Commit
·
88e273b
1
Parent(s):
b165d22
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": 0.9125000238418579,
|
| 13 |
+
"eval_loss": 0.2230108231306076,
|
| 14 |
+
"eval_runtime": 5.588,
|
| 15 |
+
"eval_samples_per_second": 200.429,
|
| 16 |
+
"eval_steps_per_second": 25.054,
|
| 17 |
+
"step": 280
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.79,
|
| 21 |
+
"learning_rate": 2.131486410954287e-05,
|
| 22 |
+
"loss": 0.3039,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.0,
|
| 27 |
+
"eval_accuracy": 0.9071428775787354,
|
| 28 |
+
"eval_loss": 0.26908189058303833,
|
| 29 |
+
"eval_runtime": 5.6261,
|
| 30 |
+
"eval_samples_per_second": 199.073,
|
| 31 |
+
"eval_steps_per_second": 24.884,
|
| 32 |
+
"step": 560
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.0,
|
| 36 |
+
"eval_accuracy": 0.9160714149475098,
|
| 37 |
+
"eval_loss": 0.28345081210136414,
|
| 38 |
+
"eval_runtime": 5.3742,
|
| 39 |
+
"eval_samples_per_second": 208.405,
|
| 40 |
+
"eval_steps_per_second": 26.051,
|
| 41 |
+
"step": 840
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 3.57,
|
| 45 |
+
"learning_rate": 1.9225171549783764e-05,
|
| 46 |
+
"loss": 0.104,
|
| 47 |
+
"step": 1000
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 4.0,
|
| 51 |
+
"eval_accuracy": 0.9241071343421936,
|
| 52 |
+
"eval_loss": 0.39826202392578125,
|
| 53 |
+
"eval_runtime": 5.6203,
|
| 54 |
+
"eval_samples_per_second": 199.277,
|
| 55 |
+
"eval_steps_per_second": 24.91,
|
| 56 |
+
"step": 1120
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 5.0,
|
| 60 |
+
"eval_accuracy": 0.9178571701049805,
|
| 61 |
+
"eval_loss": 0.463891863822937,
|
| 62 |
+
"eval_runtime": 5.6398,
|
| 63 |
+
"eval_samples_per_second": 198.588,
|
| 64 |
+
"eval_steps_per_second": 24.823,
|
| 65 |
+
"step": 1400
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 5.36,
|
| 69 |
+
"learning_rate": 1.713547899002466e-05,
|
| 70 |
+
"loss": 0.0361,
|
| 71 |
+
"step": 1500
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 6.0,
|
| 75 |
+
"eval_accuracy": 0.9205357432365417,
|
| 76 |
+
"eval_loss": 0.46613049507141113,
|
| 77 |
+
"eval_runtime": 5.5849,
|
| 78 |
+
"eval_samples_per_second": 200.541,
|
| 79 |
+
"eval_steps_per_second": 25.068,
|
| 80 |
+
"step": 1680
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 7.0,
|
| 84 |
+
"eval_accuracy": 0.9258928298950195,
|
| 85 |
+
"eval_loss": 0.4626975357532501,
|
| 86 |
+
"eval_runtime": 5.6519,
|
| 87 |
+
"eval_samples_per_second": 198.163,
|
| 88 |
+
"eval_steps_per_second": 24.77,
|
| 89 |
+
"step": 1960
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 7.14,
|
| 93 |
+
"learning_rate": 1.5045786430265557e-05,
|
| 94 |
+
"loss": 0.0228,
|
| 95 |
+
"step": 2000
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 8.0,
|
| 99 |
+
"eval_accuracy": 0.9258928298950195,
|
| 100 |
+
"eval_loss": 0.4725589156150818,
|
| 101 |
+
"eval_runtime": 5.5505,
|
| 102 |
+
"eval_samples_per_second": 201.783,
|
| 103 |
+
"eval_steps_per_second": 25.223,
|
| 104 |
+
"step": 2240
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 8.93,
|
| 108 |
+
"learning_rate": 1.2956093870506452e-05,
|
| 109 |
+
"loss": 0.0158,
|
| 110 |
+
"step": 2500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 9.0,
|
| 114 |
+
"eval_accuracy": 0.925000011920929,
|
| 115 |
+
"eval_loss": 0.5453544855117798,
|
| 116 |
+
"eval_runtime": 5.5094,
|
| 117 |
+
"eval_samples_per_second": 203.288,
|
| 118 |
+
"eval_steps_per_second": 25.411,
|
| 119 |
+
"step": 2520
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 10.0,
|
| 123 |
+
"eval_accuracy": 0.9160714149475098,
|
| 124 |
+
"eval_loss": 0.5454272627830505,
|
| 125 |
+
"eval_runtime": 5.4705,
|
| 126 |
+
"eval_samples_per_second": 204.734,
|
| 127 |
+
"eval_steps_per_second": 25.592,
|
| 128 |
+
"step": 2800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 10.71,
|
| 132 |
+
"learning_rate": 1.0866401310747346e-05,
|
| 133 |
+
"loss": 0.0146,
|
| 134 |
+
"step": 3000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 11.0,
|
| 138 |
+
"eval_accuracy": 0.9196428656578064,
|
| 139 |
+
"eval_loss": 0.5470275282859802,
|
| 140 |
+
"eval_runtime": 5.5797,
|
| 141 |
+
"eval_samples_per_second": 200.729,
