| license: other | |
| tags: | |
| - rl-forgetting | |
| - filtered-sft | |
| # Coding SFT baseline delivery, October 27 2026 | |
| Incremental delivery from the two missing Llama coding SFT baseline arms (the other 13 requested arms were previously delivered to mbpp-code-sft-artifacts). Incomplete runs must not be treated as complete studies. Check the dataset delivery_manifest.json for verified upload units and hashes. Model checkpoints are under runs/checkpoints/step_N/model; final models under runs/final. Load with the corresponding subfolder. | |
| Built with Qwen: Qwen2.5-3B derivatives retain the upstream Qwen RESEARCH LICENSE AGREEMENT. Built with Llama: Llama derivatives use Llama 3.2 and subject to the Llama 3.2 Community License and Acceptable Use Policy; see https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct. No additional rights to base models are granted. | |