Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
epoch train_loss train_trial_accuracy valid_loss valid_trial_accuracy dur | |
------- ------------ ---------------------- ------------ ---------------------- ------- | |
1 1.3710 0.2565 7.1190 0.2241 12.1600 | |
2 1.1152 0.3130 2.7722 0.2069 11.9932 | |
3 0.9749 0.6739 1.2467 0.4138 12.0441 | |
4 0.8932 0.8261 0.9292 0.6034 12.0141 | |
5 0.8167 0.8391 0.8623 0.7586 12.0072 | |
6 0.7648 0.8870 0.7970 0.7586 11.9814 | |
7 0.6920 0.8609 0.8171 0.7586 11.9903 | |
8 0.6505 0.8696 0.8143 0.7241 11.9883 |
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
def rescale(a, a_new, i_op_0, i_op_1): | |
i_op_0 = int(i_op_0) | |
i_op_1 = int(i_op_1) | |
old_p = F.softmax(a, dim=0) | |
new_p = F.softmax(a_new, dim=0) | |
old_sum = old_p[i_op_0] + old_p[i_op_1] | |
new_sum = new_p[i_op_0] + new_p[i_op_1] | |
ratio = old_sum / new_sum | |
# rescaled probabilties such that sum is same as before | |
p_r_0 = ratio * new_p[i_op_0] |