Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1008 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1008 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.985611510791367,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1386,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03597122302158273,
|
| 13 |
+
"grad_norm": 3.388711929321289,
|
| 14 |
+
"learning_rate": 2.857142857142857e-05,
|
| 15 |
+
"loss": 1.2192,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07194244604316546,
|
| 20 |
+
"grad_norm": 7.04016637802124,
|
| 21 |
+
"learning_rate": 5.714285714285714e-05,
|
| 22 |
+
"loss": 0.4355,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.1079136690647482,
|
| 27 |
+
"grad_norm": 1.9987560510635376,
|
| 28 |
+
"learning_rate": 8.571428571428571e-05,
|
| 29 |
+
"loss": 0.2583,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.14388489208633093,
|
| 34 |
+
"grad_norm": 1.4927663803100586,
|
| 35 |
+
"learning_rate": 0.00011428571428571428,
|
| 36 |
+
"loss": 0.1959,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.17985611510791366,
|
| 41 |
+
"grad_norm": 2.1035594940185547,
|
| 42 |
+
"learning_rate": 0.00014285714285714287,
|
| 43 |
+
"loss": 0.158,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.2158273381294964,
|
| 48 |
+
"grad_norm": 1.8321762084960938,
|
| 49 |
+
"learning_rate": 0.00017142857142857143,
|
| 50 |
+
"loss": 0.1593,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2517985611510791,
|
| 55 |
+
"grad_norm": 0.9496012330055237,
|
| 56 |
+
"learning_rate": 0.0002,
|
| 57 |
+
"loss": 0.1172,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.28776978417266186,
|
| 62 |
+
"grad_norm": 1.5062251091003418,
|
| 63 |
+
"learning_rate": 0.0001999715070545774,
|
| 64 |
+
"loss": 0.1394,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3237410071942446,
|
| 69 |
+
"grad_norm": 1.1900173425674438,
|
| 70 |
+
"learning_rate": 0.00019988604445526827,
|
| 71 |
+
"loss": 0.1144,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3597122302158273,
|
| 76 |
+
"grad_norm": 1.6312843561172485,
|
| 77 |
+
"learning_rate": 0.0001997436609036963,
|
| 78 |
+
"loss": 0.0954,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.39568345323741005,
|
| 83 |
+
"grad_norm": 0.7715893387794495,
|
| 84 |
+
"learning_rate": 0.00019954443753839667,
|
| 85 |
+
"loss": 0.1043,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.4316546762589928,
|
| 90 |
+
"grad_norm": 0.9805018901824951,
|
| 91 |
+
"learning_rate": 0.00019928848788857887,
|
| 92 |
+
"loss": 0.107,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4676258992805755,
|
| 97 |
+
"grad_norm": 0.79938805103302,
|
| 98 |
+
"learning_rate": 0.00019897595780943102,
|
| 99 |
+
"loss": 0.0939,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5035971223021583,
|
| 104 |
+
"grad_norm": 0.7441235184669495,
|
| 105 |
+
"learning_rate": 0.00019860702539900287,
|
| 106 |
+
"loss": 0.0876,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.539568345323741,
|
| 111 |
+
"grad_norm": 1.4985952377319336,
|
| 112 |
+
"learning_rate": 0.00019818190089671508,
|
| 113 |
+
"loss": 0.0998,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.5755395683453237,
|
| 118 |
+
"grad_norm": 0.49447914958000183,
|
| 119 |
+
"learning_rate": 0.0001977008265635525,
|
| 120 |
+
"loss": 0.084,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6115107913669064,
|
| 125 |
+
"grad_norm": 0.9330535531044006,
|
| 126 |
+
"learning_rate": 0.00019716407654400952,
|
| 127 |
+
