529 B
529 B
| 1 | key | value |
|---|---|---|
| 2 | root_checkpoint_step | 90 |
| 3 | root_checkpoint_kind | best_validation |
| 4 | root_actor_dir | checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-4B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8/_actor_archive/global_step_90/actor |
| 5 | last_checkpoint_step | 100 |
| 6 | last_actor_dir | checkpoints/datasets/sciknoweval/biology/qwen3gen-biology-RLSD_TR-Qwen-Qwen3-4B-mbs8-decay0-tr0.1-train32-rollout8-lr1e-6-vllm0.8/global_step_100/actor |
| 7 | best_validation_mean16 | 0.485000000000 |
| 8 | final_validation_mean16 | 0.428750000000 |