Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1179 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1179 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.9987492182614135,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1598,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012507817385866166,
|
| 13 |
+
"grad_norm": 11.196227689671167,
|
| 14 |
+
"learning_rate": 3.125e-06,
|
| 15 |
+
"loss": 2.214,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.025015634771732333,
|
| 20 |
+
"grad_norm": 8.94526469956846,
|
| 21 |
+
"learning_rate": 6.25e-06,
|
| 22 |
+
"loss": 1.9951,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0375234521575985,
|
| 27 |
+
"grad_norm": 4.435119438184541,
|
| 28 |
+
"learning_rate": 9.375000000000001e-06,
|
| 29 |
+
"loss": 1.7806,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.050031269543464665,
|
| 34 |
+
"grad_norm": 3.5809779955407897,
|
| 35 |
+
"learning_rate": 1.25e-05,
|
| 36 |
+
"loss": 1.6741,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.06253908692933083,
|
| 41 |
+
"grad_norm": 3.166651040106077,
|
| 42 |
+
"learning_rate": 1.5625e-05,
|
| 43 |
+
"loss": 1.6914,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.075046904315197,
|
| 48 |
+
"grad_norm": 3.9053021248979682,
|
| 49 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 50 |
+
"loss": 1.6361,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08755472170106317,
|
| 55 |
+
"grad_norm": 3.155446891106998,
|
| 56 |
+
"learning_rate": 2.1875e-05,
|
| 57 |
+
"loss": 1.6324,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.10006253908692933,
|
| 62 |
+
"grad_norm": 3.479387369477938,
|
| 63 |
+
"learning_rate": 2.5e-05,
|
| 64 |
+
"loss": 1.691,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1125703564727955,
|
| 69 |
+
"grad_norm": 3.1552136407516627,
|
| 70 |
+
"learning_rate": 2.8125000000000003e-05,
|
| 71 |
+
"loss": 1.5845,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.12507817385866166,
|
| 76 |
+
"grad_norm": 2.980287991118556,
|
| 77 |
+
"learning_rate": 3.125e-05,
|
| 78 |
+
"loss": 1.6084,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.13758599124452783,
|
| 83 |
+
"grad_norm": 3.2625386096828253,
|
| 84 |
+
"learning_rate": 3.4375e-05,
|
| 85 |
+
"loss": 1.5345,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.150093808630394,
|
| 90 |
+
"grad_norm": 3.220585292919637,
|
| 91 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 92 |
+
"loss": 1.5981,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.16260162601626016,
|
| 97 |
+
"grad_norm": 3.3487046432982237,
|
| 98 |
+
"learning_rate": 4.0625000000000005e-05,
|
| 99 |
+
"loss": 1.6322,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.17510944340212634,
|
| 104 |
+
"grad_norm": 3.497338581118587,
|
| 105 |
+
"learning_rate": 4.375e-05,
|
| 106 |
+
"loss": 1.5688,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.18761726078799248,
|
| 111 |
+
"grad_norm": 2.7183936172436955,
|
| 112 |
+
"learning_rate": 4.6875e-05,
|
| 113 |
+
"loss": 1.6114,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.20012507817385866,
|
| 118 |
+
"grad_norm": 3.204681866651314,
|
| 119 |
+
"learning_rate": 5e-05,
|
| 120 |
+
"loss": 1.5904,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.21263289555972484,
|
| 125 |
+
"grad_norm": 2.8734026340283725,
|
| 126 |
+
"learning_rate": 4.999403411726636e-05,
|
| 127 |
+
"loss": 1.5623,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.225140712945591,
|
| 132 |
+
"grad_norm": 2.731883784862566,
|
| 133 |
+
"learning_rate": 4.997613931640598e-05,
|
| 134 |
+
"loss": 1.5458,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.23764853033145716,
|
| 139 |
+
"grad_norm": 3.0795590036212364,
|
| 140 |
+
"learning_rate": 4.9946324138081534e-05,
|
| 141 |
+
"loss": 1.5823,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2501563477173233,
|
| 146 |
+
"grad_norm": 3.105095822757511,
|
| 147 |
+
"learning_rate": 4.990460281220164e-05,
|
| 148 |
+
"loss": 1.6045,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2626641651031895,
|
| 153 |
+
"grad_norm": 2.667234411289808,
|
| 154 |
+
"learning_rate": 4.98509952511293e-05,
|
| 155 |
+
"loss": 1.5437,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.27517198248905567,
|
| 160 |
+
"grad_norm": 2.8566556940170322,
|
| 161 |
+
"learning_rate": 4.978552704017835e-05,
|
| 162 |
+
"loss": 1.5735,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2876797998749218,
|
| 167 |
+
"grad_norm": 3.163297164791991,
