Training checkpoint at step 3000
Browse files- trainer_state.json +1885 -5
trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 2.
|
| 4 |
"best_model_checkpoint": "../runs/karpathy/fineweb-edu-100b-shuffle/meta-llama/Llama-3.2-1B/linear_adamw_wd1e-03_7x1024_mem32_bs64_hf_armt_dmem64/run_20/checkpoint-2000",
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 5,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -3768,6 +3768,1886 @@
|
|
| 3768 |
"eval_samples_per_second": 3.483,
|
| 3769 |
"eval_steps_per_second": 1.757,
|
| 3770 |
"step": 2000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3771 |
}
|
| 3772 |
],
|
| 3773 |
"logging_steps": 25,
|
|
@@ -3787,7 +5667,7 @@
|
|
| 3787 |
"attributes": {}
|
| 3788 |
}
|
| 3789 |
},
|
| 3790 |
-
"total_flos":
|
| 3791 |
"train_batch_size": 1,
|
| 3792 |
"trial_name": null,
|
| 3793 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 2985,
|
| 3 |
+
"best_metric": 2.4361066818237305,
|
| 4 |
"best_model_checkpoint": "../runs/karpathy/fineweb-edu-100b-shuffle/meta-llama/Llama-3.2-1B/linear_adamw_wd1e-03_7x1024_mem32_bs64_hf_armt_dmem64/run_20/checkpoint-2000",
|
| 5 |
+
"epoch": 0.06,
|
| 6 |
"eval_steps": 5,
|
| 7 |
+
"global_step": 3000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 3768 |
"eval_samples_per_second": 3.483,
|
| 3769 |
"eval_steps_per_second": 1.757,
|
| 3770 |
"step": 2000
|
| 3771 |
+
},
|
| 3772 |
+
{
|
| 3773 |
+
"epoch": 0.0401,
|
| 3774 |
+
"eval_loss": 2.449021577835083,
|
| 3775 |
+
"eval_runtime": 33.5048,
|
| 3776 |
+
"eval_samples_per_second": 3.492,
|
| 3777 |
+
"eval_steps_per_second": 1.761,
|
| 3778 |
+
"step": 2005
|
| 3779 |
+
},
|
| 3780 |
+
{
|
| 3781 |
+
"epoch": 0.0402,
|
| 3782 |
+
"eval_loss": 2.449159622192383,
|
| 3783 |
+
"eval_runtime": 33.4845,
|
| 3784 |
+
"eval_samples_per_second": 3.494,
|
| 3785 |
+
"eval_steps_per_second": 1.762,
|
| 3786 |
+
"step": 2010
|
| 3787 |
+
},
|
| 3788 |
+
{
|
| 3789 |
+
"epoch": 0.0403,
|
| 3790 |
+
"eval_loss": 2.448726177215576,
|
| 3791 |
+
"eval_runtime": 33.9926,
|
| 3792 |
+
"eval_samples_per_second": 3.442,
|
| 3793 |
+
"eval_steps_per_second": 1.736,
|
| 3794 |
+
"step": 2015
|
| 3795 |
+
},
|
| 3796 |
+
{
|
| 3797 |
+
"epoch": 0.0404,
|
| 3798 |
+
"eval_loss": 2.4484922885894775,
|
| 3799 |
+
"eval_runtime": 33.6594,
|
| 3800 |
+
"eval_samples_per_second": 3.476,
|
| 3801 |
+
"eval_steps_per_second": 1.753,
|
| 3802 |
+
"step": 2020
|
| 3803 |
+
},
|
| 3804 |
+
{
|
| 3805 |
+
"epoch": 0.0405,
|
| 3806 |
+
"grad_norm": 0.029877786947315705,
|
| 3807 |
+
"learning_rate": 4.048e-06,
|
| 3808 |
+
"loss": 2.438,
|
| 3809 |
+
"step": 2025
|
| 3810 |
+
},
|
| 3811 |
+
{
|
| 3812 |
+
"epoch": 0.0405,
|
| 3813 |
+
"eval_loss": 2.4485254287719727,
|
| 3814 |
+
"eval_runtime": 33.6812,
|
| 3815 |
+
"eval_samples_per_second": 3.474,
|
| 3816 |
+
"eval_steps_per_second": 1.752,
|
| 3817 |
+
"step": 2025
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.0406,
|
| 3821 |
+
"eval_loss": 2.448495388031006,
|
| 3822 |
+
"eval_runtime": 33.9733,
|
| 3823 |
+
"eval_samples_per_second": 3.444,
|
| 3824 |
+
"eval_steps_per_second": 1.737,
|
| 3825 |
+
"step": 2030
|
| 3826 |
+
},
|
| 3827 |
+
{
|
| 3828 |
+
"epoch": 0.0407,
|
| 3829 |
+
"eval_loss": 2.4482643604278564,
|
| 3830 |
+
"eval_runtime": 33.9957,
|
| 3831 |
+
"eval_samples_per_second": 3.442,
|
| 3832 |
+
"eval_steps_per_second": 1.736,
|
| 3833 |
+
"step": 2035
|
| 3834 |
+
},
|
| 3835 |
+
{
|
| 3836 |
+
"epoch": 0.0408,
|
| 3837 |
+
"eval_loss": 2.4481942653656006,
|
| 3838 |
+
"eval_runtime": 34.3014,
|
| 3839 |
+
"eval_samples_per_second": 3.411,
|
| 3840 |
+
"eval_steps_per_second": 1.72,
|
| 3841 |
+
"step": 2040
|
| 3842 |
+
},
|
| 3843 |
+
{
|
| 3844 |
+
"epoch": 0.0409,
|
| 3845 |
+
"eval_loss": 2.448082208633423,
|
| 3846 |
+
"eval_runtime": 34.0411,
|
| 3847 |
+
"eval_samples_per_second": 3.437,
|
| 3848 |
+
"eval_steps_per_second": 1.733,
|
| 3849 |
+
"step": 2045
|
| 3850 |
+
},
|
| 3851 |
+
{
|
| 3852 |
+
"epoch": 0.041,
|
| 3853 |
+
"grad_norm": 0.031175983773220776,
|
| 3854 |
+
"learning_rate": 4.098e-06,
|
| 3855 |
+
"loss": 2.4332,
|
| 3856 |
+
"step": 2050
|
| 3857 |
+
},
|
| 3858 |
+
{
|
| 3859 |
+
"epoch": 0.041,
|
| 3860 |
+
"eval_loss": 2.4478490352630615,
|
| 3861 |
+
"eval_runtime": 33.9245,
|
| 3862 |
+
"eval_samples_per_second": 3.449,
|
| 3863 |
+
"eval_steps_per_second": 1.739,
|
| 3864 |
+
"step": 2050
|
| 3865 |
+
},
|
| 3866 |
+
{
|
| 3867 |
+
"epoch": 0.0411,
|
| 3868 |
+
"eval_loss": 2.4480035305023193,
|
| 3869 |
+
"eval_runtime": 34.0079,
|
| 3870 |
+
"eval_samples_per_second": 3.44,
|
| 3871 |
+
"eval_steps_per_second": 1.735,
|
| 3872 |
+
"step": 2055
|
| 3873 |
+
},
|
| 3874 |
+
{
|
| 3875 |
+
"epoch": 0.0412,
|
| 3876 |
+
"eval_loss": 2.447685718536377,
|
| 3877 |
+
"eval_runtime": 33.999,
|
| 3878 |
+
"eval_samples_per_second": 3.441,
|
| 3879 |
+
"eval_steps_per_second": 1.735,
|
| 3880 |
+
"step": 2060
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.0413,
|
| 3884 |
+
"eval_loss": 2.447507619857788,
|
| 3885 |
+
"eval_runtime": 34.1446,
|
| 3886 |
+
"eval_samples_per_second": 3.427,
|
| 3887 |
+
"eval_steps_per_second": 1.728,
|
| 3888 |
+
"step": 2065
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.0414,
|
| 3892 |
+
"eval_loss": 2.447322130203247,
|
| 3893 |
+
"eval_runtime": 33.7479,
|
| 3894 |
+
"eval_samples_per_second": 3.467,
|
| 3895 |
+
"eval_steps_per_second": 1.748,
|
| 3896 |
+
"step": 2070
|
| 3897 |
+
},
|
| 3898 |
+
{
|
| 3899 |
+
"epoch": 0.0415,
|
| 3900 |
+
"grad_norm": 0.02904850084773878,
|
| 3901 |
+
"learning_rate": 4.148000000000001e-06,
|
| 3902 |
+
"loss": 2.4481,
|
| 3903 |
+
"step": 2075
|
| 3904 |
+
},
|
| 3905 |
+
{
|
| 3906 |
+
"epoch": 0.0415,
|
| 3907 |
+
"eval_loss": 2.4471347332000732,
|
| 3908 |
+
"eval_runtime": 33.917,
|
| 3909 |
+
"eval_samples_per_second": 3.45,
|
| 3910 |
+
"eval_steps_per_second": 1.74,
|
| 3911 |
+
"step": 2075
|
| 3912 |
+
},
|
| 3913 |
+
{
|
| 3914 |
+
"epoch": 0.0416,
|
| 3915 |
+
"eval_loss": 2.447152853012085,
|
| 3916 |
+
"eval_runtime": 33.8287,
|
| 3917 |
+
"eval_samples_per_second": 3.459,
|
| 3918 |
+
"eval_steps_per_second": 1.744,
|
| 3919 |
+
"step": 2080
|
| 3920 |
+
},
|
| 3921 |
+
{
|
| 3922 |
+
"epoch": 0.0417,
|
| 3923 |
+
"eval_loss": 2.4469242095947266,
|
| 3924 |
+
"eval_runtime": 33.7591,
|
| 3925 |
+
"eval_samples_per_second": 3.466,
|
| 3926 |
+
"eval_steps_per_second": 1.748,
|
| 3927 |
+
"step": 2085
|
| 3928 |
+
},
|
| 3929 |
+
{
|
| 3930 |
+
"epoch": 0.0418,
|
| 3931 |
+
"eval_loss": 2.4471774101257324,
|
| 3932 |
+
"eval_runtime": 33.7879,
|
| 3933 |
+
"eval_samples_per_second": 3.463,
|
| 3934 |
+
"eval_steps_per_second": 1.746,
|
| 3935 |
+
"step": 2090
|
| 3936 |
+
},
|
| 3937 |
+
{
|
| 3938 |
+
"epoch": 0.0419,
|
| 3939 |
+
"eval_loss": 2.447988986968994,
|
| 3940 |
+
"eval_runtime": 33.6878,
|
| 3941 |
+
"eval_samples_per_second": 3.473,
|
| 3942 |
+
"eval_steps_per_second": 1.751,
|
| 3943 |
+
"step": 2095
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.042,
|
| 3947 |
+
"grad_norm": 0.033838990669225626,
|
| 3948 |
+
"learning_rate": 4.198e-06,
|
| 3949 |
+
"loss": 2.4386,
|
| 3950 |
+
"step": 2100
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.042,
|
| 3954 |
+
"eval_loss": 2.4477100372314453,
|
| 3955 |
+
"eval_runtime": 33.6345,
|
| 3956 |
+
"eval_samples_per_second": 3.479,
|
| 3957 |
+
"eval_steps_per_second": 1.754,
|
| 3958 |
+
"step": 2100
|
| 3959 |
+
},
|
| 3960 |
+
{
|
| 3961 |
+
"epoch": 0.0421,
|
| 3962 |
+
"eval_loss": 2.447394847869873,
|
| 3963 |
+
"eval_runtime": 33.6221,
|
| 3964 |
+
"eval_samples_per_second": 3.48,
|
| 3965 |
+
"eval_steps_per_second": 1.755,
|
| 3966 |
+
"step": 2105
|
| 3967 |
+
},
|
| 3968 |
+
{
|
| 3969 |
+
"epoch": 0.0422,
|
| 3970 |
+
"eval_loss": 2.4470951557159424,
