训练成本最低 0.5B R1-Zero
active 2025-02-11 → 2026-05-02 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 446 days from 2025-02-11 to 2026-05-02. Pushes: 17 total, peak 5 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 44 total, peak 7 in a day. Comments: 83 total, peak 19 in a day. Stars: 390 total, peak 88 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| dhcode-cpp | 77 | 17 | 0 | 45 |
| hengzhi40 | 5 | 0 | 0 | 4 |
| Jason8Kang | 5 | 0 | 0 | 4 |
| lky-violet | 4 | 0 | 0 | 1 |
| glennccc | 4 | 0 | 0 | 3 |
| haohaotian1011 | 4 | 0 | 0 | 2 |
| nkqiaolin | 3 | 0 | 0 | 2 |
| Minami-su | 3 | 0 | 0 | 2 |
| tsrigo | 3 | 0 | 0 | 3 |
| KOBEliever | 2 | 0 | 0 | 0 |
| zhen0qiang | 2 | 0 | 0 | 1 |
| wulixianming | 2 | 0 | 0 | 1 |
| women1995 | 2 | 0 | 0 | 2 |
| Kryptonites | 2 | 0 | 0 | 1 |
| loxs123 | 2 | 0 | 0 | 1 |
| liuqiangh | 2 | 0 | 0 | 1 |
| mmxuan18 | 2 | 0 | 0 | 1 |
| veolata | 1 | 0 | 0 | 0 |
| hwei-hw | 1 | 0 | 0 | 0 |
| KyrieXiao25 | 1 | 0 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#72DengNingyuan2025-11-11 06:22Problem with code
- Issue comment#66Minami-su2025-07-15 06:50Using lora for training is not effective
- Issue comment#66Minami-su2025-07-11 01:59Using lora for training is not effective
- Issue#71Minami-su2025-07-11 01:46The hostname of the client socket cannot be retrieved. err=-3
- Issue comment#64ShuoZheLi2025-03-18 04:59So many bugs in the new version code
- Issue comment#65hengzhi402025-03-13 03:23RuntimeError: 'weight' must be 2-D
- Issue comment#64lyh10282025-03-13 02:59So many bugs in the new version code
- Issue#62gumanchang2025-03-09 01:30SCQ5K，训练出来的曲线看不懂啊？
- Issue comment#58wulixianming2025-03-02 08:57无法复现X-R1-3B在Math500数据集上精度
- Issue#61mysuns2025-03-02 04:43self.accelerator.unwrap_model(self.model).disable_adapters() ??
- Issue#60KyrieXiao252025-02-26 06:57Use Qwen2.5-3B-Instruct get warning: flash attention 2.0 with a model not initialized on GPU
- Issue comment#58dhcode-cpp2025-02-26 03:25无法复现X-R1-3B在Math500数据集上精度
- Issue comment#58Jason8Kang2025-02-26 03:19无法复现X-R1-3B在Math500数据集上精度
- Issue comment#58dhcode-cpp2025-02-26 03:06无法复现X-R1-3B在Math500数据集上精度
- Issue comment#58dhcode-cpp2025-02-26 03:05无法复现X-R1-3B在Math500数据集上精度
- Issue comment#59dhcode-cpp2025-02-26 03:02训练过程中Qwen-7B模型输出无法获得有效结果，使得Reward始终为0.
- Issue comment#30dhcode-cpp2025-02-26 03:01看readme里面中文的ahamoment，但训练数据都是全英文的，这个是模型自主学习到中文的推理能力？
- Issue comment#58women19952025-02-26 02:48无法复现X-R1-3B在Math500数据集上精度
- Issue comment#58Jason8Kang2025-02-26 02:33无法复现X-R1-3B在Math500数据集上精度
- Issue#59taobiaoli13142025-02-26 02:32训练过程中Qwen-7B模型输出无法获得有效结果，使得Reward始终为0.
- Issue comment#58dhcode-cpp2025-02-26 02:31无法复现X-R1-3B在Math500数据集上精度
- Issue comment#56tsrigo2025-02-26 02:283B X-R1的math-500的benchmark上微调后性能是下降的
- Issue comment#58Jason8Kang2025-02-26 02:25无法复现X-R1-3B在Math500数据集上精度
- Issue comment#56dhcode-cpp2025-02-26 02:183B X-R1的math-500的benchmark上微调后性能是下降的
- Issue comment#58dhcode-cpp2025-02-26 02:15无法复现X-R1-3B在Math500数据集上精度
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 390 stars here means stars gained during the window, not the repo's star count.