|
| 168 |
+
"learning_rate": 4.9708229425402354e-05,
|
| 169 |
+
"loss": 1.4995,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.300187617260788,
|
| 174 |
+
"grad_norm": 2.469748730613264,
|
| 175 |
+
"learning_rate": 4.961913929868172e-05,
|
| 176 |
+
"loss": 1.5849,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.31269543464665417,
|
| 181 |
+
"grad_norm": 2.3807543392076562,
|
| 182 |
+
"learning_rate": 4.9518299180116366e-05,
|
| 183 |
+
"loss": 1.542,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3252032520325203,
|
| 188 |
+
"grad_norm": 2.6158605188730752,
|
| 189 |
+
"learning_rate": 4.940575719773205e-05,
|
| 190 |
+
"loss": 1.5421,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.33771106941838647,
|
| 195 |
+
"grad_norm": 2.418711564788159,
|
| 196 |
+
"learning_rate": 4.928156706451035e-05,
|
| 197 |
+
"loss": 1.519,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.35021888680425267,
|
| 202 |
+
"grad_norm": 2.5688271193206056,
|
| 203 |
+
"learning_rate": 4.9145788052752975e-05,
|
| 204 |
+
"loss": 1.534,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3627267041901188,
|
| 209 |
+
"grad_norm": 2.749288049559472,
|
| 210 |
+
"learning_rate": 4.899848496579287e-05,
|
| 211 |
+
"loss": 1.5005,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.37523452157598497,
|
| 216 |
+
"grad_norm": 2.535758124519615,
|
| 217 |
+
"learning_rate": 4.883972810706548e-05,
|
| 218 |
+
"loss": 1.4939,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3877423389618512,
|
| 223 |
+
"grad_norm": 2.7195432729416833,
|
| 224 |
+
"learning_rate": 4.866959324655501e-05,
|
| 225 |
+
"loss": 1.5848,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4002501563477173,
|
| 230 |
+
"grad_norm": 2.573563516367268,
|
| 231 |
+
"learning_rate": 4.8488161584631574e-05,
|
| 232 |
+
"loss": 1.5405,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.41275797373358347,
|
| 237 |
+
"grad_norm": 2.3636405411668773,
|
| 238 |
+
"learning_rate": 4.829551971329672e-05,
|
| 239 |
+
"loss": 1.4888,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.4252657911194497,
|
| 244 |
+
"grad_norm": 2.662627405108673,
|
| 245 |
+
"learning_rate": 4.809175957485558e-05,
|
| 246 |
+
"loss": 1.5268,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.4377736085053158,
|
| 251 |
+
"grad_norm": 2.9176522067958586,
|
| 252 |
+
"learning_rate": 4.7876978418035455e-05,
|
| 253 |
+
"loss": 1.5558,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.450281425891182,
|
| 258 |
+
"grad_norm": 2.6802852707789944,
|
| 259 |
+
"learning_rate": 4.7651278751571984e-05,
|
| 260 |
+
"loss": 1.4693,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4627892432770482,
|
| 265 |
+
"grad_norm": 2.3192718381618747,
|
| 266 |
+
"learning_rate": 4.7414768295284594e-05,
|
| 267 |
+
"loss": 1.4593,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.47529706066291433,
|
| 272 |
+
"grad_norm": 1.9806583859678688,
|
| 273 |
+
"learning_rate": 4.716755992866509e-05,
|
| 274 |
+
"loss": 1.4264,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4878048780487805,
|
| 279 |
+
"grad_norm": 2.3557329695543774,
|
| 280 |
+
"learning_rate": 4.6909771637003556e-05,
|
| 281 |
+
"loss": 1.4341,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.5003126954346466,
|
| 286 |
+
"grad_norm": 2.461600528378801,
|
| 287 |
+
"learning_rate": 4.664152645507744e-05,
|
| 288 |
+
"loss": 1.4021,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.5128205128205128,
|
| 293 |
+
"grad_norm": 2.1270176374651935,
|
| 294 |
+
"learning_rate": 4.636295240843068e-05,
|
| 295 |
+
"loss": 1.3945,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.525328330206379,
|
| 300 |
+
"grad_norm": 2.3665847282413575,
|
| 301 |
+
"learning_rate": 4.607418245227089e-05,
|
| 302 |
+
"loss": 1.4461,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5378361475922452,
|
| 307 |
+
"grad_norm": 2.4752815650228093,
|
| 308 |
+
"learning_rate": 4.577535440801368e-05,
|
| 309 |
+
"loss": 1.4968,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5503439649781113,
|
| 314 |
+
"grad_norm": 2.442249949806804,
|
| 315 |
+
"learning_rate": 4.546661089750462e-05,
|
| 316 |
+
"loss": 1.4575,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5628517823639775,
|
| 321 |
+
"grad_norm": 2.7248721839556795,
|
| 322 |
+
"learning_rate": 4.5148099274950005e-05,
|
| 323 |
+
"loss": 1.4477,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5753595997498436,
|
| 328 |
+