|
| 3971 |
+
"eval_runtime": 33.6689,
|
| 3972 |
+
"eval_samples_per_second": 3.475,
|
| 3973 |
+
"eval_steps_per_second": 1.752,
|
| 3974 |
+
"step": 2110
|
| 3975 |
+
},
|
| 3976 |
+
{
|
| 3977 |
+
"epoch": 0.0423,
|
| 3978 |
+
"eval_loss": 2.4467623233795166,
|
| 3979 |
+
"eval_runtime": 33.6979,
|
| 3980 |
+
"eval_samples_per_second": 3.472,
|
| 3981 |
+
"eval_steps_per_second": 1.751,
|
| 3982 |
+
"step": 2115
|
| 3983 |
+
},
|
| 3984 |
+
{
|
| 3985 |
+
"epoch": 0.0424,
|
| 3986 |
+
"eval_loss": 2.4469833374023438,
|
| 3987 |
+
"eval_runtime": 33.8632,
|
| 3988 |
+
"eval_samples_per_second": 3.455,
|
| 3989 |
+
"eval_steps_per_second": 1.742,
|
| 3990 |
+
"step": 2120
|
| 3991 |
+
},
|
| 3992 |
+
{
|
| 3993 |
+
"epoch": 0.0425,
|
| 3994 |
+
"grad_norm": 0.0382703849144026,
|
| 3995 |
+
"learning_rate": 4.248000000000001e-06,
|
| 3996 |
+
"loss": 2.4313,
|
| 3997 |
+
"step": 2125
|
| 3998 |
+
},
|
| 3999 |
+
{
|
| 4000 |
+
"epoch": 0.0425,
|
| 4001 |
+
"eval_loss": 2.447753667831421,
|
| 4002 |
+
"eval_runtime": 33.7269,
|
| 4003 |
+
"eval_samples_per_second": 3.469,
|
| 4004 |
+
"eval_steps_per_second": 1.749,
|
| 4005 |
+
"step": 2125
|
| 4006 |
+
},
|
| 4007 |
+
{
|
| 4008 |
+
"epoch": 0.0426,
|
| 4009 |
+
"eval_loss": 2.447281837463379,
|
| 4010 |
+
"eval_runtime": 33.7037,
|
| 4011 |
+
"eval_samples_per_second": 3.471,
|
| 4012 |
+
"eval_steps_per_second": 1.751,
|
| 4013 |
+
"step": 2130
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.0427,
|
| 4017 |
+
"eval_loss": 2.4472267627716064,
|
| 4018 |
+
"eval_runtime": 33.6873,
|
| 4019 |
+
"eval_samples_per_second": 3.473,
|
| 4020 |
+
"eval_steps_per_second": 1.751,
|
| 4021 |
+
"step": 2135
|
| 4022 |
+
},
|
| 4023 |
+
{
|
| 4024 |
+
"epoch": 0.0428,
|
| 4025 |
+
"eval_loss": 2.446859836578369,
|
| 4026 |
+
"eval_runtime": 33.6738,
|
| 4027 |
+
"eval_samples_per_second": 3.475,
|
| 4028 |
+
"eval_steps_per_second": 1.752,
|
| 4029 |
+
"step": 2140
|
| 4030 |
+
},
|
| 4031 |
+
{
|
| 4032 |
+
"epoch": 0.0429,
|
| 4033 |
+
"eval_loss": 2.446655035018921,
|
| 4034 |
+
"eval_runtime": 33.6536,
|
| 4035 |
+
"eval_samples_per_second": 3.477,
|
| 4036 |
+
"eval_steps_per_second": 1.753,
|
| 4037 |
+
"step": 2145
|
| 4038 |
+
},
|
| 4039 |
+
{
|
| 4040 |
+
"epoch": 0.043,
|
| 4041 |
+
"grad_norm": 0.027126678960545086,
|
| 4042 |
+
"learning_rate": 4.298e-06,
|
| 4043 |
+
"loss": 2.4298,
|
| 4044 |
+
"step": 2150
|
| 4045 |
+
},
|
| 4046 |
+
{
|
| 4047 |
+
"epoch": 0.043,
|
| 4048 |
+
"eval_loss": 2.4463651180267334,
|
| 4049 |
+
"eval_runtime": 33.6454,
|
| 4050 |
+
"eval_samples_per_second": 3.477,
|
| 4051 |
+
"eval_steps_per_second": 1.754,
|
| 4052 |
+
"step": 2150
|
| 4053 |
+
},
|
| 4054 |
+
{
|
| 4055 |
+
"epoch": 0.0431,
|
| 4056 |
+
"eval_loss": 2.4461581707000732,
|
| 4057 |
+
"eval_runtime": 33.6166,
|
| 4058 |
+
"eval_samples_per_second": 3.48,
|
| 4059 |
+
"eval_steps_per_second": 1.755,
|
| 4060 |
+
"step": 2155
|
| 4061 |
+
},
|
| 4062 |
+
{
|
| 4063 |
+
"epoch": 0.0432,
|
| 4064 |
+
"eval_loss": 2.4461660385131836,
|
| 4065 |
+
"eval_runtime": 33.5484,
|
| 4066 |
+
"eval_samples_per_second": 3.488,
|
| 4067 |
+
"eval_steps_per_second": 1.759,
|
| 4068 |
+
"step": 2160
|
| 4069 |
+
},
|
| 4070 |
+
{
|
| 4071 |
+
"epoch": 0.0433,
|
| 4072 |
+
"eval_loss": 2.4458513259887695,
|
| 4073 |
+
"eval_runtime": 33.6579,
|
| 4074 |
+
"eval_samples_per_second": 3.476,
|
| 4075 |
+
"eval_steps_per_second": 1.753,
|
| 4076 |
+
"step": 2165
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 0.0434,
|
| 4080 |
+
"eval_loss": 2.4454855918884277,
|
| 4081 |
+
"eval_runtime": 33.5647,
|
| 4082 |
+
"eval_samples_per_second": 3.486,
|
| 4083 |
+
"eval_steps_per_second": 1.758,
|
| 4084 |
+
"step": 2170
|
| 4085 |
+
},
|
| 4086 |
+
{
|
| 4087 |
+
"epoch": 0.0435,
|
| 4088 |
+
"grad_norm": 0.030565328679921875,
|
| 4089 |
+
"learning_rate": 4.3480000000000006e-06,
|
| 4090 |
+
"loss": 2.4387,
|
| 4091 |
+
"step": 2175
|
| 4092 |
+
},
|
| 4093 |
+
{
|
| 4094 |
+
"epoch": 0.0435,
|
| 4095 |
+
"eval_loss": 2.445688009262085,
|
| 4096 |
+
"eval_runtime": 33.5164,
|
| 4097 |
+
"eval_samples_per_second": 3.491,
|
| 4098 |
+
"eval_steps_per_second": 1.76,
|
| 4099 |
+
"step": 2175
|
| 4100 |
+
},
|
| 4101 |
+
{
|
| 4102 |
+
"epoch": 0.0436,
|
| 4103 |
+
"eval_loss": 2.4456729888916016,
|
| 4104 |
+
"eval_runtime": 33.4724,
|
| 4105 |
+
"eval_samples_per_second": 3.495,
|
| 4106 |
+
"eval_steps_per_second": 1.763,
|
| 4107 |
+
"step": 2180
|
| 4108 |
+
},
|
| 4109 |
+
{
|
| 4110 |
+
"epoch": 0.0437,
|
| 4111 |
+
"eval_loss": 2.4460015296936035,
|
| 4112 |
+
"eval_runtime": 33.3984,
|
| 4113 |
+
"eval_samples_per_second": 3.503,
|
| 4114 |
+
"eval_steps_per_second": 1.767,
|
| 4115 |
+
"step": 2185
|
| 4116 |
+
},
|
| 4117 |
+
{
|
| 4118 |
+
"epoch": 0.0438,
|
| 4119 |
+
"eval_loss": 2.4460256099700928,
|
| 4120 |
+
"eval_runtime": 33.4582,
|
| 4121 |
+
"eval_samples_per_second": 3.497,
|
| 4122 |
+
"eval_steps_per_second": 1.763,
|
| 4123 |
+
"step": 2190
|
| 4124 |
+
},
|
| 4125 |
+
{
|
| 4126 |
+
"epoch": 0.0439,
|
| 4127 |
+
"eval_loss": 2.4456872940063477,
|
| 4128 |
+
"eval_runtime": 33.444,
|
| 4129 |
+
"eval_samples_per_second": 3.498,
|
| 4130 |
+
"eval_steps_per_second": 1.764,
|
| 4131 |
+
"step": 2195
|
| 4132 |
+
},
|
| 4133 |
+
{
|
| 4134 |
+
"epoch": 0.044,
|
| 4135 |
+
"grad_norm": 0.03864046787827566,
|
| 4136 |
+
"learning_rate": 4.398000000000001e-06,
|
| 4137 |
+
"loss": 2.445,
|
| 4138 |
+
"step": 2200
|
| 4139 |
+
},
|
| 4140 |
+
{
|
| 4141 |
+
"epoch": 0.044,
|
| 4142 |
+
"eval_loss": 2.4454870223999023,
|
| 4143 |
+
"eval_runtime": 33.4474,
|
| 4144 |
+
"eval_samples_per_second": 3.498,
|
| 4145 |
+
"eval_steps_per_second": 1.764,
|
| 4146 |
+
"step": 2200
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 0.0441,
|
| 4150 |
+
"eval_loss": 2.4453113079071045,
|
| 4151 |
+
"eval_runtime": 33.4062,
|
| 4152 |
+
"eval_samples_per_second": 3.502,
|
| 4153 |
+
"eval_steps_per_second": 1.766,
|
| 4154 |
+
"step": 2205
|
| 4155 |
+
},
|
| 4156 |
+
{
|
| 4157 |
+
"epoch": 0.0442,
|
| 4158 |
+
"eval_loss": 2.4448771476745605,
|
| 4159 |
+
"eval_runtime": 33.3542,
|
| 4160 |
+
"eval_samples_per_second": 3.508,
|
| 4161 |
+
"eval_steps_per_second": 1.769,
|
| 4162 |
+
"step": 2210
|
| 4163 |
+
},
|
| 4164 |
+
{
|
| 4165 |
+
"epoch": 0.0443,
|
| 4166 |
+
"eval_loss": 2.444946765899658,
|
| 4167 |
+
"eval_runtime": 33.3997,
|
| 4168 |
+
"eval_samples_per_second": 3.503,
|
| 4169 |
+
"eval_steps_per_second": 1.766,
|
| 4170 |
+
"step": 2215
|
| 4171 |
+
},
|
| 4172 |
+
{
|
| 4173 |
+
"epoch": 0.0444,
|
| 4174 |
+
"eval_loss": 2.445194959640503,
|
| 4175 |
+
"eval_runtime": 33.3669,
|
| 4176 |
+
"eval_samples_per_second": 3.506,
|
| 4177 |
+
"eval_steps_per_second": 1.768,
|
| 4178 |
+
"step": 2220
|
| 4179 |
+
},
|
| 4180 |
+
{
|
| 4181 |
+
"epoch": 0.0445,
|
| 4182 |
+
"grad_norm": 0.026792091668494698,
|
| 4183 |
+
"learning_rate": 4.4480000000000004e-06,
|
| 4184 |
+
"loss": 2.4339,
|
| 4185 |
+
"step": 2225
|
| 4186 |
+
},
|
| 4187 |
+
{
|
| 4188 |
+
"epoch": 0.0445,
|
| 4189 |
+
"eval_loss": 2.445009469985962,
|
| 4190 |
+
"eval_runtime": 33.4467,
|
| 4191 |
+
"eval_samples_per_second": 3.498,
|
| 4192 |
+
"eval_steps_per_second": 1.764,
|
| 4193 |
+
"step": 2225
|
| 4194 |
+
},
|
| 4195 |
+
{
|
| 4196 |
+
"epoch": 0.0446,
|
| 4197 |
+
"eval_loss": 2.4450981616973877,
|
| 4198 |
+
"eval_runtime": 33.4513,
|
| 4199 |
+
"eval_samples_per_second": 3.498,
|
| 4200 |
+
"eval_steps_per_second": 1.764,
|
| 4201 |
+
"step": 2230
|
| 4202 |
+
},
|
| 4203 |
+
{
|
| 4204 |
+
"epoch": 0.0447,
|
| 4205 |
+
"eval_loss": 2.444899082183838,
|
| 4206 |
+
"eval_runtime": 33.3869,
|
| 4207 |
+
"eval_samples_per_second": 3.504,
|
| 4208 |
+
"eval_steps_per_second": 1.767,
|
| 4209 |
+
"step": 2235
|
| 4210 |
+
},
|
