10:33:36.593678 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.47211 (* 0.0454545 = 0.112368 loss)
I0405 10:33:36.593693 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.932904 (* 0.0454545 = 0.0424047 loss)
I0405 10:33:36.593708 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.522573 (* 0.0454545 = 0.0237533 loss)
I0405 10:33:36.593721 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.203749 (* 0.0454545 = 0.00926132 loss)
I0405 10:33:36.593736 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.23829 (* 0.0454545 = 0.0108314 loss)
I0405 10:33:36.593751 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.97651e-05 (* 0.0454545 = 3.17114e-06 loss)
I0405 10:33:36.593770 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 7.61813e-05 (* 0.0454545 = 3.46279e-06 loss)
I0405 10:33:36.593786 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.58681e-05 (* 0.0454545 = 2.994e-06 loss)
I0405 10:33:36.593801 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 8.09627e-05 (* 0.0454545 = 3.68012e-06 loss)
I0405 10:33:36.593814 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 7.73011e-05 (* 0.0454545 = 3.51369e-06 loss)
I0405 10:33:36.593829 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.44783e-05 (* 0.0454545 = 3.83992e-06 loss)
I0405 10:33:36.593844 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 7.21171e-05 (* 0.0454545 = 3.27805e-06 loss)
I0405 10:33:36.593876 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.92594e-05 (* 0.0454545 = 3.14816e-06 loss)
I0405 10:33:36.593891 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.75105e-05 (* 0.0454545 = 2.61411e-06 loss)
I0405 10:33:36.593906 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 6.54083e-05 (* 0.0454545 = 2.97311e-06 loss)
I0405 10:33:36.593921 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 6.45245e-05 (* 0.0454545 = 2.93293e-06 loss)
I0405 10:33:36.593935 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 7.44308e-05 (* 0.0454545 = 3.38322e-06 loss)
I0405 10:33:36.593947 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:33:36.593961 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000477502
I0405 10:33:36.593976 26022 sgd_solver.cpp:106] Iteration 3600, lr = 0.039856
I0405 10:35:25.451781 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 33.8717 > 30) by scale factor 0.885695
I0405 10:37:03.185272 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.287 > 30) by scale factor 0.990525
I0405 10:42:38.760766 26022 solver.cpp:229] Iteration 3650, loss = 0.955846
I0405 10:42:38.760866 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 10:42:38.760886 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 10:42:38.760900 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 10:42:38.760916 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.375
I0405 10:42:38.760928 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.40625
I0405 10:42:38.760941 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.65625
I0405 10:42:38.760953 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 10:42:38.760965 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.96875
I0405 10:42:38.760977 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 10:42:38.760989 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 10:42:38.761001 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:42:38.761013 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:42:38.761024 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:42:38.761035 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:42:38.761047 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:42:38.761059 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:42:38.761071 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:42:38.761082 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:42:38.761095 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:42:38.761106 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:42:38.761117 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:42:38.761129 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:42:38.761145 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.84774 (* 0.0454545 = 0.129443 loss)
I0405 10:42:38.761159 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.13071 (* 0.0454545 = 0.142305 loss)
I0405 10:42:38.761173 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.14281 (* 0.0454545 = 0.142855 loss)
I0405 10:42:38.761188 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 2.80359 (* 0.0454545 = 0.127436 loss)
I0405 10:42:38.761203 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.4667 (* 0.0454545 = 0.112123 loss)
I0405 10:42:38.761216 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 1.59202 (* 0.0454545 = 0.0723645 loss)
I0405 10:42:38.761230 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.716062 (* 0.0454545 = 0.0325483 loss)
I0405 10:42:38.761245 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.233258 (* 0.0454545 = 0.0106026 loss)
I0405 10:42:38.761260 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.214975 (* 0.0454545 = 0.00977159 loss)
I0405 10:42:38.761275 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0238485 (* 0.0454545 = 0.00108402 loss)
I0405 10:42:38.761289 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.10745e-05 (* 0.0454545 = 5.03388e-07 loss)
I0405 10:42:38.761304 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 9.74436e-06 (* 0.0454545 = 4.42926e-07 loss)
I0405 10:42:38.761319 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 9.31592e-06 (* 0.0454545 = 4.23451e-07 loss)
I0405 10:42:38.761334 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 1.10298e-05 (* 0.0454545 = 5.01354e-07 loss)
I0405 10:42:38.761348 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 9.8263e-06 (* 0.0454545 = 4.4665e-07 loss)
I0405 10:42:38.761363 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 1.03182e-05 (* 0.0454545 = 4.69009e-07 loss)
I0405 10:42:38.761379 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.04374e-05 (* 0.0454545 = 4.74428e-07 loss)
I0405 10:42:38.761410 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.05343e-05 (* 0.0454545 = 4.78833e-07 loss)
I0405 10:42:38.761428 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 9.02156e-06 (* 0.0454545 = 4.10071e-07 loss)
I0405 10:42:38.761445 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 8.9731e-06 (* 0.0454545 = 4.07868e-07 loss)
I0405 10:42:38.761461 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 8.82778e-06 (* 0.0454545 = 4.01263e-07 loss)
I0405 10:42:38.761474 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 9.37552e-06 (* 0.0454545 = 4.2616e-07 loss)
I0405 10:42:38.761487 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:42:38.761499 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.00110054
I0405 10:42:38.761513 26022 sgd_solver.cpp:106] Iteration 3650, lr = 0.039854
I0405 10:51:40.780134 26022 solver.cpp:229] Iteration 3700, loss = 0.949458
I0405 10:51:40.780318 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 10:51:40.780336 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 10:51:40.780350 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 10:51:40.780364 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 10:51:40.780376 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 10:51:40.780388 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 10:51:40.780401 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 10:51:40.780413 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.8125
I0405 10:51:40.780426 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.875
I0405 10:51:40.780438 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 10:51:40.780452 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 10:51:40.780463 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 10:51:40.780475 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 10:51:40.780488 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 10:51:40.780498 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 10:51:40.780510 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 10:51:40.780524 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 10:51:40.780536 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 10:51:40.780549 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 10:51:40.780560 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 10:51:40.780572 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 10:51:40.780585 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 10:51:40.780601 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.30095 (* 0.0454545 = 0.150043 loss)
I0405 10:51:40.780616 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.50092 (* 0.0454545 = 0.159133 loss)
I0405 10:51:40.780629 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.45271 (* 0.0454545 = 0.156941 loss)
I0405 10:51:40.780643 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.35062 (* 0.0454545 = 0.152301 loss)
I0405 10:51:40.780658 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.37354 (* 0.0454545 = 0.153343 loss)
I0405 10:51:40.780673 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.25438 (* 0.0454545 = 0.102472 loss)
I0405 10:51:40.780688 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.53811 (* 0.0454545 = 0.0699143 loss)
I0405 10:51:40.780702 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.851261 (* 0.0454545 = 0.0386937 loss)
I0405 10:51:40.780717 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.772077 (* 0.0454545 = 0.0350944 loss)
I0405 10:51:40.780732 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.24835 (* 0.0454545 = 0.0112887 loss)
I0405 10:51:40.780752 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.81264e-05 (* 0.0454545 = 8.23928e-07 loss)
I0405 10:51:40.780767 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.03341e-05 (* 0.0454545 = 9.24278e-07 loss)
I0405 10:51:40.780782 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.71914e-05 (* 0.0454545 = 7.81428e-07 loss)
I0405 10:51:40.780797 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.01664e-05 (* 0.0454545 = 9.16655e-07 loss)
I0405 10:51:40.780812 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.98163e-05 (* 0.0454545 = 9.0074e-07 loss)
I0405 10:51:40.780827 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.3218e-05 (* 0.0454545 = 1.05536e-06 loss)
I0405 10:51:40.780843 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.91362e-05 (* 0.0454545 = 8.69829e-07 loss)
I0405 10:51:40.780874 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.93375e-05 (* 0.0454545 = 8.78976e-07 loss)
I0405 10:51:40.780890 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.4846e-05 (* 0.0454545 = 6.74816e-07 loss)
I0405 10:51:40.780905 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.72025e-05 (* 0.0454545 = 7.81934e-07 loss)
I0405 10:51:40.780920 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.74075e-05 (* 0.0454545 = 7.91251e-07 loss)
I0405 10:51:40.780935 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.98499e-05 (* 0.0454545 = 9.02268e-07 loss)
I0405 10:51:40.780947 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 10:51:40.780959 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000203318
I0405 10:51:40.780974 26022 sgd_solver.cpp:106] Iteration 3700, lr = 0.039852
I0405 10:56:12.325989 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 37.4144 > 30) by scale factor 0.80183
I0405 10:56:55.700337 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.0872 > 30) by scale factor 0.965029
I0405 11:00:42.867581 26022 solver.cpp:229] Iteration 3750, loss = 0.95633
I0405 11:00:42.867707 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 11:00:42.867727 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 11:00:42.867740 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:00:42.867754 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 11:00:42.867768 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 11:00:42.867779 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 11:00:42.867791 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 11:00:42.867805 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 11:00:42.867816 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 11:00:42.867828 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 11:00:42.867841 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:00:42.867851 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:00:42.867863 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:00:42.867876 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:00:42.867887 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:00:42.867898 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:00:42.867910 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:00:42.867923 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:00:42.867934 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:00:42.867945 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:00:42.867957 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:00:42.867969 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:00:42.867985 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.89251 (* 0.0454545 = 0.131478 loss)
