Commit
•
5a635a7
1
Parent(s):
77f36f8
Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +271 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,271 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 20.0,
|
| 5 |
+
"global_step": 5600,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"eval_accuracy": 0.8955357074737549,
|
| 13 |
+
"eval_loss": 0.33214062452316284,
|
| 14 |
+
"eval_runtime": 7.5153,
|
| 15 |
+
"eval_samples_per_second": 149.029,
|
| 16 |
+
"eval_steps_per_second": 18.629,
|
| 17 |
+
"step": 280
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.79,
|
| 21 |
+
"learning_rate": 2.743006199096114e-05,
|
| 22 |
+
"loss": 0.3501,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.0,
|
| 27 |
+
"eval_accuracy": 0.9142857193946838,
|
| 28 |
+
"eval_loss": 0.29908323287963867,
|
| 29 |
+
"eval_runtime": 7.7634,
|
| 30 |
+
"eval_samples_per_second": 144.266,
|
| 31 |
+
"eval_steps_per_second": 18.033,
|
| 32 |
+
"step": 560
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.0,
|
| 36 |
+
"eval_accuracy": 0.9008928537368774,
|
| 37 |
+
"eval_loss": 0.2781011164188385,
|
| 38 |
+
"eval_runtime": 7.6341,
|
| 39 |
+
"eval_samples_per_second": 146.71,
|
| 40 |
+
"eval_steps_per_second": 18.339,
|
| 41 |
+
"step": 840
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 3.57,
|
| 45 |
+
"learning_rate": 2.474084022714142e-05,
|
| 46 |
+
"loss": 0.1794,
|
| 47 |
+
"step": 1000
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 4.0,
|
| 51 |
+
"eval_accuracy": 0.9107142686843872,
|
| 52 |
+
"eval_loss": 0.37903639674186707,
|
| 53 |
+
"eval_runtime": 7.6349,
|
| 54 |
+
"eval_samples_per_second": 146.696,
|
| 55 |
+
"eval_steps_per_second": 18.337,
|
| 56 |
+
"step": 1120
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 5.0,
|
| 60 |
+
"eval_accuracy": 0.9214285612106323,
|
| 61 |
+
"eval_loss": 0.40712061524391174,
|
| 62 |
+
"eval_runtime": 7.4748,
|
| 63 |
+
"eval_samples_per_second": 149.837,
|
| 64 |
+
"eval_steps_per_second": 18.73,
|
| 65 |
+
"step": 1400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 5.36,
|
| 69 |
+
"learning_rate": 2.2051618463321703e-05,
|
| 70 |
+
"loss": 0.0871,
|
| 71 |
+
"step": 1500
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 6.0,
|
| 75 |
+
"eval_accuracy": 0.9160714149475098,
|
| 76 |
+
"eval_loss": 0.576979398727417,
|
| 77 |
+
"eval_runtime": 7.7299,
|
| 78 |
+
"eval_samples_per_second": 144.892,
|
| 79 |
+
"eval_steps_per_second": 18.112,
|
| 80 |
+
"step": 1680
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 7.0,
|
| 84 |
+
"eval_accuracy": 0.9116071462631226,
|
| 85 |
+
"eval_loss": 0.6520564556121826,
|
| 86 |
+
"eval_runtime": 7.614,
|
| 87 |
+
"eval_samples_per_second": 147.098,
|
| 88 |
+
"eval_steps_per_second": 18.387,
|
| 89 |
+
"step": 1960
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 7.14,
|
| 93 |
+
"learning_rate": 1.9362396699501985e-05,
|
| 94 |
+
"loss": 0.0437,
|
| 95 |
+
"step": 2000
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 8.0,
|
| 99 |
+
"eval_accuracy": 0.9196428656578064,
|
| 100 |
+
"eval_loss": 0.6204714179039001,
|
| 101 |
+
"eval_runtime": 7.8377,
|
| 102 |
+
"eval_samples_per_second": 142.899,
|
| 103 |
+
"eval_steps_per_second": 17.862,
|
| 104 |
+
"step": 2240
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 8.93,
|
| 108 |
+
"learning_rate": 1.6673174935682264e-05,
|
| 109 |
+
"loss": 0.036,
|
| 110 |
+
"step": 2500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 9.0,
|
| 114 |
+
"eval_accuracy": 0.9169642925262451,
|
| 115 |
+
"eval_loss": 0.5968001484870911,
|
| 116 |
+
"eval_runtime": 7.7266,
|
| 117 |
+
"eval_samples_per_second": 144.953,
|
| 118 |
+
"eval_steps_per_second": 18.119,
|
| 119 |
+
"step": 2520
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 10.0,
|
| 123 |
+
"eval_accuracy": 0.9214285612106323,
|
| 124 |
+
"eval_loss": 0.6099338531494141,
|
| 125 |
+
"eval_runtime": 7.5404,
|
| 126 |
+
"eval_samples_per_second": 148.534,
|
| 127 |
+
"eval_steps_per_second": 18.567,
|
| 128 |
+
"step": 2800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 10.71,
|
| 132 |
+
"learning_rate": 1.3983953171862545e-05,
|
| 133 |
+
"loss": 0.0199,
|
| 134 |
+
"step": 3000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 11.0,
|
| 138 |
+
"eval_accuracy": 0.9169642925262451,
|
| 139 |
