Downloads · 30 days
11
17% of all-time downloads
mathnano-qwen1.5b-grpo
adimunot/mathnano-qwen1.5b-grpo
mathnano-qwen1.5b-grpo is a text generation model from adimunot. Use it when you need the model to write or continue text. It is set up for peft. The card lists the license as apache-2.0.
A GRPO (RL) LoRA adapter on top of the MathNano SFT model. This run degraded the model and is kept only for the write-up — do not use it for inference. Use the SFT adapter instead.
All-time downloads
65
Public
Repo size
455 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt296 MB · 52%
At a glance
- Task
- Text Generation
- Library
- peft
- License
- apache-2.0
- Access
- Public
- Created
- Jun 27, 2026
- Updated
- Jun 27, 2026
- SHA
- 7d359e36
Try a prompt
Base models
Model card
From the Hugging Face model README
MathNano — GRPO run (⚠️ failed / archived)
A GRPO (RL) LoRA adapter on top of the MathNano SFT model. This run degraded the model and is kept only for the write-up — do not use it for inference. Use the SFT adapter instead.
What happened
| GSM8K | MATH | |
|---|---|---|
| SFT (before) | 39.0% | 40.0% |
| **GRPO (this) ** | 5.5% | 5.5% |
Root cause
On a single GPU we couldn't run vLLM rollouts, so GRPO used HF generation that produced fixed-length completions which never stopped at the answer. The verifiable reward (which reads the last boxed value) then frequently graded correct solutions as wrong, so GRPO optimized a corrupted signal and collapsed the policy. The lesson — GRPO is only as good as its reward's fidelity — is the most useful outcome of the run.
Full post-mortem: https://github.com/adimunot21/mathnano/blob/main/RESULTS.md
- Task
- Text Generation
- Library
- peft
- License
- apache-2.0
- Created
- Jun 27, 2026
- Updated
- Jun 27, 2026