I0405 11:00:42.868000 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.24245 (* 0.0454545 = 0.147384 loss)
I0405 11:00:42.868013 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.30358 (* 0.0454545 = 0.150163 loss)
I0405 11:00:42.868028 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.25144 (* 0.0454545 = 0.147793 loss)
I0405 11:00:42.868042 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.96651 (* 0.0454545 = 0.134841 loss)
I0405 11:00:42.868057 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.04976 (* 0.0454545 = 0.0931707 loss)
I0405 11:00:42.868085 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.30575 (* 0.0454545 = 0.0593521 loss)
I0405 11:00:42.868103 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.39669 (* 0.0454545 = 0.0180314 loss)
I0405 11:00:42.868118 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.255106 (* 0.0454545 = 0.0115957 loss)
I0405 11:00:42.868132 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00417592 (* 0.0454545 = 0.000189814 loss)
I0405 11:00:42.868149 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 4.71108e-05 (* 0.0454545 = 2.1414e-06 loss)
I0405 11:00:42.868165 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.27118e-05 (* 0.0454545 = 2.85053e-06 loss)
I0405 11:00:42.868180 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.03374e-05 (* 0.0454545 = 2.28806e-06 loss)
I0405 11:00:42.868194 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.69425e-05 (* 0.0454545 = 2.58829e-06 loss)
I0405 11:00:42.868208 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.47043e-05 (* 0.0454545 = 2.48656e-06 loss)
I0405 11:00:42.868223 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.15314e-05 (* 0.0454545 = 3.70597e-06 loss)
I0405 11:00:42.868237 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.32187e-05 (* 0.0454545 = 2.41903e-06 loss)
I0405 11:00:42.868271 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.40544e-05 (* 0.0454545 = 2.91157e-06 loss)
I0405 11:00:42.868288 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 3.62224e-05 (* 0.0454545 = 1.64647e-06 loss)
I0405 11:00:42.868301 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.12376e-05 (* 0.0454545 = 2.32898e-06 loss)
I0405 11:00:42.868316 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.76591e-05 (* 0.0454545 = 2.16632e-06 loss)
I0405 11:00:42.868330 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 6.32027e-05 (* 0.0454545 = 2.87285e-06 loss)
I0405 11:00:42.868343 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:00:42.868355 26022 solver.cpp:245]     Train net output #45: total_confidence = 9.14751e-05
I0405 11:00:42.868369 26022 sgd_solver.cpp:106] Iteration 3750, lr = 0.03985
I0405 11:04:31.010694 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 38.1985 > 30) by scale factor 0.785372
I0405 11:09:44.953629 26022 solver.cpp:229] Iteration 3800, loss = 0.944967
I0405 11:09:44.953794 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 11:09:44.953824 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 11:09:44.953850 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 11:09:44.953876 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 11:09:44.953897 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.15625
I0405 11:09:44.953919 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 11:09:44.953943 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.8125
I0405 11:09:44.953966 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 11:09:44.953987 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 11:09:44.954010 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 11:09:44.954032 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:09:44.954056 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:09:44.954077 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:09:44.954103 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:09:44.954125 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:09:44.954147 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:09:44.954170 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:09:44.954195 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:09:44.954216 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:09:44.954237 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:09:44.954262 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:09:44.954284 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:09:44.954313 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.08297 (* 0.0454545 = 0.140135 loss)
I0405 11:09:44.954339 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.29308 (* 0.0454545 = 0.149685 loss)
I0405 11:09:44.954367 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.34999 (* 0.0454545 = 0.152272 loss)
I0405 11:09:44.954393 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.45689 (* 0.0454545 = 0.157131 loss)
I0405 11:09:44.954421 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.22818 (* 0.0454545 = 0.146735 loss)
I0405 11:09:44.954447 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.48101 (* 0.0454545 = 0.112773 loss)
I0405 11:09:44.954471 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 0.88543 (* 0.0454545 = 0.0402468 loss)
I0405 11:09:44.954498 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.457628 (* 0.0454545 = 0.0208013 loss)
I0405 11:09:44.954522 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.143984 (* 0.0454545 = 0.00654473 loss)
I0405 11:09:44.954550 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.151717 (* 0.0454545 = 0.00689625 loss)
I0405 11:09:44.954577 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 5.33473e-06 (* 0.0454545 = 2.42488e-07 loss)
I0405 11:09:44.954604 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.0122e-06 (* 0.0454545 = 1.82373e-07 loss)
I0405 11:09:44.954630 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 3.30811e-06 (* 0.0454545 = 1.50368e-07 loss)
I0405 11:09:44.954658 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.87437e-06 (* 0.0454545 = 1.76108e-07 loss)
I0405 11:09:44.954682 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 3.74025e-06 (* 0.0454545 = 1.70011e-07 loss)
I0405 11:09:44.954710 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 4.60455e-06 (* 0.0454545 = 2.09298e-07 loss)
I0405 11:09:44.954735 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.23201e-06 (* 0.0454545 = 1.92364e-07 loss)
I0405 11:09:44.954782 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 3.69182e-06 (* 0.0454545 = 1.6781e-07 loss)
I0405 11:09:44.954810 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 4.9175e-06 (* 0.0454545 = 2.23523e-07 loss)
I0405 11:09:44.954838 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 3.58378e-06 (* 0.0454545 = 1.62899e-07 loss)
I0405 11:09:44.954865 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 3.90044e-06 (* 0.0454545 = 1.77293e-07 loss)
I0405 11:09:44.954891 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.56516e-06 (* 0.0454545 = 1.62053e-07 loss)
I0405 11:09:44.954913 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:09:44.954934 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000174833
I0405 11:09:44.954957 26022 sgd_solver.cpp:106] Iteration 3800, lr = 0.039848
I0405 11:18:47.272553 26022 solver.cpp:229] Iteration 3850, loss = 0.937853
I0405 11:18:47.272794 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.21875
I0405 11:18:47.272815 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 11:18:47.272828 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:18:47.272841 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 11:18:47.272855 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 11:18:47.272866 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 11:18:47.272878 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 11:18:47.272891 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 11:18:47.272902 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 11:18:47.272914 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 11:18:47.272927 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:18:47.272938 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:18:47.272950 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:18:47.272961 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:18:47.272974 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:18:47.272985 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:18:47.272996 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:18:47.273010 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:18:47.273022 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:18:47.273035 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:18:47.273046 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:18:47.273057 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:18:47.273073 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.99016 (* 0.0454545 = 0.135917 loss)
I0405 11:18:47.273088 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.27905 (* 0.0454545 = 0.149048 loss)
I0405 11:18:47.273102 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.34622 (* 0.0454545 = 0.152101 loss)
I0405 11:18:47.273116 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.29326 (* 0.0454545 = 0.149694 loss)
I0405 11:18:47.273130 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.96725 (* 0.0454545 = 0.134875 loss)
I0405 11:18:47.273144 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.3763 (* 0.0454545 = 0.108014 loss)
I0405 11:18:47.273159 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.28541 (* 0.0454545 = 0.0584279 loss)
I0405 11:18:47.273174 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.72519 (* 0.0454545 = 0.0329632 loss)
I0405 11:18:47.273187 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.546886 (* 0.0454545 = 0.0248584 loss)
I0405 11:18:47.273201 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.261067 (* 0.0454545 = 0.0118667 loss)
I0405 11:18:47.273217 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.66706e-05 (* 0.0454545 = 1.2123e-06 loss)
I0405 11:18:47.273232 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.38971e-05 (* 0.0454545 = 1.08623e-06 loss)
I0405 11:18:47.273247 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.24346e-05 (* 0.0454545 = 1.01975e-06 loss)
I0405 11:18:47.273262 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.46274e-05 (* 0.0454545 = 1.11943e-06 loss)
I0405 11:18:47.273277 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.36661e-05 (* 0.0454545 = 1.07573e-06 loss)
I0405 11:18:47.273290 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.37663e-05 (* 0.0454545 = 1.08029e-06 loss)
I0405 11:18:47.273316 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.14581e-05 (* 0.0454545 = 9.75367e-07 loss)
I0405 11:18:47.273346 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.05323e-05 (* 0.0454545 = 9.33285e-07 loss)
I0405 11:18:47.273362 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.12491e-05 (* 0.0454545 = 9.65867e-07 loss)
I0405 11:18:47.273377 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.04651e-05 (* 0.0454545 = 9.3023e-07 loss)
I0405 11:18:47.273391 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.11581e-05 (* 0.0454545 = 9.61732e-07 loss)
I0405 11:18:47.273407 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.2064e-05 (* 0.0454545 = 1.00291e-06 loss)
I0405 11:18:47.273421 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:18:47.273432 26022 solver.cpp:245]     Train net output #45: total_confidence = 9.24704e-05
I0405 11:18:47.273447 26022 sgd_solver.cpp:106] Iteration 3850, lr = 0.039846
I0405 11:27:28.127835 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 53.0881 > 30) by scale factor 0.565098
I0405 11:27:49.321316 26022 solver.cpp:229] Iteration 3900, loss = 0.936974
