Upload trainer_state.json
Browse files- trainer_state.json +289 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,289 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9335728010062837,
|
| 3 |
+
"best_model_checkpoint": "/scicore/home/lauerg/cerque0000/models/nlp-job-ads/gbert-base-ft-edu-redux/checkpoint-1400",
|
| 4 |
+
"epoch": 9.0,
|
| 5 |
+
"global_step": 2664,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.34,
|
| 12 |
+
"eval_loss": 0.31717467308044434,
|
| 13 |
+
"eval_lrap": 0.575480030593511,
|
| 14 |
+
"eval_runtime": 8.2282,
|
| 15 |
+
"eval_samples_per_second": 81.671,
|
| 16 |
+
"eval_steps_per_second": 10.209,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.68,
|
| 21 |
+
"eval_loss": 0.27608808875083923,
|
| 22 |
+
"eval_lrap": 0.7146675513885474,
|
| 23 |
+
"eval_runtime": 8.1478,
|
| 24 |
+
"eval_samples_per_second": 82.476,
|
| 25 |
+
"eval_steps_per_second": 10.31,
|
| 26 |
+
"step": 200
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.01,
|
| 30 |
+
"eval_loss": 0.24148814380168915,
|
| 31 |
+
"eval_lrap": 0.8075810415317497,
|
| 32 |
+
"eval_runtime": 8.0351,
|
| 33 |
+
"eval_samples_per_second": 83.633,
|
| 34 |
+
"eval_steps_per_second": 10.454,
|
| 35 |
+
"step": 300
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 1.35,
|
| 39 |
+
"eval_loss": 0.2186277061700821,
|
| 40 |
+
"eval_lrap": 0.825712521334449,
|
| 41 |
+
"eval_runtime": 8.0683,
|
| 42 |
+
"eval_samples_per_second": 83.289,
|
| 43 |
+
"eval_steps_per_second": 10.411,
|
| 44 |
+
"step": 400
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.69,
|
| 48 |
+
"learning_rate": 2.193243243243243e-05,
|
| 49 |
+
"loss": 0.2693,
|
| 50 |
+
"step": 500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.69,
|
| 54 |
+
"eval_loss": 0.21280047297477722,
|
| 55 |
+
"eval_lrap": 0.8625814785205315,
|
| 56 |
+
"eval_runtime": 8.196,
|
| 57 |
+
"eval_samples_per_second": 81.991,
|
| 58 |
+
"eval_steps_per_second": 10.249,
|
| 59 |
+
"step": 500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 2.03,
|
| 63 |
+
"eval_loss": 0.20358851552009583,
|
| 64 |
+
"eval_lrap": 0.8505382989284863,
|
| 65 |
+
"eval_runtime": 8.1912,
|
| 66 |
+
"eval_samples_per_second": 82.039,
|
| 67 |
+
"eval_steps_per_second": 10.255,
|
| 68 |
+
"step": 600
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 2.36,
|
| 72 |
+
"eval_loss": 0.21679966151714325,
|
| 73 |
+
"eval_lrap": 0.8932179201649483,
|
| 74 |
+
"eval_runtime": 8.0797,
|
| 75 |
+
"eval_samples_per_second": 83.172,
|
| 76 |
+
"eval_steps_per_second": 10.396,
|
| 77 |
+
"step": 700
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 2.7,
|
| 81 |
+
"eval_loss": 0.18326468765735626,
|
| 82 |
+
"eval_lrap": 0.903219248223994,
|
| 83 |
+
"eval_runtime": 8.1375,
|
| 84 |
+
"eval_samples_per_second": 82.58,
|
| 85 |
+
"eval_steps_per_second": 10.323,
|
| 86 |
+
"step": 800
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 3.04,
|
| 90 |
+
"eval_loss": 0.18310672044754028,
|
| 91 |
+
"eval_lrap": 0.8821981117114492,
|
| 92 |
+
"eval_runtime": 8.0697,
|
| 93 |
+
"eval_samples_per_second": 83.274,
|
| 94 |
+
"eval_steps_per_second": 10.409,
|
| 95 |
+
"step": 900