"grad_norm": 2.318960929300146,
|
| 329 |
+
"learning_rate": 4.481997155658898e-05,
|
| 330 |
+
"loss": 1.5055,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5878674171357098,
|
| 335 |
+
"grad_norm": 2.388316725594586,
|
| 336 |
+
"learning_rate": 4.4482384348140695e-05,
|
| 337 |
+
"loss": 1.497,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.600375234521576,
|
| 342 |
+
"grad_norm": 2.3580062020206825,
|
| 343 |
+
"learning_rate": 4.413549877006099e-05,
|
| 344 |
+
"loss": 1.4085,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.6128830519074422,
|
| 349 |
+
"grad_norm": 2.30371005036015,
|
| 350 |
+
"learning_rate": 4.3779480380644337e-05,
|
| 351 |
+
"loss": 1.4317,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.6253908692933083,
|
| 356 |
+
"grad_norm": 2.277490335597429,
|
| 357 |
+
"learning_rate": 4.34144990970077e-05,
|
| 358 |
+
"loss": 1.3452,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6253908692933083,
|
| 363 |
+
"eval_loss": 1.5973767042160034,
|
| 364 |
+
"eval_runtime": 52.0268,
|
| 365 |
+
"eval_samples_per_second": 109.271,
|
| 366 |
+
"eval_steps_per_second": 13.666,
|
| 367 |
+
"step": 500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.6378986866791745,
|
| 371 |
+
"grad_norm": 2.3070851884604475,
|
| 372 |
+
"learning_rate": 4.304072911399415e-05,
|
| 373 |
+
"loss": 1.4545,
|
| 374 |
+
"step": 510
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.6504065040650406,
|
| 378 |
+
"grad_norm": 2.3084992239900246,
|
| 379 |
+
"learning_rate": 4.2658348821034724e-05,
|
| 380 |
+
"loss": 1.4045,
|
| 381 |
+
"step": 520
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.6629143214509068,
|
| 385 |
+
"grad_norm": 2.4525306744373374,
|
| 386 |
+
"learning_rate": 4.226754071700841e-05,
|
| 387 |
+
"loss": 1.3536,
|
| 388 |
+
"step": 530
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6754221388367729,
|
| 392 |
+
"grad_norm": 2.6009908546977956,
|
| 393 |
+
"learning_rate": 4.1868491323140816e-05,
|
| 394 |
+
"loss": 1.3148,
|
| 395 |
+
"step": 540
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.6879299562226392,
|
| 399 |
+
"grad_norm": 2.021929537122587,
|
| 400 |
+
"learning_rate": 4.146139109398303e-05,
|
| 401 |
+
"loss": 1.3477,
|
| 402 |
+
"step": 550
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7004377736085053,
|
| 406 |
+
"grad_norm": 2.5221876399286094,
|
| 407 |
+
"learning_rate": 4.1046434326513305e-05,
|
| 408 |
+
"loss": 1.5216,
|
| 409 |
+
"step": 560
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.7129455909943715,
|
| 413 |
+
"grad_norm": 2.16247339066387,
|
| 414 |
+
"learning_rate": 4.062381906740477e-05,
|
| 415 |
+
"loss": 1.293,
|
| 416 |
+
"step": 570
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7254534083802376,
|
| 420 |
+
"grad_norm": 2.4082164854514474,
|
| 421 |
+
"learning_rate": 4.019374701850362e-05,
|
| 422 |
+
"loss": 1.357,
|
| 423 |
+
"step": 580
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.7379612257661038,
|
| 427 |
+
"grad_norm": 2.147993492009433,
|
| 428 |
+
"learning_rate": 3.975642344056271e-05,
|
| 429 |
+
"loss": 1.307,
|
| 430 |
+
"step": 590
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7504690431519699,
|
| 434 |
+
"grad_norm": 2.327347035494655,
|
| 435 |
+
"learning_rate": 3.931205705527664e-05,
|
| 436 |
+
"loss": 1.3439,
|
| 437 |
+
"step": 600
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.7629768605378362,
|
| 441 |
+
"grad_norm": 2.3709636186020537,
|
| 442 |
+
"learning_rate": 3.886085994566506e-05,
|
| 443 |
+
"loss": 1.3583,
|
| 444 |
+
"step": 610
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7754846779237023,
|
| 448 |
+
"grad_norm": 2.479454672693421,
|
| 449 |
+
"learning_rate": 3.840304745485162e-05,
|
| 450 |
+
"loss": 1.3027,
|
| 451 |
+
"step": 620
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7879924953095685,
|
| 455 |
+
"grad_norm": 3.0231086471463398,
|
| 456 |
+
"learning_rate": 3.7938838083287056e-05,
|
| 457 |
+
"loss": 1.2713,
|
| 458 |
+
"step": 630
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.8005003126954346,
|
| 462 |
+
"grad_norm": 2.4618474144423885,
|
| 463 |
+
"learning_rate": 3.746845338446533e-05,
|
| 464 |
+
"loss": 1.2691,
|
| 465 |
+
"step": 640
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.8130081300813008,
|
| 469 |
+
"grad_norm": 2.683364877303722,
|
| 470 |
+
"learning_rate": 3.6992117859182674e-05,
|
| 471 |
+
"loss": 1.3601,
|
| 472 |
+
"step": 650
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8255159474671669,
