Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +465 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,465 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2703,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05549389567147614,
|
| 13 |
+
"grad_norm": 15.3125,
|
| 14 |
+
"learning_rate": 4.907510173880873e-05,
|
| 15 |
+
"loss": 1.186,
|
| 16 |
+
"mean_token_accuracy": 0.7049988703429699,
|
| 17 |
+
"step": 50
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.11098779134295228,
|
| 21 |
+
"grad_norm": 7.0,
|
| 22 |
+
"learning_rate": 4.8150203477617464e-05,
|
| 23 |
+
"loss": 1.1463,
|
| 24 |
+
"mean_token_accuracy": 0.7061308635771275,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.16648168701442842,
|
| 29 |
+
"grad_norm": 9.1875,
|
| 30 |
+
"learning_rate": 4.72253052164262e-05,
|
| 31 |
+
"loss": 0.9759,
|
| 32 |
+
"mean_token_accuracy": 0.7407757295668125,
|
| 33 |
+
"step": 150
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.22197558268590456,
|
| 37 |
+
"grad_norm": 11.0625,
|
| 38 |
+
"learning_rate": 4.6300406955234926e-05,
|
| 39 |
+
"loss": 1.0052,
|
| 40 |
+
"mean_token_accuracy": 0.7320078010857105,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.27746947835738067,
|
| 45 |
+
"grad_norm": 4.90625,
|
| 46 |
+
"learning_rate": 4.5375508694043653e-05,
|
| 47 |
+
"loss": 1.0511,
|
| 48 |
+
"mean_token_accuracy": 0.7330576995015144,
|
| 49 |
+
"step": 250
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.33296337402885684,
|
| 53 |
+
"grad_norm": 9.125,
|
| 54 |
+
"learning_rate": 4.445061043285239e-05,
|
| 55 |
+
"loss": 0.9694,
|
| 56 |
+
"mean_token_accuracy": 0.7509057518839836,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.38845726970033295,
|
| 61 |
+
"grad_norm": 6.9375,
|
| 62 |
+
"learning_rate": 4.352571217166112e-05,
|
| 63 |
+
"loss": 1.029,
|
| 64 |
+
"mean_token_accuracy": 0.7393789207935333,
|
| 65 |
+
"step": 350
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.4439511653718091,
|
| 69 |
+
"grad_norm": 9.25,
|
| 70 |
+
"learning_rate": 4.260081391046985e-05,
|
| 71 |
+
"loss": 0.9199,
|
| 72 |
+
"mean_token_accuracy": 0.760094360858202,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.49944506104328523,
|
| 77 |
+
"grad_norm": 9.125,
|
| 78 |
+
"learning_rate": 4.167591564927858e-05,
|
| 79 |
+
"loss": 0.9908,
|
| 80 |
+
"mean_token_accuracy": 0.7447640253603458,
|
| 81 |
+
"step": 450
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.5549389567147613,
|
| 85 |
+
"grad_norm": 8.25,
|
| 86 |
+
"learning_rate": 4.075101738808732e-05,
|
| 87 |
+
"loss": 0.8984,
|
| 88 |
+
"mean_token_accuracy": 0.7630760133266449,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.6104328523862376,
|
| 93 |
+
"grad_norm": 9.1875,
|
| 94 |
+
"learning_rate": 3.9826119126896046e-05,
|
| 95 |
+
"loss": 0.9407,
|
| 96 |
+
"mean_token_accuracy": 0.757173229753971,
|
| 97 |
+
"step": 550
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.6659267480577137,
|
| 101 |
+
"grad_norm": 6.34375,
|
| 102 |
+
"learning_rate": 3.890122086570477e-05,
|
| 103 |
+
"loss": 0.9575,
|
| 104 |
+
"mean_token_accuracy": 0.7462678501009941,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.7214206437291898,
|
| 109 |
+
"grad_norm": 7.40625,
|
| 110 |
+
"learning_rate": 3.79763226045135e-05,
|
| 111 |
+
"loss": 0.8748,
|
| 112 |
+
"mean_token_accuracy": 0.7648530495911836,
|
| 113 |
+
"step": 650
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.7769145394006659,
|
| 117 |
+
"grad_norm": 5.90625,
|
| 118 |
+
"learning_rate": 3.705142434332224e-05,
|
| 119 |
+
"loss": 0.9121,
|
| 120 |
+
"mean_token_accuracy": 0.7616761503368616,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.832408435072142,
|
| 125 |
+
"grad_norm": 5.21875,
|
| 126 |
+
"learning_rate": 3.612652608213097e-05,
|
| 127 |
+
"loss": 0.8764,
|
| 128 |
+
"mean_token_accuracy": 0.7672346414625645,
|
| 129 |
+
"step": 750
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.8879023307436182,
|
| 133 |
+
"grad_norm": 5.5625,
|
| 134 |
+
"learning_rate": 3.5201627820939697e-05,
|
| 135 |
+
"loss": 0.8334,
|
| 136 |
+
"mean_token_accuracy": 0.7711550496518612,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.9433962264150944,
|
| 141 |
+
"grad_norm": 4.90625,
|
| 142 |
+
"learning_rate": 3.4276729559748424e-05,
|
| 143 |
+
"loss": 0.8035,
|
| 144 |
+
"mean_token_accuracy": 0.7846515120565891,
|
| 145 |
+
"step": 850
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.9988901220865705,
|
| 149 |
+
"grad_norm": 4.90625,
|
| 150 |
+
"learning_rate": 3.3351831298557165e-05,
|
| 151 |
+
"loss": 0.784,
|
| 152 |
+
"mean_token_accuracy": 0.7859167304635047,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.0543840177580466,
|
| 157 |
+
"grad_norm": 6.34375,
|
| 158 |
+
"learning_rate": 3.242693303736589e-05,
|
| 159 |
+
"loss": 0.4242,
|
| 160 |
+
"mean_token_accuracy": 0.8643624025583267,
|
| 161 |
+
"step": 950
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.1098779134295227,
|
| 165 |
+
"grad_norm": 6.46875,
|
| 166 |
+
"learning_rate": 3.150203477617462e-05,
|
| 167 |
+
"loss": 0.4209,
|
| 168 |
+
"mean_token_accuracy": 0.8601007984578609,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.1653718091009988,
|
| 173 |
+
"grad_norm": 3.734375,
|
| 174 |
+
"learning_rate": 3.0577136514983354e-05,
|
| 175 |
+
"loss": 0.4225,
|
| 176 |
+
"mean_token_accuracy": 0.8565894016623496,
|
| 177 |
+
"step": 1050
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.220865704772475,
|
| 181 |
+
"grad_norm": 8.3125,
|
| 182 |
+
"learning_rate": 2.9652238253792085e-05,
|
| 183 |
+
"loss": 0.3897,
|
| 184 |
+
"mean_token_accuracy": 0.8680702060461044,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.2763596004439512,
|
| 189 |
+
"grad_norm": 4.03125,
|
| 190 |
+
"learning_rate": 2.8727339992600816e-05,
|
| 191 |
+
"loss": 0.43,
|
| 192 |
+
"mean_token_accuracy": 0.8627835646271705,
|
| 193 |
+
"step": 1150
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.3318534961154274,
|
| 197 |
+
"grad_norm": 4.09375,
|
| 198 |
+
"learning_rate": 2.7802441731409544e-05,
|
| 199 |
+
"loss": 0.4254,
|
| 200 |
+
"mean_token_accuracy": 0.8601828774809838,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.3873473917869035,
|
| 205 |
+
"grad_norm": 3.953125,
|
| 206 |
+
"learning_rate": 2.6877543470218275e-05,
|
| 207 |
+
"loss": 0.404,
|
| 208 |
+
"mean_token_accuracy": 0.8627345436811447,
|
| 209 |
+
"step": 1250
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.4428412874583796,
|
| 213 |
+
"grad_norm": 3.796875,
|
| 214 |
+
"learning_rate": 2.595264520902701e-05,
|
| 215 |
+
"loss": 0.406,
|
| 216 |
+
"mean_token_accuracy": 0.8670766687393189,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.4983351831298557,
|
| 221 |
+
"grad_norm": 7.875,
|
| 222 |
+
"learning_rate": 2.502774694783574e-05,
|
| 223 |
+
"loss": 0.4123,
|
| 224 |
+
"mean_token_accuracy": 0.8563414543867112,
|
| 225 |
+
"step": 1350
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.5538290788013318,
|
| 229 |
+
"grad_norm": 4.5625,
|
| 230 |
+
"learning_rate": 2.410284868664447e-05,
|
| 231 |
+
"loss": 0.3906,
|
| 232 |
+
"mean_token_accuracy": 0.8690173977613449,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.609322974472808,
|
| 237 |
+
"grad_norm": 3.828125,
|
| 238 |
+
"learning_rate": 2.31779504254532e-05,
|
| 239 |
+
"loss": 0.3896,
|
| 240 |
+
"mean_token_accuracy": 0.8703913018107414,
|
| 241 |
+
"step": 1450
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 1.6648168701442843,
|
| 245 |
+
"grad_norm": 4.03125,
|
| 246 |
+
"learning_rate": 2.2253052164261932e-05,
|
| 247 |
+
"loss": 0.3916,
|
| 248 |
+
"mean_token_accuracy": 0.8693778309226036,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.7203107658157601,
|
| 253 |
+
"grad_norm": 4.4375,
|
| 254 |
+
"learning_rate": 2.1328153903070663e-05,
|
| 255 |
+
"loss": 0.3955,
|
| 256 |
+
"mean_token_accuracy": 0.8697583265602589,
|
| 257 |
+
"step": 1550
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.7758046614872365,
|
| 261 |
+
"grad_norm": 5.09375,
|
| 262 |
+
"learning_rate": 2.0403255641879394e-05,
|
| 263 |
+
"loss": 0.4101,
|
| 264 |
+
"mean_token_accuracy": 0.8629062753915787,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.8312985571587126,
|
| 269 |
+
"grad_norm": 3.59375,
|
| 270 |
+
"learning_rate": 1.9478357380688125e-05,
|
| 271 |
+
"loss": 0.3577,
|
| 272 |
+
"mean_token_accuracy": 0.8781416714191437,
|
| 273 |
+
"step": 1650
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.8867924528301887,
|
| 277 |
+
"grad_norm": 2.609375,
|
| 278 |
+
"learning_rate": 1.8553459119496856e-05,
|
| 279 |
+
"loss": 0.3991,
|
| 280 |
+
"mean_token_accuracy": 0.8602015821635723,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.9422863485016648,
|
| 285 |
+
"grad_norm": 3.5,
|
| 286 |
+
"learning_rate": 1.7628560858305587e-05,
|
| 287 |
+
"loss": 0.3717,
|
| 288 |
+
"mean_token_accuracy": 0.8744880908727646,
|
| 289 |
+
"step": 1750
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.997780244173141,
|
| 293 |
+
"grad_norm": 3.671875,
|
| 294 |
+
"learning_rate": 1.670366259711432e-05,
|
| 295 |
+
"loss": 0.3798,
|
| 296 |
+
"mean_token_accuracy": 0.8699861750006675,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.0532741398446173,
|
| 301 |
+
"grad_norm": 4.3125,
|
| 302 |
+
"learning_rate": 1.577876433592305e-05,
|
| 303 |
+
"loss": 0.156,
|
| 304 |
+
"mean_token_accuracy": 0.9473638749122619,
|
| 305 |
+
"step": 1850
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 2.108768035516093,
|
| 309 |
+
"grad_norm": 3.71875,
|
| 310 |
+
"learning_rate": 1.4853866074731781e-05,
|
| 311 |
+
"loss": 0.1442,
|
| 312 |
+
"mean_token_accuracy": 0.9488543626666069,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.1642619311875695,
|
| 317 |
+
"grad_norm": 4.96875,
|
| 318 |
+
"learning_rate": 1.392896781354051e-05,
|
| 319 |
+
"loss": 0.1507,
|
| 320 |
+
"mean_token_accuracy": 0.9458903276920319,
|
| 321 |
+
"step": 1950
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.2197558268590454,
|
| 325 |
+
"grad_norm": 3.734375,
|
| 326 |
+
"learning_rate": 1.3004069552349243e-05,
|
| 327 |
+
"loss": 0.1485,
|
| 328 |
+
"mean_token_accuracy": 0.9459331405162811,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 2.2752497225305217,
|
| 333 |
+
"grad_norm": 4.40625,
|
| 334 |
+
"learning_rate": 1.2079171291157974e-05,
|
| 335 |
+
"loss": 0.137,
|
| 336 |
+
"mean_token_accuracy": 0.946778584420681,
|
| 337 |
+
"step": 2050
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 2.3307436182019976,
|
| 341 |
+
"grad_norm": 4.65625,
|
| 342 |
+
"learning_rate": 1.1154273029966705e-05,
|
| 343 |
+
"loss": 0.1302,
|
| 344 |
+
"mean_token_accuracy": 0.9501716086268425,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.386237513873474,
|
| 349 |
+
"grad_norm": 3.875,
|
| 350 |
+
"learning_rate": 1.0229374768775436e-05,