"loss": 0.0711,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.6474820143884892,
|
| 132 |
+
"grad_norm": 0.71977299451828,
|
| 133 |
+
"learning_rate": 0.00019657195670986637,
|
| 134 |
+
"loss": 0.0649,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6834532374100719,
|
| 139 |
+
"grad_norm": 0.6885129809379578,
|
| 140 |
+
"learning_rate": 0.00019592480448588542,
|
| 141 |
+
"loss": 0.0734,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7194244604316546,
|
| 146 |
+
"grad_norm": 0.8220723271369934,
|
| 147 |
+
"learning_rate": 0.0001952229886575266,
|
| 148 |
+
"loss": 0.0692,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.7553956834532374,
|
| 153 |
+
"grad_norm": 0.4193010926246643,
|
| 154 |
+
"learning_rate": 0.0001944669091607919,
|
| 155 |
+
"loss": 0.0634,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7913669064748201,
|
| 160 |
+
"grad_norm": 0.5557839274406433,
|
| 161 |
+
"learning_rate": 0.0001936569968543179,
|
| 162 |
+
"loss": 0.0617,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.8273381294964028,
|
| 167 |
+
"grad_norm": 0.6012473702430725,
|
| 168 |
+
"learning_rate": 0.0001927937132738476,
|
| 169 |
+
"loss": 0.0567,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.8633093525179856,
|
| 174 |
+
"grad_norm": 0.5065937042236328,
|
| 175 |
+
"learning_rate": 0.00019187755036921978,
|
| 176 |
+
"loss": 0.0666,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.8992805755395683,
|
| 181 |
+
"grad_norm": 0.8122988343238831,
|
| 182 |
+
"learning_rate": 0.00019090903022402729,
|
| 183 |
+
"loss": 0.0621,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.935251798561151,
|
| 188 |
+
"grad_norm": 0.600527822971344,
|
| 189 |
+
"learning_rate": 0.00018988870475810282,
|
| 190 |
+
"loss": 0.0577,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.9712230215827338,
|
| 195 |
+
"grad_norm": 0.41501474380493164,
|
| 196 |
+
"learning_rate": 0.00018881715541300276,
|
| 197 |
+
"loss": 0.0585,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.0071942446043165,
|
| 202 |
+
"grad_norm": 0.7007383108139038,
|
| 203 |
+
"learning_rate": 0.00018769499282066717,
|
| 204 |
+
"loss": 0.0727,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.0431654676258992,
|
| 209 |
+
"grad_norm": 0.4356389045715332,
|
| 210 |
+
"learning_rate": 0.00018652285645544603,
|
| 211 |
+
"loss": 0.0576,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.079136690647482,
|
| 216 |
+
"grad_norm": 0.5948470234870911,
|
| 217 |
+
"learning_rate": 0.00018530141426968902,
|
| 218 |
+
"loss": 0.0641,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.1151079136690647,
|
| 223 |
+
"grad_norm": 0.6867853403091431,
|
| 224 |
+
"learning_rate": 0.00018403136231310684,
|
| 225 |
+
"loss": 0.0535,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.1510791366906474,
|
| 230 |
+
"grad_norm": 0.5092623829841614,
|
| 231 |
+
"learning_rate": 0.00018271342433612113,
|
| 232 |
+
"loss": 0.05,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.1870503597122302,
|
| 237 |
+
"grad_norm": 0.6697306632995605,
|
| 238 |
+
"learning_rate": 0.0001813483513774289,
|
| 239 |
+
"loss": 0.0527,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.223021582733813,
|
| 244 |
+
"grad_norm": 0.5235440731048584,
|
| 245 |
+
"learning_rate": 0.0001799369213360163,
|
| 246 |
+
"loss": 0.0477,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.2589928057553956,
|
| 251 |
+
"grad_norm": 0.40158113837242126,
|
| 252 |
+
"learning_rate": 0.0001784799385278661,
|
| 253 |
+
