Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1134 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1134 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.498407643312102,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1569,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01592356687898089,
|
| 13 |
+
"grad_norm": 4.3297224044799805,
|
| 14 |
+
"learning_rate": 2.5316455696202533e-05,
|
| 15 |
+
"loss": 0.86,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03184713375796178,
|
| 20 |
+
"grad_norm": 6.637640476226807,
|
| 21 |
+
"learning_rate": 5.0632911392405066e-05,
|
| 22 |
+
"loss": 0.4332,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04777070063694268,
|
| 27 |
+
"grad_norm": 1.9116368293762207,
|
| 28 |
+
"learning_rate": 7.59493670886076e-05,
|
| 29 |
+
"loss": 0.3014,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06369426751592357,
|
| 34 |
+
"grad_norm": 1.2928426265716553,
|
| 35 |
+
"learning_rate": 0.00010126582278481013,
|
| 36 |
+
"loss": 0.2232,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07961783439490445,
|
| 41 |
+
"grad_norm": 1.4354435205459595,
|
| 42 |
+
"learning_rate": 0.00012658227848101267,
|
| 43 |
+
"loss": 0.1771,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09554140127388536,
|
| 48 |
+
"grad_norm": 1.710893154144287,
|
| 49 |
+
"learning_rate": 0.0001518987341772152,
|
| 50 |
+
"loss": 0.2136,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11146496815286625,
|
| 55 |
+
"grad_norm": 1.6920921802520752,
|
| 56 |
+
"learning_rate": 0.00017721518987341773,
|
| 57 |
+
"loss": 0.1644,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.12738853503184713,
|
| 62 |
+
"grad_norm": 1.2916326522827148,
|
| 63 |
+
"learning_rate": 0.00019999977772170748,
|
| 64 |
+
"loss": 0.1593,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.14331210191082802,
|
| 69 |
+
"grad_norm": 1.599898099899292,
|
| 70 |
+
"learning_rate": 0.00019997310552224846,
|
| 71 |
+
"loss": 0.1578,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.1592356687898089,
|
| 76 |
+
"grad_norm": 1.4364933967590332,
|
| 77 |
+
"learning_rate": 0.00019990199125039174,
|
| 78 |
+
"loss": 0.155,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1751592356687898,
|
| 83 |
+
"grad_norm": 1.7148686647415161,
|
| 84 |
+
"learning_rate": 0.00019978646651929572,
|
| 85 |
+
"loss": 0.1463,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.1910828025477707,
|
| 90 |
+
"grad_norm": 1.2754517793655396,
|
| 91 |
+
"learning_rate": 0.0001996265826843568,
|
| 92 |
+
"loss": 0.1471,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2070063694267516,
|
| 97 |
+
"grad_norm": 1.0604585409164429,
|
| 98 |
+
"learning_rate": 0.00019942241082037982,
|
| 99 |
+
"loss": 0.124,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2229299363057325,
|
| 104 |
+
"grad_norm": 0.6741822361946106,
|
| 105 |
+
"learning_rate": 0.00019917404168998256,
|
| 106 |
+
"loss": 0.0979,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.23885350318471338,
|
| 111 |
+
"grad_norm": 1.2134864330291748,
|
| 112 |
+
"learning_rate": 0.00019888158570324795,
|
| 113 |
+
"loss": 0.1105,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.25477707006369427,
|
| 118 |
+
"grad_norm": 0.6342090368270874,
|
| 119 |
+
"learning_rate": 0.00019854517286864245,
|
| 120 |
+
"loss": 0.099,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.27070063694267515,
|
| 125 |
+
"grad_norm": 1.0112156867980957,
|
| 126 |
+
"learning_rate": 0.000198164952735222,
|
| 127 |
+
"loss": 0.0928,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.28662420382165604,
|
| 132 |
+
"grad_norm": 1.0317986011505127,
|
| 133 |
+
"learning_rate": 0.00019774109432615147,
|
| 134 |
+
"loss": 0.0963,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.30254777070063693,
|
| 139 |
+
"grad_norm": 0.7419966459274292,
|
| 140 |
+
"learning_rate": 0.00019727378606356703,
|
| 141 |
+
"loss": 0.1069,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3184713375796178,