I0405 11:27:49.321370 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 11:27:49.321389 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 11:27:49.321403 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:27:49.321414 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 11:27:49.321427 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 11:27:49.321440 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 11:27:49.321452 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 11:27:49.321465 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 11:27:49.321476 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 11:27:49.321488 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 11:27:49.321501 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:27:49.321512 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:27:49.321524 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:27:49.321537 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:27:49.321548 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:27:49.321559 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:27:49.321571 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:27:49.321586 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:27:49.321599 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:27:49.321610 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:27:49.321621 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:27:49.321633 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:27:49.321650 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.01139 (* 0.0454545 = 0.136882 loss)
I0405 11:27:49.321665 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.03971 (* 0.0454545 = 0.138169 loss)
I0405 11:27:49.321678 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.27191 (* 0.0454545 = 0.148723 loss)
I0405 11:27:49.321692 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.23032 (* 0.0454545 = 0.146833 loss)
I0405 11:27:49.321707 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.38716 (* 0.0454545 = 0.153962 loss)
I0405 11:27:49.321720 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.56138 (* 0.0454545 = 0.116426 loss)
I0405 11:27:49.321734 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.44219 (* 0.0454545 = 0.0655542 loss)
I0405 11:27:49.321748 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.716164 (* 0.0454545 = 0.0325529 loss)
I0405 11:27:49.321763 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.27941 (* 0.0454545 = 0.0127005 loss)
I0405 11:27:49.321777 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0466797 (* 0.0454545 = 0.0021218 loss)
I0405 11:27:49.321792 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000222299 (* 0.0454545 = 1.01045e-05 loss)
I0405 11:27:49.321807 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000255162 (* 0.0454545 = 1.15983e-05 loss)
I0405 11:27:49.321822 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000231223 (* 0.0454545 = 1.05102e-05 loss)
I0405 11:27:49.321838 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000227261 (* 0.0454545 = 1.033e-05 loss)
I0405 11:27:49.321853 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000212654 (* 0.0454545 = 9.66607e-06 loss)
I0405 11:27:49.321867 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000294159 (* 0.0454545 = 1.33709e-05 loss)
I0405 11:27:49.321882 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.000238177 (* 0.0454545 = 1.08262e-05 loss)
I0405 11:27:49.321926 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000260077 (* 0.0454545 = 1.18217e-05 loss)
I0405 11:27:49.321943 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.00015503 (* 0.0454545 = 7.04681e-06 loss)
I0405 11:27:49.321956 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000219361 (* 0.0454545 = 9.97097e-06 loss)
I0405 11:27:49.321971 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000182378 (* 0.0454545 = 8.28993e-06 loss)
I0405 11:27:49.321990 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000245092 (* 0.0454545 = 1.11406e-05 loss)
I0405 11:27:49.322003 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:27:49.322016 26022 solver.cpp:245]     Train net output #45: total_confidence = 9.47889e-05
I0405 11:27:49.322031 26022 sgd_solver.cpp:106] Iteration 3900, lr = 0.039844
I0405 11:28:22.324827 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.2085 > 30) by scale factor 0.961277
I0405 11:32:20.855007 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 32.5775 > 30) by scale factor 0.92088
I0405 11:32:53.372162 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 37.725 > 30) by scale factor 0.795228
I0405 11:36:51.523546 26022 solver.cpp:229] Iteration 3950, loss = 0.944202
I0405 11:36:51.523743 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.25
I0405 11:36:51.523766 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 11:36:51.523778 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:36:51.523792 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 11:36:51.523803 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 11:36:51.523815 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 11:36:51.523828 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 11:36:51.523840 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 11:36:51.523852 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 11:36:51.523865 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 11:36:51.523877 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:36:51.523890 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:36:51.523901 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:36:51.523912 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:36:51.523924 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:36:51.523936 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:36:51.523947 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:36:51.523959 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:36:51.523972 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:36:51.523983 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:36:51.523994 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:36:51.524006 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:36:51.524022 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.96695 (* 0.0454545 = 0.134861 loss)
I0405 11:36:51.524037 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.25114 (* 0.0454545 = 0.147779 loss)
I0405 11:36:51.524051 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.30956 (* 0.0454545 = 0.150435 loss)
I0405 11:36:51.524065 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.14178 (* 0.0454545 = 0.142808 loss)
I0405 11:36:51.524099 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.05489 (* 0.0454545 = 0.138859 loss)
I0405 11:36:51.524114 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.82948 (* 0.0454545 = 0.128613 loss)
I0405 11:36:51.524129 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.26084 (* 0.0454545 = 0.0573107 loss)
I0405 11:36:51.524143 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.470753 (* 0.0454545 = 0.0213979 loss)
I0405 11:36:51.524158 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.247732 (* 0.0454545 = 0.0112605 loss)
I0405 11:36:51.524173 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0233278 (* 0.0454545 = 0.00106035 loss)
I0405 11:36:51.524188 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.78551e-05 (* 0.0454545 = 1.26614e-06 loss)
I0405 11:36:51.524202 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.07198e-05 (* 0.0454545 = 9.4181e-07 loss)
I0405 11:36:51.524217 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.92206e-05 (* 0.0454545 = 8.73662e-07 loss)
I0405 11:36:51.524232 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.09359e-05 (* 0.0454545 = 9.51633e-07 loss)
I0405 11:36:51.524246 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.91176e-05 (* 0.0454545 = 8.68984e-07 loss)
I0405 11:36:51.524261 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.23872e-05 (* 0.0454545 = 1.0176e-06 loss)
I0405 11:36:51.524276 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.13478e-05 (* 0.0454545 = 9.70356e-07 loss)
I0405 11:36:51.524309 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.93509e-05 (* 0.0454545 = 8.79588e-07 loss)
I0405 11:36:51.524325 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.34492e-05 (* 0.0454545 = 1.06587e-06 loss)
I0405 11:36:51.524340 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.91439e-05 (* 0.0454545 = 8.70176e-07 loss)
I0405 11:36:51.524355 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.96393e-05 (* 0.0454545 = 8.92697e-07 loss)
I0405 11:36:51.524372 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.82758e-05 (* 0.0454545 = 8.30718e-07 loss)
I0405 11:36:51.524386 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:36:51.524399 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.00050667
I0405 11:36:51.524413 26022 sgd_solver.cpp:106] Iteration 3950, lr = 0.039842
I0405 11:45:43.216516 26022 solver.cpp:338] Iteration 4000, Testing net (#0)
I0405 11:45:56.867869 26022 solver.cpp:393] Test loss: 0.820289
I0405 11:45:56.867916 26022 solver.cpp:406]     Test net output #0: loss/accuracy01 = 0.284
I0405 11:45:56.867933 26022 solver.cpp:406]     Test net output #1: loss/accuracy02 = 0.067
I0405 11:45:56.867945 26022 solver.cpp:406]     Test net output #2: loss/accuracy03 = 0.089
I0405 11:45:56.867959 26022 solver.cpp:406]     Test net output #3: loss/accuracy04 = 0.136
I0405 11:45:56.867970 26022 solver.cpp:406]     Test net output #4: loss/accuracy05 = 0.227
I0405 11:45:56.867985 26022 solver.cpp:406]     Test net output #5: loss/accuracy06 = 0.5
I0405 11:45:56.867997 26022 solver.cpp:406]     Test net output #6: loss/accuracy07 = 0.894
I0405 11:45:56.868010 26022 solver.cpp:406]     Test net output #7: loss/accuracy08 = 0.97
I0405 11:45:56.868021 26022 solver.cpp:406]     Test net output #8: loss/accuracy09 = 0.995
I0405 11:45:56.868034 26022 solver.cpp:406]     Test net output #9: loss/accuracy10 = 0.998
I0405 11:45:56.868046 26022 solver.cpp:406]     Test net output #10: loss/accuracy11 = 1
I0405 11:45:56.868058 26022 solver.cpp:406]     Test net output #11: loss/accuracy12 = 1
I0405 11:45:56.868085 26022 solver.cpp:406]     Test net output #12: loss/accuracy13 = 1
I0405 11:45:56.868103 26022 solver.cpp:406]     Test net output #13: loss/accuracy14 = 1
I0405 11:45:56.868114 26022 solver.cpp:406]     Test net output #14: loss/accuracy15 = 1
I0405 11:45:56.868126 26022 solver.cpp:406]     Test net output #15: loss/accuracy16 = 1
I0405 11:45:56.868137 26022 solver.cpp:406]     Test net output #16: loss/accuracy17 = 1
I0405 11:45:56.868149 26022 solver.cpp:406]     Test net output #17: loss/accuracy18 = 1
I0405 11:45:56.868161 26022 solver.cpp:406]     Test net output #18: loss/accuracy19 = 1
I0405 11:45:56.868172 26022 solver.cpp:406]     Test net output #19: loss/accuracy20 = 1
I0405 11:45:56.868185 26022 solver.cpp:406]     Test net output #20: loss/accuracy21 = 1
I0405 11:45:56.868196 26022 solver.cpp:406]     Test net output #21: loss/accuracy22 = 1
I0405 11:45:56.868211 26022 solver.cpp:406]     Test net output #22: loss/loss01 = 2.77252 (* 0.0454545 = 0.126023 loss)
I0405 11:45:56.868227 26022 solver.cpp:406]     Test net output #23: loss/loss02 = 3.01626 (* 0.0454545 = 0.137103 loss)
I0405 11:45:56.868240 26022 solver.cpp:406]     Test net output #24: loss/loss03 = 3.09765 (* 0.0454545 = 0.140802 loss)
I0405 11:45:56.868254 26022 solver.cpp:406]     Test net output #25: loss/loss04 = 3.02521 (* 0.0454545 = 0.13751 loss)
I0405 11:45:56.868268 26022 solver.cpp:406]     Test net output #26: loss/loss05 = 2.90839 (* 0.0454545 = 0.132199 loss)
I0405 11:45:56.868283 26022 solver.cpp:406]     Test net output #27: loss/loss06 = 2.1651 (* 0.0454545 = 0.0984135 loss)
I0405 11:45:56.868297 26022 solver.cpp:406]     Test net output #28: loss/loss07 = 0.743218 (* 0.0454545 = 0.0337826 loss)
I0405 11:45:56.868312 26022 solver.cpp:406]     Test net output #29: loss/loss08 = 0.239143 (* 0.0454545 = 0.0108701 loss)
I0405 11:45:56.868327 26022 solver.cpp:406]     Test net output #30: loss/loss09 = 0.051947 (* 0.0454545 = 0.00236123 loss)
I0405 11:45:56.868341 26022 solver.cpp:406]     Test net output #31: loss/loss10 = 0.0268643 (* 0.0454545 = 0.0012211 loss)
I0405 11:45:56.868356 26022 solver.cpp:406]     Test net output #32: loss/loss11 = 7.68983e-06 (* 0.0454545 = 3.49538e-07 loss)
I0405 11:45:56.868371 26022 solver.cpp:406]     Test net output #33: loss/loss12 = 5.86776e-06 (* 0.0454545 = 2.66716e-07 loss)
I0405 11:45:56.868386 26022 solver.cpp:406]     Test net output #34: loss/loss13 = 5.17357e-06 (* 0.0454545 = 2.35162e-07 loss)
I0405 11:45:56.868401 26022 solver.cpp:406]     Test net output #35: loss/loss14 = 6.56084e-06 (* 0.0454545 = 2.9822e-07 loss)
I0405 11:45:56.868415 26022 solver.cpp:406]     Test net output #36: loss/loss15 = 5.40185e-06 (* 0.0454545 = 2.45539e-07 loss)
I0405 11:45:56.868430 26022 solver.cpp:406]     Test net output #37: loss/loss16 = 5.09332e-06 (* 0.0454545 = 2.31514e-07 loss)
I0405 11:45:56.868444 26022 solver.cpp:406]     Test net output #38: loss/loss17 = 5.63912e-06 (* 0.0454545 = 2.56323e-07 loss)
