train_stats.json

{
 "base": "google/gemma-3-12b-it",
 "method": "fsdp2-custom-loop",
 "n_rows": 41628,
 "n_self": 0,
 "n_sequences": 20392,
 "n_tokens": 32595992,
 "world": 4,
 "micro_batch": 4,
 "accum": 4,
 "steps": 318,
 "lr": 1e-05,
 "optim": "AdamW fused, fp32 master, bf16 compute, fp32 grad reduce",
 "loss_curve": [
  2.23734188079834,
  1.8095881938934326,
  1.6590197086334229,
  1.5920462608337402,
  1.549424171447754,
  1.5527774095535278,
  1.407991886138916,
  1.5068154335021973,
  1.5217790603637695,
  1.4271552562713623,
  1.4941649436950684,
  1.4286973476409912,
  1.4256829023361206,
  1.433403491973877,
  1.5622130632400513,
  1.4556655883789062,
  1.4827250242233276,
  1.3963528871536255,
  1.4398143291473389,
  1.5313364267349243,
  1.3986570835113525,
  1.3420710563659668,
  1.4838554859161377,
  1.4589663743972778,
  1.392141342163086,
  1.3990843296051025,
  1.4269590377807617,
  1.4152247905731201,
  1.456050992012024,
  1.478327751159668,
  1.3464207649230957,
  1.3860893249511719,
  1.3910554647445679,
  1.3571606874465942,
  1.3460079431533813,
  1.3818973302841187,
  1.380621314048767,
  1.3064377307891846,
  1.3707051277160645,
  1.4148029088974,
  1.3994386196136475,
  1.4554977416992188,
  1.4585635662078857,
  1.4156562089920044,
  1.4112359285354614,
  1.4267137050628662,
  1.3385674953460693,
  1.3928370475769043,
  1.3970692157745361,
  1.3952012062072754,
  1.4118852615356445,
  1.4345245361328125,
  1.3758606910705566,
  1.336012840270996,
  1.3574304580688477,
  1.4304163455963135,
  1.389248013496399,
  1.2987174987792969,
  1.3651535511016846,
  1.4056745767593384,
  1.3437752723693848,
  1.3492906093597412,
  1.3777880668640137,
  1.4133349657058716
 ],
 "loss_first": [
  2.23734188079834
 ],
 "loss_last": [
  1.4133349657058716
 ],
 "train_seconds": 3319.9620678424835,
 "tokens_per_second": 9818.182055671163,
 "resumed_from": null
}