End of training
Browse files- README.md +1 -1
- all_results.json +17 -17
- eval_nbest_predictions.json +2 -2
- eval_null_odds.json +0 -0
- eval_predictions.json +0 -0
- eval_results.json +12 -12
- train_results.json +6 -6
- trainer_state.json +1152 -459
README.md
CHANGED
|
@@ -14,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# tiny-clinicalbert-qa
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [nlpie/tiny-clinicalbert](https://huggingface.co/nlpie/tiny-clinicalbert) on the
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 14 |
|
| 15 |
# tiny-clinicalbert-qa
|
| 16 |
|
| 17 |
+
This model is a fine-tuned version of [nlpie/tiny-clinicalbert](https://huggingface.co/nlpie/tiny-clinicalbert) on the Eladio/emrqa-msquad and the rajpurkar/squad_v2 datasets.
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
all_results.json
CHANGED
|
@@ -1,26 +1,26 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_HasAns_exact":
|
| 4 |
-
"eval_HasAns_f1": 86.
|
| 5 |
"eval_HasAns_total": 38667,
|
| 6 |
-
"eval_NoAns_exact":
|
| 7 |
-
"eval_NoAns_f1":
|
| 8 |
"eval_NoAns_total": 5945,
|
| 9 |
-
"eval_best_exact":
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
-
"eval_best_f1": 82.
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
-
"eval_exact":
|
| 14 |
-
"eval_f1": 82.
|
| 15 |
-
"eval_runtime":
|
| 16 |
"eval_samples": 78873,
|
| 17 |
-
"eval_samples_per_second":
|
| 18 |
-
"eval_steps_per_second":
|
| 19 |
"eval_total": 44612,
|
| 20 |
-
"total_flos":
|
| 21 |
-
"train_loss":
|
| 22 |
-
"train_runtime":
|
| 23 |
"train_samples": 397452,
|
| 24 |
-
"train_samples_per_second":
|
| 25 |
-
"train_steps_per_second": 5.
|
| 26 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_HasAns_exact": 78.80363100318101,
|
| 4 |
+
"eval_HasAns_f1": 86.93823568622223,
|
| 5 |
"eval_HasAns_total": 38667,
|
| 6 |
+
"eval_NoAns_exact": 52.95206055508831,
|
| 7 |
+
"eval_NoAns_f1": 52.95206055508831,
|
| 8 |
"eval_NoAns_total": 5945,
|
| 9 |
+
"eval_best_exact": 75.3586478974267,
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
+
"eval_best_f1": 82.40923427058162,
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
+
"eval_exact": 75.3586478974267,
|
| 14 |
+
"eval_f1": 82.4092342705809,
|
| 15 |
+
"eval_runtime": 363.061,
|
| 16 |
"eval_samples": 78873,
|
| 17 |
+
"eval_samples_per_second": 217.244,
|
| 18 |
+
"eval_steps_per_second": 27.158,
|
| 19 |
"eval_total": 44612,
|
| 20 |
+
"total_flos": 2.789913716232192e+16,
|
| 21 |
+
"train_loss": 0.945329581354701,
|
| 22 |
+
"train_runtime": 20900.3624,
|
| 23 |
"train_samples": 397452,
|
| 24 |
+
"train_samples_per_second": 95.083,
|
| 25 |
+
"train_steps_per_second": 5.943
|
| 26 |
}
|
eval_nbest_predictions.json
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:81997565151d57fa9c4ada6f4a37020c7f0db212ebb0471f64e0581ae9341821
|
| 3 |
+
size 216269144
|
eval_null_odds.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
eval_predictions.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
eval_results.json
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_HasAns_exact":
|
| 4 |
-
"eval_HasAns_f1": 86.
|
| 5 |
"eval_HasAns_total": 38667,
|
| 6 |
-
"eval_NoAns_exact":
|
| 7 |
-
"eval_NoAns_f1":
|
| 8 |
"eval_NoAns_total": 5945,
|
| 9 |
-
"eval_best_exact":
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
-
"eval_best_f1": 82.
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
-
"eval_exact":
|
| 14 |
-
"eval_f1": 82.
