Files
a1-quixbugs/train_results.json

16 lines
546 B
JSON
Raw Normal View History

{
"achieved_tflops_per_gpu": 0.0036459648741560212,
"achieved_tflops_per_gpu_theoretical": 833.4245224544994,
"epoch": 7.0,
"loss_nan_ranks": 0,
"loss_rank_avg": 0.15851815044879913,
"mfu_percent": 0.0002576653621311676,
"mfu_percent_theoretical": 58.89925953742045,
"total_flos": 1040604979265536.0,
"train_loss": 0.19424459490872364,
"train_runtime": 17838.2989,
"train_samples_per_second": 3.927,
"train_steps_per_second": 0.246,
"valid_targets_mean": 4008.4,
"valid_targets_min": 1389
}