I0405 11:45:56.868489 26022 solver.cpp:406]     Test net output #39: loss/loss18 = 5.3608e-06 (* 0.0454545 = 2.43673e-07 loss)
I0405 11:45:56.868505 26022 solver.cpp:406]     Test net output #40: loss/loss19 = 5.15324e-06 (* 0.0454545 = 2.34238e-07 loss)
I0405 11:45:56.868520 26022 solver.cpp:406]     Test net output #41: loss/loss20 = 4.88949e-06 (* 0.0454545 = 2.2225e-07 loss)
I0405 11:45:56.868535 26022 solver.cpp:406]     Test net output #42: loss/loss21 = 4.95463e-06 (* 0.0454545 = 2.2521e-07 loss)
I0405 11:45:56.868549 26022 solver.cpp:406]     Test net output #43: loss/loss22 = 5.15949e-06 (* 0.0454545 = 2.34522e-07 loss)
I0405 11:45:56.868561 26022 solver.cpp:406]     Test net output #44: total_accuracy = 0.001
I0405 11:45:56.868573 26022 solver.cpp:406]     Test net output #45: total_confidence = 0.000169457
I0405 11:46:07.179795 26022 solver.cpp:229] Iteration 4000, loss = 0.934202
I0405 11:46:07.179843 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 11:46:07.179862 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 11:46:07.179874 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:46:07.179888 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.03125
I0405 11:46:07.179900 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.3125
I0405 11:46:07.179913 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 11:46:07.179924 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 11:46:07.179936 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 11:46:07.179949 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 11:46:07.179960 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 11:46:07.179975 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:46:07.179986 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:46:07.179997 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:46:07.180012 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:46:07.180024 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:46:07.180037 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:46:07.180047 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:46:07.180060 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:46:07.180093 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:46:07.180107 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:46:07.180119 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:46:07.180131 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:46:07.180146 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.03965 (* 0.0454545 = 0.138166 loss)
I0405 11:46:07.180160 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.27816 (* 0.0454545 = 0.149007 loss)
I0405 11:46:07.180176 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.48663 (* 0.0454545 = 0.158483 loss)
I0405 11:46:07.180191 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.43542 (* 0.0454545 = 0.156156 loss)
I0405 11:46:07.180204 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.86348 (* 0.0454545 = 0.130158 loss)
I0405 11:46:07.180218 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.54219 (* 0.0454545 = 0.115554 loss)
I0405 11:46:07.180233 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.44292 (* 0.0454545 = 0.0655871 loss)
I0405 11:46:07.180246 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.407333 (* 0.0454545 = 0.0185151 loss)
I0405 11:46:07.180261 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.386273 (* 0.0454545 = 0.0175578 loss)
I0405 11:46:07.180305 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.389169 (* 0.0454545 = 0.0176895 loss)
I0405 11:46:07.180320 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.56514e-06 (* 0.0454545 = 1.62052e-07 loss)
I0405 11:46:07.180336 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.3842e-06 (* 0.0454545 = 1.08373e-07 loss)
I0405 11:46:07.180351 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.075e-06 (* 0.0454545 = 9.43181e-08 loss)
I0405 11:46:07.180366 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.59655e-06 (* 0.0454545 = 1.18025e-07 loss)
I0405 11:46:07.180379 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.32087e-06 (* 0.0454545 = 1.05494e-07 loss)
I0405 11:46:07.180394 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.71948e-06 (* 0.0454545 = 1.23613e-07 loss)
I0405 11:46:07.180408 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.30969e-06 (* 0.0454545 = 1.04986e-07 loss)
I0405 11:46:07.180424 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.23891e-06 (* 0.0454545 = 1.01769e-07 loss)
I0405 11:46:07.180441 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.79399e-06 (* 0.0454545 = 1.26999e-07 loss)
I0405 11:46:07.180457 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.23891e-06 (* 0.0454545 = 1.01769e-07 loss)
I0405 11:46:07.180472 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.45498e-06 (* 0.0454545 = 1.1159e-07 loss)
I0405 11:46:07.180486 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.14578e-06 (* 0.0454545 = 9.75355e-08 loss)
I0405 11:46:07.180498 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:46:07.180510 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000213462
I0405 11:46:07.180526 26022 sgd_solver.cpp:106] Iteration 4000, lr = 0.03984
I0405 11:55:09.235816 26022 solver.cpp:229] Iteration 4050, loss = 0.936761
I0405 11:55:09.235936 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 11:55:09.235957 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 11:55:09.235971 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 11:55:09.235985 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.21875
I0405 11:55:09.235997 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.34375
I0405 11:55:09.236009 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 11:55:09.236024 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 11:55:09.236037 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 11:55:09.236049 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 11:55:09.236063 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 11:55:09.236099 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 11:55:09.236124 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 11:55:09.236137 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 11:55:09.236148 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 11:55:09.236160 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 11:55:09.236171 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 11:55:09.236184 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 11:55:09.236196 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 11:55:09.236207 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 11:55:09.236218 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 11:55:09.236229 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 11:55:09.236240 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 11:55:09.236256 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.11458 (* 0.0454545 = 0.141572 loss)
I0405 11:55:09.236270 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.15351 (* 0.0454545 = 0.143342 loss)
I0405 11:55:09.236285 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.20078 (* 0.0454545 = 0.14549 loss)
I0405 11:55:09.236299 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.06874 (* 0.0454545 = 0.139488 loss)
I0405 11:55:09.236313 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.60897 (* 0.0454545 = 0.11859 loss)
I0405 11:55:09.236328 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.16617 (* 0.0454545 = 0.0984622 loss)
I0405 11:55:09.236342 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.23968 (* 0.0454545 = 0.056349 loss)
I0405 11:55:09.236356 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.715429 (* 0.0454545 = 0.0325195 loss)
I0405 11:55:09.236371 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.164382 (* 0.0454545 = 0.00747191 loss)
I0405 11:55:09.236384 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.171088 (* 0.0454545 = 0.00777672 loss)
I0405 11:55:09.236399 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.76797e-05 (* 0.0454545 = 8.03622e-07 loss)
I0405 11:55:09.236413 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 1.70335e-05 (* 0.0454545 = 7.7425e-07 loss)
I0405 11:55:09.236428 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.41497e-05 (* 0.0454545 = 6.43167e-07 loss)
I0405 11:55:09.236443 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 1.45949e-05 (* 0.0454545 = 6.63403e-07 loss)
I0405 11:55:09.236457 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.35087e-05 (* 0.0454545 = 6.14033e-07 loss)
I0405 11:55:09.236472 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 1.82946e-05 (* 0.0454545 = 8.31573e-07 loss)
I0405 11:55:09.236486 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.49525e-05 (* 0.0454545 = 6.79661e-07 loss)
I0405 11:55:09.236521 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.50737e-05 (* 0.0454545 = 6.85167e-07 loss)
I0405 11:55:09.236536 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.40525e-05 (* 0.0454545 = 6.38752e-07 loss)
I0405 11:55:09.236552 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.38366e-05 (* 0.0454545 = 6.28936e-07 loss)
I0405 11:55:09.236565 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.30951e-05 (* 0.0454545 = 5.95233e-07 loss)
I0405 11:55:09.236580 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.41515e-05 (* 0.0454545 = 6.4325e-07 loss)
I0405 11:55:09.236593 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 11:55:09.236604 26022 solver.cpp:245]     Train net output #45: total_confidence = 3.9601e-05
I0405 11:55:09.236619 26022 sgd_solver.cpp:106] Iteration 4050, lr = 0.039838
I0405 12:04:11.472285 26022 solver.cpp:229] Iteration 4100, loss = 0.930841
I0405 12:04:11.472473 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 12:04:11.472494 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 12:04:11.472507 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.15625
I0405 12:04:11.472520 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 12:04:11.472532 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.21875
I0405 12:04:11.472545 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 12:04:11.472558 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 12:04:11.472569 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.78125
I0405 12:04:11.472581 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 12:04:11.472594 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 12:04:11.472605 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:04:11.472617 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:04:11.472628 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:04:11.472640 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:04:11.472652 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:04:11.472663 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:04:11.472676 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:04:11.472687 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:04:11.472698 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:04:11.472710 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:04:11.472721 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:04:11.472733 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:04:11.472748 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.96774 (* 0.0454545 = 0.134897 loss)
I0405 12:04:11.472764 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.20847 (* 0.0454545 = 0.14584 loss)
I0405 12:04:11.472779 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.16788 (* 0.0454545 = 0.143994 loss)
I0405 12:04:11.472792 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.03697 (* 0.0454545 = 0.138044 loss)
I0405 12:04:11.472806 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.87908 (* 0.0454545 = 0.130867 loss)
I0405 12:04:11.472820 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.31877 (* 0.0454545 = 0.105399 loss)
I0405 12:04:11.472834 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.40923 (* 0.0454545 = 0.0640558 loss)
I0405 12:04:11.472849 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.945133 (* 0.0454545 = 0.0429606 loss)
I0405 12:04:11.472863 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.42097 (* 0.0454545 = 0.019135 loss)
I0405 12:04:11.472878 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.321204 (* 0.0454545 = 0.0146002 loss)
I0405 12:04:11.472894 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000646828 (* 0.0454545 = 2.94013e-05 loss)
I0405 12:04:11.472910 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000555332 (* 0.0454545 = 2.52424e-05 loss)
