{
 "setting": "ground truth SFT",
 "alpha": 0.0,
 "order": "math_code",
 "seed": 0,
 "exp": 2,
 "student": "Qwen/Qwen2.5-1.5B-Instruct",
 "gpu_student": "NVIDIA RTX A6000",
 "gpu_teachers": "NVIDIA RTX A6000",
 "load_h": 0.004249601562817892,
 "stages": {
  "stage1_math": {
   "domain": "math",
   "steps": 896,
   "train_h": 3.3275934724675285,
   "val_start": {
    "step": 0,
    "math_kl": 0.15910778235068246,
    "math_ce": 0.6515277156879916,
    "code_kl": 0.14790875576132695,
    "code_ce": 0.30340147782401555,
    "math_hard_kl": 0.24894053959908363,
    "math_hard_ce": 0.6765219455834559
   },
   "val_end": {
    "step": 896,
    "math_kl": 0.25596919708371585,
    "math_ce": 0.4875917011939159,
    "code_kl": 0.16789859193652845,
    "code_ce": 0.33237521285888244,
    "math_hard_kl": 0.3207115241939215,
    "math_hard_ce": 0.6152356422562573
   },
   "peak_gb_student": 36.754438400268555,
   "peak_gb_teachers": 29.564642429351807,
   "bench_h": 0.36716745177904764,
   "gpu_bench": "NVIDIA RTX 6000 Ada Generation"
  },
  "stage2_code": {
   "domain": "code",
   "steps": 922,
   "train_h": 3.3934424249331157,
   "val_start": {
    "step": 0,
    "math_kl": 0.2558450283886451,
    "math_ce": 0.4875823904101404,
    "code_kl": 0.16784317317180839,
    "code_ce": 0.3320483730687428,
    "math_hard_kl": 0.3206654156163548,
    "math_hard_ce": 0.6148898550851406
   },
   "val_end": {
    "step": 922,
    "math_kl": 0.22842272387506696,
    "math_ce": 0.5158765047799361,
    "code_kl": 0.375653243372234,
    "code_ce": 0.2474786401165565,
    "math_hard_kl": 0.31006984450732655,
    "math_hard_ce": 0.6471800292055488
   },
   "peak_gb_student": 33.701783180236816,
   "peak_gb_teachers": 29.222092628479004,
   "bench_h": 0.31384981168640985,
   "gpu_bench": "NVIDIA RTX A6000"
  }
 },
 "train_h": 6.721035897400644,
 "done": true
}