End of training
Browse files- all_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +302 -15
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 9.37824189087744e+18,
|
| 4 |
+
"train_loss": 0.9114908182779948,
|
| 5 |
+
"train_runtime": 5966.5057,
|
| 6 |
+
"train_samples_per_second": 20.112,
|
| 7 |
+
"train_steps_per_second": 0.629
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 9.37824189087744e+18,
|
| 4 |
+
"train_loss": 0.9114908182779948,
|
| 5 |
+
"train_runtime": 5966.5057,
|
| 6 |
+
"train_samples_per_second": 20.112,
|
| 7 |
+
"train_steps_per_second": 0.629
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,29 +1,316 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint":
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
}
|
| 19 |
],
|
| 20 |
-
"logging_steps":
|
| 21 |
-
"max_steps":
|
| 22 |
"num_input_tokens_seen": 0,
|
| 23 |
-
"num_train_epochs":
|
| 24 |
"save_steps": 500,
|
| 25 |
-
"total_flos":
|
| 26 |
-
"train_batch_size":
|
| 27 |
"trial_name": null,
|
| 28 |
"trial_params": null
|
| 29 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8254,
|
| 3 |
+
"best_model_checkpoint": "imagenet-tiger/checkpoint-3750",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3750,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.08,
|
| 13 |
+
"grad_norm": 11.370553970336914,
|
| 14 |
+
"learning_rate": 9.982464296247522e-05,
|
| 15 |
+
"loss": 1.1275,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.16,
|
| 20 |
+
"grad_norm": 11.340897560119629,
|
| 21 |
+
"learning_rate": 9.929980185352526e-05,
|
| 22 |
+
"loss": 1.1066,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.24,
|
| 27 |
+
"grad_norm": 9.576211929321289,
|
| 28 |
+
"learning_rate": 9.842915805643155e-05,
|
| 29 |
+
"loss": 1.1377,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.32,
|
| 34 |
+
"grad_norm": 10.26335334777832,
|
| 35 |
+
"learning_rate": 9.721881851187406e-05,
|
| 36 |
+
"loss": 1.1071,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.4,
|
| 41 |
+
"grad_norm": 9.477747917175293,
|
| 42 |
+
"learning_rate": 9.567727288213005e-05,
|
| 43 |
+
"loss": 1.1081,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.48,
|
| 48 |
+
"grad_norm": 7.301119804382324,
|
| 49 |
+
"learning_rate": 9.381533400219318e-05,
|
| 50 |
+
"loss": 1.0873,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.56,
|
| 55 |
+
"grad_norm": 11.125399589538574,
|
| 56 |
+
"learning_rate": 9.164606203550497e-05,
|
| 57 |
+
"loss": 1.1031,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.64,
|
| 62 |
+
"grad_norm": 13.228848457336426,
|
| 63 |
+
"learning_rate": 8.9184672866292e-05,
|
| 64 |
+
"loss": 1.0457,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.72,
|
| 69 |
+
"grad_norm": 7.878283500671387,
|
| 70 |
+
"learning_rate": 8.644843137107059e-05,
|
| 71 |
+
"loss": 1.1162,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.8,
|
| 76 |
+
"grad_norm": 5.097326755523682,
|
| 77 |
+
"learning_rate": 8.345653031794292e-05,
|
| 78 |
+
"loss": 1.1242,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.88,
|
| 83 |
+
"grad_norm": 12.475173950195312,
|
| 84 |
+
"learning_rate": 8.022995574311876e-05,
|
| 85 |
+
"loss": 1.1094,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.96,
|
| 90 |
+
"grad_norm": 12.513256072998047,
|
| 91 |
+
"learning_rate": 7.679133974894983e-05,
|
| 92 |
+
