Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +792 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,792 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 10.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1070,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.09345794392523364,
|
| 14 |
+
"grad_norm": 2.035820722579956,
|
| 15 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 16 |
+
"loss": 1.0122,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.18691588785046728,
|
| 21 |
+
"grad_norm": 2.3610706329345703,
|
| 22 |
+
"learning_rate": 3.518518518518519e-05,
|
| 23 |
+
"loss": 0.3526,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.2803738317757009,
|
| 28 |
+
"grad_norm": 1.186436653137207,
|
| 29 |
+
"learning_rate": 5.370370370370371e-05,
|
| 30 |
+
"loss": 0.2888,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.37383177570093457,
|
| 35 |
+
"grad_norm": 2.2752561569213867,
|
| 36 |
+
"learning_rate": 7.222222222222222e-05,
|
| 37 |
+
"loss": 0.2957,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.4672897196261682,
|
| 42 |
+
"grad_norm": 1.9183247089385986,
|
| 43 |
+
"learning_rate": 9.074074074074075e-05,
|
| 44 |
+
"loss": 0.2282,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.5607476635514018,
|
| 49 |
+
"grad_norm": 1.8604793548583984,
|
| 50 |
+
"learning_rate": 9.999402437003975e-05,
|
| 51 |
+
"loss": 0.169,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.6542056074766355,
|
| 56 |
+
"grad_norm": 1.4165388345718384,
|
| 57 |
+
"learning_rate": 9.99462278999732e-05,
|
| 58 |
+
"loss": 0.1587,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.7476635514018691,
|
| 63 |
+
"grad_norm": 1.3570247888565063,
|
| 64 |
+
"learning_rate": 9.985068065535225e-05,
|
| 65 |
+
"loss": 0.1176,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.8411214953271028,
|
| 70 |
+
"grad_norm": 0.8269003033638,
|
| 71 |
+
"learning_rate": 9.970747398351445e-05,
|
| 72 |
+
"loss": 0.1031,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.9345794392523364,
|
| 77 |
+
"grad_norm": 1.9058301448822021,
|
| 78 |
+
"learning_rate": 9.951674479629056e-05,
|
| 79 |
+
"loss": 0.0991,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.02803738317757,
|
| 84 |
+
"grad_norm": 1.9390472173690796,
|
| 85 |
+
"learning_rate": 9.927867543911091e-05,
|
| 86 |
+
"loss": 0.0873,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 1.1214953271028036,
|
| 91 |
+
"grad_norm": 0.9300522208213806,
|
| 92 |
+
"learning_rate": 9.899349351667522e-05,
|
| 93 |
+
"loss": 0.0946,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.2149532710280373,
|
| 98 |
+
"grad_norm": 0.9239804148674011,
|
| 99 |
+
"learning_rate": 9.866147167535254e-05,
|
| 100 |
+
"loss": 0.0794,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.308411214953271,
|
| 105 |
+
"grad_norm": 1.4692777395248413,
|
| 106 |
+
"learning_rate": 9.828292734251944e-05,
|
| 107 |
+
"loss": 0.0898,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.4018691588785046,
|
| 112 |
+
"grad_norm": 0.8343071937561035,
|
| 113 |
+
"learning_rate": 9.785822242308562e-05,
|
| 114 |
+
"loss": 0.0819,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.4953271028037383,
|
| 119 |
+
"grad_norm": 0.7223255634307861,
|
| 120 |
+
"learning_rate": 9.738776295349687e-05,
|
| 121 |
+
"loss": 0.0803,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 1.588785046728972,
