YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Tiny RL math diagnostic checkpoints
Completed, superseded optimizer and attention controls. Each run used 2,097,152 SFT predictions followed by 65,536 RL rollouts. These are archived comparisons, not the selected models.
Each case includes SFT/RL weights, a SHA256 manifest and the training request. Download files from the pinned commit in the local archive.json receipt.
Data and current results: https://huggingface.co/datasets/davidheineman/tiny-rl-refined-math-v1
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support