Upload checkpoints/checkpoint-3000/trainer_state.json with huggingface_hub
Browse files
checkpoints/checkpoint-3000/trainer_state.json
ADDED
|
@@ -0,0 +1,297 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5802612119808278,
|
| 3 |
+
"best_model_checkpoint": "whisper-tiny-luganda-v2/checkpoint-3000",
|
| 4 |
+
"epoch": 0.6525995214270176,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.021753317380900587,
|
| 13 |
+
"grad_norm": 47.60859680175781,
|
| 14 |
+
"learning_rate": 4.5238095238095235e-06,
|
| 15 |
+
"loss": 9.6365,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04350663476180117,
|
| 20 |
+
"grad_norm": 41.293460845947266,
|
| 21 |
+
"learning_rate": 9.285714285714286e-06,
|
| 22 |
+
"loss": 4.282,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06525995214270176,
|
| 27 |
+
"grad_norm": 29.167564392089844,
|
| 28 |
+
"learning_rate": 1.4047619047619048e-05,
|
| 29 |
+
"loss": 2.8974,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08701326952360235,
|
| 34 |
+
"grad_norm": 30.911415100097656,
|
| 35 |
+
"learning_rate": 1.880952380952381e-05,
|
| 36 |
+
"loss": 2.2716,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.10876658690450294,
|
| 41 |
+
"grad_norm": 25.824464797973633,
|
| 42 |
+
"learning_rate": 2.357142857142857e-05,
|
| 43 |
+
"loss": 2.0498,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10876658690450294,
|
| 48 |
+
"eval_loss": 0.9947264194488525,
|
| 49 |
+
"eval_runtime": 2111.5938,
|
| 50 |
+
"eval_samples_per_second": 6.458,
|
| 51 |
+
"eval_steps_per_second": 1.615,
|
| 52 |
+
"eval_wer": 0.8350676193361011,
|
| 53 |
+
"step": 500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.13051990428540353,
|
| 57 |
+
"grad_norm": 22.513357162475586,
|
| 58 |
+
"learning_rate": 2.8333333333333335e-05,
|
| 59 |
+
"loss": 1.8357,
|
| 60 |
+
"step": 600
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.1522732216663041,
|
| 64 |
+
"grad_norm": 22.64017105102539,
|
| 65 |
+
"learning_rate": 3.309523809523809e-05,
|
| 66 |
+
"loss": 1.7019,
|
| 67 |
+
"step": 700
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.1740265390472047,
|
| 71 |
+
"grad_norm": 22.659271240234375,
|
| 72 |
+
"learning_rate": 3.785714285714286e-05,
|
| 73 |
+
"loss": 1.6458,
|
| 74 |
+
"step": 800
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.19577985642810528,
|
| 78 |
+
"grad_norm": 19.53955841064453,
|
| 79 |
+
"learning_rate": 4.261904761904762e-05,
|
| 80 |
+
"loss": 1.6001,
|
| 81 |
+
"step": 900
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.2175331738090059,
|
| 85 |
+
"grad_norm": 21.89055061340332,
|
| 86 |
+
"learning_rate": 4.738095238095238e-05,
|
| 87 |
+
"loss": 1.5134,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.2175331738090059,
|
| 92 |
+
"eval_loss": 0.7854410409927368,
|
| 93 |
+
"eval_runtime": 2131.7562,
|
| 94 |
+
"eval_samples_per_second": 6.397,
|
| 95 |
+
"eval_steps_per_second": 1.6,
|
| 96 |
+
"eval_wer": 0.7198385688576876,
|
| 97 |
+
"step": 1000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.23928649118990647,
|
| 101 |
+
"grad_norm": 24.862058639526367,
|
| 102 |
+
