Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1202 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.992407324698526,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 837,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01786511835640911,
|
| 13 |
+
"grad_norm": 0.5766569375991821,
|
| 14 |
+
"learning_rate": 4.999559763441755e-05,
|
| 15 |
+
"loss": 1.5453,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03573023671281822,
|
| 20 |
+
"grad_norm": 0.4148772656917572,
|
| 21 |
+
"learning_rate": 4.998239208813602e-05,
|
| 22 |
+
"loss": 1.4724,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.053595355069227336,
|
| 27 |
+
"grad_norm": 0.3131537139415741,
|
| 28 |
+
"learning_rate": 4.9960388012006784e-05,
|
| 29 |
+
"loss": 1.2985,
|
| 30 |
+
"step": 15
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07146047342563644,
|
| 34 |
+
"grad_norm": 0.26158425211906433,
|
| 35 |
+
"learning_rate": 4.992959315562887e-05,
|
| 36 |
+
"loss": 1.3255,
|
| 37 |
+
"step": 20
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08932559178204555,
|
| 41 |
+
"grad_norm": 0.23404298722743988,
|
| 42 |
+
"learning_rate": 4.9890018364619516e-05,
|
| 43 |
+
"loss": 1.245,
|
| 44 |
+
"step": 25
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10719071013845467,
|
| 48 |
+
"grad_norm": 0.19169265031814575,
|
| 49 |
+
"learning_rate": 4.984167757679458e-05,
|
| 50 |
+
"loss": 1.2509,
|
| 51 |
+
"step": 30
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12505582849486377,
|
| 55 |
+
"grad_norm": 0.29275861382484436,
|
| 56 |
+
"learning_rate": 4.9784587817259674e-05,
|
| 57 |
+
"loss": 1.2937,
|
| 58 |
+
"step": 35
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14292094685127288,
|
| 62 |
+
"grad_norm": 0.1952601969242096,
|
| 63 |
+
"learning_rate": 4.971876919241422e-05,
|
| 64 |
+
"loss": 1.1818,
|
| 65 |
+
"step": 40
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.160786065207682,
|
| 69 |
+
"grad_norm": 0.223506361246109,
|
| 70 |
+
"learning_rate": 4.964424488287009e-05,
|
| 71 |
+
"loss": 1.2293,
|
| 72 |
+
"step": 45
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1786511835640911,
|
| 76 |
+
"grad_norm": 0.23478031158447266,
|
| 77 |
+
"learning_rate": 4.956104113528776e-05,
|
| 78 |
+
"loss": 1.2531,
|
| 79 |
+
"step": 50
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19651630192050024,
|
| 83 |
+
"grad_norm": 0.26763561367988586,
|
| 84 |
+
"learning_rate": 4.9469187253132356e-05,
|
| 85 |
+
"loss": 1.2131,
|
| 86 |
+
"step": 55
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21438142027690935,
|
| 90 |
+
"grad_norm": 0.2691376507282257,
|
| 91 |
+
"learning_rate": 4.936871558635346e-05,
|
| 92 |
+
"loss": 1.2497,
|
| 93 |
+
"step": 60
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23224653863331846,
|
| 97 |
+
"grad_norm": 0.2545244097709656,
|
| 98 |
+
"learning_rate": 4.925966151999168e-05,
|
| 99 |
+
"loss": 1.1778,
|
| 100 |
+
"step": 65
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25011165698972754,
|
| 104 |
+
"grad_norm": 0.30710119009017944,
|
| 105 |
+
"learning_rate": 4.914206346171651e-05,
|
| 106 |
+
"loss": 1.2123,
|
| 107 |
+
"step": 70
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.2679767753461367,
|
| 111 |
+
"grad_norm": 0.23636460304260254,
|
| 112 |
+
"learning_rate": 4.901596282829948e-05,
|
| 113 |
+
"loss": 1.2007,
|
| 114 |
+
"step": 75
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28584189370254576,
|
| 118 |
+
"grad_norm": 0.2662898600101471,
|
| 119 |
+
"learning_rate": 4.888140403102768e-05,
|
| 120 |
+
"loss": 1.1745,
|
| 121 |
+
"step": 80
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3037070120589549,
|
| 125 |
+
"grad_norm": 0.308378130197525,
|
| 126 |
+
"learning_rate": 4.8738434460062524e-05,
|
| 127 |
+
"loss": 1.1981,
|
| 128 |
+
"step": 85
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.321572130415364,
|
| 132 |
+
"grad_norm": 0.23143641650676727,
|
| 133 |
+
"learning_rate": 4.858710446774951e-05,
|
| 134 |
+
"loss": 1.1297,
|
| 135 |
+
"step": 90
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3394372487717731,
|
| 139 |
+
"grad_norm": 0.2781310975551605,
|
| 140 |
+
"learning_rate": 4.842746735088461e-05,
|
| 141 |
+
"loss": 1.2359,
|
| 142 |
+
"step": 95
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3573023671281822,
|
| 146 |
+
"grad_norm": 0.29989567399024963,
|
| 147 |
+
"learning_rate": 4.825957933194376e-05,
|
| 148 |
+
"loss": 1.1798,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37516748548459133,
|
| 153 |
+
"grad_norm": 0.3551138937473297,
|
| 154 |
+
"learning_rate": 4.808349953928184e-05,
|
| 155 |
+
"loss": 1.2055,
|
| 156 |
+
"step": 105
