deepspeedai / deepspeedai/DeepSpeedExamples
[ERROR] [launch.py:324:sigkill_handler]:Deepspeed trainng CLM issue
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 6.8k
- Forks
- 1.1k
- Avg merge
- 2d 16h
- Merged PRs (30d)
- 1
Description
[2023-07-10 04:09:42,841] [INFO] [comm.py:652:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
Using cuda_amp half precision backend
[2023-07-10 04:09:43,963] [INFO] [logging.py:93:log_dist] [Rank 0] DeepSpeed info: version=0.8.3, git-hash=unknown, git-branch=unknown
fdsui[2023-07-10 04:26:05,876] [INFO] [launch.py:318:sigkill_handler] Killing subprocess 94984
[2023-07-10 04:26:05,876] [INFO] [launch.py:318:sigkill_handler] Killing subprocess 94985
[2023-07-10 04:26:35,999] [ERROR] [launch.py:324:sigkill_handler] ['/global/D1/homes/sushant/sushanta/venv/bin/python3.10', '-u', 'trainCLM.py', '--local_rank=1', '--deepspeed', '/home/sushant/D1/sushanta/try/nepaligpt/clm/config/deepspeed_config.json'] exits with return code = -11
deepspeed --num_gpus 2 trainCLM.py --deepspeed /home/sushant/D1/sushanta/try/nepaligpt/clm/config/deepspeed_config.json
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the reported command for trainCLM.py and the referenced config/deepspeed_config.json. Reproduce the two-GPU run and inspect the launch.py:324 sigkill_handler output and return code -11. Done means identifying the cause of the subprocess crash and documenting or fixing it with a reproducible validation.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- distributed-systems, machine-learning
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100