Upload trainer_state.json
Browse files- trainer_state.json +376 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,376 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.2075,
|
| 3 |
+
"best_model_checkpoint": "/content/model_asr/checkpoint-1100",
|
| 4 |
+
"epoch": 80.0,
|
| 5 |
+
"global_step": 1200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.33,
|
| 12 |
+
"learning_rate": 4.600000000000001e-05,
|
| 13 |
+
"loss": 8.6183,
|
| 14 |
+
"step": 50
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 3.33,
|
| 18 |
+
"eval_loss": 4.025334358215332,
|
| 19 |
+
"eval_runtime": 15.3335,
|
| 20 |
+
"eval_samples_per_second": 31.304,
|
| 21 |
+
"eval_steps_per_second": 3.913,
|
| 22 |
+
"eval_wer": 1.0,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 6.67,
|
| 27 |
+
"learning_rate": 9.6e-05,
|
| 28 |
+
"loss": 3.2794,
|
| 29 |
+
"step": 100
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 6.67,
|
| 33 |
+
"eval_loss": 3.0245492458343506,
|
| 34 |
+
"eval_runtime": 15.5923,
|
| 35 |
+
"eval_samples_per_second": 30.784,
|
| 36 |
+
"eval_steps_per_second": 3.848,
|
| 37 |
+
"eval_wer": 1.0,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 10.0,
|
| 42 |
+
"learning_rate": 9.956913093678348e-05,
|
| 43 |
+
"loss": 2.9692,
|
| 44 |
+
"step": 150
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 10.0,
|
| 48 |
+
"eval_loss": 2.8865509033203125,
|
| 49 |
+
"eval_runtime": 15.6178,
|
| 50 |
+
"eval_samples_per_second": 30.734,
|
| 51 |
+
"eval_steps_per_second": 3.842,
|
| 52 |
+
"eval_wer": 1.0,
|
| 53 |
+
"step": 150
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 13.33,
|
| 57 |
+
"learning_rate": 9.813243995692098e-05,
|
| 58 |
+
"loss": 2.8042,
|
| 59 |
+
"step": 200
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 13.33,
|
| 63 |
+
"eval_loss": 2.635584831237793,
|
| 64 |
+
"eval_runtime": 15.4284,
|
| 65 |
+
"eval_samples_per_second": 31.112,
|
| 66 |
+
"eval_steps_per_second": 3.889,
|
| 67 |
+
"eval_wer": 1.0,
|
| 68 |
+
"step": 200
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 16.67,
|
| 72 |
+
"learning_rate": 9.57159113020304e-05,
|
| 73 |
+
"loss": 2.5114,
|
| 74 |
+
"step": 250
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 16.67,
|
| 78 |
+
"eval_loss": 2.116630792617798,
|
| 79 |
+
"eval_runtime": 15.5962,
|
| 80 |
+
"eval_samples_per_second": 30.777,
|
| 81 |
+
"eval_steps_per_second": 3.847,
|
| 82 |
+
"eval_wer": 1.0,
|
| 83 |
+
"step": 250
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 20.0,
|
| 87 |
+
"learning_rate": 9.236873852683213e-05,
|
| 88 |
+
"loss": 1.7677,
|
| 89 |
+
"step": 300
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 20.0,
|
| 93 |
+
"eval_loss": 1.0035134553909302,
|
| 94 |
+
"eval_runtime": 15.4247,
|
| 95 |
+
"eval_samples_per_second": 31.119,
|
| 96 |
+
"eval_steps_per_second": 3.89,
|
| 97 |
+
"eval_wer": 0.89375,
|
| 98 |
+
"step": 300
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 23.33,
|
| 102 |
+
"learning_rate": 8.815906041658002e-05,
|
| 103 |
+
"loss": 0.8802,
|
| 104 |
+
"step": 350
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 23.33,
|
| 108 |
+
"eval_loss": 0.3643040060997009,
|
| 109 |
+
"eval_runtime": 15.6657,
|
| 110 |
+
"eval_samples_per_second": 30.64,
|
| 111 |
+
"eval_steps_per_second": 3.83,
|
| 112 |
+
"eval_wer": 0.5025,
|
| 113 |
+
"step": 350
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 26.67,
|
| 117 |
+
"learning_rate": 8.31725738778896e-05,
|
| 118 |
+
"loss": 0.4318,
|
| 119 |
+
"step": 400
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 26.67,
|
| 123 |
+
"eval_loss": 0.14456388354301453,
|
| 124 |
+
"eval_runtime": 15.4959,
|
| 125 |
+
"eval_samples_per_second": 30.976,
|
| 126 |
+
"eval_steps_per_second": 3.872,
|
| 127 |
+
"eval_wer": 0.3075,