| 4211 |
+
{
|
| 4212 |
+
"epoch": 0.0448,
|
| 4213 |
+
"eval_loss": 2.4448494911193848,
|
| 4214 |
+
"eval_runtime": 33.486,
|
| 4215 |
+
"eval_samples_per_second": 3.494,
|
| 4216 |
+
"eval_steps_per_second": 1.762,
|
| 4217 |
+
"step": 2240
|
| 4218 |
+
},
|
| 4219 |
+
{
|
| 4220 |
+
"epoch": 0.0449,
|
| 4221 |
+
"eval_loss": 2.444640636444092,
|
| 4222 |
+
"eval_runtime": 33.4202,
|
| 4223 |
+
"eval_samples_per_second": 3.501,
|
| 4224 |
+
"eval_steps_per_second": 1.765,
|
| 4225 |
+
"step": 2245
|
| 4226 |
+
},
|
| 4227 |
+
{
|
| 4228 |
+
"epoch": 0.045,
|
| 4229 |
+
"grad_norm": 0.027104711228224686,
|
| 4230 |
+
"learning_rate": 4.498e-06,
|
| 4231 |
+
"loss": 2.4326,
|
| 4232 |
+
"step": 2250
|
| 4233 |
+
},
|
| 4234 |
+
{
|
| 4235 |
+
"epoch": 0.045,
|
| 4236 |
+
"eval_loss": 2.444633722305298,
|
| 4237 |
+
"eval_runtime": 33.4154,
|
| 4238 |
+
"eval_samples_per_second": 3.501,
|
| 4239 |
+
"eval_steps_per_second": 1.766,
|
| 4240 |
+
"step": 2250
|
| 4241 |
+
},
|
| 4242 |
+
{
|
| 4243 |
+
"epoch": 0.0451,
|
| 4244 |
+
"eval_loss": 2.44467830657959,
|
| 4245 |
+
"eval_runtime": 33.4237,
|
| 4246 |
+
"eval_samples_per_second": 3.501,
|
| 4247 |
+
"eval_steps_per_second": 1.765,
|
| 4248 |
+
"step": 2255
|
| 4249 |
+
},
|
| 4250 |
+
{
|
| 4251 |
+
"epoch": 0.0452,
|
| 4252 |
+
"eval_loss": 2.444413900375366,
|
| 4253 |
+
"eval_runtime": 33.3694,
|
| 4254 |
+
"eval_samples_per_second": 3.506,
|
| 4255 |
+
"eval_steps_per_second": 1.768,
|
| 4256 |
+
"step": 2260
|
| 4257 |
+
},
|
| 4258 |
+
{
|
| 4259 |
+
"epoch": 0.0453,
|
| 4260 |
+
"eval_loss": 2.444222927093506,
|
| 4261 |
+
"eval_runtime": 33.3585,
|
| 4262 |
+
"eval_samples_per_second": 3.507,
|
| 4263 |
+
"eval_steps_per_second": 1.769,
|
| 4264 |
+
"step": 2265
|
| 4265 |
+
},
|
| 4266 |
+
{
|
| 4267 |
+
"epoch": 0.0454,
|
| 4268 |
+
"eval_loss": 2.444108724594116,
|
| 4269 |
+
"eval_runtime": 33.3346,
|
| 4270 |
+
"eval_samples_per_second": 3.51,
|
| 4271 |
+
"eval_steps_per_second": 1.77,
|
| 4272 |
+
"step": 2270
|
| 4273 |
+
},
|
| 4274 |
+
{
|
| 4275 |
+
"epoch": 0.0455,
|
| 4276 |
+
"grad_norm": 0.033569645173308425,
|
| 4277 |
+
"learning_rate": 4.548e-06,
|
| 4278 |
+
"loss": 2.4342,
|
| 4279 |
+
"step": 2275
|
| 4280 |
+
},
|
| 4281 |
+
{
|
| 4282 |
+
"epoch": 0.0455,
|
| 4283 |
+
"eval_loss": 2.443859577178955,
|
| 4284 |
+
"eval_runtime": 33.3636,
|
| 4285 |
+
"eval_samples_per_second": 3.507,
|
| 4286 |
+
"eval_steps_per_second": 1.768,
|
| 4287 |
+
"step": 2275
|
| 4288 |
+
},
|
| 4289 |
+
{
|
| 4290 |
+
"epoch": 0.0456,
|
| 4291 |
+
"eval_loss": 2.4441120624542236,
|
| 4292 |
+
"eval_runtime": 33.2442,
|
| 4293 |
+
"eval_samples_per_second": 3.519,
|
| 4294 |
+
"eval_steps_per_second": 1.775,
|
| 4295 |
+
"step": 2280
|
| 4296 |
+
},
|
| 4297 |
+
{
|
| 4298 |
+
"epoch": 0.0457,
|
| 4299 |
+
"eval_loss": 2.4439260959625244,
|
| 4300 |
+
"eval_runtime": 33.2924,
|
| 4301 |
+
"eval_samples_per_second": 3.514,
|
| 4302 |
+
"eval_steps_per_second": 1.772,
|
| 4303 |
+
"step": 2285
|
| 4304 |
+
},
|
| 4305 |
+
{
|
| 4306 |
+
"epoch": 0.0458,
|
| 4307 |
+
"eval_loss": 2.4439032077789307,
|
| 4308 |
+
"eval_runtime": 33.4004,
|
| 4309 |
+
"eval_samples_per_second": 3.503,
|
| 4310 |
+
"eval_steps_per_second": 1.766,
|
| 4311 |
+
"step": 2290
|
| 4312 |
+
},
|
| 4313 |
+
{
|
| 4314 |
+
"epoch": 0.0459,
|
| 4315 |
+
"eval_loss": 2.443621873855591,
|
| 4316 |
+
"eval_runtime": 33.3314,
|
| 4317 |
+
"eval_samples_per_second": 3.51,
|
| 4318 |
+
"eval_steps_per_second": 1.77,
|
| 4319 |
+
"step": 2295
|
| 4320 |
+
},
|
| 4321 |
+
{
|
| 4322 |
+
"epoch": 0.046,
|
| 4323 |
+
"grad_norm": 0.02648413187023774,
|
| 4324 |
+
"learning_rate": 4.598e-06,
|
| 4325 |
+
"loss": 2.4368,
|
| 4326 |
+
"step": 2300
|
| 4327 |
+
},
|
| 4328 |
+
{
|
| 4329 |
+
"epoch": 0.046,
|
| 4330 |
+
"eval_loss": 2.4436306953430176,
|
| 4331 |
+
"eval_runtime": 33.372,
|
| 4332 |
+
"eval_samples_per_second": 3.506,
|
| 4333 |
+
"eval_steps_per_second": 1.768,
|
| 4334 |
+
"step": 2300
|
| 4335 |
+
},
|
| 4336 |
+
{
|
| 4337 |
+
"epoch": 0.0461,
|
| 4338 |
+
"eval_loss": 2.4436404705047607,
|
| 4339 |
+
"eval_runtime": 33.3039,
|
| 4340 |
+
"eval_samples_per_second": 3.513,
|
| 4341 |
+
"eval_steps_per_second": 1.772,
|
| 4342 |
+
"step": 2305
|
| 4343 |
+
},
|
| 4344 |
+
{
|
| 4345 |
+
"epoch": 0.0462,
|
| 4346 |
+
"eval_loss": 2.44333815574646,
|
| 4347 |
+
"eval_runtime": 33.3059,
|
| 4348 |
+
"eval_samples_per_second": 3.513,
|
| 4349 |
+
"eval_steps_per_second": 1.771,
|
| 4350 |
+
"step": 2310
|
| 4351 |
+
},
|
| 4352 |
+
{
|
| 4353 |
+
"epoch": 0.0463,
|
| 4354 |
+
"eval_loss": 2.443415880203247,
|
| 4355 |
+
"eval_runtime": 33.4065,
|
| 4356 |
+
"eval_samples_per_second": 3.502,
|
| 4357 |
+
"eval_steps_per_second": 1.766,
|
| 4358 |
+
"step": 2315
|
| 4359 |
+
},
|
| 4360 |
+
{
|
| 4361 |
+
"epoch": 0.0464,
|
| 4362 |
+
"eval_loss": 2.443068742752075,
|
| 4363 |
+
"eval_runtime": 33.2818,
|
| 4364 |
+
"eval_samples_per_second": 3.515,
|
| 4365 |
+
"eval_steps_per_second": 1.773,
|
| 4366 |
+
"step": 2320
|
| 4367 |
+
},
|
| 4368 |
+
{
|
| 4369 |
+
"epoch": 0.0465,
|
| 4370 |
+
"grad_norm": 0.0351440602227012,
|
| 4371 |
+
"learning_rate": 4.648e-06,
|
| 4372 |
+
"loss": 2.4381,
|
| 4373 |
+
"step": 2325
|
| 4374 |
+
},
|
| 4375 |
+
{
|
| 4376 |
+
"epoch": 0.0465,
|
| 4377 |
+
"eval_loss": 2.443199634552002,
|
| 4378 |
+
"eval_runtime": 33.3538,
|
| 4379 |
+
"eval_samples_per_second": 3.508,
|
| 4380 |
+
"eval_steps_per_second": 1.769,
|
| 4381 |
+
"step": 2325
|
| 4382 |
+
},
|
| 4383 |
+
{
|
| 4384 |
+
"epoch": 0.0466,
|
| 4385 |
+
"eval_loss": 2.4433047771453857,
|
| 4386 |
+
"eval_runtime": 33.4816,
|
| 4387 |
+
"eval_samples_per_second": 3.494,
|
| 4388 |
+
"eval_steps_per_second": 1.762,
|
| 4389 |
+
"step": 2330
|
| 4390 |
+
},
|
| 4391 |
+
{
|
| 4392 |
+
"epoch": 0.0467,
|
| 4393 |
+
"eval_loss": 2.443272113800049,
|
| 4394 |
+
"eval_runtime": 33.5015,
|
| 4395 |
+
"eval_samples_per_second": 3.492,
|
| 4396 |
+
"eval_steps_per_second": 1.761,
|
| 4397 |
+
"step": 2335
|
| 4398 |
+
},
|
| 4399 |
+
{
|
| 4400 |
+
"epoch": 0.0468,
|
| 4401 |
+
"eval_loss": 2.443246603012085,
|
| 4402 |
+
"eval_runtime": 33.5753,
|
| 4403 |
+
"eval_samples_per_second": 3.485,
|
| 4404 |
+
"eval_steps_per_second": 1.757,
|
| 4405 |
+
"step": 2340
|
| 4406 |
+
},
|
| 4407 |
+
{
|
| 4408 |
+
"epoch": 0.0469,
|
| 4409 |
+
"eval_loss": 2.4432363510131836,
|
| 4410 |
+
"eval_runtime": 33.2869,
|
| 4411 |
+
"eval_samples_per_second": 3.515,
|
| 4412 |
+
"eval_steps_per_second": 1.772,
|
| 4413 |
+
"step": 2345
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 0.047,
|
| 4417 |
+
"grad_norm": 0.02695670446644145,
|
| 4418 |
+
"learning_rate": 4.698000000000001e-06,
|
| 4419 |
+
"loss": 2.4303,
|
| 4420 |
+
"step": 2350
|
| 4421 |
+
},
|
| 4422 |
+
{
|
| 4423 |
+
"epoch": 0.047,
|
| 4424 |
+
"eval_loss": 2.4429421424865723,
|
| 4425 |
+
"eval_runtime": 33.3556,
|
| 4426 |
+
"eval_samples_per_second": 3.508,
|
| 4427 |
+
"eval_steps_per_second": 1.769,
|
| 4428 |
+
"step": 2350
|
| 4429 |
+
},
|
| 4430 |
+
{
|
| 4431 |
+
"epoch": 0.0471,
|
| 4432 |
+
"eval_loss": 2.4427566528320312,
|
| 4433 |
+
"eval_runtime": 33.3612,
|
| 4434 |
+
"eval_samples_per_second": 3.507,
|
| 4435 |
+
"eval_steps_per_second": 1.769,
|
| 4436 |
+
"step": 2355
|
| 4437 |
+
},
|
| 4438 |
+
{
|
| 4439 |
+
"epoch": 0.0472,
|
| 4440 |
+
"eval_loss": 2.4425995349884033,
|
| 4441 |
+
"eval_runtime": 33.353,
|
| 4442 |
+
"eval_samples_per_second": 3.508,
|
| 4443 |
+
"eval_steps_per_second": 1.769,
|
| 4444 |
+
"step": 2360
|
| 4445 |
+
},
|
| 4446 |
+
{
|
| 4447 |
+
"epoch": 0.0473,
|
| 4448 |
+
"eval_loss": 2.4426395893096924,
|
| 4449 |
+
"eval_runtime": 33.4669,
|
| 4450 |
+
"eval_samples_per_second": 3.496,
|
| 4451 |
+
"eval_steps_per_second": 1.763,
|
| 4452 |
