Jeremiah Zhou
commited on
Commit
•
1c05f36
1
Parent(s):
e51d2cf
End of training
Browse files- all_results.json +11 -5
- eval_results.json +6 -5
- runs/Jun15_10-53-33_pikachu/events.out.tfevents.1655263882.pikachu.165858.2 +3 -0
- train_results.json +8 -0
- trainer_state.json +259 -0
all_results.json
CHANGED
|
@@ -1,8 +1,14 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
|
|
|
| 5 |
"eval_samples": 872,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9357798165137615,
|
| 4 |
+
"eval_loss": 0.23143097758293152,
|
| 5 |
+
"eval_runtime": 1.8684,
|
| 6 |
"eval_samples": 872,
|
| 7 |
+
"eval_samples_per_second": 466.7,
|
| 8 |
+
"eval_steps_per_second": 58.337,
|
| 9 |
+
"train_loss": 0.19015042130567683,
|
| 10 |
+
"train_runtime": 1734.8381,
|
| 11 |
+
"train_samples": 67349,
|
| 12 |
+
"train_samples_per_second": 388.215,
|
| 13 |
+
"train_steps_per_second": 24.267
|
| 14 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
"
|
| 4 |
-
"
|
|
|
|
| 5 |
"eval_samples": 872,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"eval_accuracy": 0.9357798165137615,
|
| 4 |
+
"eval_loss": 0.23143097758293152,
|
| 5 |
+
"eval_runtime": 1.8684,
|
| 6 |
"eval_samples": 872,
|
| 7 |
+
"eval_samples_per_second": 466.7,
|
| 8 |
+
"eval_steps_per_second": 58.337
|
| 9 |
}
|
runs/Jun15_10-53-33_pikachu/events.out.tfevents.1655263882.pikachu.165858.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d51c6079c2bd67bfc531c6cc3695f0b620ddd81c139390fecc06b4bfbd183785
|
| 3 |
+
size 369
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 0.19015042130567683,
|
| 4 |
+
"train_runtime": 1734.8381,
|
| 5 |
+
"train_samples": 67349,
|
| 6 |
+
"train_samples_per_second": 388.215,
|
| 7 |
+
"train_steps_per_second": 24.267
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,259 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9357798165137615,
|
| 3 |
+
"best_model_checkpoint": "./fine-tune/roberta-base/sst2/checkpoint-4210",
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"global_step": 16840,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.12,
|
| 12 |
+
"learning_rate": 3.9588281868566905e-06,
|
| 13 |
+
"loss": 0.5645,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.24,
|
| 18 |
+
"learning_rate": 7.917656373713381e-06,
|
| 19 |
+
"loss": 0.3021,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.36,
|
| 24 |
+
"learning_rate": 1.1876484560570072e-05,
|
| 25 |
+
"loss": 0.2658,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.48,
|
| 30 |
+
"learning_rate": 1.5835312747426762e-05,
|
| 31 |
+
"loss": 0.2706,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.59,
|
| 36 |
+
"learning_rate": 1.9794140934283453e-05,
|
| 37 |
+
"loss": 0.2508,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.71,
|
| 42 |
+
"learning_rate": 1.9760448779501697e-05,
|
| 43 |
+
"loss": 0.2557,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.83,
|
| 48 |
+
"learning_rate": 1.95077576186385e-05,
|
| 49 |
+
"loss": 0.2578,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.95,
|
| 54 |
+
"learning_rate": 1.925506645777531e-05,
|
| 55 |
+
"loss": 0.2287,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.0,
|
| 60 |
+
"eval_accuracy": 0.9357798165137615,
|
| 61 |
+
"eval_loss": 0.23143097758293152,
|
| 62 |
+
"eval_runtime": 1.9026,
|
| 63 |
+
"eval_samples_per_second": 458.311,
|
| 64 |
+
"eval_steps_per_second": 57.289,
|
| 65 |
+
"step": 4210
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.07,
|
| 69 |
+
"learning_rate": 1.9002375296912114e-05,
|
| 70 |
+
"loss": 0.2229,
|
| 71 |
+
"step": 4500
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 1.19,
|
| 75 |
+
"learning_rate": 1.874968413604892e-05,
|
| 76 |
+
"loss": 0.2084,
|
| 77 |
+
"step": 5000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.31,
|
| 81 |
+
"learning_rate": 1.849699297518573e-05,
|
| 82 |
+
"loss": 0.2047,
|
| 83 |
+
"step": 5500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 1.43,
|
| 87 |
+
"learning_rate": 1.8244301814322537e-05,
|
| 88 |
+
"loss": 0.189,
|
| 89 |
+
"step": 6000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.54,
|
| 93 |
+
"learning_rate": 1.7991610653459345e-05,
|
| 94 |
+
"loss": 0.1914,
|
| 95 |
+
"step": 6500
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 1.66,
|
| 99 |
+
"learning_rate": 1.773891949259615e-05,
|
| 100 |
+
"loss": 0.1874,
|
| 101 |
+
"step": 7000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 1.78,
|
| 105 |
+
