Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1057 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1057 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.931506849315067,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1455,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.136986301369863,
|
| 13 |
+
"grad_norm": 7.088261127471924,
|
| 14 |
+
"learning_rate": 2.7397260273972603e-05,
|
| 15 |
+
"loss": 1.4774,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.273972602739726,
|
| 20 |
+
"grad_norm": 3.0658137798309326,
|
| 21 |
+
"learning_rate": 5.479452054794521e-05,
|
| 22 |
+
"loss": 0.5535,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.410958904109589,
|
| 27 |
+
"grad_norm": 1.415347695350647,
|
| 28 |
+
"learning_rate": 8.219178082191781e-05,
|
| 29 |
+
"loss": 0.2769,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.547945205479452,
|
| 34 |
+
"grad_norm": 1.2462260723114014,
|
| 35 |
+
"learning_rate": 0.00010958904109589041,
|
| 36 |
+
"loss": 0.2071,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.684931506849315,
|
| 41 |
+
"grad_norm": 1.1219278573989868,
|
| 42 |
+
"learning_rate": 0.000136986301369863,
|
| 43 |
+
"loss": 0.179,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.821917808219178,
|
| 48 |
+
"grad_norm": 2.4104228019714355,
|
| 49 |
+
"learning_rate": 0.00016438356164383562,
|
| 50 |
+
"loss": 0.1587,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.958904109589041,
|
| 55 |
+
"grad_norm": 1.2239787578582764,
|
| 56 |
+
"learning_rate": 0.0001917808219178082,
|
| 57 |
+
"loss": 0.1366,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.095890410958904,
|
| 62 |
+
"grad_norm": 0.9942715167999268,
|
| 63 |
+
"learning_rate": 0.00019998733979961563,
|
| 64 |
+
"loss": 0.1218,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.2328767123287672,
|
| 69 |
+
"grad_norm": 0.6293880939483643,
|
| 70 |
+
"learning_rate": 0.0001999253383717226,
|
| 71 |
+
"loss": 0.1168,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.36986301369863,
|
| 76 |
+
"grad_norm": 0.7170248031616211,
|
| 77 |
+
"learning_rate": 0.00019981170237143067,
|
| 78 |
+
"loss": 0.1052,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.5068493150684932,
|
| 83 |
+
"grad_norm": 0.7464343905448914,
|
| 84 |
+
"learning_rate": 0.00019964649051804355,
|
| 85 |
+
"loss": 0.1066,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.643835616438356,
|
| 90 |
+
"grad_norm": 0.6828764081001282,
|
| 91 |
+
"learning_rate": 0.000199429788181734,
|
| 92 |
+
"loss": 0.1057,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.7808219178082192,
|
| 97 |
+
"grad_norm": 0.6028720736503601,
|
| 98 |
+
"learning_rate": 0.0001991617073394306,
|
| 99 |
+
"loss": 0.0843,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.9178082191780823,
|
| 104 |
+
"grad_norm": 0.5440357327461243,
|
| 105 |
+
"learning_rate": 0.00019884238651695556,
|
| 106 |
+
"loss": 0.0948,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.0547945205479454,
|
| 111 |
+
"grad_norm": 0.8612964749336243,
|
| 112 |
+
"learning_rate": 0.00019847199071744415,
|
| 113 |
+
"loss": 0.085,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.191780821917808,
|
| 118 |
+
"grad_norm": 0.889124870300293,
|
| 119 |
+
"learning_rate": 0.00019805071133608242,
|
| 120 |
+
"loss": 0.0962,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.328767123287671,
|
| 125 |
+
"grad_norm": 0.45466411113739014,
|
| 126 |
+
"learning_rate": 0.0001975787660612072,
|
| 127 |
+
"loss": 0.0763,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.4657534246575343,
|
| 132 |
+
"grad_norm": 0.42088282108306885,
|
| 133 |
+
"learning_rate": 0.00019705639876181969,
|
| 134 |
+