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39303260384100047,
|
| 160 |
+
"grad_norm": 0.3111821413040161,
|
| 161 |
+
"learning_rate": 4.789928998630838e-05,
|
| 162 |
+
"loss": 1.2145,
|
| 163 |
+
"step": 110
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41089772219740955,
|
| 167 |
+
"grad_norm": 0.27980926632881165,
|
| 168 |
+
"learning_rate": 4.770701554964706e-05,
|
| 169 |
+
"loss": 1.1673,
|
| 170 |
+
"step": 115
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4287628405538187,
|
| 174 |
+
"grad_norm": 0.2745010256767273,
|
| 175 |
+
"learning_rate": 4.750674394628687e-05,
|
| 176 |
+
"loss": 1.1494,
|
| 177 |
+
"step": 120
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4466279589102278,
|
| 181 |
+
"grad_norm": 0.2852526307106018,
|
| 182 |
+
"learning_rate": 4.729854570973289e-05,
|
| 183 |
+
"loss": 1.1219,
|
| 184 |
+
"step": 125
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4644930772666369,
|
| 188 |
+
"grad_norm": 0.35505807399749756,
|
| 189 |
+
"learning_rate": 4.7082494165165216e-05,
|
| 190 |
+
"loss": 1.1953,
|
| 191 |
+
"step": 130
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.482358195623046,
|
| 195 |
+
"grad_norm": 0.32421571016311646,
|
| 196 |
+
"learning_rate": 4.685866540361456e-05,
|
| 197 |
+
"loss": 1.0871,
|
| 198 |
+
"step": 135
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5002233139794551,
|
| 202 |
+
"grad_norm": 0.27323105931282043,
|
| 203 |
+
"learning_rate": 4.662713825516379e-05,
|
| 204 |
+
"loss": 1.1604,
|
| 205 |
+
"step": 140
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5180884323358642,
|
| 209 |
+
"grad_norm": 0.3120843470096588,
|
| 210 |
+
"learning_rate": 4.638799426118492e-05,
|
| 211 |
+
"loss": 1.1202,
|
| 212 |
+
"step": 145
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5359535506922734,
|
| 216 |
+
"grad_norm": 0.26333290338516235,
|
| 217 |
+
"learning_rate": 4.6141317645621e-05,
|
| 218 |
+
"loss": 1.1539,
|
| 219 |
+
"step": 150
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5538186690486825,
|
| 223 |
+
"grad_norm": 0.270474374294281,
|
| 224 |
+
"learning_rate": 4.588719528532342e-05,
|
| 225 |
+
"loss": 1.1343,
|
| 226 |
+
"step": 155
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5716837874050915,
|
| 230 |
+
"grad_norm": 0.2665104866027832,
|
| 231 |
+
"learning_rate": 4.5625716679454787e-05,
|
| 232 |
+
"loss": 1.2117,
|
| 233 |
+
"step": 160
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5895489057615007,
|
| 237 |
+
"grad_norm": 0.29763564467430115,
|
| 238 |
+
"learning_rate": 4.535697391796832e-05,
|
| 239 |
+
"loss": 1.1323,
|
| 240 |
+
"step": 165
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6074140241179098,
|
| 244 |
+
"grad_norm": 0.36548730731010437,
|
| 245 |
+
"learning_rate": 4.50810616491747e-05,
|
| 246 |
+
"loss": 1.1848,
|
| 247 |
+
"step": 170
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6252791424743189,
|
| 251 |
+
"grad_norm": 0.32450002431869507,
|
| 252 |
+
"learning_rate": 4.479807704640802e-05,
|
| 253 |
+
"loss": 1.1829,
|
| 254 |
+
"step": 175
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.643144260830728,
|
| 258 |
+
"grad_norm": 0.4330536425113678,
|
| 259 |
+
"learning_rate": 4.45081197738023e-05,
|
| 260 |
+
"loss": 1.1881,
|
| 261 |
+
"step": 180
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6610093791871371,
|
| 265 |
+
"grad_norm": 0.3612164258956909,
|
| 266 |
+
"learning_rate": 4.421129195119094e-05,
|
| 267 |
+
"loss": 1.23,
|
| 268 |
+
"step": 185
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6788744975435462,
|
| 272 |
+
"grad_norm": 0.3616299033164978,
|
| 273 |
+
"learning_rate": 4.390769811814116e-05,
|
| 274 |
+
"loss": 1.1593,
|
| 275 |
+
"step": 190
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6967396158999554,
|
| 279 |
+
"grad_norm": 0.3271953761577606,
|
| 280 |
+
"learning_rate": 4.359744519713628e-05,
|
| 281 |
+
"loss": 1.1454,
|
| 282 |
+
"step": 195
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7146047342563644,
|
| 286 |
+
"grad_norm": 0.3824130594730377,
|
| 287 |
+
"learning_rate": 4.3280642455918806e-05,
|
| 288 |
+
"loss": 1.1633,
|
| 289 |
+
"step": 200
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7324698526127735,
|
| 293 |
+
"grad_norm": 0.36359068751335144,
|
| 294 |
+
"learning_rate": 4.2957401469007495e-05,
|
| 295 |
+
"loss": 1.1841,
|
| 296 |
+
"step": 205
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7503349709691827,
|
| 300 |
+
"grad_norm": 0.41856512427330017,
|
| 301 |
+
"learning_rate": 4.262783607840199e-05,
|
| 302 |
+
"loss": 1.1567,
|
| 303 |
+
"step": 210
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7682000893255918,
|
| 307 |
+
"grad_norm": 0.2762792408466339,
|
| 308 |
+