|
| 351 |
+
"loss": 0.1487,
|
| 352 |
+
"mean_token_accuracy": 0.9474012264609337,
|
| 353 |
+
"step": 2150
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 2.44173140954495,
|
| 357 |
+
"grad_norm": 6.1875,
|
| 358 |
+
"learning_rate": 9.304476507584166e-06,
|
| 359 |
+
"loss": 0.1371,
|
| 360 |
+
"mean_token_accuracy": 0.9514653950929641,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.497225305216426,
|
| 365 |
+
"grad_norm": 6.1875,
|
| 366 |
+
"learning_rate": 8.379578246392897e-06,
|
| 367 |
+
"loss": 0.1384,
|
| 368 |
+
"mean_token_accuracy": 0.9439902834594249,
|
| 369 |
+
"step": 2250
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.5527192008879025,
|
| 373 |
+
"grad_norm": 2.71875,
|
| 374 |
+
"learning_rate": 7.454679985201627e-06,
|
| 375 |
+
"loss": 0.1357,
|
| 376 |
+
"mean_token_accuracy": 0.9507204574346543,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.6082130965593784,
|
| 381 |
+
"grad_norm": 5.71875,
|
| 382 |
+
"learning_rate": 6.529781724010358e-06,
|
| 383 |
+
"loss": 0.1284,
|
| 384 |
+
"mean_token_accuracy": 0.9505923983454704,
|
| 385 |
+
"step": 2350
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 2.6637069922308547,
|
| 389 |
+
"grad_norm": 2.90625,
|
| 390 |
+
"learning_rate": 5.60488346281909e-06,
|
| 391 |
+
"loss": 0.1327,
|
| 392 |
+
"mean_token_accuracy": 0.9514957949519157,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 2.7192008879023306,
|
| 397 |
+
"grad_norm": 3.796875,
|
| 398 |
+
"learning_rate": 4.679985201627821e-06,
|
| 399 |
+
"loss": 0.1422,
|
| 400 |
+
"mean_token_accuracy": 0.950286613702774,
|
| 401 |
+
"step": 2450
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 2.774694783573807,
|
| 405 |
+
"grad_norm": 4.03125,
|
| 406 |
+
"learning_rate": 3.7550869404365522e-06,
|
| 407 |
+
"loss": 0.1351,
|
| 408 |
+
"mean_token_accuracy": 0.9470736330747604,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.830188679245283,
|
| 413 |
+
"grad_norm": 5.71875,
|
| 414 |
+
"learning_rate": 2.830188679245283e-06,
|
| 415 |
+
"loss": 0.1247,
|
| 416 |
+
"mean_token_accuracy": 0.9515706104040146,
|
| 417 |
+
"step": 2550
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.885682574916759,
|
| 421 |
+
"grad_norm": 4.28125,
|
| 422 |
+
"learning_rate": 1.9052904180540142e-06,
|
| 423 |
+
"loss": 0.1368,
|
| 424 |
+
"mean_token_accuracy": 0.9486922469735145,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 2.9411764705882355,
|
| 429 |
+
"grad_norm": 3.796875,
|
| 430 |
+
"learning_rate": 9.80392156862745e-07,
|
| 431 |
+
"loss": 0.1282,
|
| 432 |
+
"mean_token_accuracy": 0.9518065723776817,
|
| 433 |
+
"step": 2650
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.9966703662597114,
|
| 437 |
+
"grad_norm": 4.15625,
|
| 438 |
+
"learning_rate": 5.549389567147614e-08,
|
| 439 |
+
"loss": 0.1399,
|
| 440 |
+
"mean_token_accuracy": 0.9513461437821388,
|
| 441 |
+
"step": 2700
|
| 442 |
+
}
|
| 443 |
+
],
|
| 444 |
+
"logging_steps": 50,
|
| 445 |
+
"max_steps": 2703,
|
| 446 |
+
"num_input_tokens_seen": 0,
|
| 447 |
+
"num_train_epochs": 3,
|
| 448 |
+
"save_steps": 500,
|
| 449 |
+
"stateful_callbacks": {
|
| 450 |
+
"TrainerControl": {
|
| 451 |
+
"args": {
|
| 452 |
+
"should_epoch_stop": false,
|
| 453 |
+
"should_evaluate": false,
|
| 454 |
+
"should_log": false,
|
| 455 |
+
"should_save": true,
|
| 456 |
+
"should_training_stop": true
|
| 457 |
+
},
|
| 458 |
+
"attributes": {}
|
| 459 |
+
}
|
| 460 |
+
},
|
| 461 |
+
"total_flos": 3.0206819147904e+16,
|
| 462 |
+
"train_batch_size": 2,
|
| 463 |
+
"trial_name": null,
|
| 464 |
+
"trial_params": null
|
| 465 |
+
}
|