"loss": 0.0635,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.602739726027397,
|
| 139 |
+
"grad_norm": 0.5170985460281372,
|
| 140 |
+
"learning_rate": 0.00019648387936157068,
|
| 141 |
+
"loss": 0.0726,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.73972602739726,
|
| 146 |
+
"grad_norm": 0.4313249886035919,
|
| 147 |
+
"learning_rate": 0.00019586150369928245,
|
| 148 |
+
"loss": 0.0669,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.8767123287671232,
|
| 153 |
+
"grad_norm": 0.3355115056037903,
|
| 154 |
+
"learning_rate": 0.00019518959337607957,
|
| 155 |
+
"loss": 0.0682,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.0136986301369864,
|
| 160 |
+
"grad_norm": 0.34427109360694885,
|
| 161 |
+
"learning_rate": 0.0001944684955892075,
|
| 162 |
+
"loss": 0.0638,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 3.1506849315068495,
|
| 167 |
+
"grad_norm": 0.2929873466491699,
|
| 168 |
+
"learning_rate": 0.0001936985829526247,
|
| 169 |
+
"loss": 0.0632,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.287671232876712,
|
| 174 |
+
"grad_norm": 0.3884938657283783,
|
| 175 |
+
"learning_rate": 0.00019288025330446126,
|
| 176 |
+
"loss": 0.0655,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.4246575342465753,
|
| 181 |
+
"grad_norm": 0.27399152517318726,
|
| 182 |
+
"learning_rate": 0.00019201392950144363,
|
| 183 |
+
"loss": 0.0533,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.5616438356164384,
|
| 188 |
+
"grad_norm": 0.2924444079399109,
|
| 189 |
+
"learning_rate": 0.0001911000592003909,
|
| 190 |
+
"loss": 0.0589,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.6986301369863015,
|
| 195 |
+
"grad_norm": 0.43013861775398254,
|
| 196 |
+
"learning_rate": 0.00019013911462689668,
|
| 197 |
+
"loss": 0.0615,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.8356164383561646,
|
| 202 |
+
"grad_norm": 0.5247001647949219,
|
| 203 |
+
"learning_rate": 0.000189131592331315,
|
| 204 |
+
"loss": 0.0583,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.9726027397260273,
|
| 209 |
+
"grad_norm": 0.5796880722045898,
|
| 210 |
+
"learning_rate": 0.00018807801293217735,
|
| 211 |
+
"loss": 0.0556,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.109589041095891,
|
| 216 |
+
"grad_norm": 0.5179729461669922,
|
| 217 |
+
"learning_rate": 0.00018697892084717238,
|
| 218 |
+
"loss": 0.056,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 4.2465753424657535,
|
| 223 |
+
"grad_norm": 0.42960262298583984,
|
| 224 |
+
"learning_rate": 0.00018583488401182843,
|
| 225 |
+
"loss": 0.0637,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 4.383561643835616,
|
| 230 |
+
"grad_norm": 0.3196163773536682,
|
| 231 |
+
"learning_rate": 0.0001846464935860431,
|
| 232 |
+
"loss": 0.0518,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.52054794520548,
|
| 237 |
+
"grad_norm": 0.4424096643924713,
|
| 238 |
+
"learning_rate": 0.0001834143636486124,
|
| 239 |
+
"loss": 0.0524,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.657534246575342,
|
| 244 |
+
"grad_norm": 0.50010746717453,
|
| 245 |
+
"learning_rate": 0.00018213913087991685,
|
| 246 |
+
"loss": 0.0629,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 4.794520547945205,
|
| 251 |
+
"grad_norm": 0.4036540389060974,
|
| 252 |
+
"learning_rate": 0.00018082145423292868,
|
| 253 |
+
"loss": 0.0531,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.931506849315069,
|
| 258 |
+
"grad_norm": 0.36036092042922974,
|
| 259 |
+
"learning_rate": 0.0001794620145927101,
|
| 260 |
+
"loss": 0.0556,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 5.068493150684931,
|
| 265 |
+
"grad_norm": 0.22472509741783142,
|
| 266 |
+
"learning_rate": 0.00017806151442457827,
|
| 267 |
+