"loss": 0.0587,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.2949640287769784,
|
| 258 |
+
"grad_norm": 0.36186593770980835,
|
| 259 |
+
"learning_rate": 0.000176978233227611,
|
| 260 |
+
"loss": 0.0523,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.330935251798561,
|
| 265 |
+
"grad_norm": 0.4480399489402771,
|
| 266 |
+
"learning_rate": 0.00017543266119539422,
|
| 267 |
+
"loss": 0.0413,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.3669064748201438,
|
| 272 |
+
"grad_norm": 0.5076740980148315,
|
| 273 |
+
"learning_rate": 0.00017384410318920697,
|
| 274 |
+
"loss": 0.0465,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.4028776978417266,
|
| 279 |
+
"grad_norm": 0.8270596265792847,
|
| 280 |
+
"learning_rate": 0.0001722134644629807,
|
| 281 |
+
"loss": 0.056,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.4388489208633093,
|
| 286 |
+
"grad_norm": 0.520577609539032,
|
| 287 |
+
"learning_rate": 0.00017054167425071995,
|
| 288 |
+
"loss": 0.0401,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.474820143884892,
|
| 293 |
+
"grad_norm": 0.7387260794639587,
|
| 294 |
+
"learning_rate": 0.00016882968523697028,
|
| 295 |
+
"loss": 0.0444,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.5107913669064748,
|
| 300 |
+
"grad_norm": 0.34361258149147034,
|
| 301 |
+
"learning_rate": 0.00016707847301392236,
|
| 302 |
+
"loss": 0.0573,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.5467625899280577,
|
| 307 |
+
"grad_norm": 0.49156174063682556,
|
| 308 |
+
"learning_rate": 0.00016528903552546207,
|
| 309 |
+
"loss": 0.0499,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.5827338129496402,
|
| 314 |
+
"grad_norm": 0.37369754910469055,
|
| 315 |
+
"learning_rate": 0.0001634623924984833,
|
| 316 |
+
"loss": 0.047,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.6187050359712232,
|
| 321 |
+
"grad_norm": 0.4199368357658386,
|
| 322 |
+
"learning_rate": 0.0001615995848617876,
|
| 323 |
+
"loss": 0.0422,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.6546762589928057,
|
| 328 |
+
"grad_norm": 0.40282875299453735,
|
| 329 |
+
"learning_rate": 0.0001597016741529014,
|
| 330 |
+
"loss": 0.0388,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.6906474820143886,
|
| 335 |
+
"grad_norm": 0.7056338787078857,
|
| 336 |
+
"learning_rate": 0.0001577697419131497,
|
| 337 |
+
"loss": 0.0453,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.7266187050359711,
|
| 342 |
+
"grad_norm": 0.512008786201477,
|
| 343 |
+
"learning_rate": 0.00015580488907132974,
|
| 344 |
+
"loss": 0.0419,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.762589928057554,
|
| 349 |
+
"grad_norm": 0.41075655817985535,
|
| 350 |
+
"learning_rate": 0.00015380823531633729,
|
| 351 |
+
"loss": 0.0465,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.7985611510791366,
|
| 356 |
+
"grad_norm": 0.48174160718917847,
|
| 357 |
+
"learning_rate": 0.0001517809184591017,
|
| 358 |
+
"loss": 0.0386,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.8345323741007196,
|
| 363 |
+
"grad_norm": 0.5647067427635193,
|
| 364 |
+
"learning_rate": 0.0001497240937841944,
|
| 365 |
+
"loss": 0.0425,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.870503597122302,
|
| 370 |
+
"grad_norm": 0.4203431308269501,
|
| 371 |
+
"learning_rate": 0.0001476389333914794,
|
| 372 |
+
"loss": 0.0377,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.906474820143885,
|
| 377 |
+
"grad_norm": 0.5694783329963684,
|
| 378 |
+
"learning_rate": 0.0001455266255281821,
|
| 379 |
+
"loss": 0.0374,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.9424460431654675,
|
| 384 |
+
"grad_norm": 0.6108015179634094,
|
| 385 |
+
"learning_rate": 0.00014338837391175582,
|
| 386 |
+
"loss": 0.0417,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.9784172661870505,
|
| 391 |
+
"grad_norm": 0.43549835681915283,
|
| 392 |
+
"learning_rate": 0.00014122539704393265,
|
| 393 |
+
"loss": 0.0339,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.014388489208633,
|
| 398 |
+
"grad_norm": 0.4553022086620331,
|
| 399 |
+
"learning_rate": 0.00013903892751634947,
|
| 400 |
+
"loss": 0.0423,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.050359712230216,
|
| 405 |
+
"grad_norm": 0.48167282342910767,
|
| 406 |
+
"learning_rate": 0.0001368302113081447,
|
| 407 |
+
"loss": 0.0411,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.0863309352517985,
|
| 412 |
+
"grad_norm": 0.5101397037506104,
|
| 413 |
+
"learning_rate": 0.0001346005070759258,
|
| 414 |
+
"loss": 0.0495,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.1223021582733814,
|
| 419 |
+
"grad_norm": 0.47834643721580505,
|
| 420 |
+
"learning_rate": 0.00013235108543651272,
|
| 421 |
+
"loss": 0.0352,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.158273381294964,
|
| 426 |
+
"grad_norm": 0.4055221676826477,
|
| 427 |
+
"learning_rate": 0.00013008322824286555,
|
| 428 |
+
"loss": 0.0321,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.194244604316547,
|
| 433 |
+
"grad_norm": 0.4231523275375366,
|
| 434 |
+
"learning_rate": 0.00012779822785360912,
|
| 435 |
+
"loss": 0.0467,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.2302158273381294,
|
| 440 |
+
"grad_norm": 0.6262776851654053,
|
| 441 |
+
"learning_rate": 0.00012549738639657115,
|
| 442 |
+
"loss": 0.0363,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.2661870503597124,
|
| 447 |
+
"grad_norm": 0.5341355800628662,
|
| 448 |
+
"learning_rate": 0.00012318201502675285,
|
| 449 |
+
"loss": 0.0328,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.302158273381295,
|
| 454 |
+
"grad_norm": 0.2906142473220825,
|
| 455 |
+
"learning_rate": 0.00012085343317915565,
|
| 456 |
+
"loss": 0.0316,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.338129496402878,
|
| 461 |
+
"grad_norm": 0.3438158929347992,
|
| 462 |
+
"learning_rate": 0.00011851296781688952,
|
| 463 |
+
"loss": 0.0315,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.3741007194244603,
|
| 468 |
+
"grad_norm": 0.43991851806640625,
|
| 469 |
+
"learning_rate": 0.00011616195267499102,
|
| 470 |
+
"loss": 0.031,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.4100719424460433,
|
| 475 |
+
"grad_norm": 0.3710590600967407,
|
| 476 |
+
"learning_rate": 0.00011380172750038269,
|
| 477 |
+
"loss": 0.0252,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.446043165467626,
|
| 482 |
+
"grad_norm": 0.4207635223865509,
|
| 483 |
+
"learning_rate": 0.00011143363728840625,
|
| 484 |
+
"loss": 0.0409,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.4820143884892087,
|
| 489 |
+
"grad_norm": 0.48165836930274963,
|
| 490 |
+
"learning_rate": 0.00010905903151636501,
|
| 491 |
+
"loss": 0.0282,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.5179856115107913,
|
| 496 |
+
"grad_norm": 0.45046257972717285,
|
| 497 |
+
"learning_rate": 0.00010667926337451217,
|
| 498 |
+
"loss": 0.0257,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.553956834532374,
|
| 503 |
+
"grad_norm": 0.4864603877067566,
|
| 504 |
+
"learning_rate": 0.00010429568899492348,
|
| 505 |
+
"loss": 0.0322,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.5899280575539567,
