fix(vllm_rollout): Filter prompts exceeding max_model_len instead of …#690
Open
Pritiks23 wants to merge 2 commits into
Open
fix(vllm_rollout): Filter prompts exceeding max_model_len instead of …#690Pritiks23 wants to merge 2 commits into
Pritiks23 wants to merge 2 commits into
Conversation
…raising error Fixes issue nvidia-cosmos#403 - vLLM now raises an error when decoder prompt length exceeds max_model_len, especially with large images/videos in VLM tasks. This fix filters out oversized prompts before passing them to vLLM, preventing errors while logging warnings about skipped samples. Changes: - Added prompt length validation in rollout_generation_single_turn() - Added prompt length validation in async rollout_generation() - Supports both LLM (string/dict prompts) and VLM (image/video inputs) - Logs warnings when samples are skipped due to exceeding max_model_len - Gracefully handles edge cases where token count cannot be determined
Pritiks23
force-pushed
the
fix/vllm-max-model-len-filtering
branch
from
5876831 to
4c2abb1
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes issue #403 - vLLM now raises an error when decoder prompt length exceeds max_model_len, especially with large images/videos in VLM tasks. This fix filters out oversized prompts before passing them to vLLM, preventing errors while logging warnings about skipped samples.
Changes: