Commit
•
3692ab1
1
Parent(s):
76eaa83
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +740 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,740 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.7226567019922543,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 50000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 5.4451772745526446e-05,
|
| 13 |
+
"grad_norm": 182.4078369140625,
|
| 14 |
+
"learning_rate": 2.7225701061802342e-08,
|
| 15 |
+
"loss": 13.9417,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.027225886372763224,
|
| 20 |
+
"grad_norm": 6.484030246734619,
|
| 21 |
+
"learning_rate": 1.3612850530901172e-05,
|
| 22 |
+
"loss": 8.9538,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05445177274552645,
|
| 27 |
+
"grad_norm": 8.51893138885498,
|
| 28 |
+
"learning_rate": 2.7225701061802345e-05,
|
| 29 |
+
"loss": 6.3731,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08167765911828967,
|
| 34 |
+
"grad_norm": 5.699058532714844,
|
| 35 |
+
"learning_rate": 4.083855159270352e-05,
|
| 36 |
+
"loss": 5.4332,
|
| 37 |
+
"step": 1500
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1089035454910529,
|
| 41 |
+
"grad_norm": 11.347241401672363,
|
| 42 |
+
"learning_rate": 5.445140212360469e-05,
|
| 43 |
+
"loss": 4.6858,
|
| 44 |
+
"step": 2000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.13612943186381612,
|
| 48 |
+
"grad_norm": 2.8045220375061035,
|
| 49 |
+
"learning_rate": 6.806425265450586e-05,
|
| 50 |
+
"loss": 4.1909,
|
| 51 |
+
"step": 2500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.16335531823657934,
|
| 55 |
+
"grad_norm": 2.2558796405792236,
|
| 56 |
+
"learning_rate": 8.167710318540703e-05,
|
| 57 |
+
"loss": 3.929,
|
| 58 |
+
"step": 3000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.19058120460934255,
|
| 62 |
+
"grad_norm": 1.9897191524505615,
|
| 63 |
+
"learning_rate": 9.52899537163082e-05,
|
| 64 |
+
"loss": 3.7863,
|
| 65 |
+
"step": 3500
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2178070909821058,
|
| 69 |
+
"grad_norm": 1.875031590461731,
|
| 70 |
+
"learning_rate": 9.999458213543955e-05,
|
| 71 |
+
"loss": 3.6905,
|
| 72 |
+
"step": 4000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.245032977354869,
|
| 76 |
+
"grad_norm": 1.7890335321426392,
|
| 77 |
+
"learning_rate": 9.996535015198725e-05,
|
| 78 |
+
"loss": 3.6033,
|
| 79 |
+
"step": 4500
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.27225886372763225,
|
| 83 |
+
"grad_norm": 1.7135872840881348,
|
| 84 |
+
"learning_rate": 9.99108024021366e-05,
|
| 85 |
+
"loss": 3.5396,
|
| 86 |
+
"step": 5000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.29948475010039544,
|
| 90 |
+
"grad_norm": 1.6878308057785034,
|
| 91 |
+
"learning_rate": 9.983096652340218e-05,
|
| 92 |
+
"loss": 3.4888,
|
| 93 |
+
"step": 5500
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3267106364731587,
|
| 97 |
+
"grad_norm": 1.6131703853607178,
|
| 98 |
+
"learning_rate": 9.972588296594496e-05,
|
| 99 |
+
"loss": 3.4457,
|
| 100 |
+
"step": 6000
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3539365228459219,
|
| 104 |
+
"grad_norm": 1.5502527952194214,
|
| 105 |
+
"learning_rate": 9.959560497207756e-05,
|
| 106 |
+
"loss": 3.414,
|
| 107 |
+
"step": 6500
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3811624092186851,
