Files
swesmith-3160__Qwen3-8B/train_results.json

12 lines
396 B
JSON
Raw Normal View History

{
"achieved_tflops_per_gpu": 25246.89276564641,
"achieved_tflops_per_gpu_theoretical": 950998.3624027147,
"epoch": 7.0,
"mfu_percent": 1784.2327042859654,
"mfu_percent_theoretical": 67208.36483411411,
"total_flos": 6.648213778192138e+17,
"train_loss": 0.0,
"train_runtime": 0.8229,
"train_samples_per_second": 26879.554,
"train_steps_per_second": 280.704
}