|
| 146 |
+
"grad_norm": 1.373571753501892,
|
| 147 |
+
"learning_rate": 0.00019676323568481498,
|
| 148 |
+
"loss": 0.1103,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3343949044585987,
|
| 153 |
+
"grad_norm": 0.6964902877807617,
|
| 154 |
+
"learning_rate": 0.00019620967015010395,
|
| 155 |
+
"loss": 0.0827,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3503184713375796,
|
| 160 |
+
"grad_norm": 0.9365194439888,
|
| 161 |
+
"learning_rate": 0.00019561333554161224,
|
| 162 |
+
"loss": 0.1046,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3662420382165605,
|
| 167 |
+
"grad_norm": 1.899908423423767,
|
| 168 |
+
"learning_rate": 0.00019497449695409408,
|
| 169 |
+
"loss": 0.1031,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.3821656050955414,
|
| 174 |
+
"grad_norm": 0.8126282095909119,
|
| 175 |
+
"learning_rate": 0.00019429343837703455,
|
| 176 |
+
"loss": 0.1069,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3980891719745223,
|
| 181 |
+
"grad_norm": 1.244973063468933,
|
| 182 |
+
"learning_rate": 0.00019357046256840473,
|
| 183 |
+
"loss": 0.082,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4140127388535032,
|
| 188 |
+
"grad_norm": 0.8132479786872864,
|
| 189 |
+
"learning_rate": 0.00019280589092007352,
|
| 190 |
+
"loss": 0.0867,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.4299363057324841,
|
| 195 |
+
"grad_norm": 0.8742226958274841,
|
| 196 |
+
"learning_rate": 0.0001920000633149362,
|
| 197 |
+
"loss": 0.0761,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.445859872611465,
|
| 202 |
+
"grad_norm": 0.6832584738731384,
|
| 203 |
+
"learning_rate": 0.00019115333797582254,
|
| 204 |
+
"loss": 0.0845,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.46178343949044587,
|
| 209 |
+
"grad_norm": 0.6030533909797668,
|
| 210 |
+
"learning_rate": 0.00019026609130625257,
|
| 211 |
+
"loss": 0.0722,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.47770700636942676,
|
| 216 |
+
"grad_norm": 0.7257299423217773,
|
| 217 |
+
"learning_rate": 0.0001893387177231099,
|
| 218 |
+
"loss": 0.0701,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.49363057324840764,
|
| 223 |
+
"grad_norm": 0.6390131711959839,
|
| 224 |
+
"learning_rate": 0.00018837162948130752,
|
| 225 |
+
"loss": 0.0696,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5095541401273885,
|
| 230 |
+
"grad_norm": 1.8211143016815186,
|
| 231 |
+
"learning_rate": 0.00018736525649052394,
|
| 232 |
+
"loss": 0.0757,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5254777070063694,
|
| 237 |
+
"grad_norm": 1.0733026266098022,
|
| 238 |
+
"learning_rate": 0.00018632004612409103,
|
| 239 |
+
"loss": 0.0835,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5414012738853503,
|
| 244 |
+
"grad_norm": 0.7683069705963135,
|
| 245 |
+
"learning_rate": 0.00018523646302011867,
|
| 246 |
+
"loss": 0.0765,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5573248407643312,
|
| 251 |
+
"grad_norm": 0.5842871069908142,
|
| 252 |
+
"learning_rate": 0.00018411498887494396,
|
| 253 |
+
"loss": 0.073,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5732484076433121,
|
| 258 |
+
"grad_norm": 0.6933338046073914,
|
| 259 |
+
"learning_rate": 0.0001829561222289984,
|
| 260 |
+
"loss": 0.0648,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.589171974522293,
|
| 265 |
+
"grad_norm": 0.42543935775756836,
|
| 266 |
+
"learning_rate": 0.000181760378245186,
|
| 267 |
+
"loss": 0.072,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6050955414012739,
|
| 272 |
+
"grad_norm": 0.72137451171875,
|
| 273 |
+
"learning_rate": 0.0001805282884798732,
|
| 274 |
+
"loss": 0.0661,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6210191082802548,
|
| 279 |
+
"grad_norm": 0.6393111944198608,
|
| 280 |
+
"learning_rate": 0.00017926040064659014,
|
| 281 |
+
"loss": 0.0841,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6369426751592356,
|
| 286 |
+
"grad_norm": 0.46522364020347595,
|
| 287 |
+
