DCAgent/a1-stack_cpp
023
1{2 "achieved_tflops_per_gpu": 0.003978066363362,3 "achieved_tflops_per_gpu_theoretical": 538.6768254888838,4 "epoch": 7.0,5 "loss_nan_ranks": 0,6 "loss_rank_avg": 0.15846839547157288,7 "mfu_percent": 0.0002811354320397173,8 "mfu_percent_theoretical": 38.069033603454685,9 "total_flos": 1765744004825088.0,10 "train_loss": 0.17469855595608147,11 "train_runtime": 27741.8701,12 "train_samples_per_second": 2.541,13 "train_steps_per_second": 0.159,14 "valid_targets_mean": 5295.1,15 "valid_targets_min": 125816}