|
| 142 |
+
"eval_steps_per_second": 25.091,
|
| 143 |
+
"step": 3080
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 12.0,
|
| 147 |
+
"eval_accuracy": 0.9178571701049805,
|
| 148 |
+
"eval_loss": 0.6207221746444702,
|
| 149 |
+
"eval_runtime": 5.4687,
|
| 150 |
+
"eval_samples_per_second": 204.802,
|
| 151 |
+
"eval_steps_per_second": 25.6,
|
| 152 |
+
"step": 3360
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 12.5,
|
| 156 |
+
"learning_rate": 8.776708750988241e-06,
|
| 157 |
+
"loss": 0.0106,
|
| 158 |
+
"step": 3500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 13.0,
|
| 162 |
+
"eval_accuracy": 0.9142857193946838,
|
| 163 |
+
"eval_loss": 0.6711627840995789,
|
| 164 |
+
"eval_runtime": 5.5881,
|
| 165 |
+
"eval_samples_per_second": 200.427,
|
| 166 |
+
"eval_steps_per_second": 25.053,
|
| 167 |
+
"step": 3640
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 14.0,
|
| 171 |
+
"eval_accuracy": 0.918749988079071,
|
| 172 |
+
"eval_loss": 0.6122187972068787,
|
| 173 |
+
"eval_runtime": 5.7115,
|
| 174 |
+
"eval_samples_per_second": 196.094,
|
| 175 |
+
"eval_steps_per_second": 24.512,
|
| 176 |
+
"step": 3920
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 14.29,
|
| 180 |
+
"learning_rate": 6.687016191229136e-06,
|
| 181 |
+
"loss": 0.0101,
|
| 182 |
+
"step": 4000
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 15.0,
|
| 186 |
+
"eval_accuracy": 0.9196428656578064,
|
| 187 |
+
"eval_loss": 0.6081933975219727,
|
| 188 |
+
"eval_runtime": 5.6756,
|
| 189 |
+
"eval_samples_per_second": 197.336,
|
| 190 |
+
"eval_steps_per_second": 24.667,
|
| 191 |
+
"step": 4200
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 16.0,
|
| 195 |
+
"eval_accuracy": 0.9214285612106323,
|
| 196 |
+
"eval_loss": 0.6368692517280579,
|
| 197 |
+
"eval_runtime": 5.4708,
|
| 198 |
+
"eval_samples_per_second": 204.723,
|
| 199 |
+
"eval_steps_per_second": 25.59,
|
| 200 |
+
"step": 4480
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 16.07,
|
| 204 |
+
"learning_rate": 4.59732363147003e-06,
|
| 205 |
+
"loss": 0.0108,
|
| 206 |
+
"step": 4500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 17.0,
|
| 210 |
+
"eval_accuracy": 0.9205357432365417,
|
| 211 |
+
"eval_loss": 0.6168231964111328,
|
| 212 |
+
"eval_runtime": 5.8239,
|
| 213 |
+
"eval_samples_per_second": 192.312,
|
| 214 |
+
"eval_steps_per_second": 24.039,
|
| 215 |
+
"step": 4760
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 17.86,
|
| 219 |
+
"learning_rate": 2.507631071710926e-06,
|
| 220 |
+
"loss": 0.0069,
|
| 221 |
+
"step": 5000
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 18.0,
|
| 225 |
+
"eval_accuracy": 0.9196428656578064,
|
| 226 |
+
"eval_loss": 0.6445040106773376,
|
| 227 |
+
"eval_runtime": 5.6834,
|
| 228 |
+
"eval_samples_per_second": 197.066,
|
| 229 |
+
"eval_steps_per_second": 24.633,
|
| 230 |
+
"step": 5040
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 19.0,
|
| 234 |
+
"eval_accuracy": 0.9178571701049805,
|
| 235 |
+
"eval_loss": 0.6451646089553833,
|
| 236 |
+
"eval_runtime": 5.6811,
|
| 237 |
+
"eval_samples_per_second": 197.145,
|
| 238 |
+
"eval_steps_per_second": 24.643,
|
| 239 |
+
"step": 5320
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 19.64,
|
| 243 |
+
"learning_rate": 4.17938511951821e-07,
|
| 244 |
+
"loss": 0.0097,
|
| 245 |
+
"step": 5500
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 20.0,
|
| 249 |
+
"eval_accuracy": 0.9205357432365417,
|
| 250 |
+
"eval_loss": 0.6383888721466064,
|
| 251 |
+
"eval_runtime": 5.7353,
|
| 252 |
+
"eval_samples_per_second": 195.283,
|
| 253 |
+
"eval_steps_per_second": 24.41,
|
| 254 |
+
"step": 5600
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 20.0,
|
| 258 |
+
"step": 5600,
|
| 259 |
+
"total_flos": 5893687640064000.0,
|
| 260 |
+
"train_loss": 0.04879493878356048,
|
| 261 |
+
"train_runtime": 2058.2783,
|
| 262 |
+
"train_samples_per_second": 43.532,
|
| 263 |
+
"train_steps_per_second": 2.721
|
| 264 |
+
}
|
| 265 |
+
],
|
| 266 |
+
"max_steps": 5600,
|
| 267 |
+
"num_train_epochs": 20,
|
| 268 |
+
"total_flos": 5893687640064000.0,
|
| 269 |
+
"trial_name": null,
|
| 270 |
+
"trial_params": null
|
| 271 |
+
}
|