|
| 476 |
+
"grad_norm": 2.2275611309820915,
|
| 477 |
+
"learning_rate": 3.651005884838995e-05,
|
| 478 |
+
"loss": 1.3015,
|
| 479 |
+
"step": 660
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.8380237648530331,
|
| 483 |
+
"grad_norm": 2.103409392935606,
|
| 484 |
+
"learning_rate": 3.6022506424689485e-05,
|
| 485 |
+
"loss": 1.2854,
|
| 486 |
+
"step": 670
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8505315822388994,
|
| 490 |
+
"grad_norm": 2.0900193989619416,
|
| 491 |
+
"learning_rate": 3.552969328252817e-05,
|
| 492 |
+
"loss": 1.2873,
|
| 493 |
+
"step": 680
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.8630393996247655,
|
| 497 |
+
"grad_norm": 2.4106505955260005,
|
| 498 |
+
"learning_rate": 3.5031854627139284e-05,
|
| 499 |
+
"loss": 1.2996,
|
| 500 |
+
"step": 690
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8755472170106317,
|
| 504 |
+
"grad_norm": 2.4059236606171717,
|
| 505 |
+
"learning_rate": 3.452922806228587e-05,
|
| 506 |
+
"loss": 1.3185,
|
| 507 |
+
"step": 700
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8880550343964978,
|
| 511 |
+
"grad_norm": 2.5241974809557175,
|
| 512 |
+
"learning_rate": 3.402205347685953e-05,
|
| 513 |
+
"loss": 1.2231,
|
| 514 |
+
"step": 710
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.900562851782364,
|
| 518 |
+
"grad_norm": 2.055185555826128,
|
| 519 |
+
"learning_rate": 3.351057293038841e-05,
|
| 520 |
+
"loss": 1.3374,
|
| 521 |
+
"step": 720
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.9130706691682301,
|
| 525 |
+
"grad_norm": 2.2760930492977054,
|
| 526 |
+
"learning_rate": 3.2995030537509396e-05,
|
| 527 |
+
"loss": 1.3444,
|
| 528 |
+
"step": 730
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.9255784865540964,
|
| 532 |
+
"grad_norm": 2.2552997614472914,
|
| 533 |
+
"learning_rate": 3.247567235145929e-05,
|
| 534 |
+
"loss": 1.2456,
|
| 535 |
+
"step": 740
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.9380863039399625,
|
| 539 |
+
"grad_norm": 2.433944058150599,
|
| 540 |
+
"learning_rate": 3.195274624664086e-05,
|
| 541 |
+
"loss": 1.1923,
|
| 542 |
+
"step": 750
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9505941213258287,
|
| 546 |
+
"grad_norm": 2.1846841795842202,
|
| 547 |
+
"learning_rate": 3.14265018003197e-05,
|
| 548 |
+
"loss": 1.2513,
|
| 549 |
+
"step": 760
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.9631019387116948,
|
| 553 |
+
"grad_norm": 2.4960158537732773,
|
| 554 |
+
"learning_rate": 3.089719017350828e-05,
|
| 555 |
+
"loss": 1.2928,
|
| 556 |
+
"step": 770
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.975609756097561,
|
| 560 |
+
"grad_norm": 2.2820418756697665,
|
| 561 |
+
"learning_rate": 3.03650639910942e-05,
|
| 562 |
+
"loss": 1.2285,
|
| 563 |
+
"step": 780
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.9881175734834271,
|
| 567 |
+
"grad_norm": 2.2304674808935503,
|
| 568 |
+
"learning_rate": 2.983037722126978e-05,
|
| 569 |
+
"loss": 1.2109,
|
| 570 |
+
"step": 790
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.0006253908692933,
|
| 574 |
+
"grad_norm": 2.4884834373076257,
|
| 575 |
+
"learning_rate": 2.929338505432045e-05,
|
| 576 |
+
"loss": 1.2071,
|
| 577 |
+
"step": 800
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.0131332082551594,
|
| 581 |
+
"grad_norm": 2.2652235105915866,
|
| 582 |
+
"learning_rate": 2.8754343780829968e-05,
|
| 583 |
+
"loss": 0.9707,
|
| 584 |
+
"step": 810
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.0256410256410255,
|
| 588 |
+
"grad_norm": 2.7829211847796436,
|
| 589 |
+
"learning_rate": 2.821351066936043e-05,
|
| 590 |
+
"loss": 0.9281,
|
| 591 |
+
"step": 820
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.038148843026892,
|
| 595 |
+
"grad_norm": 2.3690248065827264,
|
| 596 |
+
"learning_rate": 2.767114384366555e-05,
|
| 597 |
+
"loss": 0.9399,
|
| 598 |
+
"step": 830
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.050656660412758,
|
| 602 |
+
"grad_norm": 2.5409598116910734,
|
| 603 |
+
"learning_rate": 2.7127502159495782e-05,
|
| 604 |
+
"loss": 0.8951,
|
| 605 |
+
"step": 840
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.0631644777986242,
|
| 609 |
+
"grad_norm": 2.5575660182733673,
|
| 610 |
+
"learning_rate": 2.6582845081054096e-05,
|
| 611 |
+
"loss": 0.8859,
|
| 612 |
+
"step": 850
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.0756722951844904,
|
| 616 |
+
"grad_norm": 2.243946653312665,
|