"learning_rate": 0.00017795727837255015,
|
| 288 |
+
"loss": 0.0668,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6528662420382165,
|
| 293 |
+
"grad_norm": 0.790262758731842,
|
| 294 |
+
"learning_rate": 0.0001766195009480949,
|
| 295 |
+
"loss": 0.0702,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6687898089171974,
|
| 300 |
+
"grad_norm": 0.7343723177909851,
|
| 301 |
+
"learning_rate": 0.00017524766306917618,
|
| 302 |
+
"loss": 0.0751,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.6847133757961783,
|
| 307 |
+
"grad_norm": 0.497380793094635,
|
| 308 |
+
"learning_rate": 0.00017384237457298987,
|
| 309 |
+
"loss": 0.0663,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7006369426751592,
|
| 314 |
+
"grad_norm": 0.5845641493797302,
|
| 315 |
+
"learning_rate": 0.00017240426016687863,
|
| 316 |
+
"loss": 0.0656,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7165605095541401,
|
| 321 |
+
"grad_norm": 1.0296058654785156,
|
| 322 |
+
"learning_rate": 0.00017093395915062428,
|
| 323 |
+
"loss": 0.0664,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.732484076433121,
|
| 328 |
+
"grad_norm": 0.7704136371612549,
|
| 329 |
+
"learning_rate": 0.00016943212513225345,
|
| 330 |
+
"loss": 0.0597,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7484076433121019,
|
| 335 |
+
"grad_norm": 0.7734206318855286,
|
| 336 |
+
"learning_rate": 0.00016789942573748232,
|
| 337 |
+
"loss": 0.0638,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7643312101910829,
|
| 342 |
+
"grad_norm": 0.6289462447166443,
|
| 343 |
+
"learning_rate": 0.00016633654231293013,
|
| 344 |
+
"loss": 0.0573,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7802547770700637,
|
| 349 |
+
"grad_norm": 0.6616397500038147,
|
| 350 |
+
"learning_rate": 0.00016474416962323325,
|
| 351 |
+
"loss": 0.0506,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7961783439490446,
|
| 356 |
+
"grad_norm": 0.3910081386566162,
|
| 357 |
+
"learning_rate": 0.00016312301554219426,
|
| 358 |
+
"loss": 0.0457,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.8121019108280255,
|
| 363 |
+
"grad_norm": 0.6165742874145508,
|
| 364 |
+
"learning_rate": 0.00016147380073810346,
|
| 365 |
+
"loss": 0.0503,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.8280254777070064,
|
| 370 |
+
"grad_norm": 0.6931295990943909,
|
| 371 |
+
"learning_rate": 0.00015979725835337294,
|
| 372 |
+
"loss": 0.0481,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.8439490445859873,
|
| 377 |
+
"grad_norm": 0.7499469518661499,
|
| 378 |
+
"learning_rate": 0.00015809413367862512,
|
| 379 |
+
"loss": 0.0725,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.8598726114649682,
|
| 384 |
+
"grad_norm": 0.4135420620441437,
|
| 385 |
+
"learning_rate": 0.00015636518382138107,
|
| 386 |
+
"loss": 0.0646,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.8757961783439491,
|
| 391 |
+
"grad_norm": 0.4710856080055237,
|
| 392 |
+
"learning_rate": 0.00015461117736949577,
|
| 393 |
+
"loss": 0.0538,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.89171974522293,
|
| 398 |
+
"grad_norm": 0.428799033164978,
|
| 399 |
+
"learning_rate": 0.00015283289404948976,
|
| 400 |
+
"loss": 0.0593,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.9076433121019108,
|
| 405 |
+
"grad_norm": 0.4142407774925232,
|
| 406 |
+
"learning_rate": 0.0001510311243799295,
|
| 407 |
+
"loss": 0.0605,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.9235668789808917,
|
| 412 |
+
"grad_norm": 0.46757104992866516,
|
| 413 |
+
"learning_rate": 0.0001492066693200096,
|
| 414 |
+
"loss": 0.0634,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.9394904458598726,
|
| 419 |
+
"grad_norm": 0.4138467609882355,
|
| 420 |
+
"learning_rate": 0.0001473603399134948,
|
| 421 |
+
"loss": 0.0586,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.9554140127388535,
|
| 426 |
+
"grad_norm": 0.43409624695777893,
|
| 427 |
+
"learning_rate": 0.00014549295692817778,
|
| 428 |
+
