Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +924 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,924 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.984251968503937,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1266,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03937007874015748,
|
| 13 |
+
"grad_norm": 6.157460689544678,
|
| 14 |
+
"learning_rate": 3.125e-05,
|
| 15 |
+
"loss": 1.4374,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.07874015748031496,
|
| 20 |
+
"grad_norm": 8.225975036621094,
|
| 21 |
+
"learning_rate": 6.25e-05,
|
| 22 |
+
"loss": 0.5746,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.11811023622047244,
|
| 27 |
+
"grad_norm": 2.298436403274536,
|
| 28 |
+
"learning_rate": 9.375e-05,
|
| 29 |
+
"loss": 0.3466,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.15748031496062992,
|
| 34 |
+
"grad_norm": 1.926019549369812,
|
| 35 |
+
"learning_rate": 0.000125,
|
| 36 |
+
"loss": 0.3016,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1968503937007874,
|
| 41 |
+
"grad_norm": 1.7295136451721191,
|
| 42 |
+
"learning_rate": 0.00015625,
|
| 43 |
+
"loss": 0.2241,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.23622047244094488,
|
| 48 |
+
"grad_norm": 1.6235859394073486,
|
| 49 |
+
"learning_rate": 0.0001875,
|
| 50 |
+
"loss": 0.2474,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.2755905511811024,
|
| 55 |
+
"grad_norm": 2.1525580883026123,
|
| 56 |
+
"learning_rate": 0.00019998770426725326,
|
| 57 |
+
"loss": 0.1819,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.31496062992125984,
|
| 62 |
+
"grad_norm": 1.2614878416061401,
|
| 63 |
+
"learning_rate": 0.00019991257462799398,
|
| 64 |
+
"loss": 0.1753,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.3543307086614173,
|
| 69 |
+
"grad_norm": 1.5198956727981567,
|
| 70 |
+
"learning_rate": 0.00019976919756880076,
|
| 71 |
+
"loss": 0.1734,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3937007874015748,
|
| 76 |
+
"grad_norm": 1.3017441034317017,
|
| 77 |
+
"learning_rate": 0.00019955767102643884,
|
| 78 |
+
"loss": 0.1558,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.4330708661417323,
|
| 83 |
+
"grad_norm": 1.3634755611419678,
|
| 84 |
+
"learning_rate": 0.00019927813948863483,
|
| 85 |
+
"loss": 0.1396,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.47244094488188976,
|
| 90 |
+
"grad_norm": 1.0837377309799194,
|
| 91 |
+
"learning_rate": 0.00019893079389538115,
|
| 92 |
+
"loss": 0.151,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.5118110236220472,
|
| 97 |
+
"grad_norm": 0.7570620179176331,
|
| 98 |
+
"learning_rate": 0.0001985158715085105,
|
| 99 |
+
"loss": 0.1345,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.5511811023622047,
|
| 104 |
+
"grad_norm": 1.1616774797439575,
|
| 105 |
+
"learning_rate": 0.00019803365574962904,
|
| 106 |
+
"loss": 0.1193,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.5905511811023622,
|
| 111 |
+
"grad_norm": 1.0171620845794678,
|
| 112 |
+
"learning_rate": 0.00019748447600651922,
|
| 113 |
+
"loss": 0.1168,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.6299212598425197,
|
| 118 |
+
"grad_norm": 1.0031143426895142,
|
| 119 |
+
"learning_rate": 0.00019686870740814438,
|
| 120 |
+
"loss": 0.1325,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.6692913385826772,
|
| 125 |
+
"grad_norm": 0.6090546250343323,
|
| 126 |
+
"learning_rate": 0.00019618677056840896,
|
| 127 |
+
"loss": 0.1154,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.7086614173228346,
|
| 132 |
+
"grad_norm": 1.0556284189224243,
|
| 133 |
+
"learning_rate": 0.00019543913129884918,
|
| 134 |
+
"loss": 0.1006,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.7480314960629921,
