New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Error in Batch Size #6

Open

rohitx007 opened this issue Sep 9, 2021 · 0 comments

rohitx007 commented Sep 9, 2021

If I take batch size to be close to Data size then line #129 in [irl_coco.py]
(https://github.com/FudanNLP/Irl_gen/blob/master/imagecoco/irl_coco.py)
_, r_loss = rewarder.reward_train_step(sess, x_text, np.ones(BATCH_SIZE), 1.0, re_dropout_keep_prob, 0.01)

throws, Value error: cannot feed value of shape (426, 9) for Tensor 'Placeholder_1:0', which has shape '(427, 9).

The code works only when (batch size << 1/2x Data Size).

If I want to use different batch size for pre-training and adversarial-training then is it possible?

The text was updated successfully, but these errors were encountered:

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment