Model: laion/sft__Kimi-2-5-swesmith-oracle-maxeps-32k__Qwen3-8B Source: Original Platform
16 lines
543 B
JSON
16 lines
543 B
JSON
{
|
|
"achieved_tflops_per_gpu": 5.0966199872969975,
|
|
"achieved_tflops_per_gpu_theoretical": 145.67989620139622,
|
|
"epoch": 7.0,
|
|
"loss_nan_ranks": 0,
|
|
"loss_rank_avg": 0.11108480393886566,
|
|
"mfu_percent": 0.360185158112862,
|
|
"mfu_percent_theoretical": 10.295399024833655,
|
|
"total_flos": 2.601520842818978e+18,
|
|
"train_loss": 0.2925965665424183,
|
|
"train_runtime": 15951.2631,
|
|
"train_samples_per_second": 4.106,
|
|
"train_steps_per_second": 0.043,
|
|
"valid_targets_mean": 4441.2,
|
|
"valid_targets_min": 1797
|
|
} |