|
| 128 |
+
"step": 400
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 30.0,
|
| 132 |
+
"learning_rate": 7.751078939684885e-05,
|
| 133 |
+
"loss": 0.2275,
|
| 134 |
+
"step": 450
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 30.0,
|
| 138 |
+
"eval_loss": 0.07944890111684799,
|
| 139 |
+
"eval_runtime": 16.2382,
|
| 140 |
+
"eval_samples_per_second": 29.56,
|
| 141 |
+
"eval_steps_per_second": 3.695,
|
| 142 |
+
"eval_wer": 0.25,
|
| 143 |
+
"step": 450
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 33.33,
|
| 147 |
+
"learning_rate": 7.128896457825364e-05,
|
| 148 |
+
"loss": 0.1407,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 33.33,
|
| 153 |
+
"eval_loss": 0.05435527116060257,
|
| 154 |
+
"eval_runtime": 15.4181,
|
| 155 |
+
"eval_samples_per_second": 31.132,
|
| 156 |
+
"eval_steps_per_second": 3.892,
|
| 157 |
+
"eval_wer": 0.22125,
|
| 158 |
+
"step": 500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 36.67,
|
| 162 |
+
"learning_rate": 6.463375783314937e-05,
|
| 163 |
+
"loss": 0.1191,
|
| 164 |
+
"step": 550
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 36.67,
|
| 168 |
+
"eval_loss": 0.046384669840335846,
|
| 169 |
+
"eval_runtime": 15.8291,
|
| 170 |
+
"eval_samples_per_second": 30.324,
|
| 171 |
+
"eval_steps_per_second": 3.79,
|
| 172 |
+
"eval_wer": 0.23375,
|
| 173 |
+
"step": 550
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 40.0,
|
| 177 |
+
"learning_rate": 5.768064997883498e-05,
|
| 178 |
+
"loss": 0.0825,
|
| 179 |
+
"step": 600
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 40.0,
|
| 183 |
+
"eval_loss": 0.03952096030116081,
|
| 184 |
+
"eval_runtime": 15.4288,
|
| 185 |
+
"eval_samples_per_second": 31.111,
|
| 186 |
+
"eval_steps_per_second": 3.889,
|
| 187 |
+
"eval_wer": 0.25625,
|
| 188 |
+
"step": 600
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 43.33,
|
| 192 |
+
"learning_rate": 5.057118624011702e-05,
|
| 193 |
+
"loss": 0.0699,
|
| 194 |
+
"step": 650
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 43.33,
|
| 198 |
+
"eval_loss": 0.03604818880558014,
|
| 199 |
+
"eval_runtime": 15.4828,
|
| 200 |
+
"eval_samples_per_second": 31.002,
|
| 201 |
+
"eval_steps_per_second": 3.875,
|
| 202 |
+
"eval_wer": 0.2125,
|
| 203 |
+
"step": 650
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 46.67,
|
| 207 |
+
"learning_rate": 4.345009479671536e-05,
|
| 208 |
+
"loss": 0.0592,
|
| 209 |
+
"step": 700
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 46.67,
|
| 213 |
+
"eval_loss": 0.02805432118475437,
|
| 214 |
+
"eval_runtime": 15.5578,
|
| 215 |
+
"eval_samples_per_second": 30.853,
|
| 216 |
+
"eval_steps_per_second": 3.857,
|
| 217 |
+
"eval_wer": 0.23875,
|
| 218 |
+
"step": 700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 50.0,
|
| 222 |
+
"learning_rate": 3.646234053488574e-05,
|
| 223 |
+
"loss": 0.0546,
|
| 224 |
+
"step": 750
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 50.0,
|
| 228 |
+
"eval_loss": 0.023771686479449272,
|
| 229 |
+
"eval_runtime": 15.8929,
|
| 230 |
+
"eval_samples_per_second": 30.202,
|
| 231 |
+
"eval_steps_per_second": 3.775,
|
| 232 |
+
"eval_wer": 0.26375,
|
| 233 |
+
"step": 750
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 53.33,
|
| 237 |
+
"learning_rate": 2.9750173980379736e-05,
|
| 238 |
+
"loss": 0.0455,
|
| 239 |
+
"step": 800
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 53.33,
|
| 243 |
+
"eval_loss": 0.021949028596282005,
|
| 244 |
+
"eval_runtime": 15.824,
|
| 245 |
+
"eval_samples_per_second": 30.334,
|
| 246 |
+
"eval_steps_per_second": 3.792,
|
| 247 |
+
"eval_wer": 0.24125,
|
| 248 |
+
"step": 800
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 56.67,
|
| 252 |
+
"learning_rate": 2.3450235487957135e-05,
|