|
| 510 |
+
"grad_norm": 0.3659379184246063,
|
| 511 |
+
"learning_rate": 0.0001019096666786931,
|
| 512 |
+
"loss": 0.0271,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.6258992805755397,
|
| 517 |
+
"grad_norm": 0.2604960799217224,
|
| 518 |
+
"learning_rate": 9.952255612189368e-05,
|
| 519 |
+
"loss": 0.0326,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.661870503597122,
|
| 524 |
+
"grad_norm": 0.4740797281265259,
|
| 525 |
+
"learning_rate": 9.713571764074152e-05,
|
| 526 |
+
"loss": 0.0306,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.697841726618705,
|
| 531 |
+
"grad_norm": 0.2611696124076843,
|
| 532 |
+
"learning_rate": 9.475051139640809e-05,
|
| 533 |
+
"loss": 0.0271,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.7338129496402876,
|
| 538 |
+
"grad_norm": 0.3587149679660797,
|
| 539 |
+
"learning_rate": 9.236829661992023e-05,
|
| 540 |
+
"loss": 0.0251,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.7697841726618706,
|
| 545 |
+
"grad_norm": 0.4013247489929199,
|
| 546 |
+
"learning_rate": 8.999043083759017e-05,
|
| 547 |
+
"loss": 0.0256,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.805755395683453,
|
| 552 |
+
"grad_norm": 0.40374016761779785,
|
| 553 |
+
"learning_rate": 8.761826909741709e-05,
|
| 554 |
+
"loss": 0.032,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.841726618705036,
|
| 559 |
+
"grad_norm": 0.2673892080783844,
|
| 560 |
+
"learning_rate": 8.525316319690092e-05,
|
| 561 |
+
"loss": 0.0201,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.8776978417266186,
|
| 566 |
+
"grad_norm": 0.44146105647087097,
|
| 567 |
+
"learning_rate": 8.289646091270849e-05,
|
| 568 |
+
"loss": 0.0305,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.9136690647482015,
|
| 573 |
+
"grad_norm": 0.33300504088401794,
|
| 574 |
+
"learning_rate": 8.054950523263096e-05,
|
| 575 |
+
"loss": 0.0219,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.949640287769784,
|
| 580 |
+
"grad_norm": 0.26247546076774597,
|
| 581 |
+
"learning_rate": 7.821363359027048e-05,
|
| 582 |
+
"loss": 0.0359,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.985611510791367,
|
| 587 |
+
"grad_norm": 0.5542314052581787,
|
| 588 |
+
"learning_rate": 7.589017710289139e-05,
|
| 589 |
+
"loss": 0.0286,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.0215827338129495,
|
| 594 |
+
"grad_norm": 0.4509626030921936,
|
| 595 |
+
"learning_rate": 7.358045981287141e-05,
|
| 596 |
+
"loss": 0.0234,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.0575539568345325,
|
| 601 |
+
"grad_norm": 0.2706989049911499,
|
| 602 |
+
"learning_rate": 7.128579793318428e-05,
|
| 603 |
+
"loss": 0.0225,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.093525179856115,
|
| 608 |
+
"grad_norm": 0.21301843225955963,
|
| 609 |
+
"learning_rate": 6.900749909734406e-05,
|
| 610 |
+
"loss": 0.0274,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.129496402877698,
|
| 615 |
+
"grad_norm": 0.32899874448776245,
|
| 616 |
+
"learning_rate": 6.674686161423843e-05,
|
| 617 |
+
"loss": 0.0234,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.1654676258992804,
|
| 622 |
+
"grad_norm": 0.4368799924850464,
|
| 623 |
+
"learning_rate": 6.450517372827591e-05,
|
| 624 |
+
"loss": 0.0199,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.2014388489208634,
|
| 629 |
+
"grad_norm": 0.36518239974975586,
|
| 630 |
+
"learning_rate": 6.22837128852683e-05,
|
| 631 |
+
"loss": 0.0218,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.237410071942446,