|
| 139 |
+
"grad_norm": 0.7367269396781921,
|
| 140 |
+
"learning_rate": 0.00019462630029045035,
|
| 141 |
+
"loss": 0.1182,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.7874015748031497,
|
| 146 |
+
"grad_norm": 17.055007934570312,
|
| 147 |
+
"learning_rate": 0.00019374883276480865,
|
| 148 |
+
"loss": 0.1464,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.8267716535433071,
|
| 153 |
+
"grad_norm": 8.562920570373535,
|
| 154 |
+
"learning_rate": 0.00019280732809487521,
|
| 155 |
+
"loss": 0.2103,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.8661417322834646,
|
| 160 |
+
"grad_norm": 1.041045069694519,
|
| 161 |
+
"learning_rate": 0.00019180242939554137,
|
| 162 |
+
"loss": 0.1697,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.905511811023622,
|
| 167 |
+
"grad_norm": 0.8244247436523438,
|
| 168 |
+
"learning_rate": 0.00019073482308434544,
|
| 169 |
+
"loss": 0.1478,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.9448818897637795,
|
| 174 |
+
"grad_norm": 0.9097272753715515,
|
| 175 |
+
"learning_rate": 0.00018960523841260053,
|
| 176 |
+
"loss": 0.1712,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.984251968503937,
|
| 181 |
+
"grad_norm": 1.7566819190979004,
|
| 182 |
+
"learning_rate": 0.00018841444696726388,
|
| 183 |
+
"loss": 0.1788,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.0236220472440944,
|
| 188 |
+
"grad_norm": 0.7728438973426819,
|
| 189 |
+
"learning_rate": 0.00018716326214388777,
|
| 190 |
+
"loss": 0.1519,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.0629921259842519,
|
| 195 |
+
"grad_norm": 1.4151078462600708,
|
| 196 |
+
"learning_rate": 0.00018585253859101247,
|
| 197 |
+
"loss": 0.178,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 1.1023622047244095,
|
| 202 |
+
"grad_norm": 0.7322055101394653,
|
| 203 |
+
"learning_rate": 0.0001844831716263801,
|
| 204 |
+
"loss": 0.1493,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.141732283464567,
|
| 209 |
+
"grad_norm": 1.238358974456787,
|
| 210 |
+
"learning_rate": 0.00018305609662536888,
|
| 211 |
+
"loss": 0.1366,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.1811023622047245,
|
| 216 |
+
"grad_norm": 1.0000638961791992,
|
| 217 |
+
"learning_rate": 0.0001815722883820649,
|
| 218 |
+
"loss": 0.1483,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.220472440944882,
|
| 223 |
+
"grad_norm": 0.9049971699714661,
|
| 224 |
+
"learning_rate": 0.00018003276044340814,
|
| 225 |
+
"loss": 0.1455,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.2598425196850394,
|
| 230 |
+
"grad_norm": 1.2286065816879272,
|
| 231 |
+
"learning_rate": 0.00017843856441686763,
|
| 232 |
+
"loss": 0.1519,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.2992125984251968,
|
| 237 |
+
"grad_norm": 0.9089654088020325,
|
| 238 |
+
"learning_rate": 0.00017679078925211851,
|
| 239 |
+
"loss": 0.1487,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.3385826771653544,
|
| 244 |
+
"grad_norm": 1.4565386772155762,
|
| 245 |
+
"learning_rate": 0.0001750905604972116,
|
| 246 |
+
"loss": 0.1299,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.3779527559055118,
|
| 251 |
+
"grad_norm": 1.1536667346954346,
|
| 252 |
+
"learning_rate": 0.00017333903952974374,
|
| 253 |
+
"loss": 0.1333,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.4173228346456692,
|
| 258 |
+
"grad_norm": 1.241052508354187,
|
| 259 |
+
"learning_rate": 0.0001715374227635541,
|
| 260 |
+
"loss": 0.1422,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.4566929133858268,
|
| 265 |
+
"grad_norm": 0.700495719909668,
|
| 266 |
+
"learning_rate": 0.00016968694083148805,
|
| 267 |
+