"learning_rate": 4.229206235348891e-05,
|
| 309 |
+
"loss": 1.1147,
|
| 310 |
+
"step": 215
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7860652076820009,
|
| 314 |
+
"grad_norm": 0.37006062269210815,
|
| 315 |
+
"learning_rate": 4.195019855016346e-05,
|
| 316 |
+
"loss": 1.1194,
|
| 317 |
+
"step": 220
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.80393032603841,
|
| 321 |
+
"grad_norm": 0.33580338954925537,
|
| 322 |
+
"learning_rate": 4.160236506918098e-05,
|
| 323 |
+
"loss": 1.1884,
|
| 324 |
+
"step": 225
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8217954443948191,
|
| 328 |
+
"grad_norm": 0.33786651492118835,
|
| 329 |
+
"learning_rate": 4.124868441375307e-05,
|
| 330 |
+
"loss": 1.1687,
|
| 331 |
+
"step": 230
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8396605627512282,
|
| 335 |
+
"grad_norm": 0.3432258665561676,
|
| 336 |
+
"learning_rate": 4.08892811464033e-05,
|
| 337 |
+
"loss": 1.1512,
|
| 338 |
+
"step": 235
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8575256811076374,
|
| 342 |
+
"grad_norm": 0.3321971595287323,
|
| 343 |
+
"learning_rate": 4.052428184509762e-05,
|
| 344 |
+
"loss": 1.1417,
|
| 345 |
+
"step": 240
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8753907994640464,
|
| 349 |
+
"grad_norm": 0.371055543422699,
|
| 350 |
+
"learning_rate": 4.0153815058664976e-05,
|
| 351 |
+
"loss": 1.1595,
|
| 352 |
+
"step": 245
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8932559178204555,
|
| 356 |
+
"grad_norm": 0.3127325773239136,
|
| 357 |
+
"learning_rate": 3.977801126152376e-05,
|
| 358 |
+
"loss": 1.1661,
|
| 359 |
+
"step": 250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9111210361768647,
|
| 363 |
+
"grad_norm": 0.28069669008255005,
|
| 364 |
+
"learning_rate": 3.9397002807730166e-05,
|
| 365 |
+
"loss": 1.1271,
|
| 366 |
+
"step": 255
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9289861545332738,
|
| 370 |
+
"grad_norm": 0.39312201738357544,
|
| 371 |
+
"learning_rate": 3.9010923884364467e-05,
|
| 372 |
+
"loss": 1.0852,
|
| 373 |
+
"step": 260
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9468512728896828,
|
| 377 |
+
"grad_norm": 0.44475504755973816,
|
| 378 |
+
"learning_rate": 3.861991046427182e-05,
|
| 379 |
+
"loss": 1.0884,
|
| 380 |
+
"step": 265
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.964716391246092,
|
| 384 |
+
"grad_norm": 0.3146402835845947,
|
| 385 |
+
"learning_rate": 3.822410025817406e-05,
|
| 386 |
+
"loss": 1.1159,
|
| 387 |
+
"step": 270
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9825815096025011,
|
| 391 |
+
"grad_norm": 0.4495191276073456,
|
| 392 |
+
"learning_rate": 3.782363266616946e-05,
|
| 393 |
+
"loss": 1.1642,
|
| 394 |
+
"step": 275
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.0013398838767307,
|
| 398 |
+
"grad_norm": 0.30039986968040466,
|
| 399 |
+
"learning_rate": 3.741864872863754e-05,
|
| 400 |
+
"loss": 1.1599,
|
| 401 |
+
"step": 280
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0192050022331398,
|
| 405 |
+
"grad_norm": 0.3826463222503662,
|
| 406 |
+
"learning_rate": 3.700929107656614e-05,
|
| 407 |
+
"loss": 1.1167,
|
| 408 |
+
"step": 285
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.037070120589549,
|
| 412 |
+
"grad_norm": 0.38993003964424133,
|
| 413 |
+
"learning_rate": 3.659570388131832e-05,
|
| 414 |
+
"loss": 1.087,
|
| 415 |
+
"step": 290
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.054935238945958,
|
| 419 |
+
"grad_norm": 0.4054009020328522,
|
| 420 |
+
"learning_rate": 3.61780328038568e-05,
|
| 421 |
+
"loss": 1.0912,
|
| 422 |
+
"step": 295
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.0728003573023672,
|
| 426 |
+
"grad_norm": 0.47545984387397766,
|
| 427 |
+
"learning_rate": 3.575642494344365e-05,
|
| 428 |
+
"loss": 1.0991,
|
| 429 |
+
"step": 300
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.0906654756587761,
|
| 433 |
+
"grad_norm": 0.4191313683986664,
|
| 434 |
+
"learning_rate": 3.533102878583361e-05,
|
| 435 |
+
"loss": 1.1219,
|
| 436 |
+
"step": 305
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.1085305940151853,
|
| 440 |
+
"grad_norm": 0.3124043643474579,
|
| 441 |
+
"learning_rate": 3.490199415097892e-05,
|
| 442 |
+
"loss": 1.1198,
|
| 443 |
+
"step": 310
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.1263957123715944,
|
| 447 |
+
"grad_norm": 0.36894044280052185,
|
| 448 |
+
"learning_rate": 3.44694721402644e-05,
|
| 449 |
+
"loss": 1.1193,
|
| 450 |
+
"step": 315
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.1442608307280036,
|
| 454 |
+
"grad_norm": 0.5683015584945679,
|
| 455 |
+
"learning_rate": 3.4033615083291135e-05,
|
| 456 |
+
"loss": 1.142,
|
| 457 |
+
"step": 320