"loss": 0.0579,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.9713375796178344,
|
| 433 |
+
"grad_norm": 0.44268661737442017,
|
| 434 |
+
"learning_rate": 0.0001436053504910151,
|
| 435 |
+
"loss": 0.0637,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.9872611464968153,
|
| 440 |
+
"grad_norm": 0.3840467929840088,
|
| 441 |
+
"learning_rate": 0.00014169835971910238,
|
| 442 |
+
"loss": 0.0511,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.0031847133757963,
|
| 447 |
+
"grad_norm": 0.4400401711463928,
|
| 448 |
+
"learning_rate": 0.00013977283234665273,
|
| 449 |
+
"loss": 0.0477,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.019108280254777,
|
| 454 |
+
"grad_norm": 0.3332836925983429,
|
| 455 |
+
"learning_rate": 0.00013782962434814492,
|
| 456 |
+
"loss": 0.0478,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.035031847133758,
|
| 461 |
+
"grad_norm": 0.3282802700996399,
|
| 462 |
+
"learning_rate": 0.00013586959955780824,
|
| 463 |
+
"loss": 0.0508,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.0509554140127388,
|
| 468 |
+
"grad_norm": 0.6200388073921204,
|
| 469 |
+
"learning_rate": 0.00013389362928561317,
|
| 470 |
+
"loss": 0.0479,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.0668789808917198,
|
| 475 |
+
"grad_norm": 0.5026119947433472,
|
| 476 |
+
"learning_rate": 0.0001319025919299394,
|
| 477 |
+
"loss": 0.0402,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.0828025477707006,
|
| 482 |
+
"grad_norm": 0.440782755613327,
|
| 483 |
+
"learning_rate": 0.00012989737258709203,
|
| 484 |
+
"loss": 0.0508,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.0987261146496816,
|
| 489 |
+
"grad_norm": 0.34029659628868103,
|
| 490 |
+
"learning_rate": 0.0001278788626578407,
|
| 491 |
+
"loss": 0.044,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.1146496815286624,
|
| 496 |
+
"grad_norm": 0.46119582653045654,
|
| 497 |
+
"learning_rate": 0.00012584795945115603,
|
| 498 |
+
"loss": 0.0426,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.1305732484076434,
|
| 503 |
+
"grad_norm": 0.47696712613105774,
|
| 504 |
+
"learning_rate": 0.0001238055657853198,
|
| 505 |
+
"loss": 0.0515,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.1464968152866242,
|
| 510 |
+
"grad_norm": 0.49719110131263733,
|
| 511 |
+
"learning_rate": 0.00012175258958658564,
|
| 512 |
+
"loss": 0.0453,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.1624203821656052,
|
| 517 |
+
"grad_norm": 0.36977869272232056,
|
| 518 |
+
"learning_rate": 0.0001196899434855693,
|
| 519 |
+
"loss": 0.0529,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.178343949044586,
|
| 524 |
+
"grad_norm": 0.3333928883075714,
|
| 525 |
+
"learning_rate": 0.00011761854441154767,
|
| 526 |
+
"loss": 0.0429,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.194267515923567,
|
| 531 |
+
"grad_norm": 0.3594333231449127,
|
| 532 |
+
"learning_rate": 0.0001155393131848467,
|
| 533 |
+
"loss": 0.044,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.2101910828025477,
|
| 538 |
+
"grad_norm": 0.33035799860954285,
|
| 539 |
+
"learning_rate": 0.00011345317410749964,
|
| 540 |
+
"loss": 0.0425,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.2261146496815287,
|
| 545 |
+
"grad_norm": 0.431429386138916,
|
| 546 |
+
"learning_rate": 0.00011136105455235766,
|
| 547 |
+
"loss": 0.0395,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.2420382165605095,
|
| 552 |
+
"grad_norm": 0.34141987562179565,
|
| 553 |
+
"learning_rate": 0.00010926388455083522,
|
| 554 |
+
"loss": 0.0455,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.2579617834394905,
|
| 559 |
+
"grad_norm": 0.48105695843696594,
|
| 560 |
+
"learning_rate": 0.00010716259637947357,
|
| 561 |
+
"loss": 0.033,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.2738853503184713,
|
| 566 |
+
"grad_norm": 0.40072181820869446,
|
| 567 |
+
"learning_rate": 0.0001050581241455064,
|
| 568 |
+
"loss": 0.0404,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.2898089171974523,
|
| 573 |