"loss": 0.0446,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 5.205479452054795,
|
| 272 |
+
"grad_norm": 0.3514921963214874,
|
| 273 |
+
"learning_rate": 0.00017662067741111974,
|
| 274 |
+
"loss": 0.0443,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 5.342465753424658,
|
| 279 |
+
"grad_norm": 0.2920095920562744,
|
| 280 |
+
"learning_rate": 0.00017514024807824055,
|
| 281 |
+
"loss": 0.0451,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 5.47945205479452,
|
| 286 |
+
"grad_norm": 0.21051590144634247,
|
| 287 |
+
"learning_rate": 0.00017362099141044626,
|
| 288 |
+
"loss": 0.0476,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 5.616438356164384,
|
| 293 |
+
"grad_norm": 0.36196619272232056,
|
| 294 |
+
"learning_rate": 0.00017206369245555036,
|
| 295 |
+
"loss": 0.0521,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.7534246575342465,
|
| 300 |
+
"grad_norm": 0.3503723442554474,
|
| 301 |
+
"learning_rate": 0.0001704691559190155,
|
| 302 |
+
"loss": 0.0472,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.890410958904109,
|
| 307 |
+
"grad_norm": 0.3881896734237671,
|
| 308 |
+
"learning_rate": 0.0001688382057481364,
|
| 309 |
+
"loss": 0.0537,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 6.027397260273973,
|
| 314 |
+
"grad_norm": 0.29409492015838623,
|
| 315 |
+
"learning_rate": 0.00016717168470628077,
|
| 316 |
+
"loss": 0.0436,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 6.164383561643835,
|
| 321 |
+
"grad_norm": 0.2455558031797409,
|
| 322 |
+
"learning_rate": 0.0001654704539374066,
|
| 323 |
+
"loss": 0.0429,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 6.301369863013699,
|
| 328 |
+
"grad_norm": 0.30749672651290894,
|
| 329 |
+
"learning_rate": 0.00016373539252108202,
|
| 330 |
+
"loss": 0.042,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 6.438356164383562,
|
| 335 |
+
"grad_norm": 0.4117829501628876,
|
| 336 |
+
"learning_rate": 0.00016196739701823716,
|
| 337 |
+
"loss": 0.0422,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 6.575342465753424,
|
| 342 |
+
"grad_norm": 0.3047957718372345,
|
| 343 |
+
"learning_rate": 0.00016016738100788297,
|
| 344 |
+
"loss": 0.0456,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.712328767123288,
|
| 349 |
+
"grad_norm": 0.3104310631752014,
|
| 350 |
+
"learning_rate": 0.00015833627461503595,
|
| 351 |
+
"loss": 0.0405,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.8493150684931505,
|
| 356 |
+
"grad_norm": 0.3713166415691376,
|
| 357 |
+
"learning_rate": 0.0001564750240300934,
|
| 358 |
+
"loss": 0.0451,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.986301369863014,
|
| 363 |
+
"grad_norm": 0.23804673552513123,
|
| 364 |
+
"learning_rate": 0.00015458459101990693,
|
| 365 |
+
"loss": 0.0387,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.123287671232877,
|
| 370 |
+
"grad_norm": 0.4476951062679291,
|
| 371 |
+
"learning_rate": 0.00015266595243080714,
|
| 372 |
+
"loss": 0.0406,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.260273972602739,
|
| 377 |
+
"grad_norm": 0.27973777055740356,
|
| 378 |
+
"learning_rate": 0.00015072009968383656,
|
| 379 |
+
"loss": 0.0464,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 7.397260273972603,
|
| 384 |
+
"grad_norm": 0.3597777783870697,
|
| 385 |
+
"learning_rate": 0.00014874803826245089,
|
| 386 |
+
"loss": 0.0459,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 7.534246575342466,
|
| 391 |
+
"grad_norm": 0.27027377486228943,
|
| 392 |
+
"learning_rate": 0.00014675078719295415,
|
| 393 |
+
"loss": 0.0375,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 7.671232876712329,
|
| 398 |
+
"grad_norm": 0.27681443095207214,
|
| 399 |
+
"learning_rate": 0.00014472937851793557,
|
| 400 |
+