+
"eval_loss": 0.6376703381538391,
|
| 140 |
+
"eval_runtime": 7.6099,
|
| 141 |
+
"eval_samples_per_second": 147.177,
|
| 142 |
+
"eval_steps_per_second": 18.397,
|
| 143 |
+
"step": 3080
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 12.0,
|
| 147 |
+
"eval_accuracy": 0.9205357432365417,
|
| 148 |
+
"eval_loss": 0.6757161021232605,
|
| 149 |
+
"eval_runtime": 7.6257,
|
| 150 |
+
"eval_samples_per_second": 146.872,
|
| 151 |
+
"eval_steps_per_second": 18.359,
|
| 152 |
+
"step": 3360
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 12.5,
|
| 156 |
+
"learning_rate": 1.1294731408042823e-05,
|
| 157 |
+
"loss": 0.0198,
|
| 158 |
+
"step": 3500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 13.0,
|
| 162 |
+
"eval_accuracy": 0.9205357432365417,
|
| 163 |
+
"eval_loss": 0.6298871636390686,
|
| 164 |
+
"eval_runtime": 7.5908,
|
| 165 |
+
"eval_samples_per_second": 147.548,
|
| 166 |
+
"eval_steps_per_second": 18.443,
|
| 167 |
+
"step": 3640
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 14.0,
|
| 171 |
+
"eval_accuracy": 0.9125000238418579,
|
| 172 |
+
"eval_loss": 0.693723201751709,
|
| 173 |
+
"eval_runtime": 7.7616,
|
| 174 |
+
"eval_samples_per_second": 144.301,
|
| 175 |
+
"eval_steps_per_second": 18.038,
|
| 176 |
+
"step": 3920
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 14.29,
|
| 180 |
+
"learning_rate": 8.605509644223102e-06,
|
| 181 |
+
"loss": 0.0119,
|
| 182 |
+
"step": 4000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 15.0,
|
| 186 |
+
"eval_accuracy": 0.9169642925262451,
|
| 187 |
+
"eval_loss": 0.6261208653450012,
|
| 188 |
+
"eval_runtime": 7.7494,
|
| 189 |
+
"eval_samples_per_second": 144.528,
|
| 190 |
+
"eval_steps_per_second": 18.066,
|
| 191 |
+
"step": 4200
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 16.0,
|
| 195 |
+
"eval_accuracy": 0.9160714149475098,
|
| 196 |
+
"eval_loss": 0.7174915671348572,
|
| 197 |
+
"eval_runtime": 7.862,
|
| 198 |
+
"eval_samples_per_second": 142.457,
|
| 199 |
+
"eval_steps_per_second": 17.807,
|
| 200 |
+
"step": 4480
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 16.07,
|
| 204 |
+
"learning_rate": 5.916287880403384e-06,
|
| 205 |
+
"loss": 0.0065,
|
| 206 |
+
"step": 4500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 17.0,
|
| 210 |
+
"eval_accuracy": 0.9169642925262451,
|
| 211 |
+
"eval_loss": 0.7007285952568054,
|
| 212 |
+
"eval_runtime": 7.7124,
|
| 213 |
+
"eval_samples_per_second": 145.221,
|
| 214 |
+
"eval_steps_per_second": 18.153,
|
| 215 |
+
"step": 4760
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 17.86,
|
| 219 |
+
"learning_rate": 3.2270661165836636e-06,
|
| 220 |
+
"loss": 0.0069,
|
| 221 |
+
"step": 5000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 18.0,
|
| 225 |
+
"eval_accuracy": 0.918749988079071,
|
| 226 |
+
"eval_loss": 0.7041569352149963,
|
| 227 |
+
"eval_runtime": 7.8136,
|
| 228 |
+
"eval_samples_per_second": 143.339,
|
| 229 |
+
"eval_steps_per_second": 17.917,
|
| 230 |
+
"step": 5040
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 19.0,
|
| 234 |
+
"eval_accuracy": 0.9107142686843872,
|
| 235 |
+
"eval_loss": 0.732833981513977,
|
| 236 |
+
"eval_runtime": 7.5262,
|
| 237 |
+
"eval_samples_per_second": 148.814,
|
| 238 |
+
"eval_steps_per_second": 18.602,
|
| 239 |
+
"step": 5320
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 19.64,
|
| 243 |
+
"learning_rate": 5.378443527639439e-07,
|
| 244 |
+
"loss": 0.0034,
|
| 245 |
+
"step": 5500
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 20.0,
|
| 249 |
+
"eval_accuracy": 0.9116071462631226,
|
| 250 |
+
"eval_loss": 0.7286208271980286,
|
| 251 |
+
"eval_runtime": 7.6221,
|
| 252 |
+
"eval_samples_per_second": 146.941,
|
| 253 |
+
"eval_steps_per_second": 18.368,
|
| 254 |
+
"step": 5600
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 20.0,
|
| 258 |
+
"step": 5600,
|
| 259 |
+
"total_flos": 5893793336524800.0,
|
| 260 |
+
"train_loss": 0.06829224105924368,
|
| 261 |
+
"train_runtime": 4007.6255,
|
| 262 |
+
"train_samples_per_second": 22.357,
|
| 263 |
+
"train_steps_per_second": 1.397
|
| 264 |
+
}
|
| 265 |
+
],
|
| 266 |
+
"max_steps": 5600,
|
| 267 |
+
"num_train_epochs": 20,
|
| 268 |
+
"total_flos": 5893793336524800.0,
|
| 269 |
+
"trial_name": null,
|
| 270 |
+
"trial_params": null
|
| 271 |
+
}
|