|
| 126 |
+
"grad_norm": 0.7587630748748779,
|
| 127 |
+
"learning_rate": 9.687199871354669e-05,
|
| 128 |
+
"loss": 0.0745,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 1.6822429906542056,
|
| 133 |
+
"grad_norm": 1.022656798362732,
|
| 134 |
+
"learning_rate": 9.631142279636706e-05,
|
| 135 |
+
"loss": 0.0894,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.7757009345794392,
|
| 140 |
+
"grad_norm": 0.8869289755821228,
|
| 141 |
+
"learning_rate": 9.570657113700985e-05,
|
| 142 |
+
"loss": 0.0859,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 1.8691588785046729,
|
| 147 |
+
"grad_norm": 0.9170877933502197,
|
| 148 |
+
"learning_rate": 9.50580220000696e-05,
|
| 149 |
+
"loss": 0.0737,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 1.9626168224299065,
|
| 154 |
+
"grad_norm": 1.0185060501098633,
|
| 155 |
+
"learning_rate": 9.436639542683727e-05,
|
| 156 |
+
"loss": 0.0696,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 2.05607476635514,
|
| 161 |
+
"grad_norm": 1.1690869331359863,
|
| 162 |
+
"learning_rate": 9.363235264251369e-05,
|
| 163 |
+
"loss": 0.0635,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.149532710280374,
|
| 168 |
+
"grad_norm": 0.940513551235199,
|
| 169 |
+
"learning_rate": 9.285659542404941e-05,
|
| 170 |
+
"loss": 0.072,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 2.2429906542056073,
|
| 175 |
+
"grad_norm": 0.5256121158599854,
|
| 176 |
+
"learning_rate": 9.203986542921532e-05,
|
| 177 |
+
"loss": 0.065,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 2.336448598130841,
|
| 182 |
+
"grad_norm": 0.7171943783760071,
|
| 183 |
+
"learning_rate": 9.11829434875454e-05,
|
| 184 |
+
"loss": 0.0594,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.4299065420560746,
|
| 189 |
+
"grad_norm": 0.6012755036354065,
|
| 190 |
+
"learning_rate": 9.02866488538296e-05,
|
| 191 |
+
"loss": 0.0657,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 2.5233644859813085,
|
| 196 |
+
"grad_norm": 0.496242493391037,
|
| 197 |
+
"learning_rate": 8.93518384248705e-05,
|
| 198 |
+
"loss": 0.0712,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 2.616822429906542,
|
| 203 |
+
"grad_norm": 0.7729496955871582,
|
| 204 |
+
"learning_rate": 8.837940592025257e-05,
|
| 205 |
+
"loss": 0.0581,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.710280373831776,
|
| 210 |
+
"grad_norm": 0.6967052817344666,
|
| 211 |
+
"learning_rate": 8.737028102790723e-05,
|
| 212 |
+
"loss": 0.065,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 2.803738317757009,
|
| 217 |
+
"grad_norm": 1.2209193706512451,
|
| 218 |
+
"learning_rate": 8.632542851529051e-05,
|
| 219 |
+
"loss": 0.0673,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 2.897196261682243,
|
| 224 |
+
"grad_norm": 0.8657803535461426,
|
| 225 |
+
"learning_rate": 8.524584730702339e-05,
|
| 226 |
+
"loss": 0.0778,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 2.9906542056074765,
|
| 231 |
+
"grad_norm": 0.6701989769935608,
|
| 232 |
+
"learning_rate": 8.413256952987611e-05,
|
| 233 |
+
"loss": 0.0617,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 3.0841121495327104,
|
| 238 |
+
"grad_norm": 0.46317973732948303,
|
| 239 |
+
"learning_rate": 8.298665952600999e-05,
|
| 240 |
+
"loss": 0.061,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 3.177570093457944,
|
| 245 |
+