| 617 |
+
"learning_rate": 2.6037432557161285e-05,
|
| 618 |
+
"loss": 0.8429,
|
| 619 |
+
"step": 860
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.0881801125703565,
|
| 623 |
+
"grad_norm": 2.727808580020501,
|
| 624 |
+
"learning_rate": 2.5491524897190068e-05,
|
| 625 |
+
"loss": 0.8958,
|
| 626 |
+
"step": 870
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.1006879299562227,
|
| 630 |
+
"grad_norm": 2.6050024232102906,
|
| 631 |
+
"learning_rate": 2.4945382646827058e-05,
|
| 632 |
+
"loss": 0.9412,
|
| 633 |
+
"step": 880
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.1131957473420888,
|
| 637 |
+
"grad_norm": 2.5841504477219135,
|
| 638 |
+
"learning_rate": 2.4399266463721993e-05,
|
| 639 |
+
"loss": 0.9089,
|
| 640 |
+
"step": 890
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.125703564727955,
|
| 644 |
+
"grad_norm": 2.5085199599814882,
|
| 645 |
+
"learning_rate": 2.3853436993083453e-05,
|
| 646 |
+
"loss": 0.9373,
|
| 647 |
+
"step": 900
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.1382113821138211,
|
| 651 |
+
"grad_norm": 3.2343558948374165,
|
| 652 |
+
"learning_rate": 2.3308154743280595e-05,
|
| 653 |
+
"loss": 0.8741,
|
| 654 |
+
"step": 910
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.1507191994996873,
|
| 658 |
+
"grad_norm": 2.881822507701942,
|
| 659 |
+
"learning_rate": 2.2763679961510137e-05,
|
| 660 |
+
"loss": 0.8959,
|
| 661 |
+
"step": 920
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.1632270168855534,
|
| 665 |
+
"grad_norm": 2.6590136429428104,
|
| 666 |
+
"learning_rate": 2.222027250958805e-05,
|
| 667 |
+
"loss": 0.8447,
|
| 668 |
+
"step": 930
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.1757348342714196,
|
| 672 |
+
"grad_norm": 2.330687894654819,
|
| 673 |
+
"learning_rate": 2.1678191739925107e-05,
|
| 674 |
+
"loss": 0.8689,
|
| 675 |
+
"step": 940
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.1882426516572857,
|
| 679 |
+
"grad_norm": 2.790243974607849,
|
| 680 |
+
"learning_rate": 2.1137696371745608e-05,
|
| 681 |
+
"loss": 0.8875,
|
| 682 |
+
"step": 950
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.200750469043152,
|
| 686 |
+
"grad_norm": 2.283914922416874,
|
| 687 |
+
"learning_rate": 2.059904436760835e-05,
|
| 688 |
+
"loss": 0.8459,
|
| 689 |
+
"step": 960
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.2132582864290182,
|
| 693 |
+
"grad_norm": 2.575042681933137,
|
| 694 |
+
"learning_rate": 2.006249281028859e-05,
|
| 695 |
+
"loss": 0.8405,
|
| 696 |
+
"step": 970
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.2257661038148844,
|
| 700 |
+
"grad_norm": 2.5311184713219,
|
| 701 |
+
"learning_rate": 1.9528297780080067e-05,
|
| 702 |
+
"loss": 0.9238,
|
| 703 |
+
"step": 980
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.2382739212007505,
|
| 707 |
+
"grad_norm": 2.7281405763745727,
|
| 708 |
+
"learning_rate": 1.899671423257534e-05,
|
| 709 |
+
"loss": 0.906,
|
| 710 |
+
"step": 990
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.2507817385866167,
|
| 714 |
+
"grad_norm": 2.273017241171145,
|
| 715 |
+
"learning_rate": 1.8467995876983002e-05,
|
| 716 |
+
"loss": 0.8293,
|
| 717 |
+
"step": 1000
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.2507817385866167,
|
| 721 |
+
"eval_loss": 1.3726614713668823,
|
| 722 |
+
"eval_runtime": 51.9356,
|
| 723 |
+
"eval_samples_per_second": 109.462,
|
| 724 |
+
"eval_steps_per_second": 13.69,
|
| 725 |
+
"step": 1000
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.2632895559724828,
|
| 729 |
+
"grad_norm": 2.8020838753185022,
|
| 730 |
+
"learning_rate": 1.7942395055039756e-05,
|
| 731 |
+
"loss": 0.818,
|
| 732 |
+
"step": 1010
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.275797373358349,
|
| 736 |
+
"grad_norm": 2.814873162280655,
|
| 737 |
+
"learning_rate": 1.742016262057507e-05,
|
| 738 |
+
"loss": 0.8782,
|
| 739 |
+
"step": 1020
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.2883051907442151,
|
| 743 |
+
"grad_norm": 2.5555727855199133,
|
| 744 |
+
"learning_rate": 1.690154781978604e-05,
|
| 745 |
+
"loss": 0.8692,
|
| 746 |
+
"step": 1030
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.3008130081300813,
|
| 750 |
+
"grad_norm": 2.9198901354453866,
|
| 751 |
+
"learning_rate": 1.638679817227951e-05,
|
| 752 |
+
"loss": 0.885,
|
| 753 |
+
"step": 1040
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.3133208255159474,
|
| 757 |
+
"grad_norm": 2.326734936151972,
|
| 758 |
+