"loss": 0.0421,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 7.808219178082192,
|
| 405 |
+
"grad_norm": 0.3312411904335022,
|
| 406 |
+
"learning_rate": 0.00014268485676298078,
|
| 407 |
+
"loss": 0.048,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.945205479452055,
|
| 412 |
+
"grad_norm": 0.2358381599187851,
|
| 413 |
+
"learning_rate": 0.0001406182783969324,
|
| 414 |
+
"loss": 0.0409,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 8.082191780821917,
|
| 419 |
+
"grad_norm": 0.19072838127613068,
|
| 420 |
+
"learning_rate": 0.00013853071128597924,
|
| 421 |
+
"loss": 0.0417,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 8.219178082191782,
|
| 426 |
+
"grad_norm": 0.3328644931316376,
|
| 427 |
+
"learning_rate": 0.0001364232341418564,
|
| 428 |
+
"loss": 0.0397,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 8.356164383561644,
|
| 433 |
+
"grad_norm": 0.27157458662986755,
|
| 434 |
+
"learning_rate": 0.00013429693596444067,
|
| 435 |
+
"loss": 0.0395,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 8.493150684931507,
|
| 440 |
+
"grad_norm": 0.2969032824039459,
|
| 441 |
+
"learning_rate": 0.00013215291547903006,
|
| 442 |
+
"loss": 0.0406,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 8.63013698630137,
|
| 447 |
+
"grad_norm": 0.2864357829093933,
|
| 448 |
+
"learning_rate": 0.00012999228056859784,
|
| 449 |
+
"loss": 0.0424,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 8.767123287671232,
|
| 454 |
+
"grad_norm": 0.25885725021362305,
|
| 455 |
+
"learning_rate": 0.00012781614770131442,
|
| 456 |
+
"loss": 0.0392,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 8.904109589041095,
|
| 461 |
+
"grad_norm": 0.2456735372543335,
|
| 462 |
+
"learning_rate": 0.00012562564135363313,
|
| 463 |
+
"loss": 0.0415,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 9.04109589041096,
|
| 468 |
+
"grad_norm": 0.41431066393852234,
|
| 469 |
+
"learning_rate": 0.0001234218934292376,
|
| 470 |
+
"loss": 0.0407,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 9.178082191780822,
|
| 475 |
+
"grad_norm": 0.260213702917099,
|
| 476 |
+
"learning_rate": 0.00012120604267415172,
|
| 477 |
+
"loss": 0.0393,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 9.315068493150685,
|
| 482 |
+
"grad_norm": 0.3395901322364807,
|
| 483 |
+
"learning_rate": 0.00011897923408831346,
|
| 484 |
+
"loss": 0.035,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 9.452054794520548,
|
| 489 |
+
"grad_norm": 0.3405311405658722,
|
| 490 |
+
"learning_rate": 0.0001167426183339174,
|
| 491 |
+
"loss": 0.0342,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 9.58904109589041,
|
| 496 |
+
"grad_norm": 0.20802819728851318,
|
| 497 |
+
"learning_rate": 0.00011449735114083127,
|
| 498 |
+
"loss": 0.0347,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 9.726027397260275,
|
| 503 |
+
"grad_norm": 0.5094506144523621,
|
| 504 |
+
"learning_rate": 0.00011224459270939384,
|
| 505 |
+
"loss": 0.0373,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 9.863013698630137,
|
| 510 |
+
"grad_norm": 0.21799403429031372,
|
| 511 |
+
"learning_rate": 0.000109985507110903,
|
| 512 |
+
"loss": 0.0392,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 10.0,
|
| 517 |
+
"grad_norm": 0.28433603048324585,
|
| 518 |
+
"learning_rate": 0.00010772126168610325,
|
| 519 |
+
"loss": 0.0373,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 10.136986301369863,
|
| 524 |
+
"grad_norm": 0.3425813913345337,
|
| 525 |
+
"learning_rate": 0.00010545302644198405,
|
| 526 |
+
"loss": 0.0385,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 10.273972602739725,
|
| 531 |
+
"grad_norm": 0.2662697434425354,
|
| 532 |
+
"learning_rate": 0.00010318197344720018,
|
| 533 |
+
