Add top_p_size step fn, StepFunctionArgs class
#206
Merged
Conversation
This file contains 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Description
This PR adds the following capabilities to the Inseq library:
A new
top_p_size_fn(identifier:"top_p_size") step function returning the number of tokens needed to reach probabilitypfor a specific generation step (e.g. 5 withp=0.95means that the top 5 tokens in the probability distribution over the vocabulary are needed to reach a CDF of 95%)The
kl_divergencestep function now supports a new parametertop_p: floatdefined in [0,1] (default 1, full distribution) to preserve only tokens in the top p of either the original or the contrastive output distributions before computing the KL divergence between the two.🔥 Breaking change: This PR introduces a new
StepFunctionArgsto better structure the inputs to step function methods. This doesn't change anything in the usage of pre-registered functions, but functions that were previously registered with explicit default params (attribution_model,forward_output,encoder_input_ids, etc.) will now break, and should be converted to useStepFunctionArgs. The step function registration tutorial will be updated accordingly.