"learning_rate": 4.988721804511278e-05,
|
| 103 |
+
"loss": 1.421,
|
| 104 |
+
"step": 1100
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.26103980857080705,
|
| 108 |
+
"grad_norm": 18.665178298950195,
|
| 109 |
+
"learning_rate": 4.963659147869674e-05,
|
| 110 |
+
"loss": 1.4254,
|
| 111 |
+
"step": 1200
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.28279312595170764,
|
| 115 |
+
"grad_norm": 16.429445266723633,
|
| 116 |
+
"learning_rate": 4.93859649122807e-05,
|
| 117 |
+
"loss": 1.3466,
|
| 118 |
+
"step": 1300
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.3045464433326082,
|
| 122 |
+
"grad_norm": 24.4464054107666,
|
| 123 |
+
"learning_rate": 4.913533834586466e-05,
|
| 124 |
+
"loss": 1.3342,
|
| 125 |
+
"step": 1400
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.3262997607135088,
|
| 129 |
+
"grad_norm": 14.964367866516113,
|
| 130 |
+
"learning_rate": 4.888471177944862e-05,
|
| 131 |
+
"loss": 1.2494,
|
| 132 |
+
"step": 1500
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.3262997607135088,
|
| 136 |
+
"eval_loss": 0.6914966106414795,
|
| 137 |
+
"eval_runtime": 2154.6429,
|
| 138 |
+
"eval_samples_per_second": 6.329,
|
| 139 |
+
"eval_steps_per_second": 1.583,
|
| 140 |
+
"eval_wer": 0.7817383247509934,
|
| 141 |
+
"step": 1500
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.3480530780944094,
|
| 145 |
+
"grad_norm": 17.042715072631836,
|
| 146 |
+
"learning_rate": 4.863408521303258e-05,
|
| 147 |
+
"loss": 1.2668,
|
| 148 |
+
"step": 1600
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.36980639547530997,
|
| 152 |
+
"grad_norm": 14.82885456085205,
|
| 153 |
+
"learning_rate": 4.838345864661654e-05,
|
| 154 |
+
"loss": 1.2279,
|
| 155 |
+
"step": 1700
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.39155971285621055,
|
| 159 |
+
"grad_norm": 19.486909866333008,
|
| 160 |
+
"learning_rate": 4.81328320802005e-05,
|
| 161 |
+
"loss": 1.2124,
|
| 162 |
+
"step": 1800
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.41331303023711113,
|
| 166 |
+
"grad_norm": 18.727062225341797,
|
| 167 |
+
"learning_rate": 4.7882205513784464e-05,
|
| 168 |
+
"loss": 1.2099,
|
| 169 |
+
"step": 1900
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.4350663476180118,
|
| 173 |
+
"grad_norm": 17.26986312866211,
|
| 174 |
+
"learning_rate": 4.7631578947368424e-05,
|
| 175 |
+
"loss": 1.1587,
|
| 176 |
+
"step": 2000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.4350663476180118,
|
| 180 |
+
"eval_loss": 0.6377026438713074,
|
| 181 |
+
"eval_runtime": 2184.0106,
|
| 182 |
+
"eval_samples_per_second": 6.244,
|
| 183 |
+
"eval_steps_per_second": 1.561,
|
| 184 |
+
"eval_wer": 0.7916005915578284,
|
| 185 |
+
"step": 2000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.45681966499891236,
|
| 189 |
+
"grad_norm": 20.09056282043457,
|
| 190 |
+
"learning_rate": 4.738095238095238e-05,
|
| 191 |
+
"loss": 1.1293,
|
| 192 |
+
"step": 2100
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.47857298237981294,
|
| 196 |
+
"grad_norm": 12.891928672790527,
|
| 197 |
+
"learning_rate": 4.713032581453634e-05,
|
| 198 |
+
"loss": 1.1347,
|