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 3.38,
|
| 99 |
+
"learning_rate": 1.6864864864864864e-05,
|
| 100 |
+
"loss": 0.1665,
|
| 101 |
+
"step": 1000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 3.38,
|
| 105 |
+
"eval_loss": 0.1836722493171692,
|
| 106 |
+
"eval_lrap": 0.9224537287097426,
|
| 107 |
+
"eval_runtime": 8.1462,
|
| 108 |
+
"eval_samples_per_second": 82.493,
|
| 109 |
+
"eval_steps_per_second": 10.312,
|
| 110 |
+
"step": 1000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 3.72,
|
| 114 |
+
"eval_loss": 0.1867215484380722,
|
| 115 |
+
"eval_lrap": 0.9292872160328117,
|
| 116 |
+
"eval_runtime": 8.1758,
|
| 117 |
+
"eval_samples_per_second": 82.194,
|
| 118 |
+
"eval_steps_per_second": 10.274,
|
| 119 |
+
"step": 1100
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 4.05,
|
| 123 |
+
"eval_loss": 0.17954795062541962,
|
| 124 |
+
"eval_lrap": 0.9233198336105406,
|
| 125 |
+
"eval_runtime": 8.1628,
|
| 126 |
+
"eval_samples_per_second": 82.325,
|
| 127 |
+
"eval_steps_per_second": 10.291,
|
| 128 |
+
"step": 1200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 4.39,
|
| 132 |
+
"eval_loss": 0.1790640652179718,
|
| 133 |
+
"eval_lrap": 0.9050188553107904,
|
| 134 |
+
"eval_runtime": 8.1622,
|
| 135 |
+
"eval_samples_per_second": 82.331,
|
| 136 |
+
"eval_steps_per_second": 10.291,
|
| 137 |
+
"step": 1300
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 4.73,
|
| 141 |
+
"eval_loss": 0.18696065247058868,
|
| 142 |
+
"eval_lrap": 0.9335728010062837,
|
| 143 |
+
"eval_runtime": 8.1263,
|
| 144 |
+
"eval_samples_per_second": 82.694,
|
| 145 |
+
"eval_steps_per_second": 10.337,
|
| 146 |
+
"step": 1400
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 5.07,
|
| 150 |
+
"learning_rate": 1.1797297297297297e-05,
|
| 151 |
+
"loss": 0.1146,
|
| 152 |
+
"step": 1500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 5.07,
|
| 156 |
+
"eval_loss": 0.1917356699705124,
|
| 157 |
+
"eval_lrap": 0.9100622279217009,
|
| 158 |
+
"eval_runtime": 8.1497,
|
| 159 |
+
"eval_samples_per_second": 82.457,
|
| 160 |
+
"eval_steps_per_second": 10.307,
|
| 161 |
+
"step": 1500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 5.41,
|
| 165 |
+
"eval_loss": 0.17960630357265472,
|
| 166 |
+
"eval_lrap": 0.9317027532386741,
|
| 167 |
+
"eval_runtime": 8.1448,
|
| 168 |
+
"eval_samples_per_second": 82.507,
|
| 169 |
+
"eval_steps_per_second": 10.313,
|
| 170 |
+
"step": 1600
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 5.74,
|
| 174 |
+
"eval_loss": 0.18476144969463348,
|
| 175 |
+
"eval_lrap": 0.9179141722192571,
|
| 176 |
+
"eval_runtime": 8.0555,
|
| 177 |
+
"eval_samples_per_second": 83.422,
|
| 178 |
+
"eval_steps_per_second": 10.428,
|
| 179 |
+
"step": 1700
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 6.08,
|
| 183 |
+
"eval_loss": 0.18720324337482452,
|
| 184 |
+
"eval_lrap": 0.9076720128841897,
|
| 185 |
+
"eval_runtime": 8.1385,
|
| 186 |
+
"eval_samples_per_second": 82.571,
|
| 187 |
+
"eval_steps_per_second": 10.321,
|
| 188 |
+
"step": 1800
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 6.42,
|
| 192 |
+
"eval_loss": 0.18699432909488678,
|