| 253 |
+
"loss": 0.0391,
|
| 254 |
+
"step": 850
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 56.67,
|
| 258 |
+
"eval_loss": 0.020865805447101593,
|
| 259 |
+
"eval_runtime": 16.0515,
|
| 260 |
+
"eval_samples_per_second": 29.904,
|
| 261 |
+
"eval_steps_per_second": 3.738,
|
| 262 |
+
"eval_wer": 0.24375,
|
| 263 |
+
"step": 850
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 60.0,
|
| 267 |
+
"learning_rate": 1.7690773637801294e-05,
|
| 268 |
+
"loss": 0.0367,
|
| 269 |
+
"step": 900
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 60.0,
|
| 273 |
+
"eval_loss": 0.022958721965551376,
|
| 274 |
+
"eval_runtime": 15.4228,
|
| 275 |
+
"eval_samples_per_second": 31.123,
|
| 276 |
+
"eval_steps_per_second": 3.89,
|
| 277 |
+
"eval_wer": 0.21625,
|
| 278 |
+
"step": 900
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 63.33,
|
| 282 |
+
"learning_rate": 1.2589034464264932e-05,
|
| 283 |
+
"loss": 0.0333,
|
| 284 |
+
"step": 950
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 63.33,
|
| 288 |
+
"eval_loss": 0.021822014823555946,
|
| 289 |
+
"eval_runtime": 15.5217,
|
| 290 |
+
"eval_samples_per_second": 30.924,
|
| 291 |
+
"eval_steps_per_second": 3.866,
|
| 292 |
+
"eval_wer": 0.23,
|
| 293 |
+
"step": 950
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 66.67,
|
| 297 |
+
"learning_rate": 8.248874664720375e-06,
|
| 298 |
+
"loss": 0.0327,
|
| 299 |
+
"step": 1000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 66.67,
|
| 303 |
+
"eval_loss": 0.02000792883336544,
|
| 304 |
+
"eval_runtime": 15.8435,
|
| 305 |
+
"eval_samples_per_second": 30.296,
|
| 306 |
+
"eval_steps_per_second": 3.787,
|
| 307 |
+
"eval_wer": 0.21625,
|
| 308 |
+
"step": 1000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 70.0,
|
| 312 |
+
"learning_rate": 4.758647376699032e-06,
|
| 313 |
+
"loss": 0.0306,
|
| 314 |
+
"step": 1050
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 70.0,
|
| 318 |
+
"eval_loss": 0.02014043927192688,
|
| 319 |
+
"eval_runtime": 15.7796,
|
| 320 |
+
"eval_samples_per_second": 30.419,
|
| 321 |
+
"eval_steps_per_second": 3.802,
|
| 322 |
+
"eval_wer": 0.21625,
|
| 323 |
+
"step": 1050
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 73.33,
|
| 327 |
+
"learning_rate": 2.1894035628006517e-06,
|
| 328 |
+
"loss": 0.0269,
|
| 329 |
+
"step": 1100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 73.33,
|
| 333 |
+
"eval_loss": 0.02098565548658371,
|
| 334 |
+
"eval_runtime": 16.1184,
|
| 335 |
+
"eval_samples_per_second": 29.78,
|
| 336 |
+
"eval_steps_per_second": 3.722,
|
| 337 |
+
"eval_wer": 0.2075,
|
| 338 |
+
"step": 1100
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 76.67,
|
| 342 |
+
"learning_rate": 5.934456179888804e-07,
|
| 343 |
+
"loss": 0.0288,
|
| 344 |
+
"step": 1150
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 76.67,
|
| 348 |
+
"eval_loss": 0.021614952012896538,
|
| 349 |
+
"eval_runtime": 15.8598,
|
| 350 |
+
"eval_samples_per_second": 30.265,
|
| 351 |
+
"eval_steps_per_second": 3.783,
|
| 352 |
+
"eval_wer": 0.21625,
|
| 353 |
+
"step": 1150
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 80.0,
|
| 357 |
+
"learning_rate": 3.262643657425679e-09,
|
| 358 |
+
"loss": 0.0289,
|
| 359 |
+
"step": 1200
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 80.0,
|
| 363 |
+
"eval_loss": 0.02152070961892605,
|
| 364 |
+
"eval_runtime": 16.1024,
|
| 365 |
+
"eval_samples_per_second": 29.809,
|
| 366 |
+
"eval_steps_per_second": 3.726,
|
| 367 |
+
"eval_wer": 0.21625,
|
| 368 |
+
"step": 1200
|
| 369 |
+
}
|
| 370 |
+
],
|
| 371 |
+
"max_steps": 1200,
|
| 372 |
+
"num_train_epochs": 80,
|
| 373 |
+
"total_flos": 3.3688058529081743e+18,
|
| 374 |
+
"trial_name": null,
|
| 375 |
+
"trial_params": null
|
| 376 |
+
}
|