"grad_norm": 0.549863338470459,
|
| 246 |
+
"learning_rate": 8.180921283541986e-05,
|
| 247 |
+
"loss": 0.0608,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 3.2710280373831777,
|
| 252 |
+
"grad_norm": 0.5789127349853516,
|
| 253 |
+
"learning_rate": 8.060135514854994e-05,
|
| 254 |
+
"loss": 0.0588,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 3.364485981308411,
|
| 259 |
+
"grad_norm": 0.5358679294586182,
|
| 260 |
+
"learning_rate": 7.936424123008464e-05,
|
| 261 |
+
"loss": 0.0584,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 3.457943925233645,
|
| 266 |
+
"grad_norm": 0.682597815990448,
|
| 267 |
+
"learning_rate": 7.809905381494316e-05,
|
| 268 |
+
"loss": 0.0585,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 3.5514018691588785,
|
| 273 |
+
"grad_norm": 0.7718446254730225,
|
| 274 |
+
"learning_rate": 7.68070024775332e-05,
|
| 275 |
+
"loss": 0.0525,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 3.6448598130841123,
|
| 280 |
+
"grad_norm": 0.5267370343208313,
|
| 281 |
+
"learning_rate": 7.548932247534506e-05,
|
| 282 |
+
"loss": 0.0537,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 3.7383177570093458,
|
| 287 |
+
"grad_norm": 0.8042123913764954,
|
| 288 |
+
"learning_rate": 7.414727356799154e-05,
|
| 289 |
+
"loss": 0.0613,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 3.831775700934579,
|
| 294 |
+
"grad_norm": 0.5799658298492432,
|
| 295 |
+
"learning_rate": 7.27821388128227e-05,
|
| 296 |
+
"loss": 0.0535,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 3.925233644859813,
|
| 301 |
+
"grad_norm": 0.8124668598175049,
|
| 302 |
+
"learning_rate": 7.139522333826707e-05,
|
| 303 |
+
"loss": 0.0493,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 4.018691588785047,
|
| 308 |
+
"grad_norm": 0.8771295547485352,
|
| 309 |
+
"learning_rate": 6.99878530960719e-05,
|
| 310 |
+
"loss": 0.0615,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 4.11214953271028,
|
| 315 |
+
"grad_norm": 0.7655673027038574,
|
| 316 |
+
"learning_rate": 6.856137359363533e-05,
|
| 317 |
+
"loss": 0.0605,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 4.205607476635514,
|
| 322 |
+
"grad_norm": 0.6487105488777161,
|
| 323 |
+
"learning_rate": 6.711714860764266e-05,
|
| 324 |
+
"loss": 0.0559,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 4.299065420560748,
|
| 329 |
+
"grad_norm": 0.7439901232719421,
|
| 330 |
+
"learning_rate": 6.565655888023618e-05,
|
| 331 |
+
"loss": 0.0562,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 4.392523364485982,
|
| 336 |
+
"grad_norm": 0.566832423210144,
|
| 337 |
+
"learning_rate": 6.418100079896556e-05,
|
| 338 |
+
"loss": 0.0535,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 4.485981308411215,
|
| 343 |
+
"grad_norm": 0.5109242796897888,
|
| 344 |
+
"learning_rate": 6.269188506178019e-05,
|
| 345 |
+
"loss": 0.0558,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 4.579439252336448,
|
| 350 |
+
"grad_norm": 0.5645363926887512,
|
| 351 |
+
"learning_rate": 6.11906353283405e-05,
|
| 352 |
+
"loss": 0.0445,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 4.672897196261682,
|
| 357 |
+
"grad_norm": 0.7012343406677246,
|
| 358 |
+
"learning_rate": 5.967868685893715e-05,
|
| 359 |
+
"loss": 0.0496,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 4.766355140186916,
|
| 364 |
+
"grad_norm": 0.47865885496139526,
|
| 365 |
+
"learning_rate": 5.815748514231944e-05,
|
| 366 |
+
"loss": 0.0486,
|
| 367 |
+
"step": 510
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 4.859813084112149,
|
| 371 |
+
"grad_norm": 0.38202741742134094,
|
| 372 |
+
"learning_rate": 5.6628484513745e-05,
|
| 373 |
+
"loss": 0.0412,
|
| 374 |
+
"step": 520
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 4.953271028037383,
|
| 378 |
+
"grad_norm": 0.7451114058494568,
|
| 379 |
+
"learning_rate": 5.5093146764571866e-05,
|
| 380 |
+
"loss": 0.0561,
|
| 381 |
+
"step": 530
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 5.046728971962617,
|
| 385 |
+
"grad_norm": 0.29692840576171875,
|
| 386 |
+
"learning_rate": 5.355293974472197e-05,
|
| 387 |
+
"loss": 0.0432,
|
| 388 |
+
"step": 540
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 5.140186915887851,
|
| 392 |
+
"grad_norm": 0.5470758080482483,
|
| 393 |
+
"learning_rate": 5.2009335959352666e-05,
|
| 394 |
+
"loss": 0.0456,
|
| 395 |
+
"step": 550
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 5.233644859813084,
|
| 399 |
+
"grad_norm": 0.3530406057834625,
|
| 400 |
+
"learning_rate": 5.046381116107742e-05,
|
| 401 |
+
"loss": 0.05,
|
| 402 |
+
"step": 560
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 5.327102803738318,
|
| 406 |
+
"grad_norm": 0.5904394388198853,
|
| 407 |
+
"learning_rate": 4.891784293908192e-05,
|
| 408 |
+
"loss": 0.0479,
|
| 409 |
+
"step": 570
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 5.420560747663552,
|
| 413 |
+
"grad_norm": 0.6387467384338379,
|
| 414 |
+
"learning_rate": 4.7372909306484276e-05,
|
| 415 |
+
"loss": 0.0419,
|
| 416 |
+
"step": 580
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 5.5140186915887845,
|
| 420 |
+
"grad_norm": 0.3460151255130768,
|
| 421 |
+
"learning_rate": 4.5830487287289966e-05,
|
| 422 |
+
"loss": 0.0457,
|
| 423 |
+
"step": 590
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 5.607476635514018,
|
| 427 |
+
"grad_norm": 0.6098641157150269,
|
| 428 |
+
"learning_rate": 4.429205150429241e-05,
|
| 429 |
+
"loss": 0.038,
|
| 430 |
+
"step": 600
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 5.700934579439252,
|
| 434 |
+
"grad_norm": 0.5131182670593262,
|
| 435 |
+
"learning_rate": 4.275907276926918e-05,
|
| 436 |
+
"loss": 0.0518,
|
| 437 |
+
"step": 610
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 5.794392523364486,
|
| 441 |
+
"grad_norm": 0.5069635510444641,
|
| 442 |
+
"learning_rate": 4.123301667682171e-05,
|
| 443 |
+
"loss": 0.0447,
|
| 444 |
+
"step": 620
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 5.88785046728972,
|
| 448 |
+
"grad_norm": 0.4461626708507538,
|
| 449 |
+
"learning_rate": 3.971534220320291e-05,
|
| 450 |
+
"loss": 0.0465,
|
| 451 |
+
"step": 630
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 5.981308411214953,
|
| 455 |
+
"grad_norm": 0.5642176866531372,
|
| 456 |
+
"learning_rate": 3.820750031147211e-05,
|
| 457 |
+
"loss": 0.0415,
|
| 458 |
+
"step": 640
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 6.074766355140187,
|
| 462 |
+
"grad_norm": 0.525589108467102,
|
| 463 |
+
"learning_rate": 3.67109325643111e-05,
|
| 464 |
+
"loss": 0.0398,
|
| 465 |
+
"step": 650
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 6.168224299065421,
|
| 469 |
+
"grad_norm": 0.5913786292076111,
|
| 470 |
+
"learning_rate": 3.522706974582717e-05,
|
| 471 |
+