+
"grad_norm": 0.5137969851493835,
|
| 574 |
+
"learning_rate": 0.00010295140337161146,
|
| 575 |
+
"loss": 0.0388,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.305732484076433,
|
| 580 |
+
"grad_norm": 0.6727393865585327,
|
| 581 |
+
"learning_rate": 0.00010084337058003303,
|
| 582 |
+
"loss": 0.0455,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.321656050955414,
|
| 587 |
+
"grad_norm": 0.36769551038742065,
|
| 588 |
+
"learning_rate": 9.873496287626019e-05,
|
| 589 |
+
"loss": 0.0416,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.3375796178343948,
|
| 594 |
+
"grad_norm": 0.40487590432167053,
|
| 595 |
+
"learning_rate": 9.662711753244551e-05,
|
| 596 |
+
"loss": 0.0394,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.3535031847133758,
|
| 601 |
+
"grad_norm": 0.3262194097042084,
|
| 602 |
+
"learning_rate": 9.452077157074994e-05,
|
| 603 |
+
"loss": 0.0371,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.3694267515923566,
|
| 608 |
+
"grad_norm": 0.49285051226615906,
|
| 609 |
+
"learning_rate": 9.241686134679867e-05,
|
| 610 |
+
"loss": 0.0434,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.3853503184713376,
|
| 615 |
+
"grad_norm": 0.4286457896232605,
|
| 616 |
+
"learning_rate": 9.031632213343339e-05,
|
| 617 |
+
"loss": 0.0388,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.4012738853503186,
|
| 622 |
+
"grad_norm": 0.4375673234462738,
|
| 623 |
+
"learning_rate": 8.822008770494572e-05,
|
| 624 |
+
"loss": 0.0419,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.4171974522292994,
|
| 629 |
+
"grad_norm": 0.34244051575660706,
|
| 630 |
+
"learning_rate": 8.612908992197705e-05,
|
| 631 |
+
"loss": 0.0478,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.4331210191082802,
|
| 636 |
+
"grad_norm": 0.325065016746521,
|
| 637 |
+
"learning_rate": 8.404425831726894e-05,
|
| 638 |
+
"loss": 0.04,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.4490445859872612,
|
| 643 |
+
"grad_norm": 0.4267027676105499,
|
| 644 |
+
"learning_rate": 8.196651968244826e-05,
|
| 645 |
+
"loss": 0.0377,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.4649681528662422,
|
| 650 |
+
"grad_norm": 0.5654813051223755,
|
| 651 |
+
"learning_rate": 7.989679765603108e-05,
|
| 652 |
+
"loss": 0.0445,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.480891719745223,
|
| 657 |
+
"grad_norm": 0.39815276861190796,
|
| 658 |
+
"learning_rate": 7.783601231282812e-05,
|
| 659 |
+
"loss": 0.0553,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.4968152866242037,
|
| 664 |
+
"grad_norm": 0.46311256289482117,
|
| 665 |
+
"learning_rate": 7.578507975493448e-05,
|
| 666 |
+
"loss": 0.0373,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.5127388535031847,
|
| 671 |
+
"grad_norm": 0.37769854068756104,
|
| 672 |
+
"learning_rate": 7.374491170448525e-05,
|
| 673 |
+
"loss": 0.0368,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.5286624203821657,
|
| 678 |
+
"grad_norm": 0.2921994924545288,
|
| 679 |
+
"learning_rate": 7.17164150983584e-05,
|
| 680 |
+
"loss": 0.0428,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.5445859872611465,
|
| 685 |
+
"grad_norm": 0.48072826862335205,
|
| 686 |
+
"learning_rate": 6.970049168500474e-05,
|
| 687 |
+
"loss": 0.0329,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.5605095541401273,
|
| 692 |
+
"grad_norm": 0.37165966629981995,
|
| 693 |
+
"learning_rate": 6.769803762358443e-05,
|
| 694 |
+
"loss": 0.0376,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.5764331210191083,
|
| 699 |
+
"grad_norm": 0.309915155172348,
|
| 700 |
+
"learning_rate": 6.570994308558812e-05,
|
| 701 |
+
"loss": 0.032,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.5923566878980893,
|
| 706 |
+
"grad_norm": 0.41205787658691406,
|
| 707 |
+
"learning_rate": 6.373709185911998e-05,
|
| 708 |
+
"loss": 0.036,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.60828025477707,
|
| 713 |
+
"grad_norm": 0.45491883158683777,
|
| 714 |
+
"learning_rate": 6.17803609560181e-05,
