Qwen2.5-3B math SFT, ordered: all checkpoints
Ten existing checkpoints: 107, 214, 322, 429, 536, 643, 750, 858, 965, 1072. Training and evaluation were already complete.
Per-checkpoint evaluation outputs: https://huggingface.co/datasets/RL-Forgetting-Experiments-3/qwen2.5-3b-math-kk-sft-artifacts/tree/main/runs/math_ordered/eval
Each checkpoints/step_N/ directory is a directly loadable Hugging Face model.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for RL-Forgetting-Experiments-3/qwen2.5-3b-math-sft-ordered-lr1e5-all-checkpoints
Base model
Qwen/Qwen2.5-3B