"loss": 0.0347,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 10.41095890410959,
|
| 538 |
+
"grad_norm": 0.2841816842556,
|
| 539 |
+
"learning_rate": 0.0001009092762264271,
|
| 540 |
+
"loss": 0.04,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 10.547945205479452,
|
| 545 |
+
"grad_norm": 0.2933363914489746,
|
| 546 |
+
"learning_rate": 9.863610915396365e-05,
|
| 547 |
+
"loss": 0.0363,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 10.684931506849315,
|
| 552 |
+
"grad_norm": 0.20692330598831177,
|
| 553 |
+
"learning_rate": 9.63636468468959e-05,
|
| 554 |
+
"loss": 0.0361,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 10.821917808219178,
|
| 559 |
+
"grad_norm": 0.24741721153259277,
|
| 560 |
+
"learning_rate": 9.409306355813529e-05,
|
| 561 |
+
"loss": 0.0341,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 10.95890410958904,
|
| 566 |
+
"grad_norm": 0.1948077529668808,
|
| 567 |
+
"learning_rate": 9.18255325696454e-05,
|
| 568 |
+
"loss": 0.0349,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 11.095890410958905,
|
| 573 |
+
"grad_norm": 0.16165360808372498,
|
| 574 |
+
"learning_rate": 8.956222558616998e-05,
|
| 575 |
+
"loss": 0.0318,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 11.232876712328768,
|
| 580 |
+
"grad_norm": 0.25702184438705444,
|
| 581 |
+
"learning_rate": 8.730431212977625e-05,
|
| 582 |
+
"loss": 0.0281,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 11.36986301369863,
|
| 587 |
+
"grad_norm": 0.27587395906448364,
|
| 588 |
+
"learning_rate": 8.505295893552594e-05,
|
| 589 |
+
"loss": 0.0349,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 11.506849315068493,
|
| 594 |
+
"grad_norm": 0.3140430152416229,
|
| 595 |
+
"learning_rate": 8.280932934858652e-05,
|
| 596 |
+
"loss": 0.0305,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 11.643835616438356,
|
| 601 |
+
"grad_norm": 0.21165433526039124,
|
| 602 |
+
"learning_rate": 8.05745827230941e-05,
|
| 603 |
+
"loss": 0.0314,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 11.780821917808218,
|
| 608 |
+
"grad_norm": 0.20445489883422852,
|
| 609 |
+
"learning_rate": 7.834987382307861e-05,
|
| 610 |
+
"loss": 0.0319,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 11.917808219178083,
|
| 615 |
+
"grad_norm": 0.27832481265068054,
|
| 616 |
+
"learning_rate": 7.613635222576072e-05,
|
| 617 |
+
"loss": 0.0334,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 12.054794520547945,
|
| 622 |
+
"grad_norm": 0.25728923082351685,
|
| 623 |
+
"learning_rate": 7.393516172752919e-05,
|
| 624 |
+
"loss": 0.033,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 12.191780821917808,
|
| 629 |
+
"grad_norm": 0.2254086136817932,
|
| 630 |
+
"learning_rate": 7.174743975290513e-05,
|
| 631 |
+
"loss": 0.0346,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 12.32876712328767,
|
| 636 |
+
"grad_norm": 0.31018713116645813,
|
| 637 |
+
"learning_rate": 6.957431676679896e-05,
|
| 638 |
+
"loss": 0.0329,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 12.465753424657533,
|
| 643 |
+
"grad_norm": 0.32662343978881836,
|
| 644 |
+
"learning_rate": 6.741691569036338e-05,
|
| 645 |
+
"loss": 0.0342,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 12.602739726027398,
|
| 650 |
+
"grad_norm": 0.2533169984817505,
|
| 651 |
+
"learning_rate": 6.527635132074493e-05,
|
| 652 |
+
"loss": 0.0264,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 12.73972602739726,
|
| 657 |
+
"grad_norm": 0.27445635199546814,
|
| 658 |
+
"learning_rate": 6.315372975503285e-05,
|
| 659 |
+
"loss": 0.0281,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 12.876712328767123,
|
| 664 |
+
"grad_norm": 0.21471256017684937,
|
| 665 |
+
"learning_rate": 6.1050147818704e-05,
|
| 666 |
+
