Feature/add remax support #234
Merged
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
Collaborator
|
Hi @liziniu Thank you for your contribution! According to our implementation, I guess ReMax can be implemented by adding a few lines to the original PPO/GRPO/Reinforce implementation instead of writing a new trainer to make maintenance easier. Correct me if this is invalid |
Collaborator
|
+1. From my understanding, Remax can be implemented similarly to reinforce++ with a different adv estimator. See reinforce++ implementation: #228 |
| - huggingface models support | ||
| - Supervised fine-tuning | ||
| - Reinforcement learning from human feedback with [PPO](https://github.com/volcengine/verl/tree/main/examples/ppo_trainer) and [GRPO](https://github.com/volcengine/verl/tree/main/examples/grpo_trainer) | ||
| - Reinforcement learning from human feedback with [PPO](https://github.com/volcengine/verl/tree/main/examples/ppo_trainer), [GRPO](https://github.com/volcengine/verl/tree/main/examples/grpo_trainer), and [ReMax](https://github.com/volcengine/verl/tree/main/examples/remax_trainer) |
Collaborator
There was a problem hiding this comment.
if you have the training log and wandb already, would you mind adding one more record to docs/experiment/ppo.rst to include remax? it would help the community to track if experiment can be reproduced in future version.
We can do that in the next PR
Contributor
Author
There was a problem hiding this comment.
Yes. A preliminary result on Qwen2.5-3B is added and more results will come later.
Contributor
Author
|
@vermouth1992 @PeterSH6 I see. Let me reformat the code with minimal changes of the PPO's trainer. |
Contributor
Author
|
I have completed the implementation of ReMaX support. The changes include:
The code follows the project's style guidelines. Please review when you have a chance. Let me know if any changes or clarifications are needed. Thank you for your time! |
PeterSH6
requested changes
Feb 10, 2025
PeterSH6
approved these changes
Feb 10, 2025
Collaborator
|
Could you add a CI to run remax with Qwen 0.5b to protect this functionality? You can follow the example here: https://github.com/volcengine/verl/blob/main/.github/workflows/e2e_gsm8k.yml#L69 |
Contributor
Author
|
I’ve updated the end-to-end (e2e) test. Since this is my first time writing such a test within the workflows, I’m not entirely confident about it. Could you please review it when you have a moment? Thank you! |
sunyi0505
pushed a commit
to sunyi0505/verl
that referenced
this pull request
Feb 20, 2025
## Description Added [ReMax](https://arxiv.org/abs/2310.10505) support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction. The [HybridFlow](https://arxiv.org/pdf/2409.19256v2) paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added. ## Changes - Added RayReMaxTrainer implementation - Added example scripts for ReMax training - Added documentation for ReMax algorithm ## Testing - Tested ReMax example scripts with Qwen models validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset: <img width="501" alt="截屏2025-02-09 20 51 14" src="https://github.com/user-attachments/assets/742c2eab-6877-4c3c-b0a2-4159bd109add" /> The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning. ## Documentation - Added ReMax documentation - Updated example configurations ## Checklist - [x] Code follows project's style guidelines (yapf formatted) - [x] Tests added/updated and passing - [x] Documentation updated - [x] Example scripts added
yuchenwang3
pushed a commit
to yuchenwang3/verl
that referenced
this pull request
Apr 25, 2025
## Description Added [ReMax](https://arxiv.org/abs/2310.10505) support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction. The [HybridFlow](https://arxiv.org/pdf/2409.19256v2) paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added. ## Changes - Added RayReMaxTrainer implementation - Added example scripts for ReMax training - Added documentation for ReMax algorithm ## Testing - Tested ReMax example scripts with Qwen models validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset: <img width="501" alt="截屏2025-02-09 20 51 14" src="https://github.com/user-attachments/assets/742c2eab-6877-4c3c-b0a2-4159bd109add" /> The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning. ## Documentation - Added ReMax documentation - Updated example configurations ## Checklist - [x] Code follows project's style guidelines (yapf formatted) - [x] Tests added/updated and passing - [x] Documentation updated - [x] Example scripts added
histmeisah
pushed a commit
to SJTU-IAAR/verl
that referenced
this pull request
Apr 27, 2025
## Description Added [ReMax](https://arxiv.org/abs/2310.10505) support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction. The [HybridFlow](https://arxiv.org/pdf/2409.19256v2) paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added. ## Changes - Added RayReMaxTrainer implementation - Added example scripts for ReMax training - Added documentation for ReMax algorithm ## Testing - Tested ReMax example scripts with Qwen models validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset: <img width="501" alt="截屏2025-02-09 20 51 14" src="https://github.com/user-attachments/assets/742c2eab-6877-4c3c-b0a2-4159bd109add" /> The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning. ## Documentation - Added ReMax documentation - Updated example configurations ## Checklist - [x] Code follows project's style guidelines (yapf formatted) - [x] Tests added/updated and passing - [x] Documentation updated - [x] Example scripts added
chenjiaoAngel
added a commit
to chenjiaoAngel/verl
that referenced
this pull request
Nov 14, 2025
## Description Added [ReMax](https://arxiv.org/abs/2310.10505) support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction. The [HybridFlow](https://arxiv.org/pdf/2409.19256v2) paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added. ## Changes - Added RayReMaxTrainer implementation - Added example scripts for ReMax training - Added documentation for ReMax algorithm ## Testing - Tested ReMax example scripts with Qwen models validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset: <img width="501" alt="截屏2025-02-09 20 51 14" src="https://github.com/user-attachments/assets/742c2eab-6877-4c3c-b0a2-4159bd109add" /> The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning. ## Documentation - Added ReMax documentation - Updated example configurations ## Checklist - [x] Code follows project's style guidelines (yapf formatted) - [x] Tests added/updated and passing - [x] Documentation updated - [x] Example scripts added
TimurTaepov
pushed a commit
to giorgossideris/verl
that referenced
this pull request
Dec 20, 2025
## Description Added [ReMax](https://arxiv.org/abs/2310.10505) support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction. The [HybridFlow](https://arxiv.org/pdf/2409.19256v2) paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added. ## Changes - Added RayReMaxTrainer implementation - Added example scripts for ReMax training - Added documentation for ReMax algorithm ## Testing - Tested ReMax example scripts with Qwen models validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset: <img width="501" alt="截屏2025-02-09 20 51 14" src="https://github.com/user-attachments/assets/742c2eab-6877-4c3c-b0a2-4159bd109add" /> The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning. ## Documentation - Added ReMax documentation - Updated example configurations ## Checklist - [x] Code follows project's style guidelines (yapf formatted) - [x] Tests added/updated and passing - [x] Documentation updated - [x] Example scripts added
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Description
Added ReMax support to verl. ReMax is a simple, efficient, and stable RL algorithm customized for LLM training, with theoretical guarantees for variance reduction.
The HybridFlow paper experimented with ReMax, but verl did not provide an implementation. Therefore, ReMax has been added.
Changes
Testing
validation reward of optimizing Qwen2.5-3B-Instruct on the GSM8K dataset:
The curve demonstrates the effectiveness of ReMax, though its performance can be further enhanced through hyperparameter fine-tuning.
Documentation
Checklist