Files
latamgpt-1b-sft/train_results.json

9 lines
238 B
JSON
Raw Normal View History

{
"epoch": 2.0,
"total_flos": 8.418445674676224e+18,
"train_loss": 0.8579012258703058,
"train_runtime": 27047.1098,
"train_samples": 175995,
"train_samples_per_second": 13.014,
"train_steps_per_second": 0.102
}