| 193 |
+
"eval_lrap": 0.924198462000043,
|
| 194 |
+
"eval_runtime": 8.1746,
|
| 195 |
+
"eval_samples_per_second": 82.206,
|
| 196 |
+
"eval_steps_per_second": 10.276,
|
| 197 |
+
"step": 1900
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 6.76,
|
| 201 |
+
"learning_rate": 6.729729729729729e-06,
|
| 202 |
+
"loss": 0.0813,
|
| 203 |
+
"step": 2000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 6.76,
|
| 207 |
+
"eval_loss": 0.17621222138404846,
|
| 208 |
+
"eval_lrap": 0.915026297261488,
|
| 209 |
+
"eval_runtime": 8.0568,
|
| 210 |
+
"eval_samples_per_second": 83.408,
|
| 211 |
+
"eval_steps_per_second": 10.426,
|
| 212 |
+
"step": 2000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 7.09,
|
| 216 |
+
"eval_loss": 0.18799513578414917,
|
| 217 |
+
"eval_lrap": 0.9157933801078312,
|
| 218 |
+
"eval_runtime": 8.2,
|
| 219 |
+
"eval_samples_per_second": 81.951,
|
| 220 |
+
"eval_steps_per_second": 10.244,
|
| 221 |
+
"step": 2100
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 7.43,
|
| 225 |
+
"eval_loss": 0.1881764978170395,
|
| 226 |
+
"eval_lrap": 0.9201832052904669,
|
| 227 |
+
"eval_runtime": 8.1918,
|
| 228 |
+
"eval_samples_per_second": 82.034,
|
| 229 |
+
"eval_steps_per_second": 10.254,
|
| 230 |
+
"step": 2200
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 7.77,
|
| 234 |
+
"eval_loss": 0.1911892592906952,
|
| 235 |
+
"eval_lrap": 0.9248369082537248,
|
| 236 |
+
"eval_runtime": 8.0575,
|
| 237 |
+
"eval_samples_per_second": 83.401,
|
| 238 |
+
"eval_steps_per_second": 10.425,
|
| 239 |
+
"step": 2300
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 8.11,
|
| 243 |
+
"eval_loss": 0.18784251809120178,
|
| 244 |
+
"eval_lrap": 0.9253503927004727,
|
| 245 |
+
"eval_runtime": 8.1128,
|
| 246 |
+
"eval_samples_per_second": 82.832,
|
| 247 |
+
"eval_steps_per_second": 10.354,
|
| 248 |
+
"step": 2400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 8.45,
|
| 252 |
+
"learning_rate": 1.662162162162162e-06,
|
| 253 |
+
"loss": 0.0624,
|
| 254 |
+
"step": 2500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 8.45,
|
| 258 |
+
"eval_loss": 0.19267971813678741,
|
| 259 |
+
"eval_lrap": 0.9211606593694309,
|
| 260 |
+
"eval_runtime": 8.201,
|
| 261 |
+
"eval_samples_per_second": 81.942,
|
| 262 |
+
"eval_steps_per_second": 10.243,
|
| 263 |
+
"step": 2500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 8.78,
|
| 267 |
+
"eval_loss": 0.1919146031141281,
|
| 268 |
+
"eval_lrap": 0.9207750746085017,
|
| 269 |
+
"eval_runtime": 8.267,
|
| 270 |
+
"eval_samples_per_second": 81.287,
|
| 271 |
+
"eval_steps_per_second": 10.161,
|
| 272 |
+
"step": 2600
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 9.0,
|
| 276 |
+
"step": 2664,
|
| 277 |
+
"total_flos": 5605306439961600.0,
|
| 278 |
+
"train_loss": 0.1333099156170636,
|
| 279 |
+
"train_runtime": 980.3632,
|
| 280 |
+
"train_samples_per_second": 21.73,
|
| 281 |
+
"train_steps_per_second": 2.717
|
| 282 |
+
}
|
| 283 |
+
],
|
| 284 |
+
"max_steps": 2664,
|
| 285 |
+
"num_train_epochs": 9,
|
| 286 |
+
"total_flos": 5605306439961600.0,
|
| 287 |
+
"trial_name": null,
|
| 288 |
+
"trial_params": null
|
| 289 |
+
}
|