+
"step": 2365
|
| 4453 |
+
},
|
| 4454 |
+
{
|
| 4455 |
+
"epoch": 0.0474,
|
| 4456 |
+
"eval_loss": 2.4425301551818848,
|
| 4457 |
+
"eval_runtime": 33.3803,
|
| 4458 |
+
"eval_samples_per_second": 3.505,
|
| 4459 |
+
"eval_steps_per_second": 1.768,
|
| 4460 |
+
"step": 2370
|
| 4461 |
+
},
|
| 4462 |
+
{
|
| 4463 |
+
"epoch": 0.0475,
|
| 4464 |
+
"grad_norm": 0.031232764672567994,
|
| 4465 |
+
"learning_rate": 4.748e-06,
|
| 4466 |
+
"loss": 2.4284,
|
| 4467 |
+
"step": 2375
|
| 4468 |
+
},
|
| 4469 |
+
{
|
| 4470 |
+
"epoch": 0.0475,
|
| 4471 |
+
"eval_loss": 2.4426214694976807,
|
| 4472 |
+
"eval_runtime": 33.3013,
|
| 4473 |
+
"eval_samples_per_second": 3.513,
|
| 4474 |
+
"eval_steps_per_second": 1.772,
|
| 4475 |
+
"step": 2375
|
| 4476 |
+
},
|
| 4477 |
+
{
|
| 4478 |
+
"epoch": 0.0476,
|
| 4479 |
+
"eval_loss": 2.442599296569824,
|
| 4480 |
+
"eval_runtime": 33.3419,
|
| 4481 |
+
"eval_samples_per_second": 3.509,
|
| 4482 |
+
"eval_steps_per_second": 1.77,
|
| 4483 |
+
"step": 2380
|
| 4484 |
+
},
|
| 4485 |
+
{
|
| 4486 |
+
"epoch": 0.0477,
|
| 4487 |
+
"eval_loss": 2.442364454269409,
|
| 4488 |
+
"eval_runtime": 33.3677,
|
| 4489 |
+
"eval_samples_per_second": 3.506,
|
| 4490 |
+
"eval_steps_per_second": 1.768,
|
| 4491 |
+
"step": 2385
|
| 4492 |
+
},
|
| 4493 |
+
{
|
| 4494 |
+
"epoch": 0.0478,
|
| 4495 |
+
"eval_loss": 2.4425458908081055,
|
| 4496 |
+
"eval_runtime": 33.3892,
|
| 4497 |
+
"eval_samples_per_second": 3.504,
|
| 4498 |
+
"eval_steps_per_second": 1.767,
|
| 4499 |
+
"step": 2390
|
| 4500 |
+
},
|
| 4501 |
+
{
|
| 4502 |
+
"epoch": 0.0479,
|
| 4503 |
+
"eval_loss": 2.4425549507141113,
|
| 4504 |
+
"eval_runtime": 33.4202,
|
| 4505 |
+
"eval_samples_per_second": 3.501,
|
| 4506 |
+
"eval_steps_per_second": 1.765,
|
| 4507 |
+
"step": 2395
|
| 4508 |
+
},
|
| 4509 |
+
{
|
| 4510 |
+
"epoch": 0.048,
|
| 4511 |
+
"grad_norm": 0.027127721086561404,
|
| 4512 |
+
"learning_rate": 4.7980000000000005e-06,
|
| 4513 |
+
"loss": 2.4291,
|
| 4514 |
+
"step": 2400
|
| 4515 |
+
},
|
| 4516 |
+
{
|
| 4517 |
+
"epoch": 0.048,
|
| 4518 |
+
"eval_loss": 2.4425251483917236,
|
| 4519 |
+
"eval_runtime": 33.3802,
|
| 4520 |
+
"eval_samples_per_second": 3.505,
|
| 4521 |
+
"eval_steps_per_second": 1.768,
|
| 4522 |
+
"step": 2400
|
| 4523 |
+
},
|
| 4524 |
+
{
|
| 4525 |
+
"epoch": 0.0481,
|
| 4526 |
+
"eval_loss": 2.4424123764038086,
|
| 4527 |
+
"eval_runtime": 33.3283,
|
| 4528 |
+
"eval_samples_per_second": 3.511,
|
| 4529 |
+
"eval_steps_per_second": 1.77,
|
| 4530 |
+
"step": 2405
|
| 4531 |
+
},
|
| 4532 |
+
{
|
| 4533 |
+
"epoch": 0.0482,
|
| 4534 |
+
"eval_loss": 2.4421849250793457,
|
| 4535 |
+
"eval_runtime": 33.4172,
|
| 4536 |
+
"eval_samples_per_second": 3.501,
|
| 4537 |
+
"eval_steps_per_second": 1.766,
|
| 4538 |
+
"step": 2410
|
| 4539 |
+
},
|
| 4540 |
+
{
|
| 4541 |
+
"epoch": 0.0483,
|
| 4542 |
+
"eval_loss": 2.4419970512390137,
|
| 4543 |
+
"eval_runtime": 33.4642,
|
| 4544 |
+
"eval_samples_per_second": 3.496,
|
| 4545 |
+
"eval_steps_per_second": 1.763,
|
| 4546 |
+
"step": 2415
|
| 4547 |
+
},
|
| 4548 |
+
{
|
| 4549 |
+
"epoch": 0.0484,
|
| 4550 |
+
"eval_loss": 2.4419567584991455,
|
| 4551 |
+
"eval_runtime": 33.3663,
|
| 4552 |
+
"eval_samples_per_second": 3.507,
|
| 4553 |
+
"eval_steps_per_second": 1.768,
|
| 4554 |
+
"step": 2420
|
| 4555 |
+
},
|
| 4556 |
+
{
|
| 4557 |
+
"epoch": 0.0485,
|
| 4558 |
+
"grad_norm": 0.026032952013136927,
|
| 4559 |
+
"learning_rate": 4.848000000000001e-06,
|
| 4560 |
+
"loss": 2.4256,
|
| 4561 |
+
"step": 2425
|
| 4562 |
+
},
|
| 4563 |
+
{
|
| 4564 |
+
"epoch": 0.0485,
|
| 4565 |
+
"eval_loss": 2.441688299179077,
|
| 4566 |
+
"eval_runtime": 33.3169,
|
| 4567 |
+
"eval_samples_per_second": 3.512,
|
| 4568 |
+
"eval_steps_per_second": 1.771,
|
| 4569 |
+
"step": 2425
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 0.0486,
|
| 4573 |
+
"eval_loss": 2.4417548179626465,
|
| 4574 |
+
"eval_runtime": 33.3476,
|
| 4575 |
+
"eval_samples_per_second": 3.508,
|
| 4576 |
+
"eval_steps_per_second": 1.769,
|
| 4577 |
+
"step": 2430
|
| 4578 |
+
},
|
| 4579 |
+
{
|
| 4580 |
+
"epoch": 0.0487,
|
| 4581 |
+
"eval_loss": 2.441769599914551,
|
| 4582 |
+
"eval_runtime": 33.4488,
|
| 4583 |
+
"eval_samples_per_second": 3.498,
|
| 4584 |
+
"eval_steps_per_second": 1.764,
|
| 4585 |
+
"step": 2435
|
| 4586 |
+
},
|
| 4587 |
+
{
|
| 4588 |
+
"epoch": 0.0488,
|
| 4589 |
+
"eval_loss": 2.4415283203125,
|
| 4590 |
+
"eval_runtime": 33.4555,
|
| 4591 |
+
"eval_samples_per_second": 3.497,
|
| 4592 |
+
"eval_steps_per_second": 1.764,
|
| 4593 |
+
"step": 2440
|
| 4594 |
+
},
|
| 4595 |
+
{
|
| 4596 |
+
"epoch": 0.0489,
|
| 4597 |
+
"eval_loss": 2.4416847229003906,
|
| 4598 |
+
"eval_runtime": 33.2459,
|
| 4599 |
+
"eval_samples_per_second": 3.519,
|
| 4600 |
+
"eval_steps_per_second": 1.775,
|
| 4601 |
+
"step": 2445
|
| 4602 |
+
},
|
| 4603 |
+
{
|
| 4604 |
+
"epoch": 0.049,
|
| 4605 |
+
"grad_norm": 0.02804626155591942,
|
| 4606 |
+
"learning_rate": 4.898e-06,
|
| 4607 |
+
"loss": 2.4334,
|
| 4608 |
+
"step": 2450
|
| 4609 |
+
},
|
| 4610 |
+
{
|
| 4611 |
+
"epoch": 0.049,
|
| 4612 |
+
"eval_loss": 2.4414188861846924,
|
| 4613 |
+
"eval_runtime": 33.2989,
|
| 4614 |
+
"eval_samples_per_second": 3.514,
|
| 4615 |
+
"eval_steps_per_second": 1.772,
|
| 4616 |
+
"step": 2450
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 0.0491,
|
| 4620 |
+
"eval_loss": 2.4416472911834717,
|
| 4621 |
+
"eval_runtime": 33.3676,
|
| 4622 |
+
"eval_samples_per_second": 3.506,
|
| 4623 |
+
"eval_steps_per_second": 1.768,
|
| 4624 |
+
"step": 2455
|
| 4625 |
+
},
|
| 4626 |
+
{
|
| 4627 |
+
"epoch": 0.0492,
|
| 4628 |
+
"eval_loss": 2.4414844512939453,
|
| 4629 |
+
"eval_runtime": 33.4116,
|
| 4630 |
+
"eval_samples_per_second": 3.502,
|
| 4631 |
+
"eval_steps_per_second": 1.766,
|
| 4632 |
+
"step": 2460
|
| 4633 |
+
},
|
| 4634 |
+
{
|
| 4635 |
+
"epoch": 0.0493,
|
| 4636 |
+
"eval_loss": 2.441408395767212,
|
| 4637 |
+
"eval_runtime": 33.6104,
|
| 4638 |
+
"eval_samples_per_second": 3.481,
|
| 4639 |
+
"eval_steps_per_second": 1.755,
|
| 4640 |
+
"step": 2465
|
| 4641 |
+
},
|
| 4642 |
+
{
|
| 4643 |
+
"epoch": 0.0494,
|
| 4644 |
+
"eval_loss": 2.4413650035858154,
|
| 4645 |
+
"eval_runtime": 33.3838,
|
| 4646 |
+
"eval_samples_per_second": 3.505,
|
| 4647 |
+
"eval_steps_per_second": 1.767,
|
| 4648 |
+
"step": 2470
|
| 4649 |
+
},
|
| 4650 |
+
{
|
| 4651 |
+
"epoch": 0.0495,
|
| 4652 |
+
"grad_norm": 0.025351866385684634,
|
| 4653 |
+
"learning_rate": 4.948000000000001e-06,
|
| 4654 |
+
"loss": 2.4356,
|
| 4655 |
+
"step": 2475
|
| 4656 |
+
},
|
| 4657 |
+
{
|
| 4658 |
+
"epoch": 0.0495,
|
| 4659 |
+
"eval_loss": 2.4411768913269043,
|
| 4660 |
+
"eval_runtime": 33.3857,
|
| 4661 |
+
"eval_samples_per_second": 3.504,
|
| 4662 |
+
"eval_steps_per_second": 1.767,
|
| 4663 |
+
"step": 2475
|
| 4664 |
+
},
|
| 4665 |
+
{
|
| 4666 |
+
"epoch": 0.0496,
|
| 4667 |
+
"eval_loss": 2.441201686859131,
|
| 4668 |
+
"eval_runtime": 33.4117,
|
| 4669 |
+
"eval_samples_per_second": 3.502,
|
| 4670 |
+
"eval_steps_per_second": 1.766,
|
| 4671 |
+
"step": 2480
|
| 4672 |
+
},
|
| 4673 |
+
{
|
| 4674 |
+
"epoch": 0.0497,
|
| 4675 |
+
"eval_loss": 2.4408698081970215,
|
| 4676 |
+
"eval_runtime": 33.3015,
|
| 4677 |
+
"eval_samples_per_second": 3.513,
|
| 4678 |
+
"eval_steps_per_second": 1.772,
|
| 4679 |
+
"step": 2485
|
| 4680 |
+
},
|
| 4681 |
+
{
|
| 4682 |
+
"epoch": 0.0498,
|
| 4683 |
+
"eval_loss": 2.440950393676758,
|
| 4684 |
+
"eval_runtime": 33.379,
|
| 4685 |
+
"eval_samples_per_second": 3.505,
|
| 4686 |
+
"eval_steps_per_second": 1.768,
|
| 4687 |
+
"step": 2490
|
| 4688 |
+
},
|
| 4689 |
+
{
|
| 4690 |
+
"epoch": 0.0499,
|
| 4691 |
+
"eval_loss": 2.4407267570495605,
|
| 4692 |
+
"eval_runtime": 33.2561,
|
| 4693 |
+