|
| 636 |
+
"grad_norm": 0.20987823605537415,
|
| 637 |
+
"learning_rate": 6.008374500446676e-05,
|
| 638 |
+
"loss": 0.0217,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.273381294964029,
|
| 643 |
+
"grad_norm": 0.23875588178634644,
|
| 644 |
+
"learning_rate": 5.790652375716652e-05,
|
| 645 |
+
"loss": 0.0237,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.3093525179856114,
|
| 650 |
+
"grad_norm": 0.320413738489151,
|
| 651 |
+
"learning_rate": 5.575328985229098e-05,
|
| 652 |
+
"loss": 0.0279,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.3453237410071943,
|
| 657 |
+
"grad_norm": 0.20259268581867218,
|
| 658 |
+
"learning_rate": 5.362527032936277e-05,
|
| 659 |
+
"loss": 0.0265,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.381294964028777,
|
| 664 |
+
"grad_norm": 0.22162136435508728,
|
| 665 |
+
"learning_rate": 5.1523677859264516e-05,
|
| 666 |
+
"loss": 0.0256,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.41726618705036,
|
| 671 |
+
"grad_norm": 0.15575292706489563,
|
| 672 |
+
"learning_rate": 4.944971005318716e-05,
|
| 673 |
+
"loss": 0.0194,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.4532374100719423,
|
| 678 |
+
"grad_norm": 0.29202401638031006,
|
| 679 |
+
"learning_rate": 4.740454878016084e-05,
|
| 680 |
+
"loss": 0.0276,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.4892086330935252,
|
| 685 |
+
"grad_norm": 0.40271732211112976,
|
| 686 |
+
"learning_rate": 4.538935949355623e-05,
|
| 687 |
+
"loss": 0.0183,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.5251798561151078,
|
| 692 |
+
"grad_norm": 0.22273583710193634,
|
| 693 |
+
"learning_rate": 4.340529056694047e-05,
|
| 694 |
+
"loss": 0.02,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.5611510791366907,
|
| 699 |
+
"grad_norm": 0.1683393120765686,
|
| 700 |
+
"learning_rate": 4.1453472639666457e-05,
|
| 701 |
+
"loss": 0.0177,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.597122302158273,
|
| 706 |
+
"grad_norm": 0.4574311375617981,
|
| 707 |
+
"learning_rate": 3.9535017972567675e-05,
|
| 708 |
+
"loss": 0.0205,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.633093525179856,
|
| 713 |
+
"grad_norm": 0.31037506461143494,
|
| 714 |
+
"learning_rate": 3.7651019814126654e-05,
|
| 715 |
+
"loss": 0.0265,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.6690647482014387,
|
| 720 |
+
"grad_norm": 0.349869966506958,
|
| 721 |
+
"learning_rate": 3.5802551777477476e-05,
|
| 722 |
+
"loss": 0.0153,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.7050359712230216,
|
| 727 |
+
"grad_norm": 0.2690473794937134,
|
| 728 |
+
"learning_rate": 3.399066722859782e-05,
|
| 729 |
+
"loss": 0.0205,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.741007194244604,
|
| 734 |
+
"grad_norm": 0.23748761415481567,
|
| 735 |
+
"learning_rate": 3.2216398686038926e-05,
|
| 736 |
+
"loss": 0.0145,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.776978417266187,
|
| 741 |
+
"grad_norm": 0.24303750693798065,
|
| 742 |
+
"learning_rate": 3.0480757232535772e-05,
|
| 743 |
+
"loss": 0.0207,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.81294964028777,
|
| 748 |
+
"grad_norm": 0.1993425041437149,
|
| 749 |
+
"learning_rate": 2.8784731938832556e-05,
|
| 750 |
+
"loss": 0.0237,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.8489208633093526,
|
| 755 |
+
"grad_norm": 0.2566690444946289,
|
| 756 |
+
"learning_rate": 2.7129289300051787e-05,
|
| 757 |
+
"loss": 0.0194,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.884892086330935,