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.1621259490844127,
|
| 461 |
+
"grad_norm": 0.371493399143219,
|
| 462 |
+
"learning_rate": 3.3594576484227655e-05,
|
| 463 |
+
"loss": 1.128,
|
| 464 |
+
"step": 325
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.1799910674408218,
|
| 468 |
+
"grad_norm": 0.38947540521621704,
|
| 469 |
+
"learning_rate": 3.315251096774737e-05,
|
| 470 |
+
"loss": 1.1046,
|
| 471 |
+
"step": 330
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.197856185797231,
|
| 475 |
+
"grad_norm": 0.3288847804069519,
|
| 476 |
+
"learning_rate": 3.2707574224571495e-05,
|
| 477 |
+
"loss": 1.1209,
|
| 478 |
+
"step": 335
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.21572130415364,
|
| 482 |
+
"grad_norm": 0.36680731177330017,
|
| 483 |
+
"learning_rate": 3.225992295663639e-05,
|
| 484 |
+
"loss": 1.0932,
|
| 485 |
+
"step": 340
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.233586422510049,
|
| 489 |
+
"grad_norm": 0.4268665909767151,
|
| 490 |
+
"learning_rate": 3.1809714821904834e-05,
|
| 491 |
+
"loss": 1.1161,
|
| 492 |
+
"step": 345
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.2514515408664582,
|
| 496 |
+
"grad_norm": 0.37560388445854187,
|
| 497 |
+
"learning_rate": 3.1357108378840616e-05,
|
| 498 |
+
"loss": 1.1264,
|
| 499 |
+
"step": 350
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.2693166592228673,
|
| 503 |
+
"grad_norm": 0.35783258080482483,
|
| 504 |
+
"learning_rate": 3.0902263030565925e-05,
|
| 505 |
+
"loss": 1.1226,
|
| 506 |
+
"step": 355
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.2871817775792764,
|
| 510 |
+
"grad_norm": 0.3974683880805969,
|
| 511 |
+
"learning_rate": 3.0445338968721287e-05,
|
| 512 |
+
"loss": 1.0651,
|
| 513 |
+
"step": 360
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3050468959356856,
|
| 517 |
+
"grad_norm": 0.36973434686660767,
|
| 518 |
+
"learning_rate": 2.9986497117047797e-05,
|
| 519 |
+
"loss": 1.1052,
|
| 520 |
+
"step": 365
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.3229120142920947,
|
| 524 |
+
"grad_norm": 0.40590983629226685,
|
| 525 |
+
"learning_rate": 2.9525899074711506e-05,
|
| 526 |
+
"loss": 1.0979,
|
| 527 |
+
"step": 370
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.3407771326485038,
|
| 531 |
+
"grad_norm": 0.3565100431442261,
|
| 532 |
+
"learning_rate": 2.906370705938991e-05,
|
| 533 |
+
"loss": 1.1254,
|
| 534 |
+
"step": 375
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.358642251004913,
|
| 538 |
+
"grad_norm": 0.4463523328304291,
|
| 539 |
+
"learning_rate": 2.8600083850140692e-05,
|
| 540 |
+
"loss": 1.1275,
|
| 541 |
+
"step": 380
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.3765073693613221,
|
| 545 |
+
"grad_norm": 0.4577709436416626,
|
| 546 |
+
"learning_rate": 2.8135192730072598e-05,
|
| 547 |
+
"loss": 1.0984,
|
| 548 |
+
"step": 385
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.3943724877177313,
|
| 552 |
+
"grad_norm": 0.3673183023929596,
|
| 553 |
+
"learning_rate": 2.7669197428838972e-05,
|
| 554 |
+
"loss": 1.1126,
|
| 555 |
+
"step": 390
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.4122376060741402,
|
| 559 |
+
"grad_norm": 0.3968072235584259,
|
| 560 |
+
"learning_rate": 2.7202262064973873e-05,
|
| 561 |
+
"loss": 1.1271,
|
| 562 |
+
"step": 395
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.4301027244305493,
|
| 566 |
+
"grad_norm": 0.5759745240211487,
|
| 567 |
+
"learning_rate": 2.6734551088091293e-05,
|
| 568 |
+
"loss": 1.1135,
|
| 569 |
+
"step": 400
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.4479678427869584,
|
| 573 |
+
"grad_norm": 0.4903964698314667,
|
| 574 |
+
"learning_rate": 2.6266229220967818e-05,
|
| 575 |
+
"loss": 1.1163,
|
| 576 |
+
"step": 405
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.4658329611433676,
|
| 580 |
+
"grad_norm": 0.4729979932308197,
|
| 581 |
+
"learning_rate": 2.5797461401529e-05,
|
| 582 |
+
"loss": 1.1011,
|
| 583 |
+
"step": 410
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.4836980794997767,
|
| 587 |
+
"grad_norm": 0.45003455877304077,
|
| 588 |
+
"learning_rate": 2.5328412724759976e-05,
|
| 589 |
+
"loss": 1.1506,
|
| 590 |
+
"step": 415
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.5015631978561856,
|
| 594 |
+
"grad_norm": 0.4313727021217346,
|
| 595 |
+
"learning_rate": 2.485924838456086e-05,
|
| 596 |
+
"loss": 1.1257,
|
| 597 |
+
"step": 420
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.5194283162125948,
|
| 601 |
+
"grad_norm": 0.39038538932800293,
|
| 602 |
+
"learning_rate": 2.4390133615567136e-05,
|
| 603 |
+
"loss": 1.0858,
|
| 604 |
+
"step": 425
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.537293434569004,