"loss": 0.1241,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.4960629921259843,
|
| 272 |
+
"grad_norm": 0.8857589960098267,
|
| 273 |
+
"learning_rate": 0.00016778885774478713,
|
| 274 |
+
"loss": 0.1312,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.5354330708661417,
|
| 279 |
+
"grad_norm": 0.8474507927894592,
|
| 280 |
+
"learning_rate": 0.00016584447002967903,
|
| 281 |
+
"loss": 0.1459,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.574803149606299,
|
| 286 |
+
"grad_norm": 1.0673331022262573,
|
| 287 |
+
"learning_rate": 0.00016385510584175785,
|
| 288 |
+
"loss": 0.1255,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.6141732283464567,
|
| 293 |
+
"grad_norm": 1.2963683605194092,
|
| 294 |
+
"learning_rate": 0.0001618221240587588,
|
| 295 |
+
"loss": 0.1443,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.6535433070866141,
|
| 300 |
+
"grad_norm": 0.6176678538322449,
|
| 301 |
+
"learning_rate": 0.0001597469133523479,
|
| 302 |
+
"loss": 0.1328,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.6929133858267718,
|
| 307 |
+
"grad_norm": 0.8610302805900574,
|
| 308 |
+
"learning_rate": 0.00015763089123956004,
|
| 309 |
+
"loss": 0.1315,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.7322834645669292,
|
| 314 |
+
"grad_norm": 0.8733415603637695,
|
| 315 |
+
"learning_rate": 0.0001554755031145336,
|
| 316 |
+
"loss": 0.1354,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.7716535433070866,
|
| 321 |
+
"grad_norm": 0.6525992155075073,
|
| 322 |
+
"learning_rate": 0.00015328222126120315,
|
| 323 |
+
"loss": 0.1306,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.811023622047244,
|
| 328 |
+
"grad_norm": 0.6734092831611633,
|
| 329 |
+
"learning_rate": 0.00015105254384762453,
|
| 330 |
+
"loss": 0.1272,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.8503937007874016,
|
| 335 |
+
"grad_norm": 0.9078968167304993,
|
| 336 |
+
"learning_rate": 0.0001487879939026189,
|
| 337 |
+
"loss": 0.1084,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.889763779527559,
|
| 342 |
+
"grad_norm": 0.6580955982208252,
|
| 343 |
+
"learning_rate": 0.0001464901182754356,
|
| 344 |
+
"loss": 0.1116,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.9291338582677167,
|
| 349 |
+
"grad_norm": 0.8749574422836304,
|
| 350 |
+
"learning_rate": 0.00014416048657914352,
|
| 351 |
+
"loss": 0.1345,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.968503937007874,
|
| 356 |
+
"grad_norm": 0.8025640845298767,
|
| 357 |
+
"learning_rate": 0.00014180069011847347,
|
| 358 |
+
"loss": 0.096,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 2.0078740157480315,
|
| 363 |
+
"grad_norm": 0.5700274705886841,
|
| 364 |
+
"learning_rate": 0.00013941234080284328,
|
| 365 |
+
"loss": 0.1201,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.047244094488189,
|
| 370 |
+
"grad_norm": 1.1415060758590698,
|
| 371 |
+
"learning_rate": 0.00013699707004530883,
|
| 372 |
+
"loss": 0.1206,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 2.0866141732283463,
|
| 377 |
+
"grad_norm": 0.8418262600898743,
|
| 378 |
+
"learning_rate": 0.0001345565276481924,
|
| 379 |
+
"loss": 0.1257,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.1259842519685037,
|
| 384 |
+
"grad_norm": 0.8720746040344238,
|
| 385 |
+
"learning_rate": 0.00013209238067614989,
|
| 386 |
+
"loss": 0.1169,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 2.1653543307086616,
|
| 391 |
+
"grad_norm": 1.166407585144043,
|
| 392 |
+
"learning_rate": 0.0001296063123174467,
|
| 393 |
+
"loss": 0.1327,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 2.204724409448819,
|
| 398 |
+
"grad_norm": 0.8546102046966553,
|
| 399 |
+
"learning_rate": 0.00012710002073421995,
|
| 400 |
+
