NVIDIA / NVIDIA/TensorRT-LLM

[Bug]: Nemotron SuperV3 BF16 crashing during GSM8K accuracy test with attention dp = off

Open
#12,226 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

bug Inference runtime
Dominant language
Python
Stars
14.7k
Forks
2.8k
Avg merge
2d 23h
Merged PRs (30d)
489

Description

System Info

See nvbug 5919796

Who can help?

No response

Information
  • The official example scripts
  • My own modified scripts
Tasks
  • An officially supported task in the examples folder (such as GLUE/SQuAD, ...)
  • My own task or dataset (give details below)
Reproduction

pytest accuracy/test_llm_api_autodeploy.py::TestNemotronSuperV3::test_accuracy[bf16-4-attn_dp_off-trtllm] -s
pytest accuracy/test_llm_api_autodeploy.py::TestNemotronSuperV3::test_accuracy[bf16-4-attn_dp_off-flashinfer] -s

Expected behavior

Test passes

actual behavior

Test crashes

additional notes

N/A

Before submitting a new issue...
  • Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by running the two listed pytest cases in accuracy/test_llm_api_autodeploy.py for Nemotron SuperV3 with BF16 and attention dp off, then inspect the resulting crash and the referenced nvbug 5919796. Done means both TRT-LLM and FlashInfer accuracy tests complete successfully instead of crashing.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
ai, testing-qa
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.