"loss": 0.0321,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 13.013698630136986,
|
| 671 |
+
"grad_norm": 0.19105984270572662,
|
| 672 |
+
"learning_rate": 5.896669249885851e-05,
|
| 673 |
+
"loss": 0.0273,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 13.150684931506849,
|
| 678 |
+
"grad_norm": 0.3308360278606415,
|
| 679 |
+
"learning_rate": 5.690444038253935e-05,
|
| 680 |
+
"loss": 0.0343,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 13.287671232876713,
|
| 685 |
+
"grad_norm": 0.1988590806722641,
|
| 686 |
+
"learning_rate": 5.4864457100425783e-05,
|
| 687 |
+
"loss": 0.028,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 13.424657534246576,
|
| 692 |
+
"grad_norm": 0.1858794391155243,
|
| 693 |
+
"learning_rate": 5.284779677618841e-05,
|
| 694 |
+
"loss": 0.0273,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 13.561643835616438,
|
| 699 |
+
"grad_norm": 0.29671627283096313,
|
| 700 |
+
"learning_rate": 5.0855501481790305e-05,
|
| 701 |
+
"loss": 0.0271,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 13.698630136986301,
|
| 706 |
+
"grad_norm": 0.17693527042865753,
|
| 707 |
+
"learning_rate": 4.8888600699015496e-05,
|
| 708 |
+
"loss": 0.034,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 13.835616438356164,
|
| 713 |
+
"grad_norm": 0.31038013100624084,
|
| 714 |
+
"learning_rate": 4.694811078750338e-05,
|
| 715 |
+
"loss": 0.0251,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 13.972602739726028,
|
| 720 |
+
"grad_norm": 0.3317829668521881,
|
| 721 |
+
"learning_rate": 4.50350344595635e-05,
|
| 722 |
+
"loss": 0.0334,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 14.10958904109589,
|
| 727 |
+
"grad_norm": 0.1818408966064453,
|
| 728 |
+
"learning_rate": 4.315036026204262e-05,
|
| 729 |
+
"loss": 0.0272,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 14.246575342465754,
|
| 734 |
+
"grad_norm": 0.2105715572834015,
|
| 735 |
+
"learning_rate": 4.129506206551138e-05,
|
| 736 |
+
"loss": 0.025,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 14.383561643835616,
|
| 741 |
+
"grad_norm": 0.18613150715827942,
|
| 742 |
+
"learning_rate": 3.947009856103465e-05,
|
| 743 |
+
"loss": 0.0238,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 14.520547945205479,
|
| 748 |
+
"grad_norm": 0.2959461212158203,
|
| 749 |
+
"learning_rate": 3.767641276478563e-05,
|
| 750 |
+
"loss": 0.0249,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 14.657534246575342,
|
| 755 |
+
"grad_norm": 0.18495745956897736,
|
| 756 |
+
"learning_rate": 3.591493153075966e-05,
|
| 757 |
+
"loss": 0.0214,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 14.794520547945206,
|
| 762 |
+
"grad_norm": 0.1501263529062271,
|
| 763 |
+
"learning_rate": 3.41865650718396e-05,
|
| 764 |
+
"loss": 0.0266,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 14.931506849315069,
|
| 769 |
+
"grad_norm": 0.3387095332145691,
|
| 770 |
+
"learning_rate": 3.24922064894601e-05,
|
| 771 |
+
"loss": 0.0268,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 15.068493150684931,
|
| 776 |
+
"grad_norm": 0.23434942960739136,
|
| 777 |
+
"learning_rate": 3.083273131211382e-05,
|
| 778 |
+
"loss": 0.0272,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 15.205479452054794,
|
| 783 |
+
"grad_norm": 0.163187175989151,
|
| 784 |
+
"learning_rate": 2.920899704293849e-05,
|
| 785 |
+
"loss": 0.0232,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 15.342465753424657,
|
| 790 |
+
"grad_norm": 0.20000265538692474,
|
| 791 |
+
"learning_rate": 2.762184271661785e-05,
|
| 792 |
+
"loss": 0.0261,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 15.479452054794521,
|
| 797 |
+
"grad_norm": 0.18943333625793457,
|
| 798 |
+
"learning_rate": 2.6072088465826038e-05,
