huggingface / huggingface/open-r1

Segmentation fault: address not mapped to object at address 0x3808

Open
#562 3 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
26.5k
Forks
2.5k
PR merge metrics
No merged PRs in 30d

Description

When I run this:
CUDA_VISIBLE_DEVICES=0 trl vllm-serve --model Qwen/Qwen2.5-1.5B-Instruct

and then run:
CUDA_VISIBLE_DEVICES=1 ACCELERATE_LOG_LEVEL=info \
accelerate launch --config_file recipes/accelerate_configs/zero2.yaml --num_processes 1 src/open_r1/grpo.py --config recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo.yaml

Got error message:
"[task-20250211152313-99256:46604:0:46767] Caught signal 11 (Segmentation fault: address not mapped to object at address 0x3808)"

How to solve it?

Contributor guide

No contributing guide indexed for this repository

Research direction

Start by reproducing the commands in the issue, then inspect src/open_r1/grpo.py, recipes/accelerate_configs/zero2.yaml, and recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo.yaml. Check the interaction between the vllm-serve command and the accelerate launch entry point; done means identifying the cause of the segmentation fault and documenting a verified way to run the configuration without it.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
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.