I0405 12:04:11.472925 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000594134 (* 0.0454545 = 2.70061e-05 loss)
I0405 12:04:11.472940 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000553366 (* 0.0454545 = 2.5153e-05 loss)
I0405 12:04:11.472954 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000482425 (* 0.0454545 = 2.19284e-05 loss)
I0405 12:04:11.472968 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.00053245 (* 0.0454545 = 2.42023e-05 loss)
I0405 12:04:11.472983 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 0.00057707 (* 0.0454545 = 2.62304e-05 loss)
I0405 12:04:11.473016 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 0.000592736 (* 0.0454545 = 2.69425e-05 loss)
I0405 12:04:11.473031 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 0.000482738 (* 0.0454545 = 2.19427e-05 loss)
I0405 12:04:11.473047 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 0.000545252 (* 0.0454545 = 2.47842e-05 loss)
I0405 12:04:11.473060 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 0.000487242 (* 0.0454545 = 2.21474e-05 loss)
I0405 12:04:11.473074 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 0.000535608 (* 0.0454545 = 2.43458e-05 loss)
I0405 12:04:11.473088 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:04:11.473099 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000232486
I0405 12:04:11.473114 26022 sgd_solver.cpp:106] Iteration 4100, lr = 0.039836
I0405 12:09:04.782937 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.9691 > 30) by scale factor 0.938405
I0405 12:13:13.621439 26022 solver.cpp:229] Iteration 4150, loss = 0.930428
I0405 12:13:13.621541 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 12:13:13.621562 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 12:13:13.621574 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 12:13:13.621587 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.125
I0405 12:13:13.621599 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 12:13:13.621611 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.4375
I0405 12:13:13.621623 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 12:13:13.621635 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 12:13:13.621647 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 12:13:13.621659 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 12:13:13.621671 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:13:13.621683 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:13:13.621695 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:13:13.621706 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:13:13.621717 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:13:13.621728 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:13:13.621740 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:13:13.621752 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:13:13.621762 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:13:13.621773 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:13:13.621784 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:13:13.621796 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:13:13.621811 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.08193 (* 0.0454545 = 0.140088 loss)
I0405 12:13:13.621826 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.40348 (* 0.0454545 = 0.154704 loss)
I0405 12:13:13.621840 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.25422 (* 0.0454545 = 0.147919 loss)
I0405 12:13:13.621853 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.05986 (* 0.0454545 = 0.139085 loss)
I0405 12:13:13.621870 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.83245 (* 0.0454545 = 0.128748 loss)
I0405 12:13:13.621886 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.38745 (* 0.0454545 = 0.108521 loss)
I0405 12:13:13.621899 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.32328 (* 0.0454545 = 0.0601489 loss)
I0405 12:13:13.621913 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.617075 (* 0.0454545 = 0.0280489 loss)
I0405 12:13:13.621927 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.261549 (* 0.0454545 = 0.0118886 loss)
I0405 12:13:13.621942 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.158553 (* 0.0454545 = 0.00720696 loss)
I0405 12:13:13.621955 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.31695e-05 (* 0.0454545 = 5.98614e-07 loss)
I0405 12:13:13.621970 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 1.65941e-05 (* 0.0454545 = 7.54275e-07 loss)
I0405 12:13:13.621984 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.29444e-05 (* 0.0454545 = 5.88384e-07 loss)
I0405 12:13:13.621999 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 1.27506e-05 (* 0.0454545 = 5.79572e-07 loss)
I0405 12:13:13.622012 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.28698e-05 (* 0.0454545 = 5.84991e-07 loss)
I0405 12:13:13.622026 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.04134e-05 (* 0.0454545 = 9.27883e-07 loss)
I0405 12:13:13.622041 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.25754e-05 (* 0.0454545 = 5.71611e-07 loss)
I0405 12:13:13.622073 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.57856e-05 (* 0.0454545 = 7.17529e-07 loss)
I0405 12:13:13.622089 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.03902e-05 (* 0.0454545 = 4.7228e-07 loss)
I0405 12:13:13.622104 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.17931e-05 (* 0.0454545 = 5.36049e-07 loss)
I0405 12:13:13.622118 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.20426e-05 (* 0.0454545 = 5.47393e-07 loss)
I0405 12:13:13.622133 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.47851e-05 (* 0.0454545 = 6.72051e-07 loss)
I0405 12:13:13.622145 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:13:13.622158 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000680223
I0405 12:13:13.622172 26022 sgd_solver.cpp:106] Iteration 4150, lr = 0.039834
I0405 12:22:15.701251 26022 solver.cpp:229] Iteration 4200, loss = 0.925432
I0405 12:22:15.701426 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 12:22:15.701445 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.15625
I0405 12:22:15.701459 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 12:22:15.701472 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 12:22:15.701484 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 12:22:15.701496 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 12:22:15.701508 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 12:22:15.701520 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 12:22:15.701532 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 12:22:15.701545 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 12:22:15.701556 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:22:15.701568 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:22:15.701580 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:22:15.701591 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:22:15.701602 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:22:15.701614 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:22:15.701625 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:22:15.701637 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:22:15.701648 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:22:15.701660 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:22:15.701673 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:22:15.701683 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:22:15.701699 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.81131 (* 0.0454545 = 0.127787 loss)
I0405 12:22:15.701714 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.01126 (* 0.0454545 = 0.136875 loss)
I0405 12:22:15.701727 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.31926 (* 0.0454545 = 0.150875 loss)
I0405 12:22:15.701742 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.2033 (* 0.0454545 = 0.145604 loss)
I0405 12:22:15.701756 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.82025 (* 0.0454545 = 0.128193 loss)
I0405 12:22:15.701771 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.36316 (* 0.0454545 = 0.107416 loss)
I0405 12:22:15.701786 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.49108 (* 0.0454545 = 0.0677765 loss)
I0405 12:22:15.701799 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.930879 (* 0.0454545 = 0.0423127 loss)
I0405 12:22:15.701814 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0416777 (* 0.0454545 = 0.00189444 loss)
I0405 12:22:15.701828 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0177292 (* 0.0454545 = 0.000805874 loss)
I0405 12:22:15.701843 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 7.34159e-05 (* 0.0454545 = 3.33709e-06 loss)
I0405 12:22:15.701859 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 6.62975e-05 (* 0.0454545 = 3.01352e-06 loss)
I0405 12:22:15.701874 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 6.47165e-05 (* 0.0454545 = 2.94166e-06 loss)
I0405 12:22:15.701892 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 6.60159e-05 (* 0.0454545 = 3.00072e-06 loss)
I0405 12:22:15.701908 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 5.39788e-05 (* 0.0454545 = 2.45358e-06 loss)
I0405 12:22:15.701922 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.59057e-05 (* 0.0454545 = 2.54117e-06 loss)
I0405 12:22:15.701937 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 6.17518e-05 (* 0.0454545 = 2.8069e-06 loss)
I0405 12:22:15.701969 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 6.07995e-05 (* 0.0454545 = 2.76361e-06 loss)
I0405 12:22:15.701985 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.70579e-05 (* 0.0454545 = 2.59354e-06 loss)
I0405 12:22:15.702011 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.53709e-05 (* 0.0454545 = 2.51686e-06 loss)
I0405 12:22:15.702040 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.35909e-05 (* 0.0454545 = 2.43595e-06 loss)
I0405 12:22:15.702059 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.95704e-05 (* 0.0454545 = 2.70775e-06 loss)
I0405 12:22:15.702070 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:22:15.702082 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000234615
I0405 12:22:15.702097 26022 sgd_solver.cpp:106] Iteration 4200, lr = 0.039832
I0405 12:31:17.903682 26022 solver.cpp:229] Iteration 4250, loss = 0.922091
I0405 12:31:17.903853 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 12:31:17.903872 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.0625
I0405 12:31:17.903887 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.09375
I0405 12:31:17.903899 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 12:31:17.903913 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 12:31:17.903924 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.28125
I0405 12:31:17.903936 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 12:31:17.903949 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 12:31:17.903961 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 12:31:17.903973 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 12:31:17.903985 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:31:17.903997 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:31:17.904009 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:31:17.904021 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:31:17.904032 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:31:17.904044 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:31:17.904057 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:31:17.904083 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:31:17.904099 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:31:17.904111 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:31:17.904124 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:31:17.904135 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:31:17.904150 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.95309 (* 0.0454545 = 0.134231 loss)
I0405 12:31:17.904165 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.08962 (* 0.0454545 = 0.140437 loss)
I0405 12:31:17.904181 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.25099 (* 0.0454545 = 0.147772 loss)
I0405 12:31:17.904196 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.43163 (* 0.0454545 = 0.155983 loss)
I0405 12:31:17.904213 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.79934 (* 0.0454545 = 0.127243 loss)
I0405 12:31:17.904228 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.58877 (* 0.0454545 = 0.117671 loss)
I0405 12:31:17.904242 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.30406 (* 0.0454545 = 0.0592754 loss)
I0405 12:31:17.904258 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.581785 (* 0.0454545 = 0.0264448 loss)