"loss": 1.0771,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.0,
|
| 97 |
+
"eval_accuracy": 0.7971,
|
| 98 |
+
"eval_loss": 0.7663370370864868,
|
| 99 |
+
"eval_runtime": 202.227,
|
| 100 |
+
"eval_samples_per_second": 49.449,
|
| 101 |
+
"eval_steps_per_second": 1.548,
|
| 102 |
+
"step": 1250
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.04,
|
| 106 |
+
"grad_norm": 10.96892261505127,
|
| 107 |
+
"learning_rate": 7.316480175599309e-05,
|
| 108 |
+
"loss": 1.1045,
|
| 109 |
+
"step": 1300
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.12,
|
| 113 |
+
"grad_norm": 12.19508171081543,
|
| 114 |
+
"learning_rate": 6.937577932260515e-05,
|
| 115 |
+
"loss": 0.9819,
|
| 116 |
+
"step": 1400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 1.2,
|
| 120 |
+
"grad_norm": 16.93886375427246,
|
| 121 |
+
"learning_rate": 6.545084971874738e-05,
|
| 122 |
+
"loss": 0.967,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 1.28,
|
| 127 |
+
"grad_norm": 12.414902687072754,
|
| 128 |
+
"learning_rate": 6.141754350553279e-05,
|
| 129 |
+
"loss": 0.9287,
|
| 130 |
+
"step": 1600
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.3599999999999999,
|
| 134 |
+
"grad_norm": 16.11931610107422,
|
| 135 |
+
"learning_rate": 5.730415142812059e-05,
|
| 136 |
+
"loss": 0.8632,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.44,
|
| 141 |
+
"grad_norm": 10.42251968383789,
|
| 142 |
+
"learning_rate": 5.313952597646568e-05,
|
| 143 |
+
"loss": 0.9125,
|
| 144 |
+
"step": 1800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.52,
|
| 148 |
+
"grad_norm": 11.504471778869629,
|
| 149 |
+
"learning_rate": 4.895287900583216e-05,
|
| 150 |
+
"loss": 0.9045,
|
| 151 |
+
"step": 1900
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.6,
|
| 155 |
+
"grad_norm": 9.115998268127441,
|
| 156 |
+
"learning_rate": 4.477357683661734e-05,
|
| 157 |
+
"loss": 0.8667,
|
| 158 |
+
"step": 2000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.6800000000000002,
|
| 162 |
+
"grad_norm": 10.520715713500977,
|
| 163 |
+
"learning_rate": 4.063093427071376e-05,
|
| 164 |
+
"loss": 0.8806,
|
| 165 |
+
"step": 2100
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.76,
|
| 169 |
+
"grad_norm": 12.112250328063965,
|
| 170 |
+
"learning_rate": 3.655400896923672e-05,
|
| 171 |
+
"loss": 0.884,
|
| 172 |
+
"step": 2200
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.8399999999999999,
|
| 176 |
+
"grad_norm": 12.806370735168457,
|
| 177 |
+
"learning_rate": 3.257139763390925e-05,
|
| 178 |
+
"loss": 0.807,
|
| 179 |
+
"step": 2300
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.92,
|
| 183 |
+
"grad_norm": 12.071837425231934,
|
| 184 |
+
"learning_rate": 2.8711035421746367e-05,
|
| 185 |
+
"loss": 0.8634,
|
| 186 |
+
"step": 2400
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.0,
|
| 190 |
+
"grad_norm": 9.2099609375,
|
| 191 |
+
"learning_rate": 2.500000000000001e-05,
|
| 192 |
+
"loss": 0.8206,
|
| 193 |
+
"step": 2500
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 2.0,
|
| 197 |
+
"eval_accuracy": 0.8207,
|
| 198 |
+
"eval_loss": 0.6771770119667053,
|
| 199 |
+
"eval_runtime": 200.5917,
|
| 200 |
+
"eval_samples_per_second": 49.853,
|
| 201 |
+
"eval_steps_per_second": 1.56,
|
| 202 |
+
"step": 2500
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 2.08,
|
| 206 |
+
"grad_norm": 9.109099388122559,
|