"loss": 0.0421,
|
| 472 |
+
"step": 660
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 6.261682242990654,
|
| 476 |
+
"grad_norm": 0.44272592663764954,
|
| 477 |
+
"learning_rate": 3.375733049366115e-05,
|
| 478 |
+
"loss": 0.0377,
|
| 479 |
+
"step": 670
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 6.355140186915888,
|
| 483 |
+
"grad_norm": 0.5991699695587158,
|
| 484 |
+
"learning_rate": 3.2303119942707796e-05,
|
| 485 |
+
"loss": 0.0401,
|
| 486 |
+
"step": 680
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 6.4485981308411215,
|
| 490 |
+
"grad_norm": 0.4788666069507599,
|
| 491 |
+
"learning_rate": 3.086582838174551e-05,
|
| 492 |
+
"loss": 0.0344,
|
| 493 |
+
"step": 690
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 6.542056074766355,
|
| 497 |
+
"grad_norm": 0.2347191423177719,
|
| 498 |
+
"learning_rate": 2.944682992425959e-05,
|
| 499 |
+
"loss": 0.0431,
|
| 500 |
+
"step": 700
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 6.635514018691588,
|
| 504 |
+
"grad_norm": 0.4982701241970062,
|
| 505 |
+
"learning_rate": 2.804748119472969e-05,
|
| 506 |
+
"loss": 0.0431,
|
| 507 |
+
"step": 710
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 6.728971962616822,
|
| 511 |
+
"grad_norm": 0.3474225103855133,
|
| 512 |
+
"learning_rate": 2.6669120031637663e-05,
|
| 513 |
+
"loss": 0.0312,
|
| 514 |
+
"step": 720
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 6.822429906542056,
|
| 518 |
+
"grad_norm": 0.484762042760849,
|
| 519 |
+
"learning_rate": 2.5313064208435423e-05,
|
| 520 |
+
"loss": 0.0402,
|
| 521 |
+
"step": 730
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 6.91588785046729,
|
| 525 |
+
"grad_norm": 0.3601996898651123,
|
| 526 |
+
"learning_rate": 2.3980610173696255e-05,
|
| 527 |
+
"loss": 0.0353,
|
| 528 |
+
"step": 740
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 7.009345794392523,
|
| 532 |
+
"grad_norm": 0.4373137056827545,
|
| 533 |
+
"learning_rate": 2.2673031811653034e-05,
|
| 534 |
+
"loss": 0.0515,
|
| 535 |
+
"step": 750
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 7.102803738317757,
|
| 539 |
+
"grad_norm": 0.4686773121356964,
|
| 540 |
+
"learning_rate": 2.139157922430956e-05,
|
| 541 |
+
"loss": 0.042,
|
| 542 |
+
"step": 760
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 7.196261682242991,
|
| 546 |
+
"grad_norm": 0.4924180507659912,
|
| 547 |
+
"learning_rate": 2.01374775362883e-05,
|
| 548 |
+
"loss": 0.0396,
|
| 549 |
+
"step": 770
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 7.289719626168225,
|
| 553 |
+
"grad_norm": 0.3904656767845154,
|
| 554 |
+
"learning_rate": 1.8911925723557806e-05,
|
| 555 |
+
"loss": 0.0399,
|
| 556 |
+
"step": 780
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 7.383177570093458,
|
| 560 |
+
"grad_norm": 0.283329039812088,
|
| 561 |
+
"learning_rate": 1.7716095467159393e-05,
|
| 562 |
+
"loss": 0.0402,
|
| 563 |
+
"step": 790
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 7.4766355140186915,
|
| 567 |
+
"grad_norm": 0.4036600887775421,
|
| 568 |
+
"learning_rate": 1.6551130033028827e-05,
|
| 569 |
+
"loss": 0.0373,
|
| 570 |
+
"step": 800
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 7.570093457943925,
|
| 574 |
+
"grad_norm": 0.3786483407020569,
|
| 575 |
+
"learning_rate": 1.541814317898425e-05,
|
| 576 |
+
"loss": 0.0399,
|
| 577 |
+
"step": 810
