Repository Issue Activity (beta)

#65 Question about Response Length

#99 Performance of Qwen-2.5-32B-SimpleRL-Zoo

#98 Actor was died.

#97 On which benchmark did the results of plot_figure1_v2.3_token_length_vs_steps.png come from?

#96 [Feedback] Replication Report: Solved Dependency Hell, Data Loading, and Hardcoding Issues

#61 can I have the training config for Qwen 0.5B and 1.5B

#80 Actor died without any reason

#95 Timeout error of math_verify

#94 Ray runtime error fix: install specific Click version

#93 Code Execution is still enabled during evaluation?

#77 math_verify and latex2sympy2_extended need python>=3.10,

#92 evaluation script to test Meta-Llama-3.1 8B(base model step 0) on GSM8K,results differ significantly acc 39.7(your) vs 24.1(mine)

#91 单机八卡a100，qwen7B，跑了半天左右到step16，mem内存占用攀升到1.6T 爆了程序终止，请问怎么处理

#88 Evaluation Details on IFEVAL, MMLU, and GPQA-Diamond

#89 aiohttp.web_exceptions.HTTPRequestEntityTooLarge: Request Entity Too Large

#90 Question: Response length not increase

#59 Question about generalization

#87 Prompt Template for training Llama Family Models

#86 Llama-3.1-8B recipe

#85 Over-Reliance on Format Rewards

#84 Actors died: CPU Out of Memory when performing actor_rollout_generate_sequences

#83 Discussion on Effective Training Strategies for Small LLMs

#82 plan to support NPU ？

#62 Math500 Dataset

#72 Question about rollout