open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-27 17:23:19 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

tutorial(training): Add a minimal example with trl (#473)

Browse source 
* v0

* 2 gpu setup

* improve parsing from yaml

* update yaml dataset example

* remove restriction on flash attn

* more comments

* first version of the readme

* pin torch

* simplify requirements

* just flash attn

* use set env instead

* simpler set env

* readme

* add wandb project to setup

* update template

* update model id

* post init to capture the config and weight

* extract metadata

* update config

* update dataset config

* move env for wandb project

* pre-commit

* remove qwen-math from training

* more instructions

* unused import

* remove trl old

* warmup ratio

* warmup ratio

* change model id

* change model_id

* add info about CUDA_VISIBLE_DEVICES
This commit is contained in:
Zafir Stojanovski 2025-06-21 00:01:31 +02:00 • committed by GitHub
parent 49f3821098
commit 56ce2e79a7
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
59 changed files with 382 additions and 155340 deletions
Download patch file Download diff file Expand all files Collapse all files

2
training/README.md
View file

@ -2,8 +2,6 @@
Training codebase for training LLMs using Reasoning Gym procedural dataset generators.
**Note**: `qwen-math/` directory contains the code from the Tina project, used for the Qwen2.5 3B RG-Math training. This is separate from the rest of our training/evaluation codebase.
This readme documents:
- Training environment setup and usage example