"learning_rate": 1.7486228331732958e-05,
|
| 106 |
+
"loss": 0.1942,
|
| 107 |
+
"step": 7500
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.9,
|
| 111 |
+
"learning_rate": 1.7233537170869766e-05,
|
| 112 |
+
"loss": 0.1959,
|
| 113 |
+
"step": 8000
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.0,
|
| 117 |
+
"eval_accuracy": 0.926605504587156,
|
| 118 |
+
"eval_loss": 0.30272433161735535,
|
| 119 |
+
"eval_runtime": 1.8407,
|
| 120 |
+
"eval_samples_per_second": 473.726,
|
| 121 |
+
"eval_steps_per_second": 59.216,
|
| 122 |
+
"step": 8420
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 2.02,
|
| 126 |
+
"learning_rate": 1.698084601000657e-05,
|
| 127 |
+
"loss": 0.1783,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.14,
|
| 132 |
+
"learning_rate": 1.672815484914338e-05,
|
| 133 |
+
"loss": 0.1453,
|
| 134 |
+
"step": 9000
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 2.26,
|
| 138 |
+
"learning_rate": 1.6475463688280183e-05,
|
| 139 |
+
"loss": 0.1481,
|
| 140 |
+
"step": 9500
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 2.38,
|
| 144 |
+
"learning_rate": 1.622277252741699e-05,
|
| 145 |
+
"loss": 0.1564,
|
| 146 |
+
"step": 10000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 2.49,
|
| 150 |
+
"learning_rate": 1.59700813665538e-05,
|
| 151 |
+
"loss": 0.1457,
|
| 152 |
+
"step": 10500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 2.61,
|
| 156 |
+
"learning_rate": 1.5717390205690607e-05,
|
| 157 |
+
"loss": 0.15,
|
| 158 |
+
"step": 11000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 2.73,
|
| 162 |
+
"learning_rate": 1.5464699044827415e-05,
|
| 163 |
+
"loss": 0.157,
|
| 164 |
+
"step": 11500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 2.85,
|
| 168 |
+
"learning_rate": 1.521200788396422e-05,
|
| 169 |
+
"loss": 0.1531,
|
| 170 |
+
"step": 12000
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.97,
|
| 174 |
+
"learning_rate": 1.4959316723101027e-05,
|
| 175 |
+
"loss": 0.1635,
|
| 176 |
+
"step": 12500
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.0,
|
| 180 |
+
"eval_accuracy": 0.930045871559633,
|
| 181 |
+
"eval_loss": 0.30217334628105164,
|
| 182 |
+
"eval_runtime": 1.7605,
|
| 183 |
+
"eval_samples_per_second": 495.309,
|
| 184 |
+
"eval_steps_per_second": 61.914,
|
| 185 |
+
"step": 12630
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 3.09,
|
| 189 |
+
"learning_rate": 1.4706625562237835e-05,
|
| 190 |
+
"loss": 0.1167,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.21,
|
| 195 |
+
"learning_rate": 1.4453934401374641e-05,
|
| 196 |
+
"loss": 0.1147,
|
| 197 |
+
"step": 13500
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 3.33,
|
| 201 |
+
"learning_rate": 1.420124324051145e-05,
|
| 202 |
+
"loss": 0.1123,
|
| 203 |
+
"step": 14000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 3.44,
|
| 207 |
+
"learning_rate": 1.3948552079648254e-05,
|
| 208 |
+
"loss": 0.1173,
|
| 209 |
+
"step": 14500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 3.56,
|
| 213 |
+
"learning_rate": 1.3695860918785062e-05,
|
| 214 |
+
"loss": 0.1105,
|
| 215 |
+
"step": 15000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 3.68,
|
| 219 |
+
"learning_rate": 1.344316975792187e-05,
|
| 220 |
+
"loss": 0.1148,
|
| 221 |
+
"step": 15500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 3.8,
|
| 225 |
+
"learning_rate": 1.3190478597058676e-05,
|
| 226 |
+
"loss": 0.1335,
|
| 227 |
+
"step": 16000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 3.92,
|
| 231 |
+
"learning_rate": 1.2937787436195484e-05,
|
| 232 |
+
"loss": 0.1148,
|
| 233 |
+
"step": 16500
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.0,
|
| 237 |
+
"eval_accuracy": 0.9288990825688074,
|
| 238 |
+
"eval_loss": 0.31621453166007996,
|
| 239 |
+
"eval_runtime": 1.6764,
|
| 240 |
+
"eval_samples_per_second": 520.163,
|
| 241 |
+
"eval_steps_per_second": 65.02,
|
| 242 |
+
"step": 16840
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 4.0,
|
| 246 |
+
"step": 16840,
|
| 247 |
+
"total_flos": 1.772026646744064e+16,
|
| 248 |
+
"train_loss": 0.19015042130567683,
|
| 249 |
+
"train_runtime": 1734.8381,
|
| 250 |
+
"train_samples_per_second": 388.215,
|
| 251 |
+
"train_steps_per_second": 24.267
|
| 252 |
+
}
|
| 253 |
+
],
|
| 254 |
+
"max_steps": 42100,
|
| 255 |
+
"num_train_epochs": 10,
|
| 256 |
+
"total_flos": 1.772026646744064e+16,
|
| 257 |
+
"trial_name": null,
|
| 258 |
+
"trial_params": null
|
| 259 |
+
}
|