|
| 799 |
+
"loss": 0.0246,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 15.616438356164384,
|
| 804 |
+
"grad_norm": 0.2833252251148224,
|
| 805 |
+
"learning_rate": 2.4560535097439108e-05,
|
| 806 |
+
"loss": 0.0253,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 15.753424657534246,
|
| 811 |
+
"grad_norm": 0.1302843540906906,
|
| 812 |
+
"learning_rate": 2.308796367873296e-05,
|
| 813 |
+
"loss": 0.0246,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 15.89041095890411,
|
| 818 |
+
"grad_norm": 0.16615238785743713,
|
| 819 |
+
"learning_rate": 2.165513513378121e-05,
|
| 820 |
+
"loss": 0.0254,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 16.027397260273972,
|
| 825 |
+
"grad_norm": 0.17113815248012543,
|
| 826 |
+
"learning_rate": 2.0262789850261798e-05,
|
| 827 |
+
"loss": 0.0288,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 16.164383561643834,
|
| 832 |
+
"grad_norm": 0.21394069492816925,
|
| 833 |
+
"learning_rate": 1.8911647296875147e-05,
|
| 834 |
+
"loss": 0.025,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 16.301369863013697,
|
| 839 |
+
"grad_norm": 0.2763649523258209,
|
| 840 |
+
"learning_rate": 1.7602405651572275e-05,
|
| 841 |
+
"loss": 0.0219,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 16.438356164383563,
|
| 846 |
+
"grad_norm": 0.13925646245479584,
|
| 847 |
+
"learning_rate": 1.6335741440784035e-05,
|
| 848 |
+
"loss": 0.0217,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 16.575342465753426,
|
| 853 |
+
"grad_norm": 0.20826192200183868,
|
| 854 |
+
"learning_rate": 1.511230918983867e-05,
|
| 855 |
+
"loss": 0.023,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 16.71232876712329,
|
| 860 |
+
"grad_norm": 0.2256271094083786,
|
| 861 |
+
"learning_rate": 1.3932741084747913e-05,
|
| 862 |
+
"loss": 0.023,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 16.84931506849315,
|
| 867 |
+
"grad_norm": 0.27016547322273254,
|
| 868 |
+
"learning_rate": 1.2797646645536566e-05,
|
| 869 |
+
"loss": 0.0211,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 16.986301369863014,
|
| 874 |
+
"grad_norm": 0.26627489924430847,
|
| 875 |
+
"learning_rate": 1.1707612411284253e-05,
|
| 876 |
+
"loss": 0.0235,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 17.123287671232877,
|
| 881 |
+
"grad_norm": 0.18498767912387848,
|
| 882 |
+
"learning_rate": 1.0663201637042252e-05,
|
| 883 |
+
"loss": 0.022,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 17.26027397260274,
|
| 888 |
+
"grad_norm": 0.23852607607841492,
|
| 889 |
+
"learning_rate": 9.664954002781745e-06,
|
| 890 |
+
"loss": 0.0228,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 17.397260273972602,
|
| 895 |
+
"grad_norm": 0.15411531925201416,
|
| 896 |
+
"learning_rate": 8.713385334524283e-06,
|
| 897 |
+
"loss": 0.0198,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 17.534246575342465,
|
| 902 |
+
"grad_norm": 0.25403866171836853,
|
| 903 |
+
"learning_rate": 7.808987337798158e-06,
|
| 904 |
+
"loss": 0.0257,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 17.671232876712327,
|
| 909 |
+
"grad_norm": 0.14403975009918213,
|
| 910 |
+
"learning_rate": 6.952227343558671e-06,
|
| 911 |
+
"loss": 0.0215,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 17.80821917808219,
|
| 916 |
+
"grad_norm": 0.188527911901474,
|
| 917 |
+
"learning_rate": 6.143548066703475e-06,
|
| 918 |
+
"loss": 0.0224,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 17.945205479452056,
|
| 923 |
+
"grad_norm": 0.1309424489736557,
|
| 924 |
+
"learning_rate": 5.383367377307857e-06,
|
| 925 |
+
"loss": 0.0215,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 18.08219178082192,
|
| 930 |
+
"grad_norm": 0.11233002692461014,
|
| 931 |