"eval_samples_per_second": 3.518,
|
| 4694 |
+
"eval_steps_per_second": 1.774,
|
| 4695 |
+
"step": 2495
|
| 4696 |
+
},
|
| 4697 |
+
{
|
| 4698 |
+
"epoch": 0.05,
|
| 4699 |
+
"grad_norm": 0.029743600833546286,
|
| 4700 |
+
"learning_rate": 4.998e-06,
|
| 4701 |
+
"loss": 2.4369,
|
| 4702 |
+
"step": 2500
|
| 4703 |
+
},
|
| 4704 |
+
{
|
| 4705 |
+
"epoch": 0.05,
|
| 4706 |
+
"eval_loss": 2.4408068656921387,
|
| 4707 |
+
"eval_runtime": 33.3807,
|
| 4708 |
+
"eval_samples_per_second": 3.505,
|
| 4709 |
+
"eval_steps_per_second": 1.767,
|
| 4710 |
+
"step": 2500
|
| 4711 |
+
},
|
| 4712 |
+
{
|
| 4713 |
+
"epoch": 0.0501,
|
| 4714 |
+
"eval_loss": 2.4407401084899902,
|
| 4715 |
+
"eval_runtime": 33.2295,
|
| 4716 |
+
"eval_samples_per_second": 3.521,
|
| 4717 |
+
"eval_steps_per_second": 1.776,
|
| 4718 |
+
"step": 2505
|
| 4719 |
+
},
|
| 4720 |
+
{
|
| 4721 |
+
"epoch": 0.0502,
|
| 4722 |
+
"eval_loss": 2.4409286975860596,
|
| 4723 |
+
"eval_runtime": 33.3925,
|
| 4724 |
+
"eval_samples_per_second": 3.504,
|
| 4725 |
+
"eval_steps_per_second": 1.767,
|
| 4726 |
+
"step": 2510
|
| 4727 |
+
},
|
| 4728 |
+
{
|
| 4729 |
+
"epoch": 0.0503,
|
| 4730 |
+
"eval_loss": 2.4407782554626465,
|
| 4731 |
+
"eval_runtime": 33.4498,
|
| 4732 |
+
"eval_samples_per_second": 3.498,
|
| 4733 |
+
"eval_steps_per_second": 1.764,
|
| 4734 |
+
"step": 2515
|
| 4735 |
+
},
|
| 4736 |
+
{
|
| 4737 |
+
"epoch": 0.0504,
|
| 4738 |
+
"eval_loss": 2.4407856464385986,
|
| 4739 |
+
"eval_runtime": 33.4899,
|
| 4740 |
+
"eval_samples_per_second": 3.494,
|
| 4741 |
+
"eval_steps_per_second": 1.762,
|
| 4742 |
+
"step": 2520
|
| 4743 |
+
},
|
| 4744 |
+
{
|
| 4745 |
+
"epoch": 0.0505,
|
| 4746 |
+
"grad_norm": 0.027292319342276494,
|
| 4747 |
+
"learning_rate": 5.048000000000001e-06,
|
| 4748 |
+
"loss": 2.4263,
|
| 4749 |
+
"step": 2525
|
| 4750 |
+
},
|
| 4751 |
+
{
|
| 4752 |
+
"epoch": 0.0505,
|
| 4753 |
+
"eval_loss": 2.440830945968628,
|
| 4754 |
+
"eval_runtime": 33.3428,
|
| 4755 |
+
"eval_samples_per_second": 3.509,
|
| 4756 |
+
"eval_steps_per_second": 1.769,
|
| 4757 |
+
"step": 2525
|
| 4758 |
+
},
|
| 4759 |
+
{
|
| 4760 |
+
"epoch": 0.0506,
|
| 4761 |
+
"eval_loss": 2.44069504737854,
|
| 4762 |
+
"eval_runtime": 33.2895,
|
| 4763 |
+
"eval_samples_per_second": 3.515,
|
| 4764 |
+
"eval_steps_per_second": 1.772,
|
| 4765 |
+
"step": 2530
|
| 4766 |
+
},
|
| 4767 |
+
{
|
| 4768 |
+
"epoch": 0.0507,
|
| 4769 |
+
"eval_loss": 2.4408159255981445,
|
| 4770 |
+
"eval_runtime": 33.3488,
|
| 4771 |
+
"eval_samples_per_second": 3.508,
|
| 4772 |
+
"eval_steps_per_second": 1.769,
|
| 4773 |
+
"step": 2535
|
| 4774 |
+
},
|
| 4775 |
+
{
|
| 4776 |
+
"epoch": 0.0508,
|
| 4777 |
+
"eval_loss": 2.440523386001587,
|
| 4778 |
+
"eval_runtime": 33.3582,
|
| 4779 |
+
"eval_samples_per_second": 3.507,
|
| 4780 |
+
"eval_steps_per_second": 1.769,
|
| 4781 |
+
"step": 2540
|
| 4782 |
+
},
|
| 4783 |
+
{
|
| 4784 |
+
"epoch": 0.0509,
|
| 4785 |
+
"eval_loss": 2.4403724670410156,
|
| 4786 |
+
"eval_runtime": 33.5287,
|
| 4787 |
+
"eval_samples_per_second": 3.49,
|
| 4788 |
+
"eval_steps_per_second": 1.76,
|
| 4789 |
+
"step": 2545
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 0.051,
|
| 4793 |
+
"grad_norm": 0.02495087994166461,
|
| 4794 |
+
"learning_rate": 5.098000000000001e-06,
|
| 4795 |
+
"loss": 2.428,
|
| 4796 |
+
"step": 2550
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 0.051,
|
| 4800 |
+
"eval_loss": 2.440495252609253,
|
| 4801 |
+
"eval_runtime": 34.4575,
|
| 4802 |
+
"eval_samples_per_second": 3.395,
|
| 4803 |
+
"eval_steps_per_second": 1.712,
|
| 4804 |
+
"step": 2550
|
| 4805 |
+
},
|
| 4806 |
+
{
|
| 4807 |
+
"epoch": 0.0511,
|
| 4808 |
+
"eval_loss": 2.440384864807129,
|
| 4809 |
+
"eval_runtime": 34.0144,
|
| 4810 |
+
"eval_samples_per_second": 3.44,
|
| 4811 |
+
"eval_steps_per_second": 1.735,
|
| 4812 |
+
"step": 2555
|
| 4813 |
+
},
|
| 4814 |
+
{
|
| 4815 |
+
"epoch": 0.0512,
|
| 4816 |
+
"eval_loss": 2.4405176639556885,
|
| 4817 |
+
"eval_runtime": 34.5852,
|
| 4818 |
+
"eval_samples_per_second": 3.383,
|
| 4819 |
+
"eval_steps_per_second": 1.706,
|
| 4820 |
+
"step": 2560
|
| 4821 |
+
},
|
| 4822 |
+
{
|
| 4823 |
+
"epoch": 0.0513,
|
| 4824 |
+
"eval_loss": 2.4402472972869873,
|
| 4825 |
+
"eval_runtime": 34.2689,
|
| 4826 |
+
"eval_samples_per_second": 3.414,
|
| 4827 |
+
"eval_steps_per_second": 1.722,
|
| 4828 |
+
"step": 2565
|
| 4829 |
+
},
|
| 4830 |
+
{
|
| 4831 |
+
"epoch": 0.0514,
|
| 4832 |
+
"eval_loss": 2.440459966659546,
|
| 4833 |
+
"eval_runtime": 33.3821,
|
| 4834 |
+
"eval_samples_per_second": 3.505,
|
| 4835 |
+
"eval_steps_per_second": 1.767,
|
| 4836 |
+
"step": 2570
|
| 4837 |
+
},
|
| 4838 |
+
{
|
| 4839 |
+
"epoch": 0.0515,
|
| 4840 |
+
"grad_norm": 0.029728034222700407,
|
| 4841 |
+
"learning_rate": 5.1480000000000005e-06,
|
| 4842 |
+
"loss": 2.439,
|
| 4843 |
+
"step": 2575
|
| 4844 |
+
},
|
| 4845 |
+
{
|
| 4846 |
+
"epoch": 0.0515,
|
| 4847 |
+
"eval_loss": 2.440525531768799,
|
| 4848 |
+
"eval_runtime": 34.3072,
|
| 4849 |
+
"eval_samples_per_second": 3.41,
|
| 4850 |
+
"eval_steps_per_second": 1.72,
|
| 4851 |
+
"step": 2575
|
| 4852 |
+
},
|
| 4853 |
+
{
|
| 4854 |
+
"epoch": 0.0516,
|
| 4855 |
+
"eval_loss": 2.440373420715332,
|
| 4856 |
+
"eval_runtime": 33.5748,
|
| 4857 |
+
"eval_samples_per_second": 3.485,
|
| 4858 |
+
"eval_steps_per_second": 1.757,
|
| 4859 |
+
"step": 2580
|
| 4860 |
+
},
|
| 4861 |
+
{
|
| 4862 |
+
"epoch": 0.0517,
|
| 4863 |
+
"eval_loss": 2.4405770301818848,
|
| 4864 |
+
"eval_runtime": 35.2655,
|
| 4865 |
+
"eval_samples_per_second": 3.318,
|
| 4866 |
+
"eval_steps_per_second": 1.673,
|
| 4867 |
+
"step": 2585
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 0.0518,
|
| 4871 |
+
"eval_loss": 2.4402198791503906,
|
| 4872 |
+
"eval_runtime": 34.9918,
|
| 4873 |
+
"eval_samples_per_second": 3.344,
|
| 4874 |
+
"eval_steps_per_second": 1.686,
|
| 4875 |
+
"step": 2590
|
| 4876 |
+
},
|
| 4877 |
+
{
|
| 4878 |
+
"epoch": 0.0519,
|
| 4879 |
+
"eval_loss": 2.440136194229126,
|
| 4880 |
+
"eval_runtime": 33.4873,
|
| 4881 |
+
"eval_samples_per_second": 3.494,
|
| 4882 |
+
"eval_steps_per_second": 1.762,
|
| 4883 |
+
"step": 2595
|
| 4884 |
+
},
|
| 4885 |
+
{
|
| 4886 |
+
"epoch": 0.052,
|
| 4887 |
+
"grad_norm": 0.02473354917836018,
|
| 4888 |
+
"learning_rate": 5.198000000000001e-06,
|
| 4889 |
+
"loss": 2.427,
|
| 4890 |
+
"step": 2600
|
| 4891 |
+
},
|
| 4892 |
+
{
|
| 4893 |
+
"epoch": 0.052,
|
| 4894 |
+
"eval_loss": 2.440282106399536,
|
| 4895 |
+
"eval_runtime": 33.4628,
|
| 4896 |
+
"eval_samples_per_second": 3.496,
|
| 4897 |
+
"eval_steps_per_second": 1.763,
|
| 4898 |
+
"step": 2600
|
| 4899 |
+
},
|
| 4900 |
+
{
|
| 4901 |
+
"epoch": 0.0521,
|
| 4902 |
+
"eval_loss": 2.440448045730591,
|
| 4903 |
+
"eval_runtime": 33.4191,
|
| 4904 |
+
"eval_samples_per_second": 3.501,
|
| 4905 |
+
"eval_steps_per_second": 1.765,
|
| 4906 |
+
"step": 2605
|
| 4907 |
+
},
|
| 4908 |
+
{
|
| 4909 |
+
"epoch": 0.0522,
|
| 4910 |
+
"eval_loss": 2.440248966217041,
|
| 4911 |
+
"eval_runtime": 33.4911,
|
| 4912 |
+
"eval_samples_per_second": 3.493,
|
| 4913 |
+
"eval_steps_per_second": 1.762,
|
| 4914 |
+
"step": 2610
|
| 4915 |
+
},
|
| 4916 |
+
{
|
| 4917 |
+
"epoch": 0.0523,
|
| 4918 |
+
"eval_loss": 2.440030336380005,
|
| 4919 |
+
"eval_runtime": 33.4921,
|
| 4920 |
+
"eval_samples_per_second": 3.493,
|
| 4921 |
+
"eval_steps_per_second": 1.762,
|
| 4922 |
+
"step": 2615
|
| 4923 |
+
},
|
| 4924 |
+
{
|
| 4925 |
+
"epoch": 0.0524,
|
| 4926 |
+
"eval_loss": 2.4397685527801514,
|
| 4927 |
+
"eval_runtime": 33.4491,
|
| 4928 |
+
"eval_samples_per_second": 3.498,
|
| 4929 |
+
"eval_steps_per_second": 1.764,
|
| 4930 |
+
"step": 2620
|
| 4931 |
+
},
|
| 4932 |
+
{
|
| 4933 |
+
"epoch": 0.0525,
|
| 4934 |
+
