Files
exp-uns-tezos-10x_glm_4_7_t…/train_results.json

16 lines
545 B
JSON
Raw Normal View History

{
"achieved_tflops_per_gpu": 3.5085839993463646,
"achieved_tflops_per_gpu_theoretical": 379.4453110361948,
"epoch": 7.0,
"loss_nan_ranks": 0,
"loss_rank_avg": 0.11891636997461319,
"mfu_percent": 0.35476076838689224,
"mfu_percent_theoretical": 38.36656329991859,
"total_flos": 1.1243312580845896e+18,
"train_loss": 0.34437558135542856,
"train_runtime": 40056.4465,
"train_samples_per_second": 1.785,
"train_steps_per_second": 0.112,
"valid_targets_mean": 3517.0,
"valid_targets_min": 1483
}