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 7.663551401869158,
|
| 581 |
+
"grad_norm": 0.34531161189079285,
|
| 582 |
+
"learning_rate": 1.4318218089924962e-05,
|
| 583 |
+
"loss": 0.04,
|
| 584 |
+
"step": 820
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 7.757009345794392,
|
| 588 |
+
"grad_norm": 0.3333149254322052,
|
| 589 |
+
"learning_rate": 1.3252406342259527e-05,
|
| 590 |
+
"loss": 0.0328,
|
| 591 |
+
"step": 830
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 7.850467289719626,
|
| 595 |
+
"grad_norm": 0.35911786556243896,
|
| 596 |
+
"learning_rate": 1.2221726898552665e-05,
|
| 597 |
+
"loss": 0.0359,
|
| 598 |
+
"step": 840
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 7.94392523364486,
|
| 602 |
+
"grad_norm": 0.37050867080688477,
|
| 603 |
+
"learning_rate": 1.122716513335262e-05,
|
| 604 |
+
"loss": 0.0468,
|
| 605 |
+
"step": 850
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 8.037383177570094,
|
| 609 |
+
"grad_norm": 0.27867504954338074,
|
| 610 |
+
"learning_rate": 1.0269671891130123e-05,
|
| 611 |
+
"loss": 0.0308,
|
| 612 |
+
"step": 860
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 8.130841121495328,
|
| 616 |
+
"grad_norm": 0.36876147985458374,
|
| 617 |
+
"learning_rate": 9.350162577229432e-06,
|
| 618 |
+
"loss": 0.0379,
|
| 619 |
+
"step": 870
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 8.22429906542056,
|
| 623 |
+
"grad_norm": 0.3114546835422516,
|
| 624 |
+
"learning_rate": 8.46951628270098e-06,
|
| 625 |
+
"loss": 0.0354,
|
| 626 |
+
"step": 880
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 8.317757009345794,
|
| 630 |
+
"grad_norm": 0.26373207569122314,
|
| 631 |
+
"learning_rate": 7.628574943851852e-06,
|
| 632 |
+
"loss": 0.0296,
|
| 633 |
+
"step": 890
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 8.411214953271028,
|
| 637 |
+
"grad_norm": 0.3879351019859314,
|
| 638 |
+
"learning_rate": 6.82814253731801e-06,
|
| 639 |
+
"loss": 0.0288,
|
| 640 |
+
"step": 900
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 8.504672897196262,
|
| 644 |
+
"grad_norm": 0.3110659718513489,
|
| 645 |
+
"learning_rate": 6.06898431142745e-06,
|
| 646 |
+
"loss": 0.0293,
|
| 647 |
+
"step": 910
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 8.598130841121495,
|
| 651 |
+
"grad_norm": 0.21370220184326172,
|
| 652 |
+
"learning_rate": 5.351826054589393e-06,
|
| 653 |
+
"loss": 0.0294,
|
| 654 |
+
"step": 920
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 8.69158878504673,
|
| 658 |
+
"grad_norm": 0.304156094789505,
|
| 659 |
+
"learning_rate": 4.677353401408974e-06,
|
| 660 |
+
"loss": 0.0323,
|
| 661 |
+
"step": 930
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 8.785046728971963,
|
| 665 |
+
"grad_norm": 0.30655941367149353,
|
| 666 |
+
"learning_rate": 4.04621117719049e-06,
|
| 667 |
+
"loss": 0.0356,
|
| 668 |
+
"step": 940
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 8.878504672897197,
|
| 672 |
+
"grad_norm": 0.43658676743507385,
|
| 673 |
+
"learning_rate": 3.459002781456344e-06,
|
| 674 |
+
"loss": 0.0332,
|
| 675 |
+
"step": 950
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 8.97196261682243,
|
| 679 |
+
"grad_norm": 0.486379474401474,
|
| 680 |
+
"learning_rate": 2.9162896110707163e-06,
|
| 681 |
+
"loss": 0.0336,
|
| 682 |
+
"step": 960
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 9.065420560747663,