"loss": 0.1287,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.2440944881889764,
|
| 405 |
+
"grad_norm": 0.6846320033073425,
|
| 406 |
+
"learning_rate": 0.00012457521790251272,
|
| 407 |
+
"loss": 0.1143,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 2.283464566929134,
|
| 412 |
+
"grad_norm": 1.0204697847366333,
|
| 413 |
+
"learning_rate": 0.00012203362844287196,
|
| 414 |
+
"loss": 0.1261,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 2.322834645669291,
|
| 419 |
+
"grad_norm": 0.7173842191696167,
|
| 420 |
+
"learning_rate": 0.00011947698844230972,
|
| 421 |
+
"loss": 0.1002,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.362204724409449,
|
| 426 |
+
"grad_norm": 0.7630149722099304,
|
| 427 |
+
"learning_rate": 0.0001169070442684318,
|
| 428 |
+
"loss": 0.1127,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 2.4015748031496065,
|
| 433 |
+
"grad_norm": 0.819685697555542,
|
| 434 |
+
"learning_rate": 0.00011432555137654386,
|
| 435 |
+
"loss": 0.1149,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 2.440944881889764,
|
| 440 |
+
"grad_norm": 0.5875028371810913,
|
| 441 |
+
"learning_rate": 0.00011173427311055041,
|
| 442 |
+
"loss": 0.1099,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.4803149606299213,
|
| 447 |
+
"grad_norm": 0.5276161432266235,
|
| 448 |
+
"learning_rate": 0.00010913497949846483,
|
| 449 |
+
"loss": 0.0935,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 2.5196850393700787,
|
| 454 |
+
"grad_norm": 0.8867918848991394,
|
| 455 |
+
"learning_rate": 0.00010652944604335413,
|
| 456 |
+
"loss": 0.1024,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.559055118110236,
|
| 461 |
+
"grad_norm": 0.7481799125671387,
|
| 462 |
+
"learning_rate": 0.00010391945251054331,
|
| 463 |
+
"loss": 0.1001,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.5984251968503935,
|
| 468 |
+
"grad_norm": 1.034694790840149,
|
| 469 |
+
"learning_rate": 0.00010130678171190893,
|
| 470 |
+
"loss": 0.1066,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.637795275590551,
|
| 475 |
+
"grad_norm": 0.744408369064331,
|
| 476 |
+
"learning_rate": 9.869321828809109e-05,
|
| 477 |
+
"loss": 0.0911,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.677165354330709,
|
| 482 |
+
"grad_norm": 0.6600897908210754,
|
| 483 |
+
"learning_rate": 9.60805474894567e-05,
|
| 484 |
+
"loss": 0.0957,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.716535433070866,
|
| 489 |
+
"grad_norm": 0.8174748420715332,
|
| 490 |
+
"learning_rate": 9.347055395664589e-05,
|
| 491 |
+
"loss": 0.1009,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.7559055118110236,
|
| 496 |
+
"grad_norm": 1.1754448413848877,
|
| 497 |
+
"learning_rate": 9.08650205015352e-05,
|
| 498 |
+
"loss": 0.0994,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.795275590551181,
|
| 503 |
+
"grad_norm": 0.6818444728851318,
|
| 504 |
+
"learning_rate": 8.826572688944966e-05,
|
| 505 |
+
"loss": 0.0978,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.8346456692913384,
|
| 510 |
+
"grad_norm": 0.7322255969047546,
|
| 511 |
+
"learning_rate": 8.56744486234562e-05,
|
| 512 |
+
"loss": 0.0917,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.8740157480314963,
|
| 517 |
+
"grad_norm": 0.4844493567943573,
|
| 518 |
+
"learning_rate": 8.309295573156825e-05,
|
| 519 |
+
"loss": 0.1054,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.9133858267716537,
|
| 524 |
+
"grad_norm": 0.4901849031448364,
|
| 525 |
+
"learning_rate": 8.052301155769029e-05,
|
| 526 |
+
"loss": 0.0933,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.952755905511811,
|
| 531 |
+
"grad_norm": 0.6136536002159119,
|
| 532 |
+
"learning_rate": 7.796637155712802e-05,
