Downloads · 30 days
10
34% of all-time downloads
autocode-rl-gptoss20b-synthetic
Yanagi-Origami/autocode-rl-gptoss20b-synthetic
autocode-rl-gptoss20b-synthetic is a reinforcement learning model from Yanagi-Origami. Use it for the reinforcement learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
GPT-OSS-20B after full-parameter GRPO against a deterministic execution judge, trained on 147 frontier-written rewrites of real Codeforces problems.
All-time downloads
29
Public
Parameters
20.9B
41.9 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors41.8 GB · 100%
At a glance
- Task
- Reinforcement Learning
- License
- apache-2.0
- Model type
- gpt_oss
- Access
- Public
- Created
- Sep 18, 2026
- Updated
- Sep 18, 2026
- SHA
- df0492a6
Base models
Model card
From the Hugging Face model README
autocode-rl-gptoss20b-synthetic
GPT-OSS-20B after full-parameter GRPO against a deterministic execution judge, trained on 147 frontier-written rewrites of real Codeforces problems.
- Expert weights are stored fused (
mlp.experts.gate_up_proj, 411 tensors), the layout the official gpt-oss release uses and the only one SGLang's disk loader understands. The per-expert layout that Megatron exporters produce loads without error and then emits gibberish, so it is deliberately not what is published here. - Endpoint checkpoint of a step budget fixed in advance: no selection on any evaluation set.
- Evaluation protocol, and the judge, are described in the accompanying paper.
| LCB-Pro easy (391) | LCB-Pro medium (251) | |
|---|---|---|
| base gpt-oss-20B | 68.0 | 18.8 |
| this checkpoint | 77.2 | 22.5 |
Pass@1, mean of 4 samples, C++ solutions graded by the benchmark's official tests.
- Task
- Reinforcement Learning
- Type
- gpt_oss
- License
- apache-2.0
- Created
- Sep 18, 2026
- Updated
- Sep 18, 2026