| 199 |
+
"step": 2200
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5003262997607135,
|
| 203 |
+
"grad_norm": 14.460548400878906,
|
| 204 |
+
"learning_rate": 4.68796992481203e-05,
|
| 205 |
+
"loss": 1.1219,
|
| 206 |
+
"step": 2300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5220796171416141,
|
| 210 |
+
"grad_norm": 14.870162963867188,
|
| 211 |
+
"learning_rate": 4.662907268170426e-05,
|
| 212 |
+
"loss": 1.0758,
|
| 213 |
+
"step": 2400
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5438329345225147,
|
| 217 |
+
"grad_norm": 15.290679931640625,
|
| 218 |
+
"learning_rate": 4.637844611528822e-05,
|
| 219 |
+
"loss": 1.0453,
|
| 220 |
+
"step": 2500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5438329345225147,
|
| 224 |
+
"eval_loss": 0.5957969427108765,
|
| 225 |
+
"eval_runtime": 2164.8981,
|
| 226 |
+
"eval_samples_per_second": 6.299,
|
| 227 |
+
"eval_steps_per_second": 1.575,
|
| 228 |
+
"eval_wer": 0.7275804928460703,
|
| 229 |
+
"step": 2500
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.5655862519034153,
|
| 233 |
+
"grad_norm": 14.45456314086914,
|
| 234 |
+
"learning_rate": 4.612781954887218e-05,
|
| 235 |
+
"loss": 1.0498,
|
| 236 |
+
"step": 2600
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.5873395692843159,
|
| 240 |
+
"grad_norm": 14.366482734680176,
|
| 241 |
+
"learning_rate": 4.587719298245614e-05,
|
| 242 |
+
"loss": 1.0699,
|
| 243 |
+
"step": 2700
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.6090928866652164,
|
| 247 |
+
"grad_norm": 20.212724685668945,
|
| 248 |
+
"learning_rate": 4.56265664160401e-05,
|
| 249 |
+
"loss": 1.0341,
|
| 250 |
+
"step": 2800
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.630846204046117,
|
| 254 |
+
"grad_norm": 14.940664291381836,
|
| 255 |
+
"learning_rate": 4.537593984962406e-05,
|
| 256 |
+
"loss": 1.0165,
|
| 257 |
+
"step": 2900
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.6525995214270176,
|
| 261 |
+
"grad_norm": 12.089710235595703,
|
| 262 |
+
"learning_rate": 4.512531328320802e-05,
|
| 263 |
+
"loss": 1.0065,
|
| 264 |
+
"step": 3000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.6525995214270176,
|
| 268 |
+
"eval_loss": 0.5708178281784058,
|
| 269 |
+
"eval_runtime": 2103.4657,
|
| 270 |
+
"eval_samples_per_second": 6.483,
|
| 271 |
+
"eval_steps_per_second": 1.621,
|
| 272 |
+
"eval_wer": 0.5802612119808278,
|
| 273 |
+
"step": 3000
|
| 274 |
+
}
|
| 275 |
+
],
|
| 276 |
+
"logging_steps": 100,
|
| 277 |
+
"max_steps": 21000,
|
| 278 |
+
"num_input_tokens_seen": 0,
|
| 279 |
+
"num_train_epochs": 5,
|
| 280 |
+
"save_steps": 500,
|
| 281 |
+
"stateful_callbacks": {
|
| 282 |
+
"TrainerControl": {
|
| 283 |
+
"args": {
|
| 284 |
+
"should_epoch_stop": false,
|
| 285 |
+
"should_evaluate": false,
|
| 286 |
+
"should_log": false,
|
| 287 |
+
"should_save": true,
|
| 288 |
+
"should_training_stop": false
|
| 289 |
+
},
|
| 290 |
+
"attributes": {}
|
| 291 |
+
}
|
| 292 |
+
},
|
| 293 |
+
"total_flos": 1.18170648576e+18,
|
| 294 |
+
"train_batch_size": 8,
|
| 295 |
+
"trial_name": null,
|
| 296 |
+
"trial_params": null
|
| 297 |
+
}
|