|
| 533 |
+
"loss": 0.0898,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.9921259842519685,
|
| 538 |
+
"grad_norm": 0.5571744441986084,
|
| 539 |
+
"learning_rate": 7.542478209748727e-05,
|
| 540 |
+
"loss": 0.0932,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 3.031496062992126,
|
| 545 |
+
"grad_norm": 0.6652244925498962,
|
| 546 |
+
"learning_rate": 7.289997926578003e-05,
|
| 547 |
+
"loss": 0.0894,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 3.0708661417322833,
|
| 552 |
+
"grad_norm": 0.4437718391418457,
|
| 553 |
+
"learning_rate": 7.039368768255335e-05,
|
| 554 |
+
"loss": 0.0862,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 3.1102362204724407,
|
| 559 |
+
"grad_norm": 0.6460803747177124,
|
| 560 |
+
"learning_rate": 6.790761932385014e-05,
|
| 561 |
+
"loss": 0.0909,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 3.1496062992125986,
|
| 566 |
+
"grad_norm": 0.7623637914657593,
|
| 567 |
+
"learning_rate": 6.544347235180761e-05,
|
| 568 |
+
"loss": 0.1056,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 3.188976377952756,
|
| 573 |
+
"grad_norm": 0.45258432626724243,
|
| 574 |
+
"learning_rate": 6.300292995469119e-05,
|
| 575 |
+
"loss": 0.0828,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 3.2283464566929134,
|
| 580 |
+
"grad_norm": 0.8195613026618958,
|
| 581 |
+
"learning_rate": 6.058765919715674e-05,
|
| 582 |
+
"loss": 0.1034,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 3.267716535433071,
|
| 587 |
+
"grad_norm": 0.5008795857429504,
|
| 588 |
+
"learning_rate": 5.819930988152658e-05,
|
| 589 |
+
"loss": 0.0997,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 3.3070866141732282,
|
| 594 |
+
"grad_norm": 0.593920886516571,
|
| 595 |
+
"learning_rate": 5.5839513420856494e-05,
|
| 596 |
+
"loss": 0.085,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 3.3464566929133857,
|
| 601 |
+
"grad_norm": 0.5025109052658081,
|
| 602 |
+
"learning_rate": 5.3509881724564426e-05,
|
| 603 |
+
"loss": 0.0909,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 3.3858267716535435,
|
| 608 |
+
"grad_norm": 0.4997495412826538,
|
| 609 |
+
"learning_rate": 5.121200609738113e-05,
|
| 610 |
+
"loss": 0.0968,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.425196850393701,
|
| 615 |
+
"grad_norm": 1.036964774131775,
|
| 616 |
+
"learning_rate": 4.894745615237551e-05,
|
| 617 |
+
"loss": 0.1025,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 3.4645669291338583,
|
| 622 |
+
"grad_norm": 0.4552832841873169,
|
| 623 |
+
"learning_rate": 4.6717778738796883e-05,
|
| 624 |
+
"loss": 0.0839,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 3.5039370078740157,
|
| 629 |
+
"grad_norm": 0.5048990845680237,
|
| 630 |
+
"learning_rate": 4.4524496885466436e-05,
|
| 631 |
+
"loss": 0.0919,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 3.543307086614173,
|
| 636 |
+
"grad_norm": 0.4812853932380676,
|
| 637 |
+
"learning_rate": 4.236910876043999e-05,
|
| 638 |
+
"loss": 0.0945,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 3.5826771653543306,
|
| 643 |
+
"grad_norm": 0.6418349146842957,
|
| 644 |
+
"learning_rate": 4.02530866476521e-05,
|
| 645 |
+
"loss": 0.0893,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 3.622047244094488,
|
| 650 |
+
"grad_norm": 0.9326638579368591,
|
| 651 |
+
"learning_rate": 3.817787594124122e-05,
|
| 652 |
+
"loss": 0.1007,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.661417322834646,
|
| 657 |
+
"grad_norm": 0.9185812473297119,
|
| 658 |
+
"learning_rate": 3.6144894158242184e-05,
|
| 659 |
+
"loss": 0.0912,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 3.7007874015748032,
|
| 664 |
+
"grad_norm": 0.6465328335762024,
|