|
| 111 |
+
"grad_norm": 1.5919506549835205,
|
| 112 |
+
"learning_rate": 9.944019854928815e-05,
|
| 113 |
+
"loss": 3.3923,
|
| 114 |
+
"step": 7000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.40838829559144835,
|
| 118 |
+
"grad_norm": 1.558945655822754,
|
| 119 |
+
"learning_rate": 9.925974243679667e-05,
|
| 120 |
+
"loss": 3.3694,
|
| 121 |
+
"step": 7500
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4356141819642116,
|
| 125 |
+
"grad_norm": 1.5559173822402954,
|
| 126 |
+
"learning_rate": 9.905432806566039e-05,
|
| 127 |
+
"loss": 3.3422,
|
| 128 |
+
"step": 8000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4628400683369748,
|
| 132 |
+
"grad_norm": 1.5621517896652222,
|
| 133 |
+
"learning_rate": 9.882405951244874e-05,
|
| 134 |
+
"loss": 3.3287,
|
| 135 |
+
"step": 8500
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.490065954709738,
|
| 139 |
+
"grad_norm": 1.4939193725585938,
|
| 140 |
+
"learning_rate": 9.856905344651121e-05,
|
| 141 |
+
"loss": 3.3063,
|
| 142 |
+
"step": 9000
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5172918410825013,
|
| 146 |
+
"grad_norm": 1.4603230953216553,
|
| 147 |
+
"learning_rate": 9.828943907086507e-05,
|
| 148 |
+
"loss": 3.291,
|
| 149 |
+
"step": 9500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5445177274552645,
|
| 153 |
+
"grad_norm": 1.4974786043167114,
|
| 154 |
+
"learning_rate": 9.798535805673234e-05,
|
| 155 |
+
"loss": 3.2838,
|
| 156 |
+
"step": 10000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5717436138280277,
|
| 160 |
+
"grad_norm": 1.5649632215499878,
|
| 161 |
+
"learning_rate": 9.765696447175982e-05,
|
| 162 |
+
"loss": 3.2695,
|
| 163 |
+
"step": 10500
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5989695002007909,
|
| 167 |
+
"grad_norm": 1.523169994354248,
|
| 168 |
+
"learning_rate": 9.73044247019582e-05,
|
| 169 |
+
"loss": 3.2554,
|
| 170 |
+
"step": 11000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6261953865735541,
|
| 174 |
+
"grad_norm": 1.4261916875839233,
|
| 175 |
+
"learning_rate": 9.692791736739979e-05,
|
| 176 |
+
"loss": 3.242,
|
| 177 |
+
"step": 11500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6534212729463174,
|
| 181 |
+
"grad_norm": 1.4566397666931152,
|
| 182 |
+
"learning_rate": 9.65276332317178e-05,
|
| 183 |
+
"loss": 3.2229,
|
| 184 |
+
"step": 12000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6806471593190806,
|
| 188 |
+
"grad_norm": 1.4012161493301392,
|
| 189 |
+
"learning_rate": 9.610377510545277e-05,
|
| 190 |
+
"loss": 3.223,
|
| 191 |
+
"step": 12500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7078730456918438,
|
| 195 |
+
"grad_norm": 1.4838244915008545,
|
| 196 |
+
"learning_rate": 9.565655774329525e-05,
|
| 197 |
+
"loss": 3.2162,
|
| 198 |
+
"step": 13000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7350989320646071,
|
| 202 |
+
"grad_norm": 1.4344415664672852,
|
| 203 |
+
"learning_rate": 9.518620773527684e-05,
|
| 204 |
+
"loss": 3.2148,
|
| 205 |
+
"step": 13500
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7623248184373702,
|
| 209 |
+
"grad_norm": 1.4415775537490845,
|
| 210 |
+
"learning_rate": 9.469296339196459e-05,
|
| 211 |
+
"loss": 3.206,
|
| 212 |
+
"step": 14000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7895507048101335,
|