I0405 12:31:17.904271 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.183036 (* 0.0454545 = 0.00831984 loss)
I0405 12:31:17.904286 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0376592 (* 0.0454545 = 0.00171178 loss)
I0405 12:31:17.904301 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.47918e-05 (* 0.0454545 = 2.94508e-06 loss)
I0405 12:31:17.904320 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 5.87884e-05 (* 0.0454545 = 2.6722e-06 loss)
I0405 12:31:17.904335 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 5.43011e-05 (* 0.0454545 = 2.46823e-06 loss)
I0405 12:31:17.904350 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 5.69638e-05 (* 0.0454545 = 2.58927e-06 loss)
I0405 12:31:17.904366 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.94458e-05 (* 0.0454545 = 2.24754e-06 loss)
I0405 12:31:17.904381 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 5.65503e-05 (* 0.0454545 = 2.57047e-06 loss)
I0405 12:31:17.904394 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 5.28461e-05 (* 0.0454545 = 2.4021e-06 loss)
I0405 12:31:17.904428 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 5.46606e-05 (* 0.0454545 = 2.48457e-06 loss)
I0405 12:31:17.904443 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 5.28988e-05 (* 0.0454545 = 2.40449e-06 loss)
I0405 12:31:17.904458 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.09237e-05 (* 0.0454545 = 2.31471e-06 loss)
I0405 12:31:17.904472 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 4.9538e-05 (* 0.0454545 = 2.25173e-06 loss)
I0405 12:31:17.904487 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 5.29775e-05 (* 0.0454545 = 2.40807e-06 loss)
I0405 12:31:17.904500 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:31:17.904512 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000435153
I0405 12:31:17.904526 26022 sgd_solver.cpp:106] Iteration 4250, lr = 0.03983
I0405 12:32:55.947969 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 32.7468 > 30) by scale factor 0.916121
I0405 12:37:26.989879 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.9032 > 30) by scale factor 0.970774
I0405 12:37:59.524092 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 33.8172 > 30) by scale factor 0.887122
I0405 12:40:20.002332 26022 solver.cpp:229] Iteration 4300, loss = 0.929997
I0405 12:40:20.002439 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.25
I0405 12:40:20.002460 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 12:40:20.002472 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 12:40:20.002485 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 12:40:20.002498 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.03125
I0405 12:40:20.002511 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.3125
I0405 12:40:20.002523 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.65625
I0405 12:40:20.002535 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.875
I0405 12:40:20.002547 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.90625
I0405 12:40:20.002560 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.9375
I0405 12:40:20.002573 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:40:20.002585 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:40:20.002596 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:40:20.002609 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:40:20.002620 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:40:20.002631 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:40:20.002643 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:40:20.002655 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:40:20.002666 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:40:20.002678 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:40:20.002689 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:40:20.002701 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:40:20.002717 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.17136 (* 0.0454545 = 0.144153 loss)
I0405 12:40:20.002730 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.30348 (* 0.0454545 = 0.150158 loss)
I0405 12:40:20.002745 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.42882 (* 0.0454545 = 0.155855 loss)
I0405 12:40:20.002760 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.31824 (* 0.0454545 = 0.150829 loss)
I0405 12:40:20.002774 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.37524 (* 0.0454545 = 0.15342 loss)
I0405 12:40:20.002789 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.74805 (* 0.0454545 = 0.124911 loss)
I0405 12:40:20.002804 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.59736 (* 0.0454545 = 0.0726074 loss)
I0405 12:40:20.002817 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.760797 (* 0.0454545 = 0.0345817 loss)
I0405 12:40:20.002831 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.617443 (* 0.0454545 = 0.0280656 loss)
I0405 12:40:20.002848 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.362134 (* 0.0454545 = 0.0164606 loss)
I0405 12:40:20.002864 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.25674e-05 (* 0.0454545 = 1.02579e-06 loss)
I0405 12:40:20.002881 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.53768e-05 (* 0.0454545 = 1.15349e-06 loss)
I0405 12:40:20.002894 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.4602e-05 (* 0.0454545 = 1.11827e-06 loss)
I0405 12:40:20.002910 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.19751e-05 (* 0.0454545 = 9.98868e-07 loss)
I0405 12:40:20.002925 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.10735e-05 (* 0.0454545 = 9.57886e-07 loss)
I0405 12:40:20.002939 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.94662e-05 (* 0.0454545 = 1.33937e-06 loss)
I0405 12:40:20.002954 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.14237e-05 (* 0.0454545 = 9.73804e-07 loss)
I0405 12:40:20.002985 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.90081e-05 (* 0.0454545 = 1.31855e-06 loss)
I0405 12:40:20.003002 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.70552e-05 (* 0.0454545 = 7.75237e-07 loss)
I0405 12:40:20.003016 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.22881e-05 (* 0.0454545 = 1.0131e-06 loss)
I0405 12:40:20.003031 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.8987e-05 (* 0.0454545 = 8.63047e-07 loss)
I0405 12:40:20.003046 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.528e-05 (* 0.0454545 = 1.14909e-06 loss)
I0405 12:40:20.003058 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:40:20.003072 26022 solver.cpp:245]     Train net output #45: total_confidence = 4.50421e-06
I0405 12:40:20.003085 26022 sgd_solver.cpp:106] Iteration 4300, lr = 0.039828
I0405 12:49:22.098310 26022 solver.cpp:229] Iteration 4350, loss = 0.926335
I0405 12:49:22.098562 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.0625
I0405 12:49:22.098584 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0
I0405 12:49:22.098598 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.03125
I0405 12:49:22.098611 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 12:49:22.098623 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.125
I0405 12:49:22.098636 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.21875
I0405 12:49:22.098649 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 12:49:22.098661 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.9375
I0405 12:49:22.098673 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 12:49:22.098688 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 12:49:22.098701 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:49:22.098713 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:49:22.098726 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:49:22.098738 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:49:22.098750 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:49:22.098762 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:49:22.098773 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:49:22.098785 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:49:22.098796 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:49:22.098809 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:49:22.098820 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:49:22.098831 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:49:22.098847 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.08135 (* 0.0454545 = 0.140061 loss)
I0405 12:49:22.098862 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.23182 (* 0.0454545 = 0.146901 loss)
I0405 12:49:22.098876 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.2365 (* 0.0454545 = 0.147113 loss)
I0405 12:49:22.098891 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.20838 (* 0.0454545 = 0.145835 loss)
I0405 12:49:22.098906 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 3.25535 (* 0.0454545 = 0.147971 loss)
I0405 12:49:22.098927 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.85899 (* 0.0454545 = 0.129954 loss)
I0405 12:49:22.098944 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.47823 (* 0.0454545 = 0.0671922 loss)
I0405 12:49:22.098958 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.30101 (* 0.0454545 = 0.0136823 loss)
I0405 12:49:22.098973 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.137203 (* 0.0454545 = 0.00623649 loss)
I0405 12:49:22.098990 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.148568 (* 0.0454545 = 0.00675307 loss)
I0405 12:49:22.099019 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.42542e-05 (* 0.0454545 = 1.10246e-06 loss)
I0405 12:49:22.099035 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.86628e-05 (* 0.0454545 = 1.30286e-06 loss)
I0405 12:49:22.099050 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.44652e-05 (* 0.0454545 = 1.11205e-06 loss)
I0405 12:49:22.099064 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.26859e-05 (* 0.0454545 = 1.03118e-06 loss)
I0405 12:49:22.099079 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.49831e-05 (* 0.0454545 = 1.1356e-06 loss)
I0405 12:49:22.099094 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 3.42009e-05 (* 0.0454545 = 1.55459e-06 loss)
I0405 12:49:22.099109 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.38875e-05 (* 0.0454545 = 1.0858e-06 loss)
I0405 12:49:22.099138 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.84061e-05 (* 0.0454545 = 1.29118e-06 loss)
I0405 12:49:22.099154 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.13923e-05 (* 0.0454545 = 9.72379e-07 loss)
I0405 12:49:22.099169 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.45508e-05 (* 0.0454545 = 1.11594e-06 loss)
I0405 12:49:22.099184 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.39042e-05 (* 0.0454545 = 1.08655e-06 loss)
I0405 12:49:22.099200 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.67217e-05 (* 0.0454545 = 1.21462e-06 loss)
I0405 12:49:22.099211 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:49:22.099223 26022 solver.cpp:245]     Train net output #45: total_confidence = 6.06346e-05
I0405 12:49:22.099238 26022 sgd_solver.cpp:106] Iteration 4350, lr = 0.039826
I0405 12:58:24.173490 26022 solver.cpp:229] Iteration 4400, loss = 0.920966
I0405 12:58:24.173683 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 12:58:24.173704 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.15625
I0405 12:58:24.173717 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 12:58:24.173730 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 12:58:24.173743 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.34375
I0405 12:58:24.173755 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.375
I0405 12:58:24.173768 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 12:58:24.173780 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 12:58:24.173791 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 12:58:24.173804 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 12:58:24.173816 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 12:58:24.173828 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 12:58:24.173840 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 12:58:24.173851 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 12:58:24.173863 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 12:58:24.173876 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 12:58:24.173887 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 12:58:24.173898 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 12:58:24.173910 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 12:58:24.173923 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 12:58:24.173933 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 12:58:24.173945 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 12:58:24.173960 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.89293 (* 0.0454545 = 0.131497 loss)