| 665 |
+
"learning_rate": 3.4155529970320986e-05,
|
| 666 |
+
"loss": 0.078,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 3.7401574803149606,
|
| 671 |
+
"grad_norm": 0.5492585897445679,
|
| 672 |
+
"learning_rate": 3.22111422552129e-05,
|
| 673 |
+
"loss": 0.078,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 3.779527559055118,
|
| 678 |
+
"grad_norm": 0.638911783695221,
|
| 679 |
+
"learning_rate": 3.031305916851196e-05,
|
| 680 |
+
"loss": 0.083,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.8188976377952755,
|
| 685 |
+
"grad_norm": 0.47551003098487854,
|
| 686 |
+
"learning_rate": 2.846257723644592e-05,
|
| 687 |
+
"loss": 0.0847,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.8582677165354333,
|
| 692 |
+
"grad_norm": 0.8244979381561279,
|
| 693 |
+
"learning_rate": 2.666096047025628e-05,
|
| 694 |
+
"loss": 0.0871,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.8976377952755907,
|
| 699 |
+
"grad_norm": 0.2906242609024048,
|
| 700 |
+
"learning_rate": 2.4909439502788423e-05,
|
| 701 |
+
"loss": 0.0757,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.937007874015748,
|
| 706 |
+
"grad_norm": 0.6135310530662537,
|
| 707 |
+
"learning_rate": 2.3209210747881515e-05,
|
| 708 |
+
"loss": 0.0812,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.9763779527559056,
|
| 713 |
+
"grad_norm": 0.5390735864639282,
|
| 714 |
+
"learning_rate": 2.1561435583132395e-05,
|
| 715 |
+
"loss": 0.0852,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 4.015748031496063,
|
| 720 |
+
"grad_norm": 0.3508215546607971,
|
| 721 |
+
"learning_rate": 1.996723955659189e-05,
|
| 722 |
+
"loss": 0.0962,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 4.05511811023622,
|
| 727 |
+
"grad_norm": 0.5338996648788452,
|
| 728 |
+
"learning_rate": 1.842771161793513e-05,
|
| 729 |
+
"loss": 0.0863,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 4.094488188976378,
|
| 734 |
+
"grad_norm": 0.45414066314697266,
|
| 735 |
+
"learning_rate": 1.6943903374631143e-05,
|
| 736 |
+
"loss": 0.0822,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 4.133858267716535,
|
| 741 |
+
"grad_norm": 0.8013415336608887,
|
| 742 |
+
"learning_rate": 1.5516828373619918e-05,
|
| 743 |
+
"loss": 0.0717,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 4.173228346456693,
|
| 748 |
+
"grad_norm": 0.8064348697662354,
|
| 749 |
+
"learning_rate": 1.4147461408987561e-05,
|
| 750 |
+
"loss": 0.0945,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 4.21259842519685,
|
| 755 |
+
"grad_norm": 0.4887343645095825,
|
| 756 |
+
"learning_rate": 1.2836737856112246e-05,
|
| 757 |
+
"loss": 0.0851,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 4.251968503937007,
|
| 762 |
+
"grad_norm": 0.39578166604042053,
|
| 763 |
+
"learning_rate": 1.158555303273614e-05,
|
| 764 |
+
"loss": 0.0715,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 4.291338582677166,
|
| 769 |
+
"grad_norm": 0.5196364521980286,
|
| 770 |
+
"learning_rate": 1.0394761587399493e-05,
|
| 771 |
+
"loss": 0.079,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 4.330708661417323,
|
| 776 |
+
"grad_norm": 0.3454650044441223,
|
| 777 |
+
"learning_rate": 9.265176915654595e-06,
|
| 778 |
+
"loss": 0.0778,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 4.3700787401574805,
|
| 783 |
+
"grad_norm": 0.44942712783813477,
|
| 784 |
+
"learning_rate": 8.19757060445865e-06,
|
| 785 |
+
"loss": 0.0725,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 4.409448818897638,
|
| 790 |
+
"grad_norm": 0.5285565257072449,
|
| 791 |
+
"learning_rate": 7.192671905124792e-06,
|
| 792 |
+
"loss": 0.0685,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 4.448818897637795,
|
| 797 |
+