"learning_rate": 1.5876159352938202e-05,
|
| 759 |
+
"loss": 0.8594,
|
| 760 |
+
"step": 1050
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.3258286429018136,
|
| 764 |
+
"grad_norm": 2.6527776239267333,
|
| 765 |
+
"learning_rate": 1.5369875074667365e-05,
|
| 766 |
+
"loss": 0.864,
|
| 767 |
+
"step": 1060
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.3383364602876797,
|
| 771 |
+
"grad_norm": 2.719137361822375,
|
| 772 |
+
"learning_rate": 1.4868186972077708e-05,
|
| 773 |
+
"loss": 0.8384,
|
| 774 |
+
"step": 1070
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.3508442776735459,
|
| 778 |
+
"grad_norm": 2.23395715978574,
|
| 779 |
+
"learning_rate": 1.4371334486160354e-05,
|
| 780 |
+
"loss": 0.7987,
|
| 781 |
+
"step": 1080
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.3633520950594122,
|
| 785 |
+
"grad_norm": 2.538354497553029,
|
| 786 |
+
"learning_rate": 1.3879554750008653e-05,
|
| 787 |
+
"loss": 0.8451,
|
| 788 |
+
"step": 1090
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.3758599124452782,
|
| 792 |
+
"grad_norm": 2.4429406057964735,
|
| 793 |
+
"learning_rate": 1.3393082475641544e-05,
|
| 794 |
+
"loss": 0.836,
|
| 795 |
+
"step": 1100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.3883677298311445,
|
| 799 |
+
"grad_norm": 2.1166767893102234,
|
| 800 |
+
"learning_rate": 1.2912149841982388e-05,
|
| 801 |
+
"loss": 0.8215,
|
| 802 |
+
"step": 1110
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.4008755472170107,
|
| 806 |
+
"grad_norm": 2.2346931729430732,
|
| 807 |
+
"learning_rate": 1.2436986384046786e-05,
|
| 808 |
+
"loss": 0.7785,
|
| 809 |
+
"step": 1120
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.4133833646028768,
|
| 813 |
+
"grad_norm": 2.494815885218028,
|
| 814 |
+
"learning_rate": 1.1967818883392306e-05,
|
| 815 |
+
"loss": 0.7624,
|
| 816 |
+
"step": 1130
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.425891181988743,
|
| 820 |
+
"grad_norm": 2.901143519765259,
|
| 821 |
+
"learning_rate": 1.150487125988224e-05,
|
| 822 |
+
"loss": 0.82,
|
| 823 |
+
"step": 1140
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.4383989993746091,
|
| 827 |
+
"grad_norm": 2.7828290624179957,
|
| 828 |
+
"learning_rate": 1.1048364464815297e-05,
|
| 829 |
+
"loss": 0.8018,
|
| 830 |
+
"step": 1150
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.4509068167604753,
|
| 834 |
+
"grad_norm": 2.4761447552027196,
|
| 835 |
+
"learning_rate": 1.0598516375471978e-05,
|
| 836 |
+
"loss": 0.8224,
|
| 837 |
+
"step": 1160
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.4634146341463414,
|
| 841 |
+
"grad_norm": 2.7960949007283014,
|
| 842 |
+
"learning_rate": 1.0155541691128212e-05,
|
| 843 |
+
"loss": 0.8362,
|
| 844 |
+
"step": 1170
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.4759224515322076,
|
| 848 |
+
"grad_norm": 2.8586426222751355,
|
| 849 |
+
"learning_rate": 9.719651830585668e-06,
|
| 850 |
+
"loss": 0.814,
|
| 851 |
+
"step": 1180
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.4884302689180737,
|
| 855 |
+
"grad_norm": 2.444704855440996,
|
| 856 |
+
"learning_rate": 9.291054831267749e-06,
|
| 857 |
+
"loss": 0.812,
|
| 858 |
+
"step": 1190
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.50093808630394,
|
| 862 |
+
"grad_norm": 2.203245268486062,
|
| 863 |
+
"learning_rate": 8.869955249929509e-06,
|
| 864 |
+
"loss": 0.756,
|
| 865 |
+
"step": 1200
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.513445903689806,
|
| 869 |
+
"grad_norm": 2.1667676187504643,
|
| 870 |
+
"learning_rate": 8.456554065028651e-06,
|
| 871 |
+
"loss": 0.7866,
|
| 872 |
+
"step": 1210
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.5259537210756724,
|
| 876 |
+
"grad_norm": 3.1729468036439368,
|
| 877 |
+
"learning_rate": 8.051048580804457e-06,
|
| 878 |
+
"loss": 0.7789,
|
| 879 |
+
"step": 1220
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.5384615384615383,
|
| 883 |
+
"grad_norm": 2.745706957042119,
|
| 884 |
+
"learning_rate": 7.65363233311028e-06,
|
| 885 |
+
"loss": 0.7651,
|
| 886 |
+
"step": 1230
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.5509693558474047,
|
| 890 |
+
"grad_norm": 2.3285021876802614,
|
| 891 |
+
"learning_rate": 7.264494997044522e-06,
|
| 892 |
+
"loss": 0.7461,
|
| 893 |
+
"step": 1240
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.5634771732332708,
|
| 897 |
+
"grad_norm": 2.703446713628546,
|
| 898 |
+
"learning_rate": 6.88382229642433e-06,
|
| 899 |
+
"loss": 0.7933,
|
| 900 |