|
| 608 |
+
"grad_norm": 0.49707311391830444,
|
| 609 |
+
"learning_rate": 2.3921233634955868e-05,
|
| 610 |
+
"loss": 1.1602,
|
| 611 |
+
"step": 430
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.555158552925413,
|
| 615 |
+
"grad_norm": 0.46587345004081726,
|
| 616 |
+
"learning_rate": 2.3452713584257955e-05,
|
| 617 |
+
"loss": 1.1144,
|
| 618 |
+
"step": 435
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.5730236712818222,
|
| 622 |
+
"grad_norm": 0.37671202421188354,
|
| 623 |
+
"learning_rate": 2.2984738471197077e-05,
|
| 624 |
+
"loss": 1.1793,
|
| 625 |
+
"step": 440
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.5908887896382313,
|
| 629 |
+
"grad_norm": 0.4020828306674957,
|
| 630 |
+
"learning_rate": 2.2517473111575725e-05,
|
| 631 |
+
"loss": 1.1206,
|
| 632 |
+
"step": 445
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.6087539079946405,
|
| 636 |
+
"grad_norm": 0.5164027214050293,
|
| 637 |
+
"learning_rate": 2.2051082071228854e-05,
|
| 638 |
+
"loss": 1.1177,
|
| 639 |
+
"step": 450
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.6266190263510496,
|
| 643 |
+
"grad_norm": 0.3768448829650879,
|
| 644 |
+
"learning_rate": 2.1585729608065595e-05,
|
| 645 |
+
"loss": 1.0802,
|
| 646 |
+
"step": 455
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.6444841447074587,
|
| 650 |
+
"grad_norm": 0.3956853747367859,
|
| 651 |
+
"learning_rate": 2.1121579614219345e-05,
|
| 652 |
+
"loss": 1.0913,
|
| 653 |
+
"step": 460
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.6623492630638679,
|
| 657 |
+
"grad_norm": 0.36919939517974854,
|
| 658 |
+
"learning_rate": 2.0658795558326743e-05,
|
| 659 |
+
"loss": 1.0896,
|
| 660 |
+
"step": 465
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.680214381420277,
|
| 664 |
+
"grad_norm": 0.45556601881980896,
|
| 665 |
+
"learning_rate": 2.0197540427955777e-05,
|
| 666 |
+
"loss": 1.1183,
|
| 667 |
+
"step": 470
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.6980794997766862,
|
| 671 |
+
"grad_norm": 0.47595128417015076,
|
| 672 |
+
"learning_rate": 1.97379766722033e-05,
|
| 673 |
+
"loss": 1.2037,
|
| 674 |
+
"step": 475
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.7159446181330953,
|
| 678 |
+
"grad_norm": 0.3943460285663605,
|
| 679 |
+
"learning_rate": 1.928026614448221e-05,
|
| 680 |
+
"loss": 1.0682,
|
| 681 |
+
"step": 480
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.7338097364895042,
|
| 685 |
+
"grad_norm": 0.40548044443130493,
|
| 686 |
+
"learning_rate": 1.8824570045518423e-05,
|
| 687 |
+
"loss": 1.0664,
|
| 688 |
+
"step": 485
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.7516748548459133,
|
| 692 |
+
"grad_norm": 0.4159083962440491,
|
| 693 |
+
"learning_rate": 1.8371048866577713e-05,
|
| 694 |
+
"loss": 1.1475,
|
| 695 |
+
"step": 490
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.7695399732023225,
|
| 699 |
+
"grad_norm": 0.4973352551460266,
|
| 700 |
+
"learning_rate": 1.79198623329424e-05,
|
| 701 |
+
"loss": 1.0675,
|
| 702 |
+
"step": 495
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.7874050915587316,
|
| 706 |
+
"grad_norm": 0.4413686692714691,
|
| 707 |
+
"learning_rate": 1.747116934765785e-05,
|
| 708 |
+
"loss": 1.0956,
|
| 709 |
+
"step": 500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.8052702099151405,
|
| 713 |
+
"grad_norm": 0.4255582094192505,
|
| 714 |
+
"learning_rate": 1.70251279355685e-05,
|
| 715 |
+
"loss": 1.071,
|
| 716 |
+
"step": 505
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.8231353282715497,
|
| 720 |
+
"grad_norm": 0.7180905938148499,
|
| 721 |
+
"learning_rate": 1.658189518766322e-05,
|
| 722 |
+
"loss": 1.1428,
|
| 723 |
+
"step": 510
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.8410004466279588,
|
| 727 |
+
"grad_norm": 0.6441544890403748,
|
| 728 |
+
"learning_rate": 1.6141627205749563e-05,
|
| 729 |
+
"loss": 1.1014,
|
| 730 |
+
"step": 515
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.858865564984368,
|
| 734 |
+
"grad_norm": 0.415781170129776,
|
| 735 |
+
"learning_rate": 1.5704479047476377e-05,
|
| 736 |
+
"loss": 1.1245,
|
| 737 |
+
"step": 520
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.876730683340777,
|
| 741 |
+
"grad_norm": 0.40880969166755676,
|
| 742 |
+
"learning_rate": 1.5270604671724188e-05,
|
| 743 |
+
"loss": 1.0354,
|
| 744 |
+
"step": 525
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.8945958016971862,
|
| 748 |
+
"grad_norm": 0.5105963349342346,
|
| 749 |
+
"learning_rate": 1.4840156884382491e-05,
|
| 750 |
+
"loss": 1.0717,
|
| 751 |
+
"step": 530
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.9124609200535954,
|
| 755 |
+
"grad_norm": 0.6085048317909241,
|
| 756 |
+
