Commit
•
65dc9a8
1
Parent(s):
ebba215
End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +331 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_loss": 0.28546908497810364,
|
| 4 |
+
"eval_runtime": 36.8636,
|
| 5 |
+
"eval_samples": 1193,
|
| 6 |
+
"eval_samples_per_second": 32.363,
|
| 7 |
+
"eval_steps_per_second": 1.031,
|
| 8 |
+
"eval_wer": 0.240081979212414,
|
| 9 |
+
"train_loss": 1.2572689837944218,
|
| 10 |
+
"train_runtime": 11083.5384,
|
| 11 |
+
"train_samples": 2606,
|
| 12 |
+
"train_samples_per_second": 23.512,
|
| 13 |
+
"train_steps_per_second": 0.74
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_loss": 0.28546908497810364,
|
| 4 |
+
"eval_runtime": 36.8636,
|
| 5 |
+
"eval_samples": 1193,
|
| 6 |
+
"eval_samples_per_second": 32.363,
|
| 7 |
+
"eval_steps_per_second": 1.031,
|
| 8 |
+
"eval_wer": 0.240081979212414
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"train_loss": 1.2572689837944218,
|
| 4 |
+
"train_runtime": 11083.5384,
|
| 5 |
+
"train_samples": 2606,
|
| 6 |
+
"train_samples_per_second": 23.512,
|
| 7 |
+
"train_steps_per_second": 0.74
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 100.0,
|
| 5 |
+
"global_step": 8200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.66,
|
| 12 |
+
"learning_rate": 2.0859999999999997e-05,
|
| 13 |
+
"loss": 6.9294,
|
| 14 |
+
"step": 300
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 6.1,
|
| 18 |
+
"eval_loss": 2.971196413040161,
|
| 19 |
+
"eval_runtime": 36.7267,
|
| 20 |
+
"eval_samples_per_second": 32.483,
|
| 21 |
+
"eval_steps_per_second": 1.035,
|
| 22 |
+
"eval_wer": 1.0,
|
| 23 |
+
"step": 500
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 7.32,
|
| 27 |
+
"learning_rate": 4.1859999999999996e-05,
|
| 28 |
+
"loss": 3.05,
|
| 29 |
+
"step": 600
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 10.98,
|
| 33 |
+
"learning_rate": 6.285999999999999e-05,
|
| 34 |
+
"loss": 2.8305,
|
| 35 |
+
"step": 900
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 12.2,
|
| 39 |
+
"eval_loss": 1.707324504852295,
|
| 40 |
+
"eval_runtime": 37.8146,
|
| 41 |
+
"eval_samples_per_second": 31.549,
|
| 42 |
+
"eval_steps_per_second": 1.005,
|
| 43 |
+
"eval_wer": 0.9478846435368175,
|
| 44 |
+
"step": 1000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 14.63,
|
| 48 |
+
"learning_rate": 6.8075e-05,
|
| 49 |
+
"loss": 2.0414,
|
| 50 |
+
"step": 1200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 18.29,
|
| 54 |
+
"learning_rate": 6.515833333333332e-05,
|
| 55 |
+
"loss": 1.4795,
|
| 56 |
+
"step": 1500
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 18.29,
|
| 60 |
+
"eval_loss": 0.5756120681762695,
|
| 61 |
+
"eval_runtime": 37.4871,
|
| 62 |
+
"eval_samples_per_second": 31.824,
|
| 63 |
+
"eval_steps_per_second": 1.014,
|
| 64 |
+
"eval_wer": 0.6397306397306397,
|
| 65 |
+
"step": 1500
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 21.95,
|
| 69 |
+
"learning_rate": 6.224166666666666e-05,
|
| 70 |
+
"loss": 1.3433,
|
| 71 |
+
"step": 1800
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 24.39,
|
| 75 |
+
"eval_loss": 0.4967685639858246,
|
| 76 |
+
"eval_runtime": 37.9583,
|
| 77 |
+
"eval_samples_per_second": 31.429,
|
| 78 |
+
"eval_steps_per_second": 1.001,
|
| 79 |
+
"eval_wer": 0.5423803249890207,
|
| 80 |
+
"step": 2000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 25.61,
|
| 84 |
+
"learning_rate": 5.9325e-05,
|
| 85 |
+
"loss": 1.2567,
|
| 86 |
+
"step": 2100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 29.27,
|
| 90 |
+
"learning_rate": 5.6408333333333327e-05,
|
| 91 |
+
"loss": 1.1766,
|
| 92 |
+