|
| 715 |
+
"loss": 0.0453,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.6242038216560508,
|
| 720 |
+
"grad_norm": 0.4330281615257263,
|
| 721 |
+
"learning_rate": 5.98406202219875e-05,
|
| 722 |
+
"loss": 0.0319,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.6401273885350318,
|
| 727 |
+
"grad_norm": 0.2400377243757248,
|
| 728 |
+
"learning_rate": 5.791873194991872e-05,
|
| 729 |
+
"loss": 0.0342,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.6560509554140128,
|
| 734 |
+
"grad_norm": 0.26502475142478943,
|
| 735 |
+
"learning_rate": 5.601555049656382e-05,
|
| 736 |
+
"loss": 0.0318,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.6719745222929936,
|
| 741 |
+
"grad_norm": 0.2682415246963501,
|
| 742 |
+
"learning_rate": 5.41319219027404e-05,
|
| 743 |
+
"loss": 0.0301,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.6878980891719744,
|
| 748 |
+
"grad_norm": 0.38560014963150024,
|
| 749 |
+
"learning_rate": 5.226868351723244e-05,
|
| 750 |
+
"loss": 0.0432,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.7038216560509554,
|
| 755 |
+
"grad_norm": 0.1889456957578659,
|
| 756 |
+
"learning_rate": 5.042666362455498e-05,
|
| 757 |
+
"loss": 0.0255,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.7197452229299364,
|
| 762 |
+
"grad_norm": 0.3188437223434448,
|
| 763 |
+
"learning_rate": 4.860668107674823e-05,
|
| 764 |
+
"loss": 0.037,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.7356687898089171,
|
| 769 |
+
"grad_norm": 0.7025167942047119,
|
| 770 |
+
"learning_rate": 4.6809544929365004e-05,
|
| 771 |
+
"loss": 0.0357,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.7515923566878981,
|
| 776 |
+
"grad_norm": 0.39571383595466614,
|
| 777 |
+
"learning_rate": 4.503605408181286e-05,
|
| 778 |
+
"loss": 0.0273,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.767515923566879,
|
| 783 |
+
"grad_norm": 0.2839336693286896,
|
| 784 |
+
"learning_rate": 4.3286996922211034e-05,
|
| 785 |
+
"loss": 0.0302,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.78343949044586,
|
| 790 |
+
"grad_norm": 0.3293486535549164,
|
| 791 |
+
"learning_rate": 4.156315097692037e-05,
|
| 792 |
+
"loss": 0.0308,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.799363057324841,
|
| 797 |
+
"grad_norm": 0.3942393958568573,
|
| 798 |
+
"learning_rate": 3.986528256490141e-05,
|
| 799 |
+
"loss": 0.0293,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.8152866242038217,
|
| 804 |
+
"grad_norm": 0.31392380595207214,
|
| 805 |
+
"learning_rate": 3.8194146457054655e-05,
|
| 806 |
+
"loss": 0.0249,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.8312101910828025,
|
| 811 |
+
"grad_norm": 0.3941175639629364,
|
| 812 |
+
"learning_rate": 3.655048554069478e-05,
|
| 813 |
+
"loss": 0.032,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.8471337579617835,
|
| 818 |
+
"grad_norm": 0.2796083092689514,
|
| 819 |
+
"learning_rate": 3.4935030489306883e-05,
|
| 820 |
+
"loss": 0.034,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.8630573248407645,
|
| 825 |
+
"grad_norm": 0.30333149433135986,
|
| 826 |
+
"learning_rate": 3.334849943773323e-05,
|
| 827 |
+
"loss": 0.0351,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.8789808917197452,
|
| 832 |
+
"grad_norm": 0.3254244029521942,
|
| 833 |
+
"learning_rate": 3.179159766293282e-05,
|
| 834 |
+
"loss": 0.0298,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.894904458598726,
|
| 839 |
+
"grad_norm": 0.20831990242004395,
|
| 840 |
+
"learning_rate": 3.0265017270457775e-05,
|
| 841 |
+
"loss": 0.0288,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.910828025477707,
|
| 846 |
+
"grad_norm": 0.31671157479286194,
|
| 847 |
+
"learning_rate": 2.8769436886784408e-05,
|
| 848 |
+
"loss": 0.0334,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.926751592356688,
|
| 853 |
+
"grad_norm": 0.5833495259284973,
|
| 854 |
+
"learning_rate": 2.730552135763632e-05,
|
| 855 |
+
"loss": 0.0425,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.9426751592356688,
