BytedTsinghua-SIA/DAPO
An Open-source RL System from ByteDance Seed and Tsinghua AIR
About the project
Open-source reinforcement learning system for large-scale LLM training, including the DAPO algorithm, code infrastructure, and dataset, plus released DAPO-Qwen-32B weights.
Useful for
- Train a reasoning LLM with the DAPO RL algorithm on your own dataset
- Evaluate DAPO-Qwen-32B on AIME 2024 using the provided Ray Serve and vLLM scripts
- Run inference with the released DAPO-Qwen-32B weights via vLLM
README summarized by DeepSeek V4.1 Flash. Details may be inaccurate.
Why it’s trending
- 5 stars so far today, about 16 expected by the end of the day. The usual pace is 1 per day, so that's 16× as much.
- Before this, the repository barely got any stars — about 1 per day.
- The spike has held for 2 days in a row — not a one-off blip.
- Hacker News: “DAPO: An Open-source RL System from ByteDance Seed and Tsinghua AIR” — 20 points, 13 h ago.
Stars per day
Bars are daily stars, the line is the usual pace. Red marks spike days.
Numbers
- Total stars
- 1,913
- Stars in a day
- 16
- Forks
- 90
- Issues and pull requests
- 40
- Watchers
- 14
- Language
- Python
- Created
- March 17, 2025
- Last push
- May 11, 2025
Star trust
Growth looks organic: forks and discussion are in line with active projects, and stars arrive unevenly, the way people give them.
These are heuristics, not a verdict: we judge by the repository’s behavior, not by a list of stargazers.
Hacker News discussions
- DAPO: An Open-source RL System from ByteDance Seed and Tsinghua AIR20 points, September 20, 2026
Spotted in
- September 21, 2026Spotted on Hacker News
Share
README badge
Paste it into your README — the badge shows the current magnitude and links to this page.
[](https://gitnova.dev/en/r/BytedTsinghua-SIA/DAPO)Similar projects
-
9.1
mizorewww/laya-mlx
Native MLX runtime for Laya typed decision models: returns probabilities for choices, scores or truth values without text generation, running locally on Apple Silicon.
-
8.3
jaredpalmer/kev
kev is a LoRA adapter with a small readout head on top of Qwen2.5-0.5B that answers many typed questions about a document in a single forward pass, returning calibrated probabilities instead of text.
-
7.2
mizorewww/laya-coreml
Local port of the Laya model to Apple Core ML and Neural Engine: returns typed decisions (choice, score, yes/no) without token generation, with speed and energy benchmarks.
-
7.2
bespokelabsai/nimble
Nimble is a model and training recipe for fast typed decisions over text: given a flat schema of enum and boolean fields, it picks an answer and returns probabilities for each option. It targets routing, condition…
-
7.1
TheoLeeCJ/SemIf
An open reproduction of the Jev-style semantic decision interface: reads typed option probabilities directly from a 4B model's logits without generating text. Runs on a single RTX 3090.
-
7.0
FareedKhan-dev/train-llm-from-scratch
Educational project: a from-scratch PyTorch Transformer plus a full LLM training pipeline — from raw text through SFT, reward modeling, PPO/DPO/GRPO to chat, without transformers, trl or peft.