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

Handle prompt-too-long error #61

Open

ccmaymay opened this issue Aug 18, 2022 · 0 comments

Labels

bug

Collaborator

ccmaymay commented Aug 18, 2022

From @nweir127:

If I feed the API text that is longer than the acceptable context window (2048 tokens) , the generator post processor currently crashes

A nice feature of the API would be to either give an appropriate error response or to perform front truncation

And a code snippet:

input_token_ids = tokenizer(text, return_tensors='pt')['input_ids']
max_context = model.config.n_positions - max_new_tokens
if hasattr(model.config, "n_positions") and input_token_ids.shape[1] > max_context:
    input_token_ids = input_token_ids[:, -max_context:]
    text = tokenizer.batch_decode(input_token_ids)[0]
    input_token_ids = tokenizer(text, return_tensors='pt')['input_ids']

ccmaymay added the bug label

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