"learning_rate": 1.4413287284533228e-05,
|
| 757 |
+
"loss": 1.1494,
|
| 758 |
+
"step": 535
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.9303260384100045,
|
| 762 |
+
"grad_norm": 0.6034563183784485,
|
| 763 |
+
"learning_rate": 1.399014621105914e-05,
|
| 764 |
+
"loss": 1.1347,
|
| 765 |
+
"step": 540
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.9481911567664136,
|
| 769 |
+
"grad_norm": 0.5377976298332214,
|
| 770 |
+
"learning_rate": 1.3570882689696127e-05,
|
| 771 |
+
"loss": 1.1074,
|
| 772 |
+
"step": 545
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.9660562751228228,
|
| 776 |
+
"grad_norm": 0.6293488144874573,
|
| 777 |
+
"learning_rate": 1.3155644380547877e-05,
|
| 778 |
+
"loss": 1.1264,
|
| 779 |
+
"step": 550
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.983921393479232,
|
| 783 |
+
"grad_norm": 0.45377108454704285,
|
| 784 |
+
"learning_rate": 1.2744577526081666e-05,
|
| 785 |
+
"loss": 1.0915,
|
| 786 |
+
"step": 555
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.0026797677534613,
|
| 790 |
+
"grad_norm": 0.38156697154045105,
|
| 791 |
+
"learning_rate": 1.2337826899623265e-05,
|
| 792 |
+
"loss": 1.1034,
|
| 793 |
+
"step": 560
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.0205448861098705,
|
| 797 |
+
"grad_norm": 0.5213829278945923,
|
| 798 |
+
"learning_rate": 1.193553575436935e-05,
|
| 799 |
+
"loss": 1.0915,
|
| 800 |
+
"step": 565
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.0384100044662796,
|
| 804 |
+
"grad_norm": 0.46200066804885864,
|
| 805 |
+
"learning_rate": 1.1537845772935279e-05,
|
| 806 |
+
"loss": 1.0797,
|
| 807 |
+
"step": 570
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.0562751228226888,
|
| 811 |
+
"grad_norm": 0.6238696575164795,
|
| 812 |
+
"learning_rate": 1.1144897017456e-05,
|
| 813 |
+
"loss": 1.086,
|
| 814 |
+
"step": 575
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.074140241179098,
|
| 818 |
+
"grad_norm": 0.45493337512016296,
|
| 819 |
+
"learning_rate": 1.0756827880257661e-05,
|
| 820 |
+
"loss": 1.1068,
|
| 821 |
+
"step": 580
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.092005359535507,
|
| 825 |
+
"grad_norm": 0.535967230796814,
|
| 826 |
+
"learning_rate": 1.0373775035117305e-05,
|
| 827 |
+
"loss": 1.0574,
|
| 828 |
+
"step": 585
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.109870477891916,
|
| 832 |
+
"grad_norm": 0.42889833450317383,
|
| 833 |
+
"learning_rate": 9.995873389127895e-06,
|
| 834 |
+
"loss": 1.1409,
|
| 835 |
+
"step": 590
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.1277355962483253,
|
| 839 |
+
"grad_norm": 0.3753083646297455,
|
| 840 |
+
"learning_rate": 9.623256035185382e-06,
|
| 841 |
+
"loss": 1.1461,
|
| 842 |
+
"step": 595
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.1456007146047344,
|
| 846 |
+
"grad_norm": 0.508271336555481,
|
| 847 |
+
"learning_rate": 9.256054205114939e-06,
|
| 848 |
+
"loss": 1.1039,
|
| 849 |
+
"step": 600
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.163465832961143,
|
| 853 |
+
"grad_norm": 0.45634639263153076,
|
| 854 |
+
"learning_rate": 8.894397223452453e-06,
|
| 855 |
+
"loss": 1.0832,
|
| 856 |
+
"step": 605
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.1813309513175523,
|
| 860 |
+
"grad_norm": 0.557949423789978,
|
| 861 |
+
"learning_rate": 8.538412461897807e-06,
|
| 862 |
+
"loss": 1.0621,
|
| 863 |
+
"step": 610
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.1991960696739614,
|
| 867 |
+
"grad_norm": 0.5253410935401917,
|
| 868 |
+
"learning_rate": 8.188225294455992e-06,
|
| 869 |
+
"loss": 1.1078,
|
| 870 |
+
"step": 615
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.2170611880303706,
|
| 874 |
+
"grad_norm": 0.4432696998119354,
|
| 875 |
+
"learning_rate": 7.843959053281663e-06,
|
| 876 |
+
"loss": 1.114,
|
| 877 |
+
"step": 620
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.2349263063867797,
|
| 881 |
+
"grad_norm": 0.4199691712856293,
|
| 882 |
+
"learning_rate": 7.505734985242927e-06,
|
| 883 |
+
"loss": 1.0363,
|
| 884 |
+
"step": 625
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.252791424743189,
|
| 888 |
+
"grad_norm": 0.4136619567871094,
|
| 889 |
+
"learning_rate": 7.173672209219495e-06,
|
| 890 |
+
"loss": 1.0577,
|
| 891 |
+
"step": 630
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.270656543099598,
|
| 895 |
+
"grad_norm": 0.589055061340332,
|
| 896 |
+
"learning_rate": 6.847887674150266e-06,
|
| 897 |
+
"loss": 1.1141,
|
| 898 |
+
"step": 635
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.288521661456007,
|
| 902 |
+
"grad_norm": 0.42672625184059143,
|
| 903 |
+
"learning_rate": 6.528496117845215e-06,
|
| 904 |
+
"loss": 1.0775,
|
| 905 |
+
"step": 640