"step": 2400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 30.49,
|
| 96 |
+
"eval_loss": 0.4184603989124298,
|
| 97 |
+
"eval_runtime": 37.222,
|
| 98 |
+
"eval_samples_per_second": 32.051,
|
| 99 |
+
"eval_steps_per_second": 1.021,
|
| 100 |
+
"eval_wer": 0.4743083003952569,
|
| 101 |
+
"step": 2500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 32.93,
|
| 105 |
+
"learning_rate": 5.349166666666666e-05,
|
| 106 |
+
"loss": 1.0943,
|
| 107 |
+
"step": 2700
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 36.59,
|
| 111 |
+
"learning_rate": 5.0574999999999996e-05,
|
| 112 |
+
"loss": 1.0017,
|
| 113 |
+
"step": 3000
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 36.59,
|
| 117 |
+
"eval_loss": 0.3302731513977051,
|
| 118 |
+
"eval_runtime": 36.8764,
|
| 119 |
+
"eval_samples_per_second": 32.351,
|
| 120 |
+
"eval_steps_per_second": 1.03,
|
| 121 |
+
"eval_wer": 0.35778070560679254,
|
| 122 |
+
"step": 3000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 40.24,
|
| 126 |
+
"learning_rate": 4.7658333333333324e-05,
|
| 127 |
+
"loss": 0.9358,
|
| 128 |
+
"step": 3300
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 42.68,
|
| 132 |
+
"eval_loss": 0.30026641488075256,
|
| 133 |
+
"eval_runtime": 37.9381,
|
| 134 |
+
"eval_samples_per_second": 31.446,
|
| 135 |
+
"eval_steps_per_second": 1.002,
|
| 136 |
+
"eval_wer": 0.3050797833406529,
|
| 137 |
+
"step": 3500
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 43.9,
|
| 141 |
+
"learning_rate": 4.4741666666666665e-05,
|
| 142 |
+
"loss": 0.8798,
|
| 143 |
+
"step": 3600
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 47.56,
|
| 147 |
+
"learning_rate": 4.183472222222222e-05,
|
| 148 |
+
"loss": 0.8358,
|
| 149 |
+
"step": 3900
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 48.78,
|
| 153 |
+
"eval_loss": 0.30453845858573914,
|
| 154 |
+
"eval_runtime": 36.3987,
|
| 155 |
+
"eval_samples_per_second": 32.776,
|
| 156 |
+
"eval_steps_per_second": 1.044,
|
| 157 |
+
"eval_wer": 0.2883911579563753,
|
| 158 |
+
"step": 4000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 51.22,
|
| 162 |
+
"learning_rate": 3.8918055555555554e-05,
|
| 163 |
+
"loss": 0.8059,
|
| 164 |
+
"step": 4200
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 54.88,
|
| 168 |
+
"learning_rate": 3.600138888888889e-05,
|
| 169 |
+
"loss": 0.7647,
|
| 170 |
+
"step": 4500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 54.88,
|
| 174 |
+
"eval_loss": 0.2865545451641083,
|
| 175 |
+
"eval_runtime": 36.6315,
|
| 176 |
+
"eval_samples_per_second": 32.568,
|
| 177 |
+
"eval_steps_per_second": 1.037,
|
| 178 |
+
"eval_wer": 0.2677499634021373,
|
| 179 |
+
"step": 4500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 58.54,
|
| 183 |
+
"learning_rate": 3.3084722222222216e-05,
|
| 184 |
+
"loss": 0.7482,
|
| 185 |
+
"step": 4800
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 60.98,
|
| 189 |
+
"eval_loss": 0.2829342782497406,
|
| 190 |
+
"eval_runtime": 36.4673,
|
| 191 |
+
"eval_samples_per_second": 32.714,
|
| 192 |
+
"eval_steps_per_second": 1.042,
|
| 193 |
+
"eval_wer": 0.2585273020055629,
|
| 194 |
+
"step": 5000
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 62.2,
|
| 198 |
+
"learning_rate": 3.016805555555555e-05,
|
| 199 |
+
"loss": 0.7223,
|
| 200 |
+
"step": 5100
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 65.85,
|
| 204 |
+
"learning_rate": 2.725138888888889e-05,
|
| 205 |
+
"loss": 0.6943,
|
| 206 |
+
"step": 5400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 67.07,
|
| 210 |
+
"eval_loss": 0.2782347798347473,
|
| 211 |
+
"eval_runtime": 37.212,
|
| 212 |
+
"eval_samples_per_second": 32.06,
|
| 213 |
+