| 216 |
+
"grad_norm": 1.469138264656067,
|
| 217 |
+
"learning_rate": 9.417707462371708e-05,
|
| 218 |
+
"loss": 3.1997,
|
| 219 |
+
"step": 14500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8167765911828967,
|
| 223 |
+
"grad_norm": 1.4184919595718384,
|
| 224 |
+
"learning_rate": 9.36388028140632e-05,
|
| 225 |
+
"loss": 3.1877,
|
| 226 |
+
"step": 15000
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8440024775556599,
|
| 230 |
+
"grad_norm": 1.4781111478805542,
|
| 231 |
+
"learning_rate": 9.307842068726786e-05,
|
| 232 |
+
"loss": 3.1893,
|
| 233 |
+
"step": 15500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8712283639284232,
|
| 237 |
+
"grad_norm": 1.4547381401062012,
|
| 238 |
+
"learning_rate": 9.249621217015181e-05,
|
| 239 |
+
"loss": 3.1728,
|
| 240 |
+
"step": 16000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.8984542503011864,
|
| 244 |
+
"grad_norm": 1.4702293872833252,
|
| 245 |
+
"learning_rate": 9.189247224823542e-05,
|
| 246 |
+
"loss": 3.1727,
|
| 247 |
+
"step": 16500
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9256801366739495,
|
| 251 |
+
"grad_norm": 1.5003573894500732,
|
| 252 |
+
"learning_rate": 9.126750681627938e-05,
|
| 253 |
+
"loss": 3.1648,
|
| 254 |
+
"step": 17000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9529060230467128,
|
| 258 |
+
"grad_norm": 1.43069589138031,
|
| 259 |
+
"learning_rate": 9.062163252329815e-05,
|
| 260 |
+
"loss": 3.1589,
|
| 261 |
+
"step": 17500
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.980131909419476,
|
| 265 |
+
"grad_norm": 1.443028450012207,
|
| 266 |
+
"learning_rate": 8.99551766121245e-05,
|
| 267 |
+
"loss": 3.1447,
|
| 268 |
+
"step": 18000
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.0073918281502052,
|
| 272 |
+
"grad_norm": 1.4987826347351074,
|
| 273 |
+
"learning_rate": 8.926847675360655e-05,
|
| 274 |
+
"loss": 3.1253,
|
| 275 |
+
"step": 18500
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.0346177145229685,
|
| 279 |
+
"grad_norm": 1.514776349067688,
|
| 280 |
+
"learning_rate": 8.856188087552123e-05,
|
| 281 |
+
"loss": 3.0292,
|
| 282 |
+
"step": 19000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.0618436008957317,
|
| 286 |
+
"grad_norm": 1.464104175567627,
|
| 287 |
+
"learning_rate": 8.783574698629109e-05,
|
| 288 |
+
"loss": 3.0248,
|
| 289 |
+
"step": 19500
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.089069487268495,
|
| 293 |
+
"grad_norm": 1.4995203018188477,
|
| 294 |
+
"learning_rate": 8.709044299359327e-05,
|
| 295 |
+
"loss": 3.0316,
|
| 296 |
+
"step": 20000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.1162953736412582,
|
| 300 |
+
"grad_norm": 1.5172643661499023,
|
| 301 |
+
"learning_rate": 8.632634651795318e-05,
|
| 302 |
+
"loss": 3.0303,
|
| 303 |
+
"step": 20500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.1435212600140212,
|
| 307 |
+
"grad_norm": 1.5371851921081543,
|
| 308 |
+
"learning_rate": 8.554384470141671e-05,
|
| 309 |
+
"loss": 3.0278,
|
| 310 |
+
"step": 21000
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.1707471463867845,
|
| 314 |
+
"grad_norm": 1.4607304334640503,
|
| 315 |
+
"learning_rate": 8.474333401139847e-05,
|
| 316 |
+
"loss": 3.0206,
|
| 317 |
+
"step": 21500
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.1979730327595477,
|
