hp_ablations_grid_mistral_bsz512_lr5e-6_scheduler-cosine-warmup0.15-mistralv0.3
/
train_results.json
{ | |
"epoch": 2.9994378864530633, | |
"total_flos": 3351540148469760.0, | |
"train_loss": 0.4023907120408921, | |
"train_runtime": 64801.3882, | |
"train_samples_per_second": 15.812, | |
"train_steps_per_second": 0.031 | |
} |