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.3063867798124162,
|
| 909 |
+
"grad_norm": 0.6587265133857727,
|
| 910 |
+
"learning_rate": 6.215610026575916e-06,
|
| 911 |
+
"loss": 1.1016,
|
| 912 |
+
"step": 645
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.3242518981688254,
|
| 916 |
+
"grad_norm": 0.4289712607860565,
|
| 917 |
+
"learning_rate": 5.909339595459123e-06,
|
| 918 |
+
"loss": 1.0545,
|
| 919 |
+
"step": 650
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.3421170165252345,
|
| 923 |
+
"grad_norm": 0.4478652775287628,
|
| 924 |
+
"learning_rate": 5.609792689647222e-06,
|
| 925 |
+
"loss": 1.1381,
|
| 926 |
+
"step": 655
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.3599821348816437,
|
| 930 |
+
"grad_norm": 0.529575765132904,
|
| 931 |
+
"learning_rate": 5.317074806339295e-06,
|
| 932 |
+
"loss": 1.0688,
|
| 933 |
+
"step": 660
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.377847253238053,
|
| 937 |
+
"grad_norm": 0.5171645879745483,
|
| 938 |
+
"learning_rate": 5.0312890376261415e-06,
|
| 939 |
+
"loss": 1.0419,
|
| 940 |
+
"step": 665
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.395712371594462,
|
| 944 |
+
"grad_norm": 0.5134549140930176,
|
| 945 |
+
"learning_rate": 4.752536034182312e-06,
|
| 946 |
+
"loss": 1.0742,
|
| 947 |
+
"step": 670
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.413577489950871,
|
| 951 |
+
"grad_norm": 0.36803004145622253,
|
| 952 |
+
"learning_rate": 4.480913969818098e-06,
|
| 953 |
+
"loss": 1.0799,
|
| 954 |
+
"step": 675
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.43144260830728,
|
| 958 |
+
"grad_norm": 0.5440959334373474,
|
| 959 |
+
"learning_rate": 4.216518506903683e-06,
|
| 960 |
+
"loss": 1.0485,
|
| 961 |
+
"step": 680
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.4493077266636893,
|
| 965 |
+
"grad_norm": 0.5819982290267944,
|
| 966 |
+
"learning_rate": 3.959442762677967e-06,
|
| 967 |
+
"loss": 1.0833,
|
| 968 |
+
"step": 685
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.467172845020098,
|
| 972 |
+
"grad_norm": 0.4522368609905243,
|
| 973 |
+
"learning_rate": 3.70977727645363e-06,
|
| 974 |
+
"loss": 1.0997,
|
| 975 |
+
"step": 690
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.485037963376507,
|
| 979 |
+
"grad_norm": 0.4565070569515228,
|
| 980 |
+
"learning_rate": 3.467609977730155e-06,
|
| 981 |
+
"loss": 1.0652,
|
| 982 |
+
"step": 695
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.5029030817329163,
|
| 986 |
+
"grad_norm": 0.46055707335472107,
|
| 987 |
+
"learning_rate": 3.233026155226046e-06,
|
| 988 |
+
"loss": 1.0679,
|
| 989 |
+
"step": 700
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.5207682000893255,
|
| 993 |
+
"grad_norm": 0.811050534248352,
|
| 994 |
+
"learning_rate": 3.0061084268410006e-06,
|
| 995 |
+
"loss": 1.1097,
|
| 996 |
+
"step": 705
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.5386333184457346,
|
| 1000 |
+
"grad_norm": 0.4883212447166443,
|
| 1001 |
+
"learning_rate": 2.786936710558821e-06,
|
| 1002 |
+
"loss": 1.1183,
|
| 1003 |
+
"step": 710
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.5564984368021437,
|
| 1007 |
+
"grad_norm": 0.6575244069099426,
|
| 1008 |
+
"learning_rate": 2.575588196301146e-06,
|
| 1009 |
+
"loss": 1.0509,
|
| 1010 |
+
"step": 715
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.574363555158553,
|
| 1014 |
+
"grad_norm": 0.4288465082645416,
|
| 1015 |
+
"learning_rate": 2.372137318741968e-06,
|
| 1016 |
+
"loss": 1.0662,
|
| 1017 |
+
"step": 720
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.592228673514962,
|
| 1021 |
+
"grad_norm": 0.628170907497406,
|
| 1022 |
+
"learning_rate": 2.1766557310925916e-06,
|
| 1023 |
+
"loss": 1.0352,
|
| 1024 |
+
"step": 725
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.610093791871371,
|
| 1028 |
+
"grad_norm": 0.4450254440307617,
|
| 1029 |
+
"learning_rate": 1.989212279866079e-06,
|
| 1030 |
+
"loss": 1.0813,
|
| 1031 |
+
"step": 730
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.6279589102277803,
|
| 1035 |
+
"grad_norm": 0.5844019651412964,
|
| 1036 |
+
"learning_rate": 1.8098729806303116e-06,
|
| 1037 |
+
"loss": 1.1297,
|
| 1038 |
+
"step": 735
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.6458240285841894,
|
| 1042 |
+
"grad_norm": 0.6221369504928589,
|
| 1043 |
+
"learning_rate": 1.638700994757955e-06,
|
| 1044 |
+
"loss": 1.0704,
|
| 1045 |
+
"step": 740
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.6636891469405986,
|
| 1049 |
+
"grad_norm": 0.49307432770729065,
|
| 1050 |
+
"learning_rate": 1.4757566071817607e-06,
|
| 1051 |
+
"loss": 1.1033,
|
| 1052 |
+
"step": 745
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.6815542652970077,