"grad_norm": 0.5806540250778198,
|
| 798 |
+
"learning_rate": 6.251167235191336e-06,
|
| 799 |
+
"loss": 0.0759,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 4.488188976377953,
|
| 804 |
+
"grad_norm": 0.592225968837738,
|
| 805 |
+
"learning_rate": 5.373699709549673e-06,
|
| 806 |
+
"loss": 0.0806,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 4.52755905511811,
|
| 811 |
+
"grad_norm": 0.7170082330703735,
|
| 812 |
+
"learning_rate": 4.560868701150844e-06,
|
| 813 |
+
"loss": 0.0735,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 4.566929133858268,
|
| 818 |
+
"grad_norm": 0.4995642304420471,
|
| 819 |
+
"learning_rate": 3.8132294315910256e-06,
|
| 820 |
+
"loss": 0.073,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 4.606299212598425,
|
| 825 |
+
"grad_norm": 0.6979544162750244,
|
| 826 |
+
"learning_rate": 3.131292591855628e-06,
|
| 827 |
+
"loss": 0.0685,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 4.645669291338582,
|
| 832 |
+
"grad_norm": 0.8452284932136536,
|
| 833 |
+
"learning_rate": 2.515523993480806e-06,
|
| 834 |
+
"loss": 0.0827,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 4.68503937007874,
|
| 839 |
+
"grad_norm": 0.5433554649353027,
|
| 840 |
+
"learning_rate": 1.966344250370977e-06,
|
| 841 |
+
"loss": 0.0839,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 4.724409448818898,
|
| 846 |
+
"grad_norm": 0.6554785370826721,
|
| 847 |
+
"learning_rate": 1.4841284914895116e-06,
|
| 848 |
+
"loss": 0.0906,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 4.7637795275590555,
|
| 853 |
+
"grad_norm": 0.7909722924232483,
|
| 854 |
+
"learning_rate": 1.0692061046188696e-06,
|
| 855 |
+
"loss": 0.0926,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 4.803149606299213,
|
| 860 |
+
"grad_norm": 0.6115571856498718,
|
| 861 |
+
"learning_rate": 7.2186051136518e-07,
|
| 862 |
+
"loss": 0.0949,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 4.84251968503937,
|
| 867 |
+
"grad_norm": 0.39121922850608826,
|
| 868 |
+
"learning_rate": 4.4232897356116795e-07,
|
| 869 |
+
"loss": 0.0866,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 4.881889763779528,
|
| 874 |
+
"grad_norm": 0.506991446018219,
|
| 875 |
+
"learning_rate": 2.308024311992618e-07,
|
| 876 |
+
"loss": 0.0713,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 4.921259842519685,
|
| 881 |
+
"grad_norm": 0.5368971228599548,
|
| 882 |
+
"learning_rate": 8.742537200603362e-08,
|
| 883 |
+
"loss": 0.0811,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 4.960629921259843,
|
| 888 |
+
"grad_norm": 0.7474204897880554,
|
| 889 |
+
"learning_rate": 1.2295732746747668e-08,
|
| 890 |
+
"loss": 0.0754,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 4.984251968503937,
|
| 895 |
+
"step": 1266,
|
| 896 |
+
"total_flos": 4.27570483010112e+16,
|
| 897 |
+
"train_loss": 0.12990388043495527,
|
| 898 |
+
"train_runtime": 571.8261,
|
| 899 |
+
"train_samples_per_second": 35.423,
|
| 900 |
+
"train_steps_per_second": 2.214
|
| 901 |
+
}
|
| 902 |
+
],
|
| 903 |
+
"logging_steps": 10,
|
| 904 |
+
"max_steps": 1266,
|
| 905 |
+
"num_input_tokens_seen": 0,
|
| 906 |
+
"num_train_epochs": 5,
|
| 907 |
+
"save_steps": 500,
|
| 908 |
+
"stateful_callbacks": {
|
| 909 |
+
"TrainerControl": {
|
| 910 |
+
"args": {
|
| 911 |
+
"should_epoch_stop": false,
|
| 912 |
+
"should_evaluate": false,
|
| 913 |
+
"should_log": false,
|
| 914 |
+
"should_save": true,
|
| 915 |
+
"should_training_stop": true
|
| 916 |
+
},
|
| 917 |
+
"attributes": {}
|
| 918 |
+
}
|
| 919 |
+
},
|
| 920 |
+
"total_flos": 4.27570483010112e+16,
|
| 921 |
+
"train_batch_size": 16,
|
| 922 |
+
"trial_name": null,
|
| 923 |
+
"trial_params": null
|
| 924 |
+
}
|