|
| 860 |
+
"grad_norm": 0.32988494634628296,
|
| 861 |
+
"learning_rate": 2.5873921452433915e-05,
|
| 862 |
+
"loss": 0.0328,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.9585987261146496,
|
| 867 |
+
"grad_norm": 0.20500314235687256,
|
| 868 |
+
"learning_rate": 2.4475273575000936e-05,
|
| 869 |
+
"loss": 0.0273,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.9745222929936306,
|
| 874 |
+
"grad_norm": 0.33419910073280334,
|
| 875 |
+
"learning_rate": 2.3110199480657525e-05,
|
| 876 |
+
"loss": 0.0264,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.9904458598726116,
|
| 881 |
+
"grad_norm": 0.20512598752975464,
|
| 882 |
+
"learning_rate": 2.1779305999824884e-05,
|
| 883 |
+
"loss": 0.0287,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.0063694267515926,
|
| 888 |
+
"grad_norm": 0.4389275312423706,
|
| 889 |
+
"learning_rate": 2.0483184768264596e-05,
|
| 890 |
+
"loss": 0.0306,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.022292993630573,
|
| 895 |
+
"grad_norm": 0.18635383248329163,
|
| 896 |
+
"learning_rate": 1.9222411964072884e-05,
|
| 897 |
+
"loss": 0.028,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.038216560509554,
|
| 902 |
+
"grad_norm": 0.32153043150901794,
|
| 903 |
+
"learning_rate": 1.799754805154603e-05,
|
| 904 |
+
"loss": 0.0274,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.054140127388535,
|
| 909 |
+
"grad_norm": 0.448691189289093,
|
| 910 |
+
"learning_rate": 1.6809137532031704e-05,
|
| 911 |
+
"loss": 0.0306,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.070063694267516,
|
| 916 |
+
"grad_norm": 0.38386714458465576,
|
| 917 |
+
"learning_rate": 1.565770870187585e-05,
|
| 918 |
+
"loss": 0.0247,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.0859872611464967,
|
| 923 |
+
"grad_norm": 0.18705415725708008,
|
| 924 |
+
"learning_rate": 1.4543773417573925e-05,
|
| 925 |
+
"loss": 0.025,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.1019108280254777,
|
| 930 |
+
"grad_norm": 0.23585732281208038,
|
| 931 |
+
"learning_rate": 1.3467826868229994e-05,
|
| 932 |
+
"loss": 0.0265,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.1178343949044587,
|
| 937 |
+
"grad_norm": 0.36259782314300537,
|
| 938 |
+
"learning_rate": 1.243034735542512e-05,
|
| 939 |
+
"loss": 0.0305,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.1337579617834397,
|
| 944 |
+
"grad_norm": 0.34997355937957764,
|
| 945 |
+
"learning_rate": 1.1431796080593283e-05,
|
| 946 |
+
"loss": 0.032,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 2.1496815286624202,
|
| 951 |
+
"grad_norm": 0.3579126298427582,
|
| 952 |
+
"learning_rate": 1.0472616939998492e-05,
|
| 953 |
+
"loss": 0.024,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.1656050955414012,
|
| 958 |
+
"grad_norm": 0.41504374146461487,
|
| 959 |
+
"learning_rate": 9.553236327405246e-06,
|
| 960 |
+
"loss": 0.0259,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 2.1815286624203822,
|
| 965 |
+
"grad_norm": 0.2632668912410736,
|
| 966 |
+
"learning_rate": 8.674062944529216e-06,
|
| 967 |
+
"loss": 0.0301,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 2.1974522292993632,
|
| 972 |
+
"grad_norm": 0.3617275059223175,
|
| 973 |
+
"learning_rate": 7.835487619352811e-06,
|
| 974 |
+
"loss": 0.0232,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 2.213375796178344,
|
| 979 |
+
"grad_norm": 0.3129204213619232,
|
| 980 |
+
"learning_rate": 7.037883132386547e-06,
|
| 981 |
+
"loss": 0.0355,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 2.229299363057325,
|
| 986 |
+
"grad_norm": 0.2824781835079193,
|
| 987 |
+
"learning_rate": 6.2816040509530165e-06,
|
| 988 |
+
"loss": 0.0269,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 2.245222929936306,
|
| 993 |
+
"grad_norm": 0.20426948368549347,
|
| 994 |
+
"learning_rate": 5.566986571567401e-06,
|
| 995 |
+
"loss": 0.0355,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.261146496815287,
|
| 1000 |
+