| 321 |
+
"grad_norm": 1.486690878868103,
|
| 322 |
+
"learning_rate": 8.392522003980496e-05,
|
| 323 |
+
"loss": 3.0251,
|
| 324 |
+
"step": 22000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.225198919132311,
|
| 328 |
+
"grad_norm": 1.4945247173309326,
|
| 329 |
+
"learning_rate": 8.308991729753484e-05,
|
| 330 |
+
"loss": 3.03,
|
| 331 |
+
"step": 22500
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.2524248055050742,
|
| 335 |
+
"grad_norm": 1.4576785564422607,
|
| 336 |
+
"learning_rate": 8.223784900446006e-05,
|
| 337 |
+
"loss": 3.0219,
|
| 338 |
+
"step": 23000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.2796506918778374,
|
| 342 |
+
"grad_norm": 1.5146127939224243,
|
| 343 |
+
"learning_rate": 8.136944687499468e-05,
|
| 344 |
+
"loss": 3.015,
|
| 345 |
+
"step": 23500
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.3068765782506007,
|
| 349 |
+
"grad_norm": 1.5249332189559937,
|
| 350 |
+
"learning_rate": 8.048515089935956e-05,
|
| 351 |
+
"loss": 3.0185,
|
| 352 |
+
"step": 24000
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.334102464623364,
|
| 356 |
+
"grad_norm": 1.4822596311569214,
|
| 357 |
+
"learning_rate": 7.958540912065415e-05,
|
| 358 |
+
"loss": 3.0271,
|
| 359 |
+
"step": 24500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3613283509961271,
|
| 363 |
+
"grad_norm": 1.5259217023849487,
|
| 364 |
+
"learning_rate": 7.867067740784801e-05,
|
| 365 |
+
"loss": 3.0188,
|
| 366 |
+
"step": 25000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.3885542373688904,
|
| 370 |
+
"grad_norm": 1.5163941383361816,
|
| 371 |
+
"learning_rate": 7.774141922480738e-05,
|
| 372 |
+
"loss": 3.0165,
|
| 373 |
+
"step": 25500
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.4157801237416536,
|
| 377 |
+
"grad_norm": 1.6021302938461304,
|
| 378 |
+
"learning_rate": 7.679810539547351e-05,
|
| 379 |
+
"loss": 3.0178,
|
| 380 |
+
"step": 26000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.4430060101144169,
|
| 384 |
+
"grad_norm": 1.4801899194717407,
|
| 385 |
+
"learning_rate": 7.584121386531205e-05,
|
| 386 |
+
"loss": 3.0167,
|
| 387 |
+
"step": 26500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.4702318964871801,
|
| 391 |
+
"grad_norm": 1.478881597518921,
|
| 392 |
+
"learning_rate": 7.487122945915404e-05,
|
| 393 |
+
"loss": 3.0134,
|
| 394 |
+
"step": 27000
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.4974577828599434,
|
| 398 |
+
"grad_norm": 1.5363502502441406,
|
| 399 |
+
"learning_rate": 7.388864363555151e-05,
|
| 400 |
+
"loss": 3.0162,
|
| 401 |
+
"step": 27500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.5246836692327066,
|
| 405 |
+
"grad_norm": 1.5000182390213013,
|
| 406 |
+
"learning_rate": 7.289395423777183e-05,
|
| 407 |
+
"loss": 3.0088,
|
| 408 |
+
"step": 28000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.5519095556054698,
|
| 412 |
+
"grad_norm": 1.4747732877731323,
|
| 413 |
+
"learning_rate": 7.188766524155734e-05,
|
| 414 |
+
"loss": 3.0004,
|
| 415 |
+
"step": 28500
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.5791354419782329,
|
| 419 |
+
"grad_norm": 1.5123411417007446,
|
| 420 |
+
"learning_rate": 7.08702864997777e-05,
|
| 421 |
+
"loss": 3.0021,
|
| 422 |
+
"step": 29000
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.606361328350996,
