TinyZero · Issues· 82 open
Open on GitHubLocally synced open issues (discussions stay on GitHub)
- #125
Did anyone try to launch this repo with a more recent version of veRL?
Updated Jun 29, 2026 - #124
Qwen2.5 1.5B with GRPO meets Entropy Collapse
Updated Jun 27, 2026 - #105
qwen2.5-3b OOM Issue on 8x V100 GPUs
Updated Mar 29, 2026 - #122
How to reproduce a qwen-3b model that supports multiplication?
Updated Dec 2, 2025 - #77
模型只会加减，不会乘除。。。
Updated Dec 2, 2025 - #92
how to run main_generation and main_eval?
Updated Dec 2, 2025 - #108
How to guarantee the problem is valid?
Updated Dec 2, 2025 - #100
How to run SFT on 2 L40 GPUs
Updated Dec 2, 2025 - #120
定义答案正确的奖励是不是应该从<answer>的标签里找
Updated Nov 7, 2025 - #57
follow the official code,got the error:because name 'global_poolverl_group_2:0' already exists
Updated Oct 2, 2025 - #7
Failed to register worker to Raylet
Updated Aug 30, 2025 - #115
RuntimeError: FlashAttention only support Ampere GPUs or newer despite setting VLLM_ATTENTION_BACKEND=XFORMERS
Updated Aug 12, 2025 - #5
1 gpu is not working , 2 gpus out of memory
Updated Aug 11, 2025 - #97
<|endoftext|>
Updated Aug 7, 2025 - #114
Support model
Updated Aug 6, 2025