"grad_norm": 0.22026273608207703,
|
| 1001 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 1002 |
+
"loss": 0.0267,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 2.2770700636942673,
|
| 1007 |
+
"grad_norm": 0.27001041173934937,
|
| 1008 |
+
"learning_rate": 4.263988462479484e-06,
|
| 1009 |
+
"loss": 0.0286,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 2.2929936305732483,
|
| 1014 |
+
"grad_norm": 0.22612400352954865,
|
| 1015 |
+
"learning_rate": 3.676187067922421e-06,
|
| 1016 |
+
"loss": 0.0263,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 2.3089171974522293,
|
| 1021 |
+
"grad_norm": 0.3205125033855438,
|
| 1022 |
+
"learning_rate": 3.131205488210409e-06,
|
| 1023 |
+
"loss": 0.0311,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 2.3248407643312103,
|
| 1028 |
+
"grad_norm": 0.2743224501609802,
|
| 1029 |
+
"learning_rate": 2.6292859896079213e-06,
|
| 1030 |
+
"loss": 0.0233,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.340764331210191,
|
| 1035 |
+
"grad_norm": 0.23394609987735748,
|
| 1036 |
+
"learning_rate": 2.170651695549786e-06,
|
| 1037 |
+
"loss": 0.0161,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.356687898089172,
|
| 1042 |
+
"grad_norm": 0.23162613809108734,
|
| 1043 |
+
"learning_rate": 1.7555064874538397e-06,
|
| 1044 |
+
"loss": 0.0243,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 2.372611464968153,
|
| 1049 |
+
"grad_norm": 0.3181140720844269,
|
| 1050 |
+
"learning_rate": 1.3840349140874619e-06,
|
| 1051 |
+
"loss": 0.0224,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 2.388535031847134,
|
| 1056 |
+
"grad_norm": 0.29145529866218567,
|
| 1057 |
+
"learning_rate": 1.0564021095281652e-06,
|
| 1058 |
+
"loss": 0.0345,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 2.404458598726115,
|
| 1063 |
+
"grad_norm": 0.321485698223114,
|
| 1064 |
+
"learning_rate": 7.727537197548707e-07,
|
| 1065 |
+
"loss": 0.0205,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 2.4203821656050954,
|
| 1070 |
+
"grad_norm": 0.2429160177707672,
|
| 1071 |
+
"learning_rate": 5.332158379024122e-07,
|
| 1072 |
+
"loss": 0.0217,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 2.4363057324840764,
|
| 1077 |
+
"grad_norm": 0.22818560898303986,
|
| 1078 |
+
"learning_rate": 3.3789494820803957e-07,
|
| 1079 |
+
"loss": 0.0219,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.4522292993630574,
|
| 1084 |
+
"grad_norm": 0.2582818865776062,
|
| 1085 |
+
"learning_rate": 1.8687787867489592e-07,
|
| 1086 |
+
"loss": 0.0186,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 2.468152866242038,
|
| 1091 |
+
"grad_norm": 0.2430548369884491,
|
| 1092 |
+
"learning_rate": 8.023176247348163e-08,
|
| 1093 |
+
"loss": 0.0344,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 2.484076433121019,
|
| 1098 |
+
"grad_norm": 0.38517120480537415,
|
| 1099 |
+
"learning_rate": 1.8004008098226887e-08,
|
| 1100 |
+
"loss": 0.0225,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 2.498407643312102,
|
| 1105 |
+
"step": 1569,
|
| 1106 |
+
"total_flos": 5.687859742190496e+16,
|
| 1107 |
+
"train_loss": 0.06483107545027846,
|
| 1108 |
+
"train_runtime": 706.8123,
|
| 1109 |
+
"train_samples_per_second": 35.517,
|
| 1110 |
+
"train_steps_per_second": 2.22
|
| 1111 |
+
}
|
| 1112 |
+
],
|
| 1113 |
+
"logging_steps": 10,
|
| 1114 |
+
"max_steps": 1569,
|
| 1115 |
+
"num_input_tokens_seen": 0,
|
| 1116 |
+
"num_train_epochs": 3,
|
| 1117 |
+
"save_steps": 500,
|
| 1118 |
+
"stateful_callbacks": {
|
| 1119 |
+
"TrainerControl": {
|
| 1120 |
+
"args": {
|
| 1121 |
+
"should_epoch_stop": false,
|
| 1122 |
+
"should_evaluate": false,
|
| 1123 |
+
"should_log": false,
|
| 1124 |
+
"should_save": true,
|
| 1125 |
+
"should_training_stop": true
|
| 1126 |
+
},
|
| 1127 |
+
"attributes": {}
|
| 1128 |
+
}
|
| 1129 |
+
},
|
| 1130 |
+
"total_flos": 5.687859742190496e+16,
|
| 1131 |
+
"train_batch_size": 16,
|
| 1132 |
+
"trial_name": null,
|
| 1133 |
+
"trial_params": null
|
| 1134 |
+
}
|