|
| 1056 |
+
"grad_norm": 0.6794072389602661,
|
| 1057 |
+
"learning_rate": 1.3210972051628328e-06,
|
| 1058 |
+
"loss": 1.0211,
|
| 1059 |
+
"step": 750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.699419383653417,
|
| 1063 |
+
"grad_norm": 0.4681510329246521,
|
| 1064 |
+
"learning_rate": 1.1747772580794348e-06,
|
| 1065 |
+
"loss": 1.0676,
|
| 1066 |
+
"step": 755
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.717284502009826,
|
| 1070 |
+
"grad_norm": 0.4018392264842987,
|
| 1071 |
+
"learning_rate": 1.0368482982435063e-06,
|
| 1072 |
+
"loss": 1.0683,
|
| 1073 |
+
"step": 760
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.7351496203662347,
|
| 1077 |
+
"grad_norm": 0.6274233460426331,
|
| 1078 |
+
"learning_rate": 9.073589027514789e-07,
|
| 1079 |
+
"loss": 1.0712,
|
| 1080 |
+
"step": 765
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.7530147387226442,
|
| 1084 |
+
"grad_norm": 0.4873206615447998,
|
| 1085 |
+
"learning_rate": 7.863546763760055e-07,
|
| 1086 |
+
"loss": 1.0447,
|
| 1087 |
+
"step": 770
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.770879857079053,
|
| 1091 |
+
"grad_norm": 0.6785792708396912,
|
| 1092 |
+
"learning_rate": 6.738782355044049e-07,
|
| 1093 |
+
"loss": 1.0881,
|
| 1094 |
+
"step": 775
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.7887449754354625,
|
| 1098 |
+
"grad_norm": 0.6314323544502258,
|
| 1099 |
+
"learning_rate": 5.699691931296463e-07,
|
| 1100 |
+
"loss": 1.0517,
|
| 1101 |
+
"step": 780
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.806610093791871,
|
| 1105 |
+
"grad_norm": 0.5061508417129517,
|
| 1106 |
+
"learning_rate": 4.746641448990785e-07,
|
| 1107 |
+
"loss": 1.0891,
|
| 1108 |
+
"step": 785
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 2.8244752121482803,
|
| 1112 |
+
"grad_norm": 0.38295778632164,
|
| 1113 |
+
"learning_rate": 3.879966562258364e-07,
|
| 1114 |
+
"loss": 1.1203,
|
| 1115 |
+
"step": 790
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 2.8423403305046895,
|
| 1119 |
+
"grad_norm": 0.7023996710777283,
|
| 1120 |
+
"learning_rate": 3.0999725046745866e-07,
|
| 1121 |
+
"loss": 1.0624,
|
| 1122 |
+
"step": 795
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.8602054488610986,
|
| 1126 |
+
"grad_norm": 0.41944465041160583,
|
| 1127 |
+
"learning_rate": 2.406933981758952e-07,
|
| 1128 |
+
"loss": 1.0512,
|
| 1129 |
+
"step": 800
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 2.8780705672175078,
|
| 1133 |
+
"grad_norm": 0.6278116703033447,
|
| 1134 |
+
"learning_rate": 1.801095074226683e-07,
|
| 1135 |
+
"loss": 1.0506,
|
| 1136 |
+
"step": 805
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.895935685573917,
|
| 1140 |
+
"grad_norm": 0.523618757724762,
|
| 1141 |
+
"learning_rate": 1.2826691520262114e-07,
|
| 1142 |
+
"loss": 1.084,
|
| 1143 |
+
"step": 810
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 2.913800803930326,
|
| 1147 |
+
"grad_norm": 0.552522599697113,
|
| 1148 |
+
"learning_rate": 8.518387991924837e-08,
|
| 1149 |
+
"loss": 1.1039,
|
| 1150 |
+
"step": 815
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 2.931665922286735,
|
| 1154 |
+
"grad_norm": 0.5657618045806885,
|
| 1155 |
+
"learning_rate": 5.0875574954287186e-08,
|
| 1156 |
+
"loss": 1.0787,
|
| 1157 |
+
"step": 820
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 2.9495310406431443,
|
| 1161 |
+
"grad_norm": 0.4286334216594696,
|
| 1162 |
+
"learning_rate": 2.535408332381417e-08,
|
| 1163 |
+
"loss": 1.0904,
|
| 1164 |
+
"step": 825
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.9673961589995534,
|
| 1168 |
+
"grad_norm": 0.5537897348403931,
|
| 1169 |
+
"learning_rate": 8.62839342274102e-09,
|
| 1170 |
+
"loss": 1.0673,
|
| 1171 |
+
"step": 830
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 2.9852612773559626,
|
| 1175 |
+
"grad_norm": 0.4285455644130707,
|
| 1176 |
+
"learning_rate": 7.043958591912425e-10,
|
| 1177 |
+
"loss": 1.0993,
|
| 1178 |
+
"step": 835
|
| 1179 |
+
}
|
| 1180 |
+
],
|
| 1181 |
+
"logging_steps": 5,
|
| 1182 |
+
"max_steps": 837,
|
| 1183 |
+
"num_input_tokens_seen": 0,
|
| 1184 |
+
"num_train_epochs": 3,
|
| 1185 |
+
"save_steps": 100,
|
| 1186 |
+
"stateful_callbacks": {
|
| 1187 |
+
"TrainerControl": {
|
| 1188 |
+
"args": {
|
| 1189 |
+
"should_epoch_stop": false,
|
| 1190 |
+
"should_evaluate": false,
|
| 1191 |
+
"should_log": false,
|
| 1192 |
+
"should_save": true,
|
| 1193 |
+
"should_training_stop": true
|
| 1194 |
+
},
|
| 1195 |
+
"attributes": {}
|
| 1196 |
+
}
|
| 1197 |
+
},
|
| 1198 |
+
"total_flos": 3.422989809751163e+17,
|
| 1199 |
+
"train_batch_size": 1,
|
| 1200 |
+
"trial_name": null,
|
| 1201 |
+
"trial_params": null
|
| 1202 |
+
}
|