train_stats.json

{
 "base": "google/gemma-4-31B-it",
 "method": "fsdp2-custom-loop",
 "n_rows": 42971,
 "n_self": 0,
 "n_sequences": 20448,
 "n_tokens": 32873760,
 "world": 8,
 "micro_batch": 4,
 "accum": 2,
 "steps": 319,
 "lr": 1e-05,
 "optim": "AdamW fused, fp32 master, bf16 compute, fp32 grad reduce",
 "loss_curve": [
  3.6044492721557617,
  2.186145544052124,
  7.073915481567383,
  1.7412365674972534,
  1.8195652961730957,
  1.5738166570663452,
  1.651289701461792,
  1.5561672449111938,
  1.6007440090179443,
  1.5938689708709717,
  1.4165937900543213,
  1.4714698791503906,
  1.5280872583389282,
  1.4857213497161865,
  1.4116941690444946,
  1.4433680772781372,
  1.5302860736846924,
  1.3890833854675293,
  1.4529991149902344,
  1.4417716264724731,
  1.4751696586608887,
  1.4375733137130737,
  1.4569019079208374,
  1.4255566596984863,
  1.3930319547653198,
  1.4071993827819824,
  1.375950574874878,
  1.3889439105987549,
  1.420447826385498,
  1.3939931392669678,
  1.4280518293380737,
  1.4625126123428345,
  1.3886096477508545,
  1.3517074584960938,
  1.3598493337631226,
  1.3748399019241333,
  1.3897030353546143,
  1.401116132736206,
  1.377851128578186,
  1.3292362689971924,
  1.35129714012146,
  1.3577051162719727,
  1.379685401916504,
  1.3450946807861328,
  1.326094388961792,
  1.3895275592803955,
  1.3360313177108765,
  1.4272863864898682,
  1.3419934511184692,
  1.4201900959014893,
  1.387103796005249,
  1.3940315246582031,
  1.3061250448226929,
  1.3201463222503662,
  1.3905746936798096,
  1.4194819927215576,
  1.266250491142273,
  1.299294352531433,
  1.3952598571777344,
  1.4587703943252563,
  1.3114347457885742,
  1.4119548797607422,
  1.3235679864883423,
  1.3996590375900269
 ],
 "loss_first": [
  3.6044492721557617
 ],
 "loss_last": [
  1.3996590375900269
 ],
 "train_seconds": 4937.522064208984,
 "tokens_per_second": 6657.946956489508,
 "resumed_from": null
}