+
"step": 1250
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.575984990619137,
|
| 904 |
+
"grad_norm": 2.4555911035434517,
|
| 905 |
+
"learning_rate": 6.51179591514505e-06,
|
| 906 |
+
"loss": 0.7563,
|
| 907 |
+
"step": 1260
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.5884928080050031,
|
| 911 |
+
"grad_norm": 2.0388567105663604,
|
| 912 |
+
"learning_rate": 6.1485934104678285e-06,
|
| 913 |
+
"loss": 0.6997,
|
| 914 |
+
"step": 1270
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.6010006253908693,
|
| 918 |
+
"grad_norm": 2.998906503082893,
|
| 919 |
+
"learning_rate": 5.794388128276798e-06,
|
| 920 |
+
"loss": 0.7702,
|
| 921 |
+
"step": 1280
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.6135084427767354,
|
| 925 |
+
"grad_norm": 2.295408095489897,
|
| 926 |
+
"learning_rate": 5.449349120346137e-06,
|
| 927 |
+
"loss": 0.7681,
|
| 928 |
+
"step": 1290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.6260162601626016,
|
| 932 |
+
"grad_norm": 2.3383927307138155,
|
| 933 |
+
"learning_rate": 5.113641063656618e-06,
|
| 934 |
+
"loss": 0.7848,
|
| 935 |
+
"step": 1300
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.6385240775484677,
|
| 939 |
+
"grad_norm": 2.535009363777371,
|
| 940 |
+
"learning_rate": 4.787424181800174e-06,
|
| 941 |
+
"loss": 0.7713,
|
| 942 |
+
"step": 1310
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.6510318949343339,
|
| 946 |
+
"grad_norm": 2.443338473601547,
|
| 947 |
+
"learning_rate": 4.470854168509819e-06,
|
| 948 |
+
"loss": 0.7888,
|
| 949 |
+
"step": 1320
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.6635397123202003,
|
| 953 |
+
"grad_norm": 2.686883685963269,
|
| 954 |
+
"learning_rate": 4.164082113351672e-06,
|
| 955 |
+
"loss": 0.7497,
|
| 956 |
+
"step": 1330
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.6760475297060662,
|
| 960 |
+
"grad_norm": 2.8068580095390674,
|
| 961 |
+
"learning_rate": 3.867254429614281e-06,
|
| 962 |
+
"loss": 0.7899,
|
| 963 |
+
"step": 1340
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.6885553470919326,
|
| 967 |
+
"grad_norm": 2.3048887226109063,
|
| 968 |
+
"learning_rate": 3.580512784429921e-06,
|
| 969 |
+
"loss": 0.7631,
|
| 970 |
+
"step": 1350
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.7010631644777985,
|
| 974 |
+
"grad_norm": 2.3839076230191307,
|
| 975 |
+
"learning_rate": 3.3039940311609868e-06,
|
| 976 |
+
"loss": 0.7005,
|
| 977 |
+
"step": 1360
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.7135709818636649,
|
| 981 |
+
"grad_norm": 2.750016598535585,
|
| 982 |
+
"learning_rate": 3.0378301440839253e-06,
|
| 983 |
+
"loss": 0.7584,
|
| 984 |
+
"step": 1370
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.726078799249531,
|
| 988 |
+
"grad_norm": 2.445902129036907,
|
| 989 |
+
"learning_rate": 2.7821481554018065e-06,
|
| 990 |
+
"loss": 0.7177,
|
| 991 |
+
"step": 1380
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.7385866166353972,
|
| 995 |
+
"grad_norm": 2.459439087514897,
|
| 996 |
+
"learning_rate": 2.5370700946155456e-06,
|
| 997 |
+
"loss": 0.7107,
|
| 998 |
+
"step": 1390
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.7510944340212633,
|
| 1002 |
+
"grad_norm": 2.3917637382771217,
|
| 1003 |
+
"learning_rate": 2.3027129302828494e-06,
|
| 1004 |
+
"loss": 0.7252,
|
| 1005 |
+
"step": 1400
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.7636022514071295,
|
| 1009 |
+
"grad_norm": 2.5825813580196977,
|
| 1010 |
+
"learning_rate": 2.0791885141925344e-06,
|
| 1011 |
+
"loss": 0.7265,
|
| 1012 |
+
"step": 1410
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.7761100687929956,
|
| 1016 |
+
"grad_norm": 2.7444524417938907,
|
| 1017 |
+
"learning_rate": 1.8666035279809535e-06,
|
| 1018 |
+
"loss": 0.7868,
|
| 1019 |
+
"step": 1420
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.7886178861788617,
|
| 1023 |
+
"grad_norm": 2.5747339796466777,
|
| 1024 |
+
"learning_rate": 1.6650594322160097e-06,
|
| 1025 |
+
"loss": 0.7461,
|
| 1026 |
+
"step": 1430
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.8011257035647281,
|
| 1030 |
+
"grad_norm": 2.814780313810251,
|
| 1031 |
+
"learning_rate": 1.4746524179729742e-06,
|
| 1032 |
+
"loss": 0.751,
|
| 1033 |
+
"step": 1440
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.813633520950594,
|
| 1037 |
+
"grad_norm": 2.491280187376505,
|
| 1038 |
+
"learning_rate": 1.295473360925345e-06,
|
| 1039 |
+
"loss": 0.7407,
|
| 1040 |
+
"step": 1450
