Skip to content

Issues: Jiayi-Pan/TinyZero

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

49 Open 19 Closed
49 Open 19 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

How to debug parallel ray
#86 opened Feb 24, 2025 by HarideP
After I trained for 500 steps, the length of think became smaller and smaller, and even disappeared.
#85 opened Feb 22, 2025 by yaxundai
Shape matching error
#83 opened Feb 21, 2025 by lixiaochuan2020
Missing trainer_state.json
#82 opened Feb 21, 2025 by anavarroa
内存最小需要多少呢？为什么500G内存还是会不够？
#79 opened Feb 19, 2025 by iaoxuesheng
1
模型只会加减，不会乘除。。。
#77 opened Feb 18, 2025 by LianhaoXue
3
ray start timeout
#75 opened Feb 16, 2025 by heningsu
Ray OOM
#74 opened Feb 16, 2025 by aivolcano
PPO vs GRPO time and space efficiency
#73 opened Feb 15, 2025 by Lineark
2
When will the TinyZero be able to support multi-node train of the GRPO algorithm?
#70 opened Feb 14, 2025 by echo-valor
Finally generated a checkpoint, how do I see use the checkpoint now to see if the training succeeded?
#69 opened Feb 14, 2025 by jaganrvce1
1
[Bug?] Evaluation failed
#68 opened Feb 14, 2025 by XuhuiZhou
Why am I generating a lot of endoftext here
#66 opened Feb 12, 2025 by ZhengChenYang
6
Questions about OOM when running Qwen2.5-0.5B, 1.5B, and 3B on RTX4090 graphics cards OOM
#64 opened Feb 12, 2025 by patrickstar-sjh
10
everything is normal in the beginning until answer after <think> suddenly all become !
#63 opened Feb 12, 2025 by momo4826
6
Minimum VRAM requirements
#62 opened Feb 11, 2025 by soulde
3
ray: Fatal Python error: Floating point exception when running on H20
#61 opened Feb 11, 2025 by AkaliKong
4
What's going on here? Why is that?
#60 opened Feb 11, 2025 by Yifei-Bi
2
follow the official code,got the error:because name 'global_poolverl_group_2:0' already exists
#57 opened Feb 10, 2025 by ArlanCooper
3
How to run this on 8 GPUs
#56 opened Feb 10, 2025 by Shiki-X
7
Reproducing on H100, 4 GPUS
#54 opened Feb 9, 2025 by karrtikiyer-tw
5
Add support for PPO/GRPO with LoRA + multi-GPU finetuning
#51 opened Feb 8, 2025 by SUMEETRM
AssertionError: only support equal chunk. Got size of DataProto 2 and chunk 4.
#47 opened Feb 7, 2025 by OKC13
4
Query Regarding GRPO Implementation
#46 opened Feb 7, 2025 by anirudhb11
执行到main_ppo.py的pprint(OmegaConf.to_container(config, resolve=True))时卡住不动
#45 opened Feb 6, 2025 by L-WWEEII
ProTip! Mix and match filters to narrow down what you’re looking for.