Fix Max New Tokens in HF's Generation Config #257
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
HuggingFace's
max_length
configuration corresponds to the total length of the prompt and the generated output, whilemax_new_tokens
corresponds to the length of generated output only.Using
args.max_length_generation
to setmax_new_tokens
fixed runtime errors for me. Usingargs.max_length_generation
to setmax_length
lead to runtime errors because the total length of prompt+generation would exceed the intended value.