Repository Issue Activity (beta)

#727 Eval datasets for reasoning

#685 Interest in dataset based on ACM SGU?

#723 New aproach to developing AGI

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

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

#719 Dependency conflicts

#341 Bug in SFT script

#239 Why does the loss start at 0 when I train GRPO, and then possibly increase?

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

#714 grad_norm nan

#713 grad_norm nan

#712 grad_norm nan

#118 Facing OOM during DRPO stage

#710 SFT Trains the Model on the Entire Sequence

#707 Forward reward always 0

#705 Checkpoint selection when using cot sft

#704 why kl = nan when grpo train?

#702 Request for License Information for Mixture-of-Thoughts Dataset

#694 GSPO update request

#403 Instead of rising steadily, the reward fluctuates wildly

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

#698 Question about evaluating AIME24 Accuracy

#695 vllm prepends two BOS for LLama

#684 Latest version of flash_attn no longer works with torch==2.6.0

#444 How to increase the context window from 4k to 32k on qwen models ?