I0405 12:58:24.173975 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.05804 (* 0.0454545 = 0.139002 loss)
I0405 12:58:24.173990 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.27827 (* 0.0454545 = 0.149012 loss)
I0405 12:58:24.174005 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.04743 (* 0.0454545 = 0.13852 loss)
I0405 12:58:24.174018 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.28458 (* 0.0454545 = 0.103845 loss)
I0405 12:58:24.174037 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.30344 (* 0.0454545 = 0.104702 loss)
I0405 12:58:24.174052 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.45469 (* 0.0454545 = 0.0661221 loss)
I0405 12:58:24.174067 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.488327 (* 0.0454545 = 0.0221967 loss)
I0405 12:58:24.174082 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.202093 (* 0.0454545 = 0.00918606 loss)
I0405 12:58:24.174095 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.163596 (* 0.0454545 = 0.00743616 loss)
I0405 12:58:24.174110 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 1.3953e-05 (* 0.0454545 = 6.34229e-07 loss)
I0405 12:58:24.174125 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 1.00686e-05 (* 0.0454545 = 4.57664e-07 loss)
I0405 12:58:24.174140 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 8.76452e-06 (* 0.0454545 = 3.98387e-07 loss)
I0405 12:58:24.174155 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 1.22504e-05 (* 0.0454545 = 5.56834e-07 loss)
I0405 12:58:24.174170 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 9.89725e-06 (* 0.0454545 = 4.49875e-07 loss)
I0405 12:58:24.174185 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 8.91348e-06 (* 0.0454545 = 4.05158e-07 loss)
I0405 12:58:24.174199 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.04934e-05 (* 0.0454545 = 4.76973e-07 loss)
I0405 12:58:24.174233 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.06947e-05 (* 0.0454545 = 4.86122e-07 loss)
I0405 12:58:24.174250 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.03517e-05 (* 0.0454545 = 4.70532e-07 loss)
I0405 12:58:24.174264 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 8.82409e-06 (* 0.0454545 = 4.01095e-07 loss)
I0405 12:58:24.174279 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 9.1222e-06 (* 0.0454545 = 4.14646e-07 loss)
I0405 12:58:24.174293 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 9.80413e-06 (* 0.0454545 = 4.45642e-07 loss)
I0405 12:58:24.174306 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 12:58:24.174319 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000472303
I0405 12:58:24.174334 26022 sgd_solver.cpp:106] Iteration 4400, lr = 0.039824
I0405 13:02:33.947335 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 39.1666 > 30) by scale factor 0.76596
I0405 13:03:17.322451 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.8404 > 30) by scale factor 0.972749
I0405 13:07:26.214232 26022 solver.cpp:229] Iteration 4450, loss = 0.928573
I0405 13:07:26.214416 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 13:07:26.214435 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 13:07:26.214448 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.125
I0405 13:07:26.214462 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.21875
I0405 13:07:26.214474 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.34375
I0405 13:07:26.214486 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.40625
I0405 13:07:26.214499 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.71875
I0405 13:07:26.214511 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 13:07:26.214524 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.9375
I0405 13:07:26.214536 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 13:07:26.214548 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 13:07:26.214560 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 13:07:26.214572 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 13:07:26.214584 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 13:07:26.214596 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 13:07:26.214607 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 13:07:26.214620 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 13:07:26.214632 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 13:07:26.214643 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 13:07:26.214654 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 13:07:26.214666 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 13:07:26.214678 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 13:07:26.214694 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.99026 (* 0.0454545 = 0.135921 loss)
I0405 13:07:26.214709 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 2.91354 (* 0.0454545 = 0.132434 loss)
I0405 13:07:26.214723 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.13748 (* 0.0454545 = 0.142613 loss)
I0405 13:07:26.214737 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.00651 (* 0.0454545 = 0.136659 loss)
I0405 13:07:26.214752 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.4992 (* 0.0454545 = 0.1136 loss)
I0405 13:07:26.214766 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.15653 (* 0.0454545 = 0.098024 loss)
I0405 13:07:26.214787 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.07777 (* 0.0454545 = 0.0489894 loss)
I0405 13:07:26.214802 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.53151 (* 0.0454545 = 0.0241595 loss)
I0405 13:07:26.214818 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.380413 (* 0.0454545 = 0.0172915 loss)
I0405 13:07:26.214831 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.36797 (* 0.0454545 = 0.0167259 loss)
I0405 13:07:26.214846 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 6.52311e-06 (* 0.0454545 = 2.96505e-07 loss)
I0405 13:07:26.214861 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 4.70883e-06 (* 0.0454545 = 2.14038e-07 loss)
I0405 13:07:26.214875 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 4.06434e-06 (* 0.0454545 = 1.84743e-07 loss)
I0405 13:07:26.214890 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 3.93395e-06 (* 0.0454545 = 1.78816e-07 loss)
I0405 13:07:26.214905 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 4.67903e-06 (* 0.0454545 = 2.12683e-07 loss)
I0405 13:07:26.214920 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 6.90313e-06 (* 0.0454545 = 3.13779e-07 loss)
I0405 13:07:26.214934 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 4.98451e-06 (* 0.0454545 = 2.26569e-07 loss)
I0405 13:07:26.214967 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 4.28041e-06 (* 0.0454545 = 1.94564e-07 loss)
I0405 13:07:26.214982 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 7.38744e-06 (* 0.0454545 = 3.35793e-07 loss)
I0405 13:07:26.214997 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 5.53216e-06 (* 0.0454545 = 2.51462e-07 loss)
I0405 13:07:26.215011 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 5.0553e-06 (* 0.0454545 = 2.29786e-07 loss)
I0405 13:07:26.215026 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 3.93767e-06 (* 0.0454545 = 1.78985e-07 loss)
I0405 13:07:26.215039 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 13:07:26.215050 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000511933
I0405 13:07:26.215065 26022 sgd_solver.cpp:106] Iteration 4450, lr = 0.039822
I0405 13:16:28.394696 26022 solver.cpp:229] Iteration 4500, loss = 0.915217
I0405 13:16:28.394942 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.09375
I0405 13:16:28.394963 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.09375
I0405 13:16:28.394978 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0
I0405 13:16:28.394990 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.1875
I0405 13:16:28.395005 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.28125
I0405 13:16:28.395018 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.5625
I0405 13:16:28.395031 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.75
I0405 13:16:28.395043 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 13:16:28.395056 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 13:16:28.395068 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 13:16:28.395081 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 13:16:28.395092 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 13:16:28.395104 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 13:16:28.395117 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 13:16:28.395128 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 13:16:28.395139 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 13:16:28.395151 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 13:16:28.395164 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 13:16:28.395175 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 13:16:28.395187 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 13:16:28.395200 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 13:16:28.395210 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 13:16:28.395226 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.03215 (* 0.0454545 = 0.137825 loss)
I0405 13:16:28.395241 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.3854 (* 0.0454545 = 0.153882 loss)
I0405 13:16:28.395256 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.36403 (* 0.0454545 = 0.15291 loss)
I0405 13:16:28.395270 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.12144 (* 0.0454545 = 0.141884 loss)
I0405 13:16:28.395285 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.66616 (* 0.0454545 = 0.121189 loss)
I0405 13:16:28.395300 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 1.81339 (* 0.0454545 = 0.082427 loss)
I0405 13:16:28.395314 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.03027 (* 0.0454545 = 0.0468304 loss)
I0405 13:16:28.395329 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.462859 (* 0.0454545 = 0.0210391 loss)
I0405 13:16:28.395344 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.232366 (* 0.0454545 = 0.0105621 loss)
I0405 13:16:28.395359 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.219943 (* 0.0454545 = 0.0099974 loss)
I0405 13:16:28.395373 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.13258e-05 (* 0.0454545 = 9.69356e-07 loss)
I0405 13:16:28.395388 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 1.8386e-05 (* 0.0454545 = 8.35726e-07 loss)
I0405 13:16:28.395404 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.50457e-05 (* 0.0454545 = 6.83894e-07 loss)
I0405 13:16:28.395419 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 1.5284e-05 (* 0.0454545 = 6.94729e-07 loss)
I0405 13:16:28.395443 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.74117e-05 (* 0.0454545 = 7.91441e-07 loss)
I0405 13:16:28.395470 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.14323e-05 (* 0.0454545 = 9.74195e-07 loss)
I0405 13:16:28.395486 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.67409e-05 (* 0.0454545 = 7.60949e-07 loss)
I0405 13:16:28.395516 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.68136e-05 (* 0.0454545 = 7.64254e-07 loss)
I0405 13:16:28.395532 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 1.94259e-05 (* 0.0454545 = 8.82997e-07 loss)
I0405 13:16:28.395547 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.82706e-05 (* 0.0454545 = 8.30484e-07 loss)
I0405 13:16:28.395561 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.77396e-05 (* 0.0454545 = 8.06346e-07 loss)
I0405 13:16:28.395576 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.78457e-05 (* 0.0454545 = 8.1117e-07 loss)
I0405 13:16:28.395589 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 13:16:28.395601 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000619774
I0405 13:16:28.395617 26022 sgd_solver.cpp:106] Iteration 4500, lr = 0.03982
I0405 13:25:30.439033 26022 solver.cpp:229] Iteration 4550, loss = 0.915567
I0405 13:25:30.439141 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.125
I0405 13:25:30.439160 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 13:25:30.439174 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.15625
I0405 13:25:30.439188 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.15625
I0405 13:25:30.439199 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 13:25:30.439211 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.46875
I0405 13:25:30.439224 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.6875