"grad_norm": 0.026533778128592735,
|
| 4935 |
+
"learning_rate": 5.248000000000001e-06,
|
| 4936 |
+
"loss": 2.4214,
|
| 4937 |
+
"step": 2625
|
| 4938 |
+
},
|
| 4939 |
+
{
|
| 4940 |
+
"epoch": 0.0525,
|
| 4941 |
+
"eval_loss": 2.43971848487854,
|
| 4942 |
+
"eval_runtime": 33.3975,
|
| 4943 |
+
"eval_samples_per_second": 3.503,
|
| 4944 |
+
"eval_steps_per_second": 1.767,
|
| 4945 |
+
"step": 2625
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 0.0526,
|
| 4949 |
+
"eval_loss": 2.4398951530456543,
|
| 4950 |
+
"eval_runtime": 33.4912,
|
| 4951 |
+
"eval_samples_per_second": 3.493,
|
| 4952 |
+
"eval_steps_per_second": 1.762,
|
| 4953 |
+
"step": 2630
|
| 4954 |
+
},
|
| 4955 |
+
{
|
| 4956 |
+
"epoch": 0.0527,
|
| 4957 |
+
"eval_loss": 2.43975830078125,
|
| 4958 |
+
"eval_runtime": 33.4071,
|
| 4959 |
+
"eval_samples_per_second": 3.502,
|
| 4960 |
+
"eval_steps_per_second": 1.766,
|
| 4961 |
+
"step": 2635
|
| 4962 |
+
},
|
| 4963 |
+
{
|
| 4964 |
+
"epoch": 0.0528,
|
| 4965 |
+
"eval_loss": 2.439666271209717,
|
| 4966 |
+
"eval_runtime": 33.4208,
|
| 4967 |
+
"eval_samples_per_second": 3.501,
|
| 4968 |
+
"eval_steps_per_second": 1.765,
|
| 4969 |
+
"step": 2640
|
| 4970 |
+
},
|
| 4971 |
+
{
|
| 4972 |
+
"epoch": 0.0529,
|
| 4973 |
+
"eval_loss": 2.439816951751709,
|
| 4974 |
+
"eval_runtime": 33.5111,
|
| 4975 |
+
"eval_samples_per_second": 3.491,
|
| 4976 |
+
"eval_steps_per_second": 1.761,
|
| 4977 |
+
"step": 2645
|
| 4978 |
+
},
|
| 4979 |
+
{
|
| 4980 |
+
"epoch": 0.053,
|
| 4981 |
+
"grad_norm": 0.024723120971366967,
|
| 4982 |
+
"learning_rate": 5.298000000000001e-06,
|
| 4983 |
+
"loss": 2.4241,
|
| 4984 |
+
"step": 2650
|
| 4985 |
+
},
|
| 4986 |
+
{
|
| 4987 |
+
"epoch": 0.053,
|
| 4988 |
+
"eval_loss": 2.4398183822631836,
|
| 4989 |
+
"eval_runtime": 33.506,
|
| 4990 |
+
"eval_samples_per_second": 3.492,
|
| 4991 |
+
"eval_steps_per_second": 1.761,
|
| 4992 |
+
"step": 2650
|
| 4993 |
+
},
|
| 4994 |
+
{
|
| 4995 |
+
"epoch": 0.0531,
|
| 4996 |
+
"eval_loss": 2.4402668476104736,
|
| 4997 |
+
"eval_runtime": 34.1298,
|
| 4998 |
+
"eval_samples_per_second": 3.428,
|
| 4999 |
+
"eval_steps_per_second": 1.729,
|
| 5000 |
+
"step": 2655
|
| 5001 |
+
},
|
| 5002 |
+
{
|
| 5003 |
+
"epoch": 0.0532,
|
| 5004 |
+
"eval_loss": 2.4400885105133057,
|
| 5005 |
+
"eval_runtime": 33.436,
|
| 5006 |
+
"eval_samples_per_second": 3.499,
|
| 5007 |
+
"eval_steps_per_second": 1.765,
|
| 5008 |
+
"step": 2660
|
| 5009 |
+
},
|
| 5010 |
+
{
|
| 5011 |
+
"epoch": 0.0533,
|
| 5012 |
+
"eval_loss": 2.439871311187744,
|
| 5013 |
+
"eval_runtime": 33.3874,
|
| 5014 |
+
"eval_samples_per_second": 3.504,
|
| 5015 |
+
"eval_steps_per_second": 1.767,
|
| 5016 |
+
"step": 2665
|
| 5017 |
+
},
|
| 5018 |
+
{
|
| 5019 |
+
"epoch": 0.0534,
|
| 5020 |
+
"eval_loss": 2.4393365383148193,
|
| 5021 |
+
"eval_runtime": 33.5258,
|
| 5022 |
+
"eval_samples_per_second": 3.49,
|
| 5023 |
+
"eval_steps_per_second": 1.76,
|
| 5024 |
+
"step": 2670
|
| 5025 |
+
},
|
| 5026 |
+
{
|
| 5027 |
+
"epoch": 0.0535,
|
| 5028 |
+
"grad_norm": 0.02173239513971497,
|
| 5029 |
+
"learning_rate": 5.348000000000001e-06,
|
| 5030 |
+
"loss": 2.4295,
|
| 5031 |
+
"step": 2675
|
| 5032 |
+
},
|
| 5033 |
+
{
|
| 5034 |
+
"epoch": 0.0535,
|
| 5035 |
+
"eval_loss": 2.439133405685425,
|
| 5036 |
+
"eval_runtime": 33.4962,
|
| 5037 |
+
"eval_samples_per_second": 3.493,
|
| 5038 |
+
"eval_steps_per_second": 1.761,
|
| 5039 |
+
"step": 2675
|
| 5040 |
+
},
|
| 5041 |
+
{
|
| 5042 |
+
"epoch": 0.0536,
|
| 5043 |
+
"eval_loss": 2.439093589782715,
|
| 5044 |
+
"eval_runtime": 33.4708,
|
| 5045 |
+
"eval_samples_per_second": 3.496,
|
| 5046 |
+
"eval_steps_per_second": 1.763,
|
| 5047 |
+
"step": 2680
|
| 5048 |
+
},
|
| 5049 |
+
{
|
| 5050 |
+
"epoch": 0.0537,
|
| 5051 |
+
"eval_loss": 2.439096212387085,
|
| 5052 |
+
"eval_runtime": 33.4284,
|
| 5053 |
+
"eval_samples_per_second": 3.5,
|
| 5054 |
+
"eval_steps_per_second": 1.765,
|
| 5055 |
+
"step": 2685
|
| 5056 |
+
},
|
| 5057 |
+
{
|
| 5058 |
+
"epoch": 0.0538,
|
| 5059 |
+
"eval_loss": 2.4389584064483643,
|
| 5060 |
+
"eval_runtime": 33.4749,
|
| 5061 |
+
"eval_samples_per_second": 3.495,
|
| 5062 |
+
"eval_steps_per_second": 1.763,
|
| 5063 |
+
"step": 2690
|
| 5064 |
+
},
|
| 5065 |
+
{
|
| 5066 |
+
"epoch": 0.0539,
|
| 5067 |
+
"eval_loss": 2.438805103302002,
|
| 5068 |
+
"eval_runtime": 33.478,
|
| 5069 |
+
"eval_samples_per_second": 3.495,
|
| 5070 |
+
"eval_steps_per_second": 1.762,
|
| 5071 |
+
"step": 2695
|
| 5072 |
+
},
|
| 5073 |
+
{
|
| 5074 |
+
"epoch": 0.054,
|
| 5075 |
+
"grad_norm": 0.023851331909406925,
|
| 5076 |
+
"learning_rate": 5.398e-06,
|
| 5077 |
+
"loss": 2.4302,
|
| 5078 |
+
"step": 2700
|
| 5079 |
+
},
|
| 5080 |
+
{
|
| 5081 |
+
"epoch": 0.054,
|
| 5082 |
+
"eval_loss": 2.4386403560638428,
|
| 5083 |
+
"eval_runtime": 33.4276,
|
| 5084 |
+
"eval_samples_per_second": 3.5,
|
| 5085 |
+
"eval_steps_per_second": 1.765,
|
| 5086 |
+
"step": 2700
|
| 5087 |
+
},
|
| 5088 |
+
{
|
| 5089 |
+
"epoch": 0.0541,
|
| 5090 |
+
"eval_loss": 2.438568115234375,
|
| 5091 |
+
"eval_runtime": 33.528,
|
| 5092 |
+
"eval_samples_per_second": 3.49,
|
| 5093 |
+
"eval_steps_per_second": 1.76,
|
| 5094 |
+
"step": 2705
|
| 5095 |
+
},
|
| 5096 |
+
{
|
| 5097 |
+
"epoch": 0.0542,
|
| 5098 |
+
"eval_loss": 2.438894510269165,
|
| 5099 |
+
"eval_runtime": 33.5228,
|
| 5100 |
+
"eval_samples_per_second": 3.49,
|
| 5101 |
+
"eval_steps_per_second": 1.76,
|
| 5102 |
+
"step": 2710
|
| 5103 |
+
},
|
| 5104 |
+
{
|
| 5105 |
+
"epoch": 0.0543,
|
| 5106 |
+
"eval_loss": 2.4387168884277344,
|
| 5107 |
+
"eval_runtime": 33.4663,
|
| 5108 |
+
"eval_samples_per_second": 3.496,
|
| 5109 |
+
"eval_steps_per_second": 1.763,
|
| 5110 |
+
"step": 2715
|
| 5111 |
+
},
|
| 5112 |
+
{
|
| 5113 |
+
"epoch": 0.0544,
|
| 5114 |
+
"eval_loss": 2.4385879039764404,
|
| 5115 |
+
"eval_runtime": 33.513,
|
| 5116 |
+
"eval_samples_per_second": 3.491,
|
| 5117 |
+
"eval_steps_per_second": 1.761,
|
| 5118 |
+
"step": 2720
|
| 5119 |
+
},
|
| 5120 |
+
{
|
| 5121 |
+
"epoch": 0.0545,
|
| 5122 |
+
"grad_norm": 0.02728082451264937,
|
| 5123 |
+
"learning_rate": 5.448e-06,
|
| 5124 |
+
"loss": 2.4308,
|
| 5125 |
+
"step": 2725
|
| 5126 |
+
},
|
| 5127 |
+
{
|
| 5128 |
+
"epoch": 0.0545,
|
| 5129 |
+
"eval_loss": 2.4388349056243896,
|
| 5130 |
+
"eval_runtime": 33.4525,
|
| 5131 |
+
"eval_samples_per_second": 3.497,
|
| 5132 |
+
"eval_steps_per_second": 1.764,
|
| 5133 |
+
"step": 2725
|
| 5134 |
+
},
|
| 5135 |
+
{
|
| 5136 |
+
"epoch": 0.0546,
|
| 5137 |
+
"eval_loss": 2.438887357711792,
|
| 5138 |
+
"eval_runtime": 33.428,
|
| 5139 |
+
"eval_samples_per_second": 3.5,
|
| 5140 |
+
"eval_steps_per_second": 1.765,
|
| 5141 |
+
"step": 2730
|
| 5142 |
+
},
|
| 5143 |
+
{
|
| 5144 |
+
"epoch": 0.0547,
|
| 5145 |
+
"eval_loss": 2.438713312149048,
|
| 5146 |
+
"eval_runtime": 33.5229,
|
| 5147 |
+
"eval_samples_per_second": 3.49,
|
| 5148 |
+
"eval_steps_per_second": 1.76,
|
| 5149 |
+
"step": 2735
|
| 5150 |
+
},
|
| 5151 |
+
{
|
| 5152 |
+
"epoch": 0.0548,
|
| 5153 |
+
"eval_loss": 2.438657283782959,
|
| 5154 |
+
"eval_runtime": 33.4169,
|
| 5155 |
+
"eval_samples_per_second": 3.501,
|
| 5156 |
+
"eval_steps_per_second": 1.766,
|
| 5157 |
+
"step": 2740
|
| 5158 |
+
},
|
| 5159 |
+
{
|
| 5160 |
+
"epoch": 0.0549,
|
| 5161 |
+
"eval_loss": 2.438544988632202,
|
| 5162 |
+
"eval_runtime": 33.4944,
|
| 5163 |
+
"eval_samples_per_second": 3.493,
|
| 5164 |
+
"eval_steps_per_second": 1.761,
|
| 5165 |
+
"step": 2745
|
| 5166 |
+
},
|
| 5167 |
+
{
|
| 5168 |
+
"epoch": 0.055,
|
| 5169 |
+
"grad_norm": 0.025461121075693184,
|
| 5170 |
+
"learning_rate": 5.498e-06,
|
| 5171 |
+
"loss": 2.4379,
|
| 5172 |
+
"step": 2750
|
| 5173 |
+
},
|
| 5174 |
+
{
|
| 5175 |
+
"epoch": 0.055,
|
| 5176 |
+
"eval_loss": 2.4386098384857178,
