mlcommons / mlcommons/inference

GPT-J_NVIDIA_TensorRT_Docker environment Failure

Open
#2,318 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.6k
Forks
650
Avg merge
1d 22h
Merged PRs (30d)
6

Description

I am running the below command :

mlcr run-mlperf,inference,_find-performance,_full,_r5.0-dev \
   --model=gptj-99 \
   --implementation=nvidia \
   --framework=tensorrt \
   --category=edge \
   --scenario=Offline \
   --execution_mode=test \
   --device=cuda  \
   --docker --quiet \
   --test_query_count=50 –rerun

Attached the Log with the error .
mlperf_inference_GPT_J.txt

As directed in the MLcommons documentation running in the python environment.
I am trying to test NVIDIA RTX 3070
Host station details:
Ubuntu 20.04

Image Please let me know what is the issue and how to solve it? Similarly I tried running Resnet as well, I am seeing an issue there as well where the Git pull command is failing. Is there a FAQ or list of issues repository which I can refer to resolve such issues?

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 with the attached mlperf_inference_GPT_J.txt log and reproduce the documented mlcr command in the Python environment on Ubuntu 20.04 with --docker. Trace the GPT-J TensorRT failure first, keeping the separately reported ResNet git-pull problem out of scope. Done means the failure has a confirmed cause, a reproducible diagnosis, and a documented resolution or clear report of the missing environment requirement.

Written by the indexing model from the issue text.

Assessment

Tech stack
docker, python
Domain
devops, machine-learning
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.