| 207 |
+
"learning_rate": 2.1464321615778422e-05,
|
| 208 |
+
"loss": 0.7565,
|
| 209 |
+
"step": 2600
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 2.16,
|
| 213 |
+
"grad_norm": 7.973142623901367,
|
| 214 |
+
"learning_rate": 1.8128800512565513e-05,
|
| 215 |
+
"loss": 0.7475,
|
| 216 |
+
"step": 2700
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.24,
|
| 220 |
+
"grad_norm": 13.218464851379395,
|
| 221 |
+
"learning_rate": 1.5016832974331724e-05,
|
| 222 |
+
"loss": 0.7201,
|
| 223 |
+
"step": 2800
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 2.32,
|
| 227 |
+
"grad_norm": 7.9871134757995605,
|
| 228 |
+
"learning_rate": 1.2150247217412186e-05,
|
| 229 |
+
"loss": 0.7449,
|
| 230 |
+
"step": 2900
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 2.4,
|
| 234 |
+
"grad_norm": 7.4544291496276855,
|
| 235 |
+
"learning_rate": 9.549150281252633e-06,
|
| 236 |
+
"loss": 0.7013,
|
| 237 |
+
"step": 3000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 2.48,
|
| 241 |
+
"grad_norm": 8.663176536560059,
|
| 242 |
+
"learning_rate": 7.2317869919746705e-06,
|
| 243 |
+
"loss": 0.7347,
|
| 244 |
+
"step": 3100
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 2.56,
|
| 248 |
+
"grad_norm": 12.630393028259277,
|
| 249 |
+
"learning_rate": 5.214411988029355e-06,
|
| 250 |
+
"loss": 0.7542,
|
| 251 |
+
"step": 3200
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 2.64,
|
| 255 |
+
"grad_norm": 8.864751815795898,
|
| 256 |
+
"learning_rate": 3.511175705587433e-06,
|
| 257 |
+
"loss": 0.6819,
|
| 258 |
+
"step": 3300
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.7199999999999998,
|
| 262 |
+
"grad_norm": 6.5524725914001465,
|
| 263 |
+
"learning_rate": 2.134025123396638e-06,
|
| 264 |
+
"loss": 0.7919,
|
| 265 |
+
"step": 3400
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.8,
|
| 269 |
+
"grad_norm": 7.316872596740723,
|
| 270 |
+
"learning_rate": 1.0926199633097157e-06,
|
| 271 |
+
"loss": 0.6864,
|
| 272 |
+
"step": 3500
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.88,
|
| 276 |
+
"grad_norm": 9.99939250946045,
|
| 277 |
+
"learning_rate": 3.9426493427611177e-07,
|
| 278 |
+
"loss": 0.7364,
|
| 279 |
+
"step": 3600
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.96,
|
| 283 |
+
"grad_norm": 12.940382957458496,
|
| 284 |
+
"learning_rate": 4.385849505708084e-08,
|
| 285 |
+
"loss": 0.7212,
|
| 286 |
+
"step": 3700
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 3.0,
|
| 290 |
+
"eval_accuracy": 0.8254,
|
| 291 |
+
"eval_loss": 0.6595176458358765,
|
| 292 |
+
"eval_runtime": 198.5571,
|
| 293 |
+
"eval_samples_per_second": 50.363,
|
| 294 |
+
"eval_steps_per_second": 1.576,
|
| 295 |
+
"step": 3750
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 3.0,
|
| 299 |
+
"step": 3750,
|
| 300 |
+
"total_flos": 9.37824189087744e+18,
|
| 301 |
+
"train_loss": 0.9114908182779948,
|
| 302 |
+
"train_runtime": 5966.5057,
|
| 303 |
+
"train_samples_per_second": 20.112,
|
| 304 |
+
"train_steps_per_second": 0.629
|
| 305 |
}
|
| 306 |
],
|
| 307 |
+
"logging_steps": 100,
|
| 308 |
+
"max_steps": 3750,
|
| 309 |
"num_input_tokens_seen": 0,
|
| 310 |
+
"num_train_epochs": 3,
|
| 311 |
"save_steps": 500,
|
| 312 |
+
"total_flos": 9.37824189087744e+18,
|
| 313 |
+
"train_batch_size": 32,
|
| 314 |
"trial_name": null,
|
| 315 |
"trial_params": null
|
| 316 |
}
|