| { |
| "format_version": 1, |
| "model": { |
| "base_model": "Qwen/Qwen3-0.6B", |
| "max_length": 512, |
| "pooling": "last_token", |
| "dropout": 0.05, |
| "freeze_backbone": true, |
| "dtype": "auto" |
| }, |
| "worker_ids": [ |
| "math_expert", |
| "code_expert", |
| "general_expert" |
| ], |
| "metadata": { |
| "stage": "sft", |
| "global_step": 15, |
| "history": [ |
| { |
| "epoch": 1, |
| "train_loss": 0.8739383816719055, |
| "validation": { |
| "examples": 6, |
| "router_utility": 0.6666659712791443, |
| "oracle_utility": 0.9999993443489075, |
| "regret": 0.3333333432674408, |
| "oracle_route_accuracy": 0.6666666865348816, |
| "best_fixed_worker": "math_expert", |
| "best_fixed_utility": 0.3333326578140259, |
| "random_utility": 0.3333326578140259, |
| "route_counts": { |
| "math_expert": 3, |
| "code_expert": 0, |
| "general_expert": 3 |
| } |
| } |
| }, |
| { |
| "epoch": 2, |
| "train_loss": 0.5408746401468912, |
| "validation": { |
| "examples": 6, |
| "router_utility": 0.8333327174186707, |
| "oracle_utility": 0.9999993443489075, |
| "regret": 0.1666666716337204, |
| "oracle_route_accuracy": 0.8333333134651184, |
| "best_fixed_worker": "math_expert", |
| "best_fixed_utility": 0.3333326578140259, |
| "random_utility": 0.3333326578140259, |
| "route_counts": { |
| "math_expert": 3, |
| "code_expert": 1, |
| "general_expert": 2 |
| } |
| } |
| }, |
| { |
| "epoch": 3, |
| "train_loss": 0.5621894498666128, |
| "validation": { |
| "examples": 6, |
| "router_utility": 0.8333327174186707, |
| "oracle_utility": 0.9999993443489075, |
| "regret": 0.1666666716337204, |
| "oracle_route_accuracy": 0.8333333134651184, |
| "best_fixed_worker": "math_expert", |
| "best_fixed_utility": 0.3333326578140259, |
| "random_utility": 0.3333326578140259, |
| "route_counts": { |
| "math_expert": 3, |
| "code_expert": 1, |
| "general_expert": 2 |
| } |
| } |
| }, |
| { |
| "epoch": 4, |
| "train_loss": 0.32038889328638714, |
| "validation": { |
| "examples": 6, |
| "router_utility": 0.9999993443489075, |
| "oracle_utility": 0.9999993443489075, |
| "regret": 0.0, |
| "oracle_route_accuracy": 1.0, |
| "best_fixed_worker": "math_expert", |
| "best_fixed_utility": 0.3333326578140259, |
| "random_utility": 0.3333326578140259, |
| "route_counts": { |
| "math_expert": 2, |
| "code_expert": 2, |
| "general_expert": 2 |
| } |
| } |
| }, |
| { |
| "epoch": 5, |
| "train_loss": 0.20881875852743784, |
| "validation": { |
| "examples": 6, |
| "router_utility": 0.9999993443489075, |
| "oracle_utility": 0.9999993443489075, |
| "regret": 0.0, |
| "oracle_route_accuracy": 1.0, |
| "best_fixed_worker": "math_expert", |
| "best_fixed_utility": 0.3333326578140259, |
| "random_utility": 0.3333326578140259, |
| "route_counts": { |
| "math_expert": 2, |
| "code_expert": 2, |
| "general_expert": 2 |
| } |
| } |
| } |
| ] |
| } |
| } |
|
|
