Llama3.1-Mamba2-8B-distill / all_results.json
Junxiong Wang
add models
9224e3e
raw
history blame contribute delete
421 Bytes
{
"epoch": 0.999995191868488,
"eval_loss": 268.4025573730469,
"eval_runtime": 162.0112,
"eval_samples": 4096,
"eval_samples_per_second": 4.296,
"eval_steps_per_second": 0.136,
"total_flos": 3.3727656990081024e+16,
"train_loss": 478.7030639354971,
"train_runtime": 1000248.8491,
"train_samples": 19473081,
"train_samples_per_second": 3.327,
"train_steps_per_second": 0.052
}