| { |
| "setting": "ground truth SFT", |
| "alpha": 0.0, |
| "order": "math_code", |
| "seed": 1, |
| "exp": 2, |
| "student": "Qwen/Qwen2.5-1.5B-Instruct", |
| "gpu_student": "NVIDIA RTX A6000", |
| "gpu_teachers": "NVIDIA RTX 6000 Ada Generation", |
| "load_h": 0.00412350767188602, |
| "stages": { |
| "stage1_math": { |
| "domain": "math", |
| "steps": 896, |
| "train_h": 3.0991316720512176, |
| "val_start": { |
| "step": 0, |
| "math_kl": 0.15874307828215575, |
| "math_ce": 0.6515277156879916, |
| "code_kl": 0.14786912634071003, |
| "code_ce": 0.30340147782401555, |
| "math_hard_kl": 0.24940142917923272, |
| "math_hard_ce": 0.6765219455834559 |
| }, |
| "val_end": { |
| "step": 896, |
| "math_kl": 0.24743005021575165, |
| "math_ce": 0.488097313558479, |
| "code_kl": 0.16872722467163054, |
| "code_ce": 0.33464992453996345, |
| "math_hard_kl": 0.32093248237493643, |
| "math_hard_ce": 0.6126785609887812 |
| }, |
| "peak_gb_student": 36.57978296279907, |
| "peak_gb_teachers": 30.618666172027588, |
| "bench_h": 0.5074694540103276, |
| "gpu_bench": "NVIDIA RTX A6000" |
| }, |
| "stage2_code": { |
| "domain": "code", |
| "steps": 922, |
| "train_h": 3.1925682120190726, |
| "val_start": { |
| "step": 0, |
| "math_kl": 0.24741131268995628, |
| "math_ce": 0.4883952099318387, |
| "code_kl": 0.16862850879186456, |
| "code_ce": 0.3346423738554089, |
| "math_hard_kl": 0.32091368792296465, |
| "math_hard_ce": 0.6125312678001616 |
| }, |
| "val_end": { |
| "step": 922, |
| "math_kl": 0.2274085356023099, |
| "math_ce": 0.5163370611763011, |
| "code_kl": 0.3787724111306608, |
| "code_ce": 0.24705130277005616, |
| "math_hard_kl": 0.3121011627936271, |
| "math_hard_ce": 0.6428350913066744 |
| }, |
| "peak_gb_student": 33.86883306503296, |
| "peak_gb_teachers": 30.005428314208984, |
| "bench_h": 0.3323956937260098, |
| "gpu_bench": "NVIDIA RTX A6000" |
| } |
| }, |
| "train_h": 6.29169988407029, |
| "done": true |
| } |