|
| 686 |
+
"grad_norm": 0.35680046677589417,
|
| 687 |
+
"learning_rate": 2.418590523519687e-06,
|
| 688 |
+
"loss": 0.0313,
|
| 689 |
+
"step": 970
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 9.158878504672897,
|
| 693 |
+
"grad_norm": 0.2002888023853302,
|
| 694 |
+
"learning_rate": 1.9663813408607845e-06,
|
| 695 |
+
"loss": 0.0357,
|
| 696 |
+
"step": 980
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 9.25233644859813,
|
| 700 |
+
"grad_norm": 0.27867501974105835,
|
| 701 |
+
"learning_rate": 1.5600943948163527e-06,
|
| 702 |
+
"loss": 0.0336,
|
| 703 |
+
"step": 990
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 9.345794392523365,
|
| 707 |
+
"grad_norm": 0.2867244780063629,
|
| 708 |
+
"learning_rate": 1.2001181134455475e-06,
|
| 709 |
+
"loss": 0.0325,
|
| 710 |
+
"step": 1000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 9.439252336448599,
|
| 714 |
+
"grad_norm": 0.284397691488266,
|
| 715 |
+
"learning_rate": 8.867966497901282e-07,
|
| 716 |
+
"loss": 0.0278,
|
| 717 |
+
"step": 1010
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 9.532710280373832,
|
| 721 |
+
"grad_norm": 0.24502161145210266,
|
| 722 |
+
"learning_rate": 6.204295528491555e-07,
|
| 723 |
+
"loss": 0.0377,
|
| 724 |
+
"step": 1020
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 9.626168224299064,
|
| 728 |
+
"grad_norm": 0.3403957784175873,
|
| 729 |
+
"learning_rate": 4.012714811970464e-07,
|
| 730 |
+
"loss": 0.0294,
|
| 731 |
+
"step": 1030
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 9.719626168224298,
|
| 735 |
+
"grad_norm": 0.311471551656723,
|
| 736 |
+
"learning_rate": 2.295319595188805e-07,
|
| 737 |
+
"loss": 0.0277,
|
| 738 |
+
"step": 1040
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 9.813084112149532,
|
| 742 |
+
"grad_norm": 0.26186737418174744,
|
| 743 |
+
"learning_rate": 1.0537517829562472e-07,
|
| 744 |
+
"loss": 0.0281,
|
| 745 |
+
"step": 1050
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 9.906542056074766,
|
| 749 |
+
"grad_norm": 0.20402012765407562,
|
| 750 |
+
"learning_rate": 2.8919836830887392e-08,
|
| 751 |
+
"loss": 0.0305,
|
| 752 |
+
"step": 1060
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 10.0,
|
| 756 |
+
"grad_norm": 0.8117493391036987,
|
| 757 |
+
"learning_rate": 2.3902976920009423e-10,
|
| 758 |
+
"loss": 0.0289,
|
| 759 |
+
"step": 1070
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 10.0,
|
| 763 |
+
"step": 1070,
|
| 764 |
+
"total_flos": 0.0,
|
| 765 |
+
"train_loss": 0.07138736554395372,
|
| 766 |
+
"train_runtime": 1117.3117,
|
| 767 |
+
"train_samples_per_second": 46.513,
|
| 768 |
+
"train_steps_per_second": 0.958
|
| 769 |
+
}
|
| 770 |
+
],
|
| 771 |
+
"logging_steps": 10,
|
| 772 |
+
"max_steps": 1070,
|
| 773 |
+
"num_input_tokens_seen": 0,
|
| 774 |
+
"num_train_epochs": 10,
|
| 775 |
+
"save_steps": 1000,
|
| 776 |
+
"stateful_callbacks": {
|
| 777 |
+
"TrainerControl": {
|
| 778 |
+
"args": {
|
| 779 |
+
"should_epoch_stop": false,
|
| 780 |
+
"should_evaluate": false,
|
| 781 |
+
"should_log": false,
|
| 782 |
+
"should_save": true,
|
| 783 |
+
"should_training_stop": true
|
| 784 |
+
},
|
| 785 |
+
"attributes": {}
|
| 786 |
+
}
|
| 787 |
+
},
|
| 788 |
+
"total_flos": 0.0,
|
| 789 |
+
"train_batch_size": 49,
|
| 790 |
+
"trial_name": null,
|
| 791 |
+
"trial_params": null
|
| 792 |
+
}
|