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.8261413383364604,
|
| 1044 |
+
"grad_norm": 2.6865531055864786,
|
| 1045 |
+
"learning_rate": 1.1276077779725335e-06,
|
| 1046 |
+
"loss": 0.7467,
|
| 1047 |
+
"step": 1460
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.8386491557223263,
|
| 1051 |
+
"grad_norm": 2.693944976015148,
|
| 1052 |
+
"learning_rate": 9.711357864251647e-07,
|
| 1053 |
+
"loss": 0.7721,
|
| 1054 |
+
"step": 1470
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.8511569731081927,
|
| 1058 |
+
"grad_norm": 3.1854255386603527,
|
| 1059 |
+
"learning_rate": 8.261320657674626e-07,
|
| 1060 |
+
"loss": 0.7555,
|
| 1061 |
+
"step": 1480
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.8636647904940586,
|
| 1065 |
+
"grad_norm": 2.3969657119662906,
|
| 1066 |
+
"learning_rate": 6.926658220148929e-07,
|
| 1067 |
+
"loss": 0.7728,
|
| 1068 |
+
"step": 1490
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.876172607879925,
|
| 1072 |
+
"grad_norm": 2.4361472021651105,
|
| 1073 |
+
"learning_rate": 5.708007546841893e-07,
|
| 1074 |
+
"loss": 0.7086,
|
| 1075 |
+
"step": 1500
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.876172607879925,
|
| 1079 |
+
"eval_loss": 1.2464503049850464,
|
| 1080 |
+
"eval_runtime": 51.9233,
|
| 1081 |
+
"eval_samples_per_second": 109.489,
|
| 1082 |
+
"eval_steps_per_second": 13.693,
|
| 1083 |
+
"step": 1500
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.8886804252657912,
|
| 1087 |
+
"grad_norm": 2.5813463203755656,
|
| 1088 |
+
"learning_rate": 4.6059502639142773e-07,
|
| 1089 |
+
"loss": 0.6916,
|
| 1090 |
+
"step": 1510
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.9011882426516573,
|
| 1094 |
+
"grad_norm": 2.8984929947738176,
|
| 1095 |
+
"learning_rate": 3.6210123509273775e-07,
|
| 1096 |
+
"loss": 0.7647,
|
| 1097 |
+
"step": 1520
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 1.9136960600375235,
|
| 1101 |
+
"grad_norm": 2.4332442721477934,
|
| 1102 |
+
"learning_rate": 2.753663889808328e-07,
|
| 1103 |
+
"loss": 0.7037,
|
| 1104 |
+
"step": 1530
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 1.9262038774233896,
|
| 1108 |
+
"grad_norm": 2.7482039407993035,
|
| 1109 |
+
"learning_rate": 2.0043188404937075e-07,
|
| 1110 |
+
"loss": 0.7467,
|
| 1111 |
+
"step": 1540
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.9387116948092558,
|
| 1115 |
+
"grad_norm": 2.6495770008884647,
|
| 1116 |
+
"learning_rate": 1.3733348433589143e-07,
|
| 1117 |
+
"loss": 0.7894,
|
| 1118 |
+
"step": 1550
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 1.951219512195122,
|
| 1122 |
+
"grad_norm": 2.6742202900643752,
|
| 1123 |
+
"learning_rate": 8.610130485265655e-08,
|
| 1124 |
+
"loss": 0.805,
|
| 1125 |
+
"step": 1560
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 1.9637273295809883,
|
| 1129 |
+
"grad_norm": 2.445319151525402,
|
| 1130 |
+
"learning_rate": 4.67597972136663e-08,
|
| 1131 |
+
"loss": 0.7277,
|
| 1132 |
+
"step": 1570
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 1.9762351469668542,
|
| 1136 |
+
"grad_norm": 2.5097808639896324,
|
| 1137 |
+
"learning_rate": 1.9327737964616733e-08,
|
| 1138 |
+
"loss": 0.7574,
|
| 1139 |
+
"step": 1580
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.9887429643527206,
|
| 1143 |
+
"grad_norm": 2.5217973584786435,
|
| 1144 |
+
"learning_rate": 3.8182196213903955e-09,
|
| 1145 |
+
"loss": 0.7318,
|
| 1146 |
+
"step": 1590
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 1.9987492182614135,
|
| 1150 |
+
"step": 1598,
|
| 1151 |
+
"total_flos": 18204158066688.0,
|
| 1152 |
+
"train_loss": 1.1366474037623973,
|
| 1153 |
+
"train_runtime": 1015.6993,
|
| 1154 |
+
"train_samples_per_second": 100.746,
|
| 1155 |
+
"train_steps_per_second": 1.573
|
| 1156 |
+
}
|
| 1157 |
+
],
|
| 1158 |
+
"logging_steps": 10,
|
| 1159 |
+
"max_steps": 1598,
|
| 1160 |
+
"num_input_tokens_seen": 0,
|
| 1161 |
+
"num_train_epochs": 2,
|
| 1162 |
+
"save_steps": 5000,
|
| 1163 |
+
"stateful_callbacks": {
|
| 1164 |
+
"TrainerControl": {
|
| 1165 |
+
"args": {
|
| 1166 |
+
"should_epoch_stop": false,
|
| 1167 |
+
"should_evaluate": false,
|
| 1168 |
+
"should_log": false,
|
| 1169 |
+
"should_save": true,
|
| 1170 |
+
"should_training_stop": true
|
| 1171 |
+
},
|
| 1172 |
+
"attributes": {}
|
| 1173 |
+
}
|
| 1174 |
+
},
|
| 1175 |
+
"total_flos": 18204158066688.0,
|
| 1176 |
+
"train_batch_size": 4,
|
| 1177 |
+
"trial_name": null,
|
| 1178 |
+
"trial_params": null
|
| 1179 |
+
}
|