mirror of
https://github.com/volcengine/verl.git
synced 2025-10-20 21:53:50 +08:00
### What does this PR do? Fix some bugs/outdated code so that we can use sandbox fusion for DAPO. ### Checklist Before Starting - [X] Search for similar PRs. Paste at least one query link here: ... - [X] Format the PR title as `[{modules}] {type}: {description}` (This will be checked by the CI) - `{modules}` include `fsdp`, `megatron`, `sglang`, `vllm`, `rollout`, `trainer`, `ci`, `training_utils`, `recipe`, `hardware`, `deployment`, `ray`, `worker`, `single_controller`, `misc`, `perf`, `model`, `algo`, `env`, `tool`, `ckpt`, `doc`, `data` - If this PR involves multiple modules, separate them with `,` like `[megatron, fsdp, doc]` - `{type}` is in `feat`, `fix`, `refactor`, `chore`, `test` - If this PR breaks any API (CLI arguments, config, function signature, etc.), add `[BREAKING]` to the beginning of the title. - Example: `[BREAKING][fsdp, megatron] feat: dynamic batching` ### Test > For changes that can not be tested by CI (e.g., algorithm implementation, new model support), validate by experiment(s) and show results like training curve plots, evaluation results, etc. ### API and Usage Example > Demonstrate how the API changes if any, and provide usage example(s) if possible. ```python # Add code snippet or script demonstrating how to use this ``` ### Design & Code Changes Use `load_reward_manager` in `verl.trainer.ppo.reward` instead of duplicating the code there. Also, set `acc` in `reward_extra_info` when the returned result is only a float number (e.g. sandbox fusion). ### Checklist Before Submitting > [!IMPORTANT] > Please check all the following items before requesting a review, otherwise the reviewer might deprioritize this PR for review. - [X] Read the [Contribute Guide](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md). - [X] Apply [pre-commit checks](https://github.com/volcengine/verl/blob/main/CONTRIBUTING.md#code-linting-and-formatting): `pre-commit install && pre-commit run --all-files --show-diff-on-failure --color=always` - [X] Add / Update [the documentation](https://github.com/volcengine/verl/tree/main/docs). - [X] Add unit or end-to-end test(s) to [the CI workflow](https://github.com/volcengine/verl/tree/main/.github/workflows) to cover all the code. If not feasible, explain why: ... - [X] Once your PR is ready for CI, send a message in [the `ci-request` channel](https://verl-project.slack.com/archives/C091TCESWB1) in [the `verl` Slack workspace](https://join.slack.com/t/verl-project/shared_invite/zt-3855yhg8g-CTkqXu~hKojPCmo7k_yXTQ). Signed-off-by: Hollow Man <hollowman@opensuse.org>
Recipe
The examples under recipes/
are representative extensions to verl for specific end-to-end RL training recipes.
The help the community reproduce experiments, verl team provides a snapshot of the codebase when each recipe is initially PR'ed to verl main. You can find them via github branches
Awesome work using verl
- Logic-RL: a reproduction of DeepSeek R1 Zero on 2K Tiny Logic Puzzle Dataset.
- Seed-Coder: RL training of Seed-Coder boosts performance on competitive programming
- all-hands/openhands-lm-32b-v0.1: A strong, open coding agent model, trained with multi-turn fine-tuning
- s3 Efficient Yet Effective Search Agent Training via RL
- Rec-R1: Bridging Generative Large Language Models and Recommendation Systems via Reinforcement Learning
- Explore RL Data Scaling: Exploring Data Scaling Trends and Effects in Reinforcement Learning from Human Feedback
- FIRE: Flaming-hot initiation with regular execution sampling for large language models
- DQO: Enhancing multi-Step reasoning abilities of language models through direct Q-function optimization
- ProRL: Prolonged Reinforcement Learning Expands Reasoning Boundaries in Large Language Models
- cognition-engineering: Test time scaling drives cognition engineering.
- Trust Region Preference Approximation: A simple and stable reinforcement learning algorithm for LLM reasoning.
- AdaRFT: Efficient Reinforcement Finetuning via Adaptive Curriculum Learning
- critic-rl: LLM critics for code generation
- self-rewarding-reasoning-LLM: self-rewarding and correction with generative reward models
- DeepEnlighten: Reproduce R1 with social reasoning tasks and analyze key findings
- MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse
- PURE: Credit assignment is the key to successful reinforcement fine-tuning using process reward model
- cognitive-behaviors: Cognitive Behaviors that Enable Self-Improving Reasoners, or, Four Habits of Highly Effective STaRs
- deepscaler: iterative context scaling with GRPO
- DAPO: the fully open source SOTA RL algorithm that beats DeepSeek-R1-zero-32B