|
| 426 |
+
"grad_norm": 1.5053893327713013,
|
| 427 |
+
"learning_rate": 6.984233348410447e-05,
|
| 428 |
+
"loss": 3.0063,
|
| 429 |
+
"step": 29500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.6335872147237593,
|
| 433 |
+
"grad_norm": 1.4705040454864502,
|
| 434 |
+
"learning_rate": 6.88043270238391e-05,
|
| 435 |
+
"loss": 2.992,
|
| 436 |
+
"step": 30000
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.6608131010965226,
|
| 440 |
+
"grad_norm": 1.4783118963241577,
|
| 441 |
+
"learning_rate": 6.775679304202598e-05,
|
| 442 |
+
"loss": 2.9957,
|
| 443 |
+
"step": 30500
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.6880389874692858,
|
| 447 |
+
"grad_norm": 1.4604405164718628,
|
| 448 |
+
"learning_rate": 6.670026228898501e-05,
|
| 449 |
+
"loss": 3.0012,
|
| 450 |
+
"step": 31000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.715264873842049,
|
| 454 |
+
"grad_norm": 1.4742037057876587,
|
| 455 |
+
"learning_rate": 6.563527007339827e-05,
|
| 456 |
+
"loss": 2.9946,
|
| 457 |
+
"step": 31500
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.742490760214812,
|
| 461 |
+
"grad_norm": 1.5092357397079468,
|
| 462 |
+
"learning_rate": 6.456235599108689e-05,
|
| 463 |
+
"loss": 2.9953,
|
| 464 |
+
"step": 32000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.7697166465875753,
|
| 468 |
+
"grad_norm": 1.4896336793899536,
|
| 469 |
+
"learning_rate": 6.348206365161614e-05,
|
| 470 |
+
"loss": 2.982,
|
| 471 |
+
"step": 32500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.7969425329603386,
|
| 475 |
+
"grad_norm": 1.5187143087387085,
|
| 476 |
+
"learning_rate": 6.239494040286657e-05,
|
| 477 |
+
"loss": 2.9954,
|
| 478 |
+
"step": 33000
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.8241684193331018,
|
| 482 |
+
"grad_norm": 1.5126278400421143,
|
| 483 |
+
"learning_rate": 6.13015370537113e-05,
|
| 484 |
+
"loss": 2.9847,
|
| 485 |
+
"step": 33500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.851394305705865,
|
| 489 |
+
"grad_norm": 1.4799253940582275,
|
| 490 |
+
"learning_rate": 6.0202407594939615e-05,
|
| 491 |
+
"loss": 2.9843,
|
| 492 |
+
"step": 34000
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.8786201920786283,
|
| 496 |
+
"grad_norm": 1.5327789783477783,
|
| 497 |
+
"learning_rate": 5.9098108918568354e-05,
|
| 498 |
+
"loss": 2.9804,
|
| 499 |
+
"step": 34500
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.9058460784513915,
|
| 503 |
+
"grad_norm": 1.544638991355896,
|
| 504 |
+
"learning_rate": 5.798920053568351e-05,
|
| 505 |
+
"loss": 2.977,
|
| 506 |
+
"step": 35000
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.9330719648241548,
|
| 510 |
+
"grad_norm": 1.5565699338912964,
|
| 511 |
+
"learning_rate": 5.687624429295468e-05,
|
| 512 |
+
"loss": 2.9804,
|
| 513 |
+
"step": 35500
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.960297851196918,
|
| 517 |
+
"grad_norm": 1.4633305072784424,
|
| 518 |
+
"learning_rate": 5.575980408796623e-05,
|
| 519 |
+
"loss": 2.9758,
|
| 520 |
+
"step": 36000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.9875237375696813,
|
| 524 |
+
"grad_norm": 1.519126534461975,
|
| 525 |
+
"learning_rate": 5.464044558350929e-05,
|
| 526 |
+
"loss": 2.9697,
|
| 527 |
+
"step": 36500
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.0147836563004105,
