Trinity-RFT is a general-purpose, flexible and scalable framework designed for reinforcement fine-tuning (RFT) of large language models (LLM).
active 2026-01-16 → 2026-07-20 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 186 days from 2026-01-16 to 2026-07-20. Pushes: 83 total, peak 5 in a day. Pull requests: 23 total, peak 3 in a day. Issues: 4 total, peak 1 in a day. Comments: 149 total, peak 12 in a day. Stars: 34 total, peak 3 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 |
|---|---|---|---|---|
| pan-x-c | 95 | 26 | 9 | 41 |
| github-actions[bot] | 77 | 39 | 0 | 38 |
| chenyushuo | 44 | 11 | 4 | 19 |
| gemini-code-assist[bot] | 39 | 0 | 0 | 30 |
| hiyuchang | 15 | 1 | 3 | 9 |
| Copilot | 9 | 1 | 0 | 5 |
| yanxi-chen | 6 | 4 | 0 | 2 |
| shiweijiezero | 3 | 0 | 3 | 0 |
| AdnanQureshi3 | 3 | 0 | 1 | 1 |
| SeoMP | 2 | 0 | 0 | 1 |
| yysxjz | 2 | 0 | 0 | 1 |
| kokolerk | 1 | 0 | 1 | 0 |
| MengsD | 1 | 1 | 0 | 0 |
| aretaafandi16-ui | 1 | 0 | 1 | 0 |
| artaasd95 | 1 | 0 | 1 | 0 |
| devnomad-byte | 1 | 0 | 0 | 1 |
| garyzhang99 | 1 | 0 | 0 | 0 |
| NielsRogge | 1 | 0 | 0 | 0 |
| songyang86 | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#591github-actions[bot]2026-07-02 05:24Rollout Models support recording experience data
- Issue comment#591pan-x-c2026-06-30 12:24vLLM OpenAI API server supports recording experience data
- Issue comment#597github-actions[bot]2026-06-30 05:58Fix `build_mm_input_for_training` for video
- Pull request#587pan-x-c2026-06-25 03:15
- Pull request#580shiweijiezero2026-06-18 04:50
- Issue comment#570github-actions[bot]2026-06-06 12:38Update VERL to 0.8.0
- Issue comment#570pan-x-c2026-06-06 06:59Update VERL to 0.8.0
- Issue comment#562pan-x-c2026-06-01 02:03fix: propagate the optional 'group' config field across paths, namespaces & docs
- Pull request#562shiweijiezero2026-05-30 01:25
- Pull request#555pan-x-c2026-05-26 12:18
- Issue comment#547github-actions[bot]2026-05-26 08:05Enhance multi-modal for vllm model
- Issue comment#553pan-x-c2026-05-26 07:07Fix Tinker API
- Issue comment#547chenyushuo2026-05-26 05:29Enhance multi-modal for vllm model
- Issue comment#553pan-x-c2026-05-26 05:21Fix Tinker API
- Pull request#552artaasd952026-05-25 17:19
- Pull request#546pan-x-c2026-05-25 07:23
- Issue comment#546github-actions[bot]2026-05-25 06:41Update Megatron
- Pull request#549chenyushuo2026-05-25 03:53
- Issue comment#546pan-x-c2026-05-22 06:43Update Megatron
- Issue comment#470pan-x-c2026-05-22 02:13Good First Issue List
- Issue comment#545github-actions[bot]2026-05-21 02:33Enhance multi-turn multi-modal sft
- Issue#544pan-x-c2026-05-20 11:23请问支持VLM吗
- Pull request#538pan-x-c2026-05-18 03:10
- Issue comment#536github-actions[bot]2026-05-12 13:44SGLang OpenAI API Server returns `prompt_token_ids` and `token_ids`
- Pull request#535shiweijiezero2026-05-09 11:35
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 34 stars here means stars gained during the window, not the repo's star count.