|
| 5177 |
+
"eval_runtime": 33.6782,
|
| 5178 |
+
"eval_samples_per_second": 3.474,
|
| 5179 |
+
"eval_steps_per_second": 1.752,
|
| 5180 |
+
"step": 2750
|
| 5181 |
+
},
|
| 5182 |
+
{
|
| 5183 |
+
"epoch": 0.0551,
|
| 5184 |
+
"eval_loss": 2.438521146774292,
|
| 5185 |
+
"eval_runtime": 33.5161,
|
| 5186 |
+
"eval_samples_per_second": 3.491,
|
| 5187 |
+
"eval_steps_per_second": 1.76,
|
| 5188 |
+
"step": 2755
|
| 5189 |
+
},
|
| 5190 |
+
{
|
| 5191 |
+
"epoch": 0.0552,
|
| 5192 |
+
"eval_loss": 2.438474178314209,
|
| 5193 |
+
"eval_runtime": 33.4773,
|
| 5194 |
+
"eval_samples_per_second": 3.495,
|
| 5195 |
+
"eval_steps_per_second": 1.762,
|
| 5196 |
+
"step": 2760
|
| 5197 |
+
},
|
| 5198 |
+
{
|
| 5199 |
+
"epoch": 0.0553,
|
| 5200 |
+
"eval_loss": 2.4382379055023193,
|
| 5201 |
+
"eval_runtime": 33.4869,
|
| 5202 |
+
"eval_samples_per_second": 3.494,
|
| 5203 |
+
"eval_steps_per_second": 1.762,
|
| 5204 |
+
"step": 2765
|
| 5205 |
+
},
|
| 5206 |
+
{
|
| 5207 |
+
"epoch": 0.0554,
|
| 5208 |
+
"eval_loss": 2.438157796859741,
|
| 5209 |
+
"eval_runtime": 33.543,
|
| 5210 |
+
"eval_samples_per_second": 3.488,
|
| 5211 |
+
"eval_steps_per_second": 1.759,
|
| 5212 |
+
"step": 2770
|
| 5213 |
+
},
|
| 5214 |
+
{
|
| 5215 |
+
"epoch": 0.0555,
|
| 5216 |
+
"grad_norm": 0.0234055445054481,
|
| 5217 |
+
"learning_rate": 5.548e-06,
|
| 5218 |
+
"loss": 2.4326,
|
| 5219 |
+
"step": 2775
|
| 5220 |
+
},
|
| 5221 |
+
{
|
| 5222 |
+
"epoch": 0.0555,
|
| 5223 |
+
"eval_loss": 2.438048839569092,
|
| 5224 |
+
"eval_runtime": 33.5073,
|
| 5225 |
+
"eval_samples_per_second": 3.492,
|
| 5226 |
+
"eval_steps_per_second": 1.761,
|
| 5227 |
+
"step": 2775
|
| 5228 |
+
},
|
| 5229 |
+
{
|
| 5230 |
+
"epoch": 0.0556,
|
| 5231 |
+
"eval_loss": 2.4379706382751465,
|
| 5232 |
+
"eval_runtime": 33.4567,
|
| 5233 |
+
"eval_samples_per_second": 3.497,
|
| 5234 |
+
"eval_steps_per_second": 1.763,
|
| 5235 |
+
"step": 2780
|
| 5236 |
+
},
|
| 5237 |
+
{
|
| 5238 |
+
"epoch": 0.0557,
|
| 5239 |
+
"eval_loss": 2.4379332065582275,
|
| 5240 |
+
"eval_runtime": 33.5172,
|
| 5241 |
+
"eval_samples_per_second": 3.491,
|
| 5242 |
+
"eval_steps_per_second": 1.76,
|
| 5243 |
+
"step": 2785
|
| 5244 |
+
},
|
| 5245 |
+
{
|
| 5246 |
+
"epoch": 0.0558,
|
| 5247 |
+
"eval_loss": 2.4380111694335938,
|
| 5248 |
+
"eval_runtime": 33.5913,
|
| 5249 |
+
"eval_samples_per_second": 3.483,
|
| 5250 |
+
"eval_steps_per_second": 1.756,
|
| 5251 |
+
"step": 2790
|
| 5252 |
+
},
|
| 5253 |
+
{
|
| 5254 |
+
"epoch": 0.0559,
|
| 5255 |
+
"eval_loss": 2.4379403591156006,
|
| 5256 |
+
"eval_runtime": 33.5223,
|
| 5257 |
+
"eval_samples_per_second": 3.49,
|
| 5258 |
+
"eval_steps_per_second": 1.76,
|
| 5259 |
+
"step": 2795
|
| 5260 |
+
},
|
| 5261 |
+
{
|
| 5262 |
+
"epoch": 0.056,
|
| 5263 |
+
"grad_norm": 0.024691045411267393,
|
| 5264 |
+
"learning_rate": 5.5980000000000004e-06,
|
| 5265 |
+
"loss": 2.4297,
|
| 5266 |
+
"step": 2800
|
| 5267 |
+
},
|
| 5268 |
+
{
|
| 5269 |
+
"epoch": 0.056,
|
| 5270 |
+
"eval_loss": 2.43778657913208,
|
| 5271 |
+
"eval_runtime": 33.524,
|
| 5272 |
+
"eval_samples_per_second": 3.49,
|
| 5273 |
+
"eval_steps_per_second": 1.76,
|
| 5274 |
+
"step": 2800
|
| 5275 |
+
},
|
| 5276 |
+
{
|
| 5277 |
+
"epoch": 0.0561,
|
| 5278 |
+
"eval_loss": 2.4376559257507324,
|
| 5279 |
+
"eval_runtime": 33.58,
|
| 5280 |
+
"eval_samples_per_second": 3.484,
|
| 5281 |
+
"eval_steps_per_second": 1.757,
|
| 5282 |
+
"step": 2805
|
| 5283 |
+
},
|
| 5284 |
+
{
|
| 5285 |
+
"epoch": 0.0562,
|
| 5286 |
+
"eval_loss": 2.437596559524536,
|
| 5287 |
+
"eval_runtime": 33.5756,
|
| 5288 |
+
"eval_samples_per_second": 3.485,
|
| 5289 |
+
"eval_steps_per_second": 1.757,
|
| 5290 |
+
"step": 2810
|
| 5291 |
+
},
|
| 5292 |
+
{
|
| 5293 |
+
"epoch": 0.0563,
|
| 5294 |
+
"eval_loss": 2.437690496444702,
|
| 5295 |
+
"eval_runtime": 33.5056,
|
| 5296 |
+
"eval_samples_per_second": 3.492,
|
| 5297 |
+
"eval_steps_per_second": 1.761,
|
| 5298 |
+
"step": 2815
|
| 5299 |
+
},
|
| 5300 |
+
{
|
| 5301 |
+
"epoch": 0.0564,
|
| 5302 |
+
"eval_loss": 2.437558174133301,
|
| 5303 |
+
"eval_runtime": 33.4948,
|
| 5304 |
+
"eval_samples_per_second": 3.493,
|
| 5305 |
+
"eval_steps_per_second": 1.761,
|
| 5306 |
+
"step": 2820
|
| 5307 |
+
},
|
| 5308 |
+
{
|
| 5309 |
+
"epoch": 0.0565,
|
| 5310 |
+
"grad_norm": 0.02500330428035899,
|
| 5311 |
+
"learning_rate": 5.648e-06,
|
| 5312 |
+
"loss": 2.4281,
|
| 5313 |
+
"step": 2825
|
| 5314 |
+
},
|
| 5315 |
+
{
|
| 5316 |
+
"epoch": 0.0565,
|
| 5317 |
+
"eval_loss": 2.437875747680664,
|
| 5318 |
+
"eval_runtime": 33.4492,
|
| 5319 |
+
"eval_samples_per_second": 3.498,
|
| 5320 |
+
"eval_steps_per_second": 1.764,
|
| 5321 |
+
"step": 2825
|
| 5322 |
+
},
|
| 5323 |
+
{
|
| 5324 |
+
"epoch": 0.0566,
|
| 5325 |
+
"eval_loss": 2.438183546066284,
|
| 5326 |
+
"eval_runtime": 33.5208,
|
| 5327 |
+
"eval_samples_per_second": 3.49,
|
| 5328 |
+
"eval_steps_per_second": 1.76,
|
| 5329 |
+
"step": 2830
|
| 5330 |
+
},
|
| 5331 |
+
{
|
| 5332 |
+
"epoch": 0.0567,
|
| 5333 |
+
"eval_loss": 2.4375228881835938,
|
| 5334 |
+
"eval_runtime": 33.5319,
|
| 5335 |
+
"eval_samples_per_second": 3.489,
|
| 5336 |
+
"eval_steps_per_second": 1.76,
|
| 5337 |
+
"step": 2835
|
| 5338 |
+
},
|
| 5339 |
+
{
|
| 5340 |
+
"epoch": 0.0568,
|
| 5341 |
+
"eval_loss": 2.437365770339966,
|
| 5342 |
+
"eval_runtime": 33.4734,
|
| 5343 |
+
"eval_samples_per_second": 3.495,
|
| 5344 |
+
"eval_steps_per_second": 1.763,
|
| 5345 |
+
"step": 2840
|
| 5346 |
+
},
|
| 5347 |
+
{
|
| 5348 |
+
"epoch": 0.0569,
|
| 5349 |
+
"eval_loss": 2.4376399517059326,
|
| 5350 |
+
"eval_runtime": 33.4578,
|
| 5351 |
+
"eval_samples_per_second": 3.497,
|
| 5352 |
+
"eval_steps_per_second": 1.763,
|
| 5353 |
+
"step": 2845
|
| 5354 |
+
},
|
| 5355 |
+
{
|
| 5356 |
+
"epoch": 0.057,
|
| 5357 |
+
"grad_norm": 0.023953363978697285,
|
| 5358 |
+
"learning_rate": 5.698e-06,
|
| 5359 |
+
"loss": 2.4341,
|
| 5360 |
+
"step": 2850
|
| 5361 |
+
},
|
| 5362 |
+
{
|
| 5363 |
+
"epoch": 0.057,
|
| 5364 |
+
"eval_loss": 2.437318801879883,
|
| 5365 |
+
"eval_runtime": 33.4551,
|
| 5366 |
+
"eval_samples_per_second": 3.497,
|
| 5367 |
+
"eval_steps_per_second": 1.764,
|
| 5368 |
+
"step": 2850
|
| 5369 |
+
},
|
| 5370 |
+
{
|
| 5371 |
+
"epoch": 0.0571,
|
| 5372 |
+
"eval_loss": 2.437349319458008,
|
| 5373 |
+
"eval_runtime": 33.4482,
|
| 5374 |
+
"eval_samples_per_second": 3.498,
|
| 5375 |
+
"eval_steps_per_second": 1.764,
|
| 5376 |
+
"step": 2855
|
| 5377 |
+
},
|
| 5378 |
+
{
|
| 5379 |
+
"epoch": 0.0572,
|
| 5380 |
+
"eval_loss": 2.437500476837158,
|
| 5381 |
+
"eval_runtime": 33.5179,
|
| 5382 |
+
"eval_samples_per_second": 3.491,
|
| 5383 |
+
"eval_steps_per_second": 1.76,
|
| 5384 |
+
"step": 2860
|
| 5385 |
+
},
|
| 5386 |
+
{
|
| 5387 |
+
"epoch": 0.0573,
|
| 5388 |
+
"eval_loss": 2.4371414184570312,
|
| 5389 |
+
"eval_runtime": 33.4246,
|
| 5390 |
+
"eval_samples_per_second": 3.5,
|
| 5391 |
+
"eval_steps_per_second": 1.765,
|
| 5392 |
+
"step": 2865
|
| 5393 |
+
},
|
| 5394 |
+
{
|
| 5395 |
+
"epoch": 0.0574,
|
| 5396 |
+
"eval_loss": 2.4371588230133057,
|
| 5397 |
+
"eval_runtime": 33.5686,
|
| 5398 |
+
"eval_samples_per_second": 3.485,
|
| 5399 |
+
"eval_steps_per_second": 1.758,
|
| 5400 |
+
"step": 2870
|
| 5401 |
+
},
|
| 5402 |
+
{
|
| 5403 |
+
"epoch": 0.0575,
|
| 5404 |
+
"grad_norm": 0.023037224733864405,
|
| 5405 |
+
"learning_rate": 5.748e-06,
|
| 5406 |
+
"loss": 2.4201,
|
| 5407 |
+
"step": 2875
|
| 5408 |
+
},
|
| 5409 |
+
{
|
| 5410 |
+
"epoch": 0.0575,
|
| 5411 |
+
"eval_loss": 2.4373178482055664,
|
| 5412 |
+
"eval_runtime": 33.4813,
|
| 5413 |
+
"eval_samples_per_second": 3.494,
|
| 5414 |
+
"eval_steps_per_second": 1.762,
|
| 5415 |
+
"step": 2875
|
| 5416 |
+
},
|
| 5417 |
+
{
|
| 5418 |
+
"epoch": 0.0576,
|