|
| 531 |
+
"grad_norm": 1.5512584447860718,
|
| 532 |
+
"learning_rate": 5.3518735920979434e-05,
|
| 533 |
+
"loss": 2.8931,
|
| 534 |
+
"step": 37000
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.0420095426731737,
|
| 538 |
+
"grad_norm": 1.5433844327926636,
|
| 539 |
+
"learning_rate": 5.239524343302511e-05,
|
| 540 |
+
"loss": 2.8394,
|
| 541 |
+
"step": 37500
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.069235429045937,
|
| 545 |
+
"grad_norm": 1.5760645866394043,
|
| 546 |
+
"learning_rate": 5.127053735559263e-05,
|
| 547 |
+
"loss": 2.8356,
|
| 548 |
+
"step": 38000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.0964613154187,
|
| 552 |
+
"grad_norm": 1.6160222291946411,
|
| 553 |
+
"learning_rate": 5.0145187539513326e-05,
|
| 554 |
+
"loss": 2.8333,
|
| 555 |
+
"step": 38500
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.1236872017914634,
|
| 559 |
+
"grad_norm": 1.5705721378326416,
|
| 560 |
+
"learning_rate": 4.901976416177933e-05,
|
| 561 |
+
"loss": 2.8355,
|
| 562 |
+
"step": 39000
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.1509130881642267,
|
| 566 |
+
"grad_norm": 1.5703595876693726,
|
| 567 |
+
"learning_rate": 4.7894837436654e-05,
|
| 568 |
+
"loss": 2.8399,
|
| 569 |
+
"step": 39500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.17813897453699,
|
| 573 |
+
"grad_norm": 1.5459660291671753,
|
| 574 |
+
"learning_rate": 4.677097732676345e-05,
|
| 575 |
+
"loss": 2.8474,
|
| 576 |
+
"step": 40000
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.205364860909753,
|
| 580 |
+
"grad_norm": 1.6205192804336548,
|
| 581 |
+
"learning_rate": 4.56487532543157e-05,
|
| 582 |
+
"loss": 2.8455,
|
| 583 |
+
"step": 40500
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.2325907472825164,
|
| 587 |
+
"grad_norm": 1.546263337135315,
|
| 588 |
+
"learning_rate": 4.4528733812593465e-05,
|
| 589 |
+
"loss": 2.8378,
|
| 590 |
+
"step": 41000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.2598166336552796,
|
| 594 |
+
"grad_norm": 1.641790509223938,
|
| 595 |
+
"learning_rate": 4.341148647786699e-05,
|
| 596 |
+
"loss": 2.8439,
|
| 597 |
+
"step": 41500
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.2870425200280424,
|
| 601 |
+
"grad_norm": 1.5860189199447632,
|
| 602 |
+
"learning_rate": 4.2297577321873025e-05,
|
| 603 |
+
"loss": 2.8485,
|
| 604 |
+
"step": 42000
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.3142684064008057,
|
| 608 |
+
"grad_norm": 1.578778624534607,
|
| 609 |
+
"learning_rate": 4.118757072500509e-05,
|
| 610 |
+
"loss": 2.8428,
|
| 611 |
+
"step": 42500
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.341494292773569,
|
| 615 |
+
"grad_norm": 1.6200742721557617,
|
| 616 |
+
"learning_rate": 4.008202909036093e-05,
|
| 617 |
+
"loss": 2.8439,
|
| 618 |
+
"step": 43000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.368720179146332,
|
| 622 |
+
"grad_norm": 1.5700623989105225,
|
| 623 |
+
"learning_rate": 3.898151255879183e-05,
|
| 624 |
+
"loss": 2.8421,
|
| 625 |
+
"step": 43500
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.3959460655190954,
|
| 629 |
+
"grad_norm": 1.608221173286438,
|
| 630 |
+
"learning_rate": 3.788657872509786e-05,
|
| 631 |
+
"loss": 2.8472,
|
| 632 |
+
"step": 44000
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.4231719518918586,