I0405 13:25:30.439235 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 13:25:30.439249 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 1
I0405 13:25:30.439260 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 13:25:30.439272 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 13:25:30.439285 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 13:25:30.439296 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 13:25:30.439306 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 13:25:30.439318 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 13:25:30.439329 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 13:25:30.439342 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 13:25:30.439352 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 13:25:30.439364 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 13:25:30.439376 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 13:25:30.439388 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 13:25:30.439399 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 13:25:30.439414 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.98712 (* 0.0454545 = 0.135778 loss)
I0405 13:25:30.439429 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.25005 (* 0.0454545 = 0.14773 loss)
I0405 13:25:30.439443 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.00737 (* 0.0454545 = 0.136699 loss)
I0405 13:25:30.439457 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.03764 (* 0.0454545 = 0.138074 loss)
I0405 13:25:30.439472 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.8733 (* 0.0454545 = 0.130605 loss)
I0405 13:25:30.439486 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.11233 (* 0.0454545 = 0.0960152 loss)
I0405 13:25:30.439501 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.6259 (* 0.0454545 = 0.0739047 loss)
I0405 13:25:30.439515 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.442157 (* 0.0454545 = 0.020098 loss)
I0405 13:25:30.439529 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.0213351 (* 0.0454545 = 0.000969776 loss)
I0405 13:25:30.439545 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.0073743 (* 0.0454545 = 0.000335195 loss)
I0405 13:25:30.439560 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 3.32308e-05 (* 0.0454545 = 1.51049e-06 loss)
I0405 13:25:30.439575 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 2.7164e-05 (* 0.0454545 = 1.23473e-06 loss)
I0405 13:25:30.439589 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 2.57878e-05 (* 0.0454545 = 1.17217e-06 loss)
I0405 13:25:30.439604 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.75967e-05 (* 0.0454545 = 1.25439e-06 loss)
I0405 13:25:30.439618 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 2.43695e-05 (* 0.0454545 = 1.10771e-06 loss)
I0405 13:25:30.439633 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 2.59753e-05 (* 0.0454545 = 1.1807e-06 loss)
I0405 13:25:30.439648 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 2.80757e-05 (* 0.0454545 = 1.27617e-06 loss)
I0405 13:25:30.439680 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 2.53517e-05 (* 0.0454545 = 1.15235e-06 loss)
I0405 13:25:30.439697 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.68793e-05 (* 0.0454545 = 1.22179e-06 loss)
I0405 13:25:30.439712 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 2.45857e-05 (* 0.0454545 = 1.11753e-06 loss)
I0405 13:25:30.439725 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 2.39526e-05 (* 0.0454545 = 1.08875e-06 loss)
I0405 13:25:30.439740 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 2.46454e-05 (* 0.0454545 = 1.12025e-06 loss)
I0405 13:25:30.439752 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 13:25:30.439764 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000333271
I0405 13:25:30.439779 26022 sgd_solver.cpp:106] Iteration 4550, lr = 0.039818
I0405 13:34:00.457092 26022 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 30.3319 > 30) by scale factor 0.989056
I0405 13:34:32.500744 26022 solver.cpp:229] Iteration 4600, loss = 0.921218
I0405 13:34:32.500872 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.15625
I0405 13:34:32.500892 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.03125
I0405 13:34:32.500905 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 13:34:32.500919 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.09375
I0405 13:34:32.500931 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.25
I0405 13:34:32.500943 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.46875
I0405 13:34:32.500957 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 13:34:32.500968 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.90625
I0405 13:34:32.500980 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 13:34:32.500993 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 1
I0405 13:34:32.501005 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 13:34:32.501018 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 13:34:32.501029 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 13:34:32.501040 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 13:34:32.501052 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 13:34:32.501063 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 13:34:32.501075 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 13:34:32.501087 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 13:34:32.501099 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 13:34:32.501111 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 13:34:32.501123 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 13:34:32.501134 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 13:34:32.501150 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 3.09621 (* 0.0454545 = 0.140737 loss)
I0405 13:34:32.501165 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.24516 (* 0.0454545 = 0.147507 loss)
I0405 13:34:32.501180 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.3129 (* 0.0454545 = 0.150586 loss)
I0405 13:34:32.501194 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.09695 (* 0.0454545 = 0.14077 loss)
I0405 13:34:32.501209 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.89739 (* 0.0454545 = 0.131699 loss)
I0405 13:34:32.501224 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.08266 (* 0.0454545 = 0.0946663 loss)
I0405 13:34:32.501238 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.06792 (* 0.0454545 = 0.0485419 loss)
I0405 13:34:32.501252 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.511324 (* 0.0454545 = 0.023242 loss)
I0405 13:34:32.501267 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.279843 (* 0.0454545 = 0.0127201 loss)
I0405 13:34:32.501281 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.00912159 (* 0.0454545 = 0.000414618 loss)
I0405 13:34:32.501297 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 2.26137e-05 (* 0.0454545 = 1.02789e-06 loss)
I0405 13:34:32.501312 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 1.82596e-05 (* 0.0454545 = 8.29981e-07 loss)
I0405 13:34:32.501327 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 1.86012e-05 (* 0.0454545 = 8.4551e-07 loss)
I0405 13:34:32.501341 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 2.05835e-05 (* 0.0454545 = 9.35616e-07 loss)
I0405 13:34:32.501356 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 1.744e-05 (* 0.0454545 = 7.92726e-07 loss)
I0405 13:34:32.501371 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 1.88071e-05 (* 0.0454545 = 8.54866e-07 loss)
I0405 13:34:32.501385 26022 solver.cpp:245]     Train net output #38: loss/loss17 = 1.97373e-05 (* 0.0454545 = 8.97151e-07 loss)
I0405 13:34:32.501418 26022 solver.cpp:245]     Train net output #39: loss/loss18 = 1.94432e-05 (* 0.0454545 = 8.83782e-07 loss)
I0405 13:34:32.501435 26022 solver.cpp:245]     Train net output #40: loss/loss19 = 2.00891e-05 (* 0.0454545 = 9.13139e-07 loss)
I0405 13:34:32.501451 26022 solver.cpp:245]     Train net output #41: loss/loss20 = 1.83268e-05 (* 0.0454545 = 8.33035e-07 loss)
I0405 13:34:32.501466 26022 solver.cpp:245]     Train net output #42: loss/loss21 = 1.90088e-05 (* 0.0454545 = 8.64037e-07 loss)
I0405 13:34:32.501479 26022 solver.cpp:245]     Train net output #43: loss/loss22 = 1.84798e-05 (* 0.0454545 = 8.39991e-07 loss)
I0405 13:34:32.501492 26022 solver.cpp:245]     Train net output #44: total_accuracy = 0
I0405 13:34:32.501504 26022 solver.cpp:245]     Train net output #45: total_confidence = 0.000100249
I0405 13:34:32.501520 26022 sgd_solver.cpp:106] Iteration 4600, lr = 0.039816
I0405 13:43:34.672369 26022 solver.cpp:229] Iteration 4650, loss = 0.915752
I0405 13:43:34.672549 26022 solver.cpp:245]     Train net output #0: loss/accuracy01 = 0.1875
I0405 13:43:34.672567 26022 solver.cpp:245]     Train net output #1: loss/accuracy02 = 0.125
I0405 13:43:34.672580 26022 solver.cpp:245]     Train net output #2: loss/accuracy03 = 0.0625
I0405 13:43:34.672593 26022 solver.cpp:245]     Train net output #3: loss/accuracy04 = 0.0625
I0405 13:43:34.672606 26022 solver.cpp:245]     Train net output #4: loss/accuracy05 = 0.1875
I0405 13:43:34.672618 26022 solver.cpp:245]     Train net output #5: loss/accuracy06 = 0.34375
I0405 13:43:34.672631 26022 solver.cpp:245]     Train net output #6: loss/accuracy07 = 0.78125
I0405 13:43:34.672642 26022 solver.cpp:245]     Train net output #7: loss/accuracy08 = 0.84375
I0405 13:43:34.672655 26022 solver.cpp:245]     Train net output #8: loss/accuracy09 = 0.96875
I0405 13:43:34.672667 26022 solver.cpp:245]     Train net output #9: loss/accuracy10 = 0.96875
I0405 13:43:34.672680 26022 solver.cpp:245]     Train net output #10: loss/accuracy11 = 1
I0405 13:43:34.672693 26022 solver.cpp:245]     Train net output #11: loss/accuracy12 = 1
I0405 13:43:34.672703 26022 solver.cpp:245]     Train net output #12: loss/accuracy13 = 1
I0405 13:43:34.672715 26022 solver.cpp:245]     Train net output #13: loss/accuracy14 = 1
I0405 13:43:34.672726 26022 solver.cpp:245]     Train net output #14: loss/accuracy15 = 1
I0405 13:43:34.672739 26022 solver.cpp:245]     Train net output #15: loss/accuracy16 = 1
I0405 13:43:34.672750 26022 solver.cpp:245]     Train net output #16: loss/accuracy17 = 1
I0405 13:43:34.672761 26022 solver.cpp:245]     Train net output #17: loss/accuracy18 = 1
I0405 13:43:34.672773 26022 solver.cpp:245]     Train net output #18: loss/accuracy19 = 1
I0405 13:43:34.672785 26022 solver.cpp:245]     Train net output #19: loss/accuracy20 = 1
I0405 13:43:34.672796 26022 solver.cpp:245]     Train net output #20: loss/accuracy21 = 1
I0405 13:43:34.672807 26022 solver.cpp:245]     Train net output #21: loss/accuracy22 = 1
I0405 13:43:34.672823 26022 solver.cpp:245]     Train net output #22: loss/loss01 = 2.82916 (* 0.0454545 = 0.128598 loss)
I0405 13:43:34.672837 26022 solver.cpp:245]     Train net output #23: loss/loss02 = 3.07397 (* 0.0454545 = 0.139726 loss)
I0405 13:43:34.672852 26022 solver.cpp:245]     Train net output #24: loss/loss03 = 3.23519 (* 0.0454545 = 0.147054 loss)
I0405 13:43:34.672868 26022 solver.cpp:245]     Train net output #25: loss/loss04 = 3.20028 (* 0.0454545 = 0.145467 loss)
I0405 13:43:34.672881 26022 solver.cpp:245]     Train net output #26: loss/loss05 = 2.953 (* 0.0454545 = 0.134227 loss)
I0405 13:43:34.672899 26022 solver.cpp:245]     Train net output #27: loss/loss06 = 2.37553 (* 0.0454545 = 0.107979 loss)
I0405 13:43:34.672914 26022 solver.cpp:245]     Train net output #28: loss/loss07 = 1.25725 (* 0.0454545 = 0.0571478 loss)
I0405 13:43:34.672929 26022 solver.cpp:245]     Train net output #29: loss/loss08 = 0.663931 (* 0.0454545 = 0.0301787 loss)
I0405 13:43:34.672942 26022 solver.cpp:245]     Train net output #30: loss/loss09 = 0.147865 (* 0.0454545 = 0.00672116 loss)
I0405 13:43:34.672960 26022 solver.cpp:245]     Train net output #31: loss/loss10 = 0.134979 (* 0.0454545 = 0.00613542 loss)
I0405 13:43:34.672976 26022 solver.cpp:245]     Train net output #32: loss/loss11 = 0.000136223 (* 0.0454545 = 6.19197e-06 loss)
I0405 13:43:34.672991 26022 solver.cpp:245]     Train net output #33: loss/loss12 = 0.000113989 (* 0.0454545 = 5.18134e-06 loss)
I0405 13:43:34.673005 26022 solver.cpp:245]     Train net output #34: loss/loss13 = 0.000105062 (* 0.0454545 = 4.77554e-06 loss)
I0405 13:43:34.673020 26022 solver.cpp:245]     Train net output #35: loss/loss14 = 0.000128425 (* 0.0454545 = 5.8375e-06 loss)
I0405 13:43:34.673035 26022 solver.cpp:245]     Train net output #36: loss/loss15 = 0.000111066 (* 0.0454545 = 5.04844e-06 loss)
I0405 13:43:34.673050 26022 solver.cpp:245]     Train net output #37: loss/loss16 = 0.000117955 (* 0.0454545 = 5.36161e-06 loss)
I0405 13:43:34.673064 26022 solver.cpp:245]     Train net output #38: lo