Update rl_replacements.py returned hidden states from logprobs #3
Conversation
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
Datta0
reviewed
May 25, 2025
unsloth/models/rl_replacements.py
Outdated
| # For transformers<=4.48, logits_to_keep argument isn't supported, so here we drop logits ourselves. | ||
| # See https://github.com/huggingface/trl/issues/2770 | ||
| logits = logits[:, -logits_to_keep:] | ||
| logits = logits.to(torch.float32) |
| hidden_states = model(input_ids=input_ids, attention_mask=attention_mask, logits_to_keep=logits_to_keep + 1).logits | ||
| #breakpoint() | ||
| #logits = logits[:, :-1, :] # (B, L-1, V), exclude the last logit: it corresponds to the next token pred | ||
| return hidden_states |
There was a problem hiding this comment.
If we're returning here, we don't need the code below. Remove perhaps?
There was a problem hiding this comment.
I want to leave it there for debugging purposes, dan had the old code below because of that
| logits = model(input_ids=input_ids, attention_mask=attention_mask, logits_to_keep=logits_to_keep + 1).logits | ||
| logits = logits[:, :-1, :] # (B, L-1, V), exclude the last logit: it corresponds to the next token pred | ||
| logits = logits.to(torch.float32) | ||
| hidden_states = model(input_ids=input_ids, attention_mask=attention_mask, logits_to_keep=logits_to_keep + 1).logits |
There was a problem hiding this comment.
NIT: These are logits. can we rename appropriately
There was a problem hiding this comment.
I actually return those as hidden states so we save memory and can calculate logits on the fly over here:
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
No description provided.