| 5419 |
+
"eval_loss": 2.4371204376220703,
|
| 5420 |
+
"eval_runtime": 33.5096,
|
| 5421 |
+
"eval_samples_per_second": 3.492,
|
| 5422 |
+
"eval_steps_per_second": 1.761,
|
| 5423 |
+
"step": 2880
|
| 5424 |
+
},
|
| 5425 |
+
{
|
| 5426 |
+
"epoch": 0.0577,
|
| 5427 |
+
"eval_loss": 2.43719482421875,
|
| 5428 |
+
"eval_runtime": 33.4709,
|
| 5429 |
+
"eval_samples_per_second": 3.496,
|
| 5430 |
+
"eval_steps_per_second": 1.763,
|
| 5431 |
+
"step": 2885
|
| 5432 |
+
},
|
| 5433 |
+
{
|
| 5434 |
+
"epoch": 0.0578,
|
| 5435 |
+
"eval_loss": 2.4369635581970215,
|
| 5436 |
+
"eval_runtime": 33.5125,
|
| 5437 |
+
"eval_samples_per_second": 3.491,
|
| 5438 |
+
"eval_steps_per_second": 1.761,
|
| 5439 |
+
"step": 2890
|
| 5440 |
+
},
|
| 5441 |
+
{
|
| 5442 |
+
"epoch": 0.0579,
|
| 5443 |
+
"eval_loss": 2.4367122650146484,
|
| 5444 |
+
"eval_runtime": 33.5349,
|
| 5445 |
+
"eval_samples_per_second": 3.489,
|
| 5446 |
+
"eval_steps_per_second": 1.759,
|
| 5447 |
+
"step": 2895
|
| 5448 |
+
},
|
| 5449 |
+
{
|
| 5450 |
+
"epoch": 0.058,
|
| 5451 |
+
"grad_norm": 0.023843041578218274,
|
| 5452 |
+
"learning_rate": 5.798e-06,
|
| 5453 |
+
"loss": 2.4322,
|
| 5454 |
+
"step": 2900
|
| 5455 |
+
},
|
| 5456 |
+
{
|
| 5457 |
+
"epoch": 0.058,
|
| 5458 |
+
"eval_loss": 2.436885118484497,
|
| 5459 |
+
"eval_runtime": 33.5038,
|
| 5460 |
+
"eval_samples_per_second": 3.492,
|
| 5461 |
+
"eval_steps_per_second": 1.761,
|
| 5462 |
+
"step": 2900
|
| 5463 |
+
},
|
| 5464 |
+
{
|
| 5465 |
+
"epoch": 0.0581,
|
| 5466 |
+
"eval_loss": 2.4368388652801514,
|
| 5467 |
+
"eval_runtime": 33.4337,
|
| 5468 |
+
"eval_samples_per_second": 3.499,
|
| 5469 |
+
"eval_steps_per_second": 1.765,
|
| 5470 |
+
"step": 2905
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 0.0582,
|
| 5474 |
+
"eval_loss": 2.436776638031006,
|
| 5475 |
+
"eval_runtime": 33.5783,
|
| 5476 |
+
"eval_samples_per_second": 3.484,
|
| 5477 |
+
"eval_steps_per_second": 1.757,
|
| 5478 |
+
"step": 2910
|
| 5479 |
+
},
|
| 5480 |
+
{
|
| 5481 |
+
"epoch": 0.0583,
|
| 5482 |
+
"eval_loss": 2.4369046688079834,
|
| 5483 |
+
"eval_runtime": 33.5764,
|
| 5484 |
+
"eval_samples_per_second": 3.485,
|
| 5485 |
+
"eval_steps_per_second": 1.757,
|
| 5486 |
+
"step": 2915
|
| 5487 |
+
},
|
| 5488 |
+
{
|
| 5489 |
+
"epoch": 0.0584,
|
| 5490 |
+
"eval_loss": 2.4369351863861084,
|
| 5491 |
+
"eval_runtime": 33.5715,
|
| 5492 |
+
"eval_samples_per_second": 3.485,
|
| 5493 |
+
"eval_steps_per_second": 1.757,
|
| 5494 |
+
"step": 2920
|
| 5495 |
+
},
|
| 5496 |
+
{
|
| 5497 |
+
"epoch": 0.0585,
|
| 5498 |
+
"grad_norm": 0.030212978437899864,
|
| 5499 |
+
"learning_rate": 5.848000000000001e-06,
|
| 5500 |
+
"loss": 2.4318,
|
| 5501 |
+
"step": 2925
|
| 5502 |
+
},
|
| 5503 |
+
{
|
| 5504 |
+
"epoch": 0.0585,
|
| 5505 |
+
"eval_loss": 2.4367170333862305,
|
| 5506 |
+
"eval_runtime": 33.455,
|
| 5507 |
+
"eval_samples_per_second": 3.497,
|
| 5508 |
+
"eval_steps_per_second": 1.764,
|
| 5509 |
+
"step": 2925
|
| 5510 |
+
},
|
| 5511 |
+
{
|
| 5512 |
+
"epoch": 0.0586,
|
| 5513 |
+
"eval_loss": 2.4367101192474365,
|
| 5514 |
+
"eval_runtime": 33.3973,
|
| 5515 |
+
"eval_samples_per_second": 3.503,
|
| 5516 |
+
"eval_steps_per_second": 1.767,
|
| 5517 |
+
"step": 2930
|
| 5518 |
+
},
|
| 5519 |
+
{
|
| 5520 |
+
"epoch": 0.0587,
|
| 5521 |
+
"eval_loss": 2.436723470687866,
|
| 5522 |
+
"eval_runtime": 33.4183,
|
| 5523 |
+
"eval_samples_per_second": 3.501,
|
| 5524 |
+
"eval_steps_per_second": 1.766,
|
| 5525 |
+
"step": 2935
|
| 5526 |
+
},
|
| 5527 |
+
{
|
| 5528 |
+
"epoch": 0.0588,
|
| 5529 |
+
"eval_loss": 2.4368371963500977,
|
| 5530 |
+
"eval_runtime": 33.5269,
|
| 5531 |
+
"eval_samples_per_second": 3.49,
|
| 5532 |
+
"eval_steps_per_second": 1.76,
|
| 5533 |
+
"step": 2940
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 0.0589,
|
| 5537 |
+
"eval_loss": 2.436763286590576,
|
| 5538 |
+
"eval_runtime": 33.4623,
|
| 5539 |
+
"eval_samples_per_second": 3.496,
|
| 5540 |
+
"eval_steps_per_second": 1.763,
|
| 5541 |
+
"step": 2945
|
| 5542 |
+
},
|
| 5543 |
+
{
|
| 5544 |
+
"epoch": 0.059,
|
| 5545 |
+
"grad_norm": 0.024293450378328845,
|
| 5546 |
+
"learning_rate": 5.898e-06,
|
| 5547 |
+
"loss": 2.4221,
|
| 5548 |
+
"step": 2950
|
| 5549 |
+
},
|
| 5550 |
+
{
|
| 5551 |
+
"epoch": 0.059,
|
| 5552 |
+
"eval_loss": 2.436692714691162,
|
| 5553 |
+
"eval_runtime": 33.523,
|
| 5554 |
+
"eval_samples_per_second": 3.49,
|
| 5555 |
+
"eval_steps_per_second": 1.76,
|
| 5556 |
+
"step": 2950
|
| 5557 |
+
},
|
| 5558 |
+
{
|
| 5559 |
+
"epoch": 0.0591,
|
| 5560 |
+
"eval_loss": 2.436657667160034,
|
| 5561 |
+
"eval_runtime": 34.902,
|
| 5562 |
+
"eval_samples_per_second": 3.352,
|
| 5563 |
+
"eval_steps_per_second": 1.69,
|
| 5564 |
+
"step": 2955
|
| 5565 |
+
},
|
| 5566 |
+
{
|
| 5567 |
+
"epoch": 0.0592,
|
| 5568 |
+
"eval_loss": 2.436432123184204,
|
| 5569 |
+
"eval_runtime": 33.4808,
|
| 5570 |
+
"eval_samples_per_second": 3.495,
|
| 5571 |
+
"eval_steps_per_second": 1.762,
|
| 5572 |
+
"step": 2960
|
| 5573 |
+
},
|
| 5574 |
+
{
|
| 5575 |
+
"epoch": 0.0593,
|
| 5576 |
+
"eval_loss": 2.436782121658325,
|
| 5577 |
+
"eval_runtime": 34.5166,
|
| 5578 |
+
"eval_samples_per_second": 3.39,
|
| 5579 |
+
"eval_steps_per_second": 1.709,
|
| 5580 |
+
"step": 2965
|
| 5581 |
+
},
|
| 5582 |
+
{
|
| 5583 |
+
"epoch": 0.0594,
|
| 5584 |
+
"eval_loss": 2.4366602897644043,
|
| 5585 |
+
"eval_runtime": 33.7416,
|
| 5586 |
+
"eval_samples_per_second": 3.468,
|
| 5587 |
+
"eval_steps_per_second": 1.749,
|
| 5588 |
+
"step": 2970
|
| 5589 |
+
},
|
| 5590 |
+
{
|
| 5591 |
+
"epoch": 0.0595,
|
| 5592 |
+
"grad_norm": 0.028294127858427973,
|
| 5593 |
+
"learning_rate": 5.9480000000000005e-06,
|
| 5594 |
+
"loss": 2.4196,
|
| 5595 |
+
"step": 2975
|
| 5596 |
+
},
|
| 5597 |
+
{
|
| 5598 |
+
"epoch": 0.0595,
|
| 5599 |
+
"eval_loss": 2.436668872833252,
|
| 5600 |
+
"eval_runtime": 35.1904,
|
| 5601 |
+
"eval_samples_per_second": 3.325,
|
| 5602 |
+
"eval_steps_per_second": 1.677,
|
| 5603 |
+
"step": 2975
|
| 5604 |
+
},
|
| 5605 |
+
{
|
| 5606 |
+
"epoch": 0.0596,
|
| 5607 |
+
"eval_loss": 2.436310052871704,
|
| 5608 |
+
"eval_runtime": 33.583,
|
| 5609 |
+
"eval_samples_per_second": 3.484,
|
| 5610 |
+
"eval_steps_per_second": 1.757,
|
| 5611 |
+
"step": 2980
|
| 5612 |
+
},
|
| 5613 |
+
{
|
| 5614 |
+
"epoch": 0.0597,
|
| 5615 |
+
"eval_loss": 2.4361066818237305,
|
| 5616 |
+
"eval_runtime": 34.1148,
|
| 5617 |
+
"eval_samples_per_second": 3.43,
|
| 5618 |
+
"eval_steps_per_second": 1.729,
|
| 5619 |
+
"step": 2985
|
| 5620 |
+
},
|
| 5621 |
+
{
|
| 5622 |
+
"epoch": 0.0598,
|
| 5623 |
+
"eval_loss": 2.436128854751587,
|
| 5624 |
+
"eval_runtime": 33.7895,
|
| 5625 |
+
"eval_samples_per_second": 3.463,
|
| 5626 |
+
"eval_steps_per_second": 1.746,
|
| 5627 |
+
"step": 2990
|
| 5628 |
+
},
|
| 5629 |
+
{
|
| 5630 |
+
"epoch": 0.0599,
|
| 5631 |
+
"eval_loss": 2.436457872390747,
|
| 5632 |
+
"eval_runtime": 34.0525,
|
| 5633 |
+
"eval_samples_per_second": 3.436,
|
| 5634 |
+
"eval_steps_per_second": 1.733,
|
| 5635 |
+
"step": 2995
|
| 5636 |
+
},
|
| 5637 |
+
{
|
| 5638 |
+
"epoch": 0.06,
|
| 5639 |
+
"grad_norm": 0.02242795270420928,
|
| 5640 |
+
"learning_rate": 5.998000000000001e-06,
|
| 5641 |
+
"loss": 2.4245,
|
| 5642 |
+
"step": 3000
|
| 5643 |
+
},
|
| 5644 |
+
{
|
| 5645 |
+
"epoch": 0.06,
|
| 5646 |
+
"eval_loss": 2.436203718185425,
|
| 5647 |
+
"eval_runtime": 33.6471,
|
| 5648 |
+
"eval_samples_per_second": 3.477,
|
| 5649 |
+
"eval_steps_per_second": 1.753,
|
| 5650 |
+
"step": 3000
|
| 5651 |
}
|
| 5652 |
],
|
| 5653 |
"logging_steps": 25,
|
|
|
|
| 5667 |
"attributes": {}
|
| 5668 |
}
|
| 5669 |
},
|
| 5670 |
+
"total_flos": 8.355905264309764e+18,
|
| 5671 |
"train_batch_size": 1,
|
| 5672 |
"trial_name": null,
|
| 5673 |
"trial_params": null
|