{
 "setting": "ground truth SFT",
 "alpha": 0.0,
 "order": "code_math",
 "seed": 1,
 "exp": 2,
 "student": "Qwen/Qwen2.5-1.5B-Instruct",
 "gpu_student": "NVIDIA RTX A6000",
 "gpu_teachers": "NVIDIA RTX A6000",
 "load_h": 0.010375746157434251,
 "stages": {
  "stage1_code": {
   "domain": "code",
   "steps": 922,
   "train_h": 3.1777428225013944,
   "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": 922,
    "math_kl": 0.18684361685993747,
    "math_ce": 0.8154262037384751,
    "code_kl": 0.343750217179818,
    "code_ce": 0.24876551347447873,
    "math_hard_kl": 0.28188513747568716,
    "math_hard_ce": 0.7975115962293889
   },
   "peak_gb_student": 33.696606159210205,
   "peak_gb_teachers": 29.222092628479004
  },
  "stage2_math": {
   "domain": "math",
   "steps": 896,
   "train_h": 2.9764579196108714,
   "val_start": {
    "step": 0,
    "math_kl": 0.18707215095927757,
    "math_ce": 0.8169582523867784,
    "code_kl": 0.34373369645959484,
    "code_ce": 0.248868452603892,
    "math_hard_kl": 0.281878406746077,
    "math_hard_ce": 0.7980098331233478
   },
   "val_end": {
    "step": 896,
    "math_kl": 0.2757146853520857,
    "math_ce": 0.48866650442357035,
    "code_kl": 0.23322471331621716,
    "code_ce": 0.2632293531111412,
    "math_hard_kl": 0.3334636116516183,
    "math_hard_ce": 0.6151642707250928
   },
   "peak_gb_student": 36.51224708557129,
   "peak_gb_teachers": 29.517569541931152,
   "bench_h": 0.3692218764623006,
   "gpu_bench": "NVIDIA RTX A6000"
  }
 },
 "train_h": 6.154200742112266,
 "done": true
}