|
| 636 |
+
"grad_norm": 1.5805708169937134,
|
| 637 |
+
"learning_rate": 3.679778235551349e-05,
|
| 638 |
+
"loss": 2.8437,
|
| 639 |
+
"step": 44500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.450397838264622,
|
| 643 |
+
"grad_norm": 1.5437690019607544,
|
| 644 |
+
"learning_rate": 3.571567510662597e-05,
|
| 645 |
+
"loss": 2.839,
|
| 646 |
+
"step": 45000
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.477623724637385,
|
| 650 |
+
"grad_norm": 1.5659681558609009,
|
| 651 |
+
"learning_rate": 3.464080524586959e-05,
|
| 652 |
+
"loss": 2.8338,
|
| 653 |
+
"step": 45500
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.5048496110101484,
|
| 657 |
+
"grad_norm": 1.5578417778015137,
|
| 658 |
+
"learning_rate": 3.357371737373694e-05,
|
| 659 |
+
"loss": 2.8417,
|
| 660 |
+
"step": 46000
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.5320754973829116,
|
| 664 |
+
"grad_norm": 1.6232887506484985,
|
| 665 |
+
"learning_rate": 3.2514952147848014e-05,
|
| 666 |
+
"loss": 2.8424,
|
| 667 |
+
"step": 46500
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.559301383755675,
|
| 671 |
+
"grad_norm": 1.580432415008545,
|
| 672 |
+
"learning_rate": 3.146504600901725e-05,
|
| 673 |
+
"loss": 2.8325,
|
| 674 |
+
"step": 47000
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.586527270128438,
|
| 678 |
+
"grad_norm": 1.589545488357544,
|
| 679 |
+
"learning_rate": 3.0424530909456973e-05,
|
| 680 |
+
"loss": 2.8342,
|
| 681 |
+
"step": 47500
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.6137531565012013,
|
| 685 |
+
"grad_norm": 1.5450959205627441,
|
| 686 |
+
"learning_rate": 2.9393934043254943e-05,
|
| 687 |
+
"loss": 2.838,
|
| 688 |
+
"step": 48000
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.6409790428739646,
|
| 692 |
+
"grad_norm": 1.59015953540802,
|
| 693 |
+
"learning_rate": 2.8373777579262905e-05,
|
| 694 |
+
"loss": 2.8379,
|
| 695 |
+
"step": 48500
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.668204929246728,
|
| 699 |
+
"grad_norm": 1.586267113685608,
|
| 700 |
+
"learning_rate": 2.7364578396531094e-05,
|
| 701 |
+
"loss": 2.8316,
|
| 702 |
+
"step": 49000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.695430815619491,
|
| 706 |
+
"grad_norm": 1.6080377101898193,
|
| 707 |
+
"learning_rate": 2.636684782242303e-05,
|
| 708 |
+
"loss": 2.8333,
|
| 709 |
+
"step": 49500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.7226567019922543,
|
| 713 |
+
"grad_norm": 1.6002167463302612,
|
| 714 |
+
"learning_rate": 2.5381091373543046e-05,
|
| 715 |
+
"loss": 2.8373,
|
| 716 |
+
"step": 50000
|
| 717 |
+
}
|
| 718 |
+
],
|
| 719 |
+
"logging_steps": 500,
|
| 720 |
+
"max_steps": 73456,
|
| 721 |
+
"num_input_tokens_seen": 0,
|
| 722 |
+
"num_train_epochs": 4,
|
| 723 |
+
"save_steps": 50000,
|
| 724 |
+
"stateful_callbacks": {
|
| 725 |
+
"TrainerControl": {
|
| 726 |
+
"args": {
|
| 727 |
+
"should_epoch_stop": false,
|
| 728 |
+
"should_evaluate": false,
|
| 729 |
+
"should_log": false,
|
| 730 |
+
"should_save": true,
|
| 731 |
+
"should_training_stop": false
|
| 732 |
+
},
|
| 733 |
+
"attributes": {}
|
| 734 |
+
}
|
| 735 |
+
},
|
| 736 |
+
"total_flos": 4.074111485812618e+19,
|
| 737 |
+
"train_batch_size": 1,
|
| 738 |
+
"trial_name": null,
|
| 739 |
+
"trial_params": null
|
| 740 |
+
}
|