Provide a minimal reproducible experiment using GRPO for mathematical reasoning on base model, referencing the approach from SimpleRL-Reason #197
Conversation
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
edbeeching
reviewed
Feb 6, 2025
| We provide a minimal reproducible experiment using GRPO for mathematical reasoning, referencing the approach from [SimpleRL-Reason](https://hkust-nlp.notion.site/simplerl-reason) which uses a 7B model trained on 8K examples. Running this on 8 H100 80G GPU takes about 3 hours: | ||
|
|
||
| ```shell | ||
| ACCELERATE_LOG_LEVEL=info accelerate launch --config_file recipes/accelerate_configs/zero2.yaml --num_processes=7 src/open_r1/grpo.py --config recipes/deepseek/DeepSeek-R1-Distill-Qwen-7B/grpo/config_base_math_smalllr.yaml |
There was a problem hiding this comment.
I am curious why you went with zero2 and not zero3?
edbeeching
approved these changes
Feb 6, 2025
|
Merging as I would like to run some benchmarks with the recipe. |
|
@Some-random @edbeeching, I have reproducible setup and zero3 tested on clean Debian with CUDA 12.1, vllm, 8x H100 in #199. It needed specific package versions pinned. I have to update branch though. I could test this one for zero3 if we want, after branch update. |
GitMonkey0
pushed a commit
to GitMonkey0/open-r1
that referenced
this pull request
Feb 24, 2025
… reasoning on base model, referencing the approach from SimpleRL-Reason (huggingface#197) * Create config_base_math_smalllr.yaml * Update README.md * Update README.md
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Also solve the issue referenced here. We should release a stable recipe that can run end-to-end, even with a small model and dataset? This would help users better understand the workflow and facilitate learning.