+
"learning_rate": 4.672078084698095e-06,
|
| 932 |
+
"loss": 0.0211,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 18.21917808219178,
|
| 937 |
+
"grad_norm": 0.22869743406772614,
|
| 938 |
+
"learning_rate": 4.010047734474454e-06,
|
| 939 |
+
"loss": 0.0215,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 18.356164383561644,
|
| 944 |
+
"grad_norm": 0.11979719996452332,
|
| 945 |
+
"learning_rate": 3.397618418588877e-06,
|
| 946 |
+
"loss": 0.0273,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 18.493150684931507,
|
| 951 |
+
"grad_norm": 0.2112375795841217,
|
| 952 |
+
"learning_rate": 2.8351065985751766e-06,
|
| 953 |
+
"loss": 0.0228,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 18.63013698630137,
|
| 958 |
+
"grad_norm": 0.14134034514427185,
|
| 959 |
+
"learning_rate": 2.322802942023461e-06,
|
| 960 |
+
"loss": 0.0247,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 18.767123287671232,
|
| 965 |
+
"grad_norm": 0.09884881973266602,
|
| 966 |
+
"learning_rate": 1.8609721723830132e-06,
|
| 967 |
+
"loss": 0.0196,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 18.904109589041095,
|
| 972 |
+
"grad_norm": 0.14044946432113647,
|
| 973 |
+
"learning_rate": 1.4498529321713584e-06,
|
| 974 |
+
"loss": 0.0198,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 19.041095890410958,
|
| 979 |
+
"grad_norm": 0.13853876292705536,
|
| 980 |
+
"learning_rate": 1.0896576596600705e-06,
|
| 981 |
+
"loss": 0.0182,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 19.17808219178082,
|
| 986 |
+
"grad_norm": 0.1654110848903656,
|
| 987 |
+
"learning_rate": 7.80572479101327e-07,
|
| 988 |
+
"loss": 0.0229,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 19.315068493150687,
|
| 993 |
+
"grad_norm": 0.15151838958263397,
|
| 994 |
+
"learning_rate": 5.227571045515633e-07,
|
| 995 |
+
"loss": 0.0202,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 19.45205479452055,
|
| 1000 |
+
"grad_norm": 0.2258201688528061,
|
| 1001 |
+
"learning_rate": 3.163447573422351e-07,
|
| 1002 |
+
"loss": 0.0197,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 19.589041095890412,
|
| 1007 |
+
"grad_norm": 0.24640779197216034,
|
| 1008 |
+
"learning_rate": 1.614420972401165e-07,
|
| 1009 |
+
"loss": 0.0187,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 19.726027397260275,
|
| 1014 |
+
"grad_norm": 0.21181590855121613,
|
| 1015 |
+
"learning_rate": 5.812916733284324e-08,
|
| 1016 |
+
"loss": 0.0198,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 19.863013698630137,
|
| 1021 |
+
"grad_norm": 0.14787183701992035,
|
| 1022 |
+
"learning_rate": 6.459352668164442e-09,
|
| 1023 |
+
"loss": 0.0186,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 19.931506849315067,
|
| 1028 |
+
"step": 1455,
|
| 1029 |
+
"total_flos": 1.1504025698630573e+17,
|
| 1030 |
+
"train_loss": 0.05927258820058554,
|
| 1031 |
+
"train_runtime": 1048.7401,
|
| 1032 |
+
"train_samples_per_second": 88.792,
|
| 1033 |
+
"train_steps_per_second": 1.387
|
| 1034 |
+
}
|
| 1035 |
+
],
|
| 1036 |
+
"logging_steps": 10,
|
| 1037 |
+
"max_steps": 1455,
|
| 1038 |
+
"num_input_tokens_seen": 0,
|
| 1039 |
+
"num_train_epochs": 20,
|
| 1040 |
+
"save_steps": 500,
|
| 1041 |
+
"stateful_callbacks": {
|
| 1042 |
+
"TrainerControl": {
|
| 1043 |
+
"args": {
|
| 1044 |
+
"should_epoch_stop": false,
|
| 1045 |
+
"should_evaluate": false,
|
| 1046 |
+
"should_log": false,
|
| 1047 |
+
"should_save": true,
|
| 1048 |
+
"should_training_stop": true
|
| 1049 |
+
},
|
| 1050 |
+
"attributes": {}
|
| 1051 |
+
}
|
| 1052 |
+
},
|
| 1053 |
+
"total_flos": 1.1504025698630573e+17,
|
| 1054 |
+
"train_batch_size": 64,
|
| 1055 |
+
"trial_name": null,
|
| 1056 |
+
"trial_params": null
|
| 1057 |
+
}
|