open-r1 · Issues· 340 open

New aproach to developing AGI

in SFT script, distributed training got stuck if set `packing=false`

RuntimeError: The size of tensor a (0) must match the size of tensor b (5120) at non-singleton dimension 1

Dependency conflicts

Bug in SFT script

Is there a Chinese version of the dataset in this project?

SFT Trains the Model on the Entire Sequence

Forward reward always 0

Checkpoint selection when using cot sft

why kl = nan when grpo train?

Request for License Information for Mixture-of-Thoughts Dataset

Instead of rising steadily, the reward fluctuates wildly

Does the SFT Framework Support Fine-Tuning for DeepSeek-R1-Distill-Qwen-7B

Question about evaluating AIME24 Accuracy

vllm prepends two BOS for LLama