|
| 762 |
+
"grad_norm": 0.2915847599506378,
|
| 763 |
+
"learning_rate": 2.5515372684928683e-05,
|
| 764 |
+
"loss": 0.0204,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.920863309352518,
|
| 769 |
+
"grad_norm": 0.21342326700687408,
|
| 770 |
+
"learning_rate": 2.394390179822382e-05,
|
| 771 |
+
"loss": 0.0157,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.956834532374101,
|
| 776 |
+
"grad_norm": 0.23817431926727295,
|
| 777 |
+
"learning_rate": 2.2415772156621382e-05,
|
| 778 |
+
"loss": 0.0224,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.9928057553956835,
|
| 783 |
+
"grad_norm": 0.2008386105298996,
|
| 784 |
+
"learning_rate": 2.0931854578410905e-05,
|
| 785 |
+
"loss": 0.0128,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.028776978417266,
|
| 790 |
+
"grad_norm": 0.16650381684303284,
|
| 791 |
+
"learning_rate": 1.9492994687243714e-05,
|
| 792 |
+
"loss": 0.0144,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 4.0647482014388485,
|
| 797 |
+
"grad_norm": 0.20245954394340515,
|
| 798 |
+
"learning_rate": 1.8100012430246837e-05,
|
| 799 |
+
"loss": 0.0188,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.100719424460432,
|
| 804 |
+
"grad_norm": 0.21814769506454468,
|
| 805 |
+
"learning_rate": 1.6753701610768724e-05,
|
| 806 |
+
"loss": 0.0128,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.136690647482014,
|
| 811 |
+
"grad_norm": 0.10709039866924286,
|
| 812 |
+
"learning_rate": 1.545482943602341e-05,
|
| 813 |
+
"loss": 0.0191,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.172661870503597,
|
| 818 |
+
"grad_norm": 0.14134275913238525,
|
| 819 |
+
"learning_rate": 1.4204136079890584e-05,
|
| 820 |
+
"loss": 0.0124,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.2086330935251794,
|
| 825 |
+
"grad_norm": 0.33789917826652527,
|
| 826 |
+
"learning_rate": 1.300233426112103e-05,
|
| 827 |
+
"loss": 0.0152,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.244604316546763,
|
| 832 |
+
"grad_norm": 0.17584295570850372,
|
| 833 |
+
"learning_rate": 1.1850108837187335e-05,
|
| 834 |
+
"loss": 0.0162,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.280575539568345,
|
| 839 |
+
"grad_norm": 0.17790471017360687,
|
| 840 |
+
"learning_rate": 1.0748116414011888e-05,
|
| 841 |
+
"loss": 0.0189,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.316546762589928,
|
| 846 |
+
"grad_norm": 0.2450784593820572,
|
| 847 |
+
"learning_rate": 9.696984971794065e-06,
|
| 848 |
+
"loss": 0.0151,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.35251798561151,
|
| 853 |
+
"grad_norm": 0.4205743670463562,
|
| 854 |
+
"learning_rate": 8.697313507150184e-06,
|
| 855 |
+
"loss": 0.0228,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.388489208633094,
|
| 860 |
+
"grad_norm": 0.15447357296943665,
|
| 861 |
+
"learning_rate": 7.749671691769911e-06,
|
| 862 |
+
"loss": 0.0174,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.424460431654676,
|
| 867 |
+
"grad_norm": 0.1472749412059784,
|
| 868 |
+
"learning_rate": 6.854599547783736e-06,
|
| 869 |
+
"loss": 0.0191,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.460431654676259,
|
| 874 |
+
"grad_norm": 0.16755138337612152,
|
| 875 |
+
"learning_rate": 6.012607140026605e-06,
|
| 876 |
+
"loss": 0.0153,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.496402877697841,
|
| 881 |
+
"grad_norm": 0.16813568770885468,
|
| 882 |
+
"learning_rate": 5.224174285372974e-06,
|
| 883 |
+
"loss": 0.0181,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.532374100719425,