"eval_steps_per_second": 1.021,
|
| 214 |
+
"eval_wer": 0.24784072610159566,
|
| 215 |
+
"step": 5500
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 69.51,
|
| 219 |
+
"learning_rate": 2.433472222222222e-05,
|
| 220 |
+
"loss": 0.6758,
|
| 221 |
+
"step": 5700
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 73.17,
|
| 225 |
+
"learning_rate": 2.1418055555555552e-05,
|
| 226 |
+
"loss": 0.6586,
|
| 227 |
+
"step": 6000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 73.17,
|
| 231 |
+
"eval_loss": 0.2911244034767151,
|
| 232 |
+
"eval_runtime": 36.7087,
|
| 233 |
+
"eval_samples_per_second": 32.499,
|
| 234 |
+
"eval_steps_per_second": 1.035,
|
| 235 |
+
"eval_wer": 0.25369638413116674,
|
| 236 |
+
"step": 6000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 76.83,
|
| 240 |
+
"learning_rate": 1.850138888888889e-05,
|
| 241 |
+
"loss": 0.6425,
|
| 242 |
+
"step": 6300
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 79.27,
|
| 246 |
+
"eval_loss": 0.2817089259624481,
|
| 247 |
+
"eval_runtime": 36.6478,
|
| 248 |
+
"eval_samples_per_second": 32.553,
|
| 249 |
+
"eval_steps_per_second": 1.037,
|
| 250 |
+
"eval_wer": 0.24623042014346364,
|
| 251 |
+
"step": 6500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 80.49,
|
| 255 |
+
"learning_rate": 1.558472222222222e-05,
|
| 256 |
+
"loss": 0.6313,
|
| 257 |
+
"step": 6600
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 84.15,
|
| 261 |
+
"learning_rate": 1.2677777777777776e-05,
|
| 262 |
+
"loss": 0.6067,
|
| 263 |
+
"step": 6900
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 85.37,
|
| 267 |
+
"eval_loss": 0.29103317856788635,
|
| 268 |
+
"eval_runtime": 37.7988,
|
| 269 |
+
"eval_samples_per_second": 31.562,
|
| 270 |
+
"eval_steps_per_second": 1.005,
|
| 271 |
+
"eval_wer": 0.24359537403015663,
|
| 272 |
+
"step": 7000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 87.8,
|
| 276 |
+
"learning_rate": 9.761111111111111e-06,
|
| 277 |
+
"loss": 0.6064,
|
| 278 |
+
"step": 7200
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 91.46,
|
| 282 |
+
"learning_rate": 6.844444444444444e-06,
|
| 283 |
+
"loss": 0.5974,
|
| 284 |
+
"step": 7500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 91.46,
|
| 288 |
+
"eval_loss": 0.28752732276916504,
|
| 289 |
+
"eval_runtime": 38.6504,
|
| 290 |
+
"eval_samples_per_second": 30.866,
|
| 291 |
+
"eval_steps_per_second": 0.983,
|
| 292 |
+
"eval_wer": 0.24300980822719953,
|
| 293 |
+
"step": 7500
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 95.12,
|
| 297 |
+
"learning_rate": 3.927777777777777e-06,
|
| 298 |
+
"loss": 0.5812,
|
| 299 |
+
"step": 7800
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 97.56,
|
| 303 |
+
"eval_loss": 0.2852196991443634,
|
| 304 |
+
"eval_runtime": 38.817,
|
| 305 |
+
"eval_samples_per_second": 30.734,
|
| 306 |
+
"eval_steps_per_second": 0.979,
|
| 307 |
+
"eval_wer": 0.23964280486019615,
|
| 308 |
+
"step": 8000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 98.78,
|
| 312 |
+
"learning_rate": 1.011111111111111e-06,
|
| 313 |
+
"loss": 0.5804,
|
| 314 |
+
"step": 8100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 100.0,
|
| 318 |
+
"step": 8200,
|
| 319 |
+
"total_flos": 2.6094292924080824e+19,
|
| 320 |
+
"train_loss": 1.2572689837944218,
|
| 321 |
+
"train_runtime": 11083.5384,
|
| 322 |
+
"train_samples_per_second": 23.512,
|
| 323 |
+
"train_steps_per_second": 0.74
|
| 324 |
+
}
|
| 325 |
+
],
|
| 326 |
+
"max_steps": 8200,
|
| 327 |
+
"num_train_epochs": 100,
|
| 328 |
+
"total_flos": 2.6094292924080824e+19,
|
| 329 |
+
"trial_name": null,
|
| 330 |
+
"trial_params": null
|
| 331 |
+
}
|