|
| 888 |
+
"grad_norm": 0.10466030985116959,
|
| 889 |
+
"learning_rate": 4.489750279308757e-06,
|
| 890 |
+
"loss": 0.0105,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.568345323741007,
|
| 895 |
+
"grad_norm": 0.1699487268924713,
|
| 896 |
+
"learning_rate": 3.8097536398963963e-06,
|
| 897 |
+
"loss": 0.0117,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 4.60431654676259,
|
| 902 |
+
"grad_norm": 0.19925230741500854,
|
| 903 |
+
"learning_rate": 3.184571869278574e-06,
|
| 904 |
+
"loss": 0.0155,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.640287769784173,
|
| 909 |
+
"grad_norm": 0.19352638721466064,
|
| 910 |
+
"learning_rate": 2.6145612328566717e-06,
|
| 911 |
+
"loss": 0.0118,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.676258992805756,
|
| 916 |
+
"grad_norm": 0.3062892556190491,
|
| 917 |
+
"learning_rate": 2.1000465562697856e-06,
|
| 918 |
+
"loss": 0.0145,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.712230215827338,
|
| 923 |
+
"grad_norm": 0.22001531720161438,
|
| 924 |
+
"learning_rate": 1.6413210402898893e-06,
|
| 925 |
+
"loss": 0.0134,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.748201438848921,
|
| 930 |
+
"grad_norm": 0.2228756546974182,
|
| 931 |
+
"learning_rate": 1.2386460937387822e-06,
|
| 932 |
+
"loss": 0.0106,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.784172661870503,
|
| 937 |
+
"grad_norm": 0.3254740834236145,
|
| 938 |
+
"learning_rate": 8.922511845219971e-07,
|
| 939 |
+
"loss": 0.0217,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.820143884892087,
|
| 944 |
+
"grad_norm": 0.23890721797943115,
|
| 945 |
+
"learning_rate": 6.023337088643665e-07,
|
| 946 |
+
"loss": 0.0217,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.856115107913669,
|
| 951 |
+
"grad_norm": 0.1692313402891159,
|
| 952 |
+
"learning_rate": 3.6905887882213717e-07,
|
| 953 |
+
"loss": 0.0171,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.892086330935252,
|
| 958 |
+
"grad_norm": 0.1608191728591919,
|
| 959 |
+
"learning_rate": 1.925596281353026e-07,
|
| 960 |
+
"loss": 0.0124,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.928057553956835,
|
| 965 |
+
"grad_norm": 0.1591498851776123,
|
| 966 |
+
"learning_rate": 7.293653647421073e-08,
|
| 967 |
+
"loss": 0.0223,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.9640287769784175,
|
| 972 |
+
"grad_norm": 0.16027827560901642,
|
| 973 |
+
"learning_rate": 1.0257772123312137e-08,
|
| 974 |
+
"loss": 0.0149,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.985611510791367,
|
| 979 |
+
"step": 1386,
|
| 980 |
+
"total_flos": 4.876041943922688e+16,
|
| 981 |
+
"train_loss": 0.05376770623662599,
|
| 982 |
+
"train_runtime": 634.0011,
|
| 983 |
+
"train_samples_per_second": 34.978,
|
| 984 |
+
"train_steps_per_second": 2.186
|
| 985 |
+
}
|
| 986 |
+
],
|
| 987 |
+
"logging_steps": 10,
|
| 988 |
+
"max_steps": 1386,
|
| 989 |
+
"num_input_tokens_seen": 0,
|
| 990 |
+
"num_train_epochs": 5,
|
| 991 |
+
"save_steps": 500,
|
| 992 |
+
"stateful_callbacks": {
|
| 993 |
+
"TrainerControl": {
|
| 994 |
+
"args": {
|
| 995 |
+
"should_epoch_stop": false,
|
| 996 |
+
"should_evaluate": false,
|
| 997 |
+
"should_log": false,
|
| 998 |
+
"should_save": true,
|
| 999 |
+
"should_training_stop": true
|
| 1000 |
+
},
|
| 1001 |
+
"attributes": {}
|
| 1002 |
+
}
|
| 1003 |
+
},
|
| 1004 |
+
"total_flos": 4.876041943922688e+16,
|
| 1005 |
+
"train_batch_size": 16,
|
| 1006 |
+
"trial_name": null,
|
| 1007 |
+
"trial_params": null
|
| 1008 |
+
}
|