|
| 15 |
-
"eval_runtime":
|
| 16 |
"eval_samples": 78873,
|
| 17 |
-
"eval_samples_per_second":
|
| 18 |
-
"eval_steps_per_second":
|
| 19 |
"eval_total": 44612
|
| 20 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_HasAns_exact": 78.80363100318101,
|
| 4 |
+
"eval_HasAns_f1": 86.93823568622223,
|
| 5 |
"eval_HasAns_total": 38667,
|
| 6 |
+
"eval_NoAns_exact": 52.95206055508831,
|
| 7 |
+
"eval_NoAns_f1": 52.95206055508831,
|
| 8 |
"eval_NoAns_total": 5945,
|
| 9 |
+
"eval_best_exact": 75.3586478974267,
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
+
"eval_best_f1": 82.40923427058162,
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
+
"eval_exact": 75.3586478974267,
|
| 14 |
+
"eval_f1": 82.4092342705809,
|
| 15 |
+
"eval_runtime": 363.061,
|
| 16 |
"eval_samples": 78873,
|
| 17 |
+
"eval_samples_per_second": 217.244,
|
| 18 |
+
"eval_steps_per_second": 27.158,
|
| 19 |
"eval_total": 44612
|
| 20 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 397452,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 5.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 2.789913716232192e+16,
|
| 4 |
+
"train_loss": 0.945329581354701,
|
| 5 |
+
"train_runtime": 20900.3624,
|
| 6 |
"train_samples": 397452,
|
| 7 |
+
"train_samples_per_second": 95.083,
|
| 8 |
+
"train_steps_per_second": 5.943
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -2,1070 +2,1763 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.020128014170121975,
|
| 14 |
-
"grad_norm": 4.
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss": 3.
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.04025602834024395,
|
| 21 |
-
"grad_norm":
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss": 3.
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.06038404251036593,
|
| 28 |
-
"grad_norm": 6.
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss": 2.
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.0805120566804879,
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 2.
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.10064007085060987,
|
| 42 |
-
"grad_norm": 9.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.12076808502073186,
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 2.
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.14089609919085383,
|
| 56 |
-
"grad_norm": 9.
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 2.
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.1610241133609758,
|
| 63 |
-
"grad_norm":
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 2.
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.18115212753109777,
|
| 70 |
-
"grad_norm":
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 2.
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.20128014170121974,
|
| 77 |
-
"grad_norm": 9.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 2.
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.22140815587134174,
|
| 84 |
-
"grad_norm":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.24153617004146372,
|
| 91 |
-
"grad_norm": 13.
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss":
|
| 94 |
"step": 6000
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.26166418421158566,
|
| 98 |
-
"grad_norm":
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss":
|
| 101 |
"step": 6500
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.28179219838170766,
|
| 105 |
-
"grad_norm": 10.
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss": 1.
|
| 108 |
"step": 7000
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.30192021255182966,
|
| 112 |
-
"grad_norm":
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss": 1.
|
| 115 |
"step": 7500
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.3220482267219516,
|
| 119 |
-
"grad_norm":
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 1.
|
| 122 |
"step": 8000
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.3421762408920736,
|
| 126 |
-
"grad_norm":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 1.
|
| 129 |
"step": 8500
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.36230425506219555,
|
| 133 |
-
"grad_norm": 10.
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss": 1.
|
| 136 |
"step": 9000
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.38243226923231755,
|
| 140 |
-
"grad_norm":
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss": 1.
|
| 143 |
"step": 9500
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.4025602834024395,
|
| 147 |
-
"grad_norm":
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss": 1.
|
| 150 |
"step": 10000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.4226882975725615,
|
| 154 |
-
"grad_norm": 11.
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss": 1.
|
| 157 |
"step": 10500
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.4428163117426835,
|
| 161 |
-
"grad_norm":
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 1.
|
| 164 |
"step": 11000
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.46294432591280543,
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 11500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.48307234008292743,
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 1.
|
| 178 |
"step": 12000
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.5032003542530494,
|
| 182 |
-
"grad_norm":
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 12500
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.5233283684231713,
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss": 1.
|
| 192 |
"step": 13000
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.5434563825932933,
|
| 196 |
-
"grad_norm": 13.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 1.
|
| 199 |
"step": 13500
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.5635843967634153,
|
| 203 |
-
"grad_norm":
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 14000
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.5837124109335373,
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 14500
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.6038404251036593,
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 15000
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.6239684392737812,
|
| 224 |
-
"grad_norm":
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss": 1.
|
| 227 |
"step": 15500
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.6440964534439032,
|
| 231 |
-
"grad_norm": 11.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss": 1.
|
| 234 |
"step": 16000
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.6642244676140252,
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss": 1.
|
| 241 |
"step": 16500
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.6843524817841472,
|
| 245 |
-
"grad_norm":
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 17000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.7044804959542692,
|
| 252 |
-
"grad_norm": 14.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss": 1.
|
| 255 |
"step": 17500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.7246085101243911,
|
| 259 |
-
"grad_norm":
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss": 1.
|
| 262 |
"step": 18000
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.7447365242945131,
|
| 266 |
-
"grad_norm":
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss": 1.
|
| 269 |
"step": 18500
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.7648645384646351,
|
| 273 |
-
"grad_norm":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 1.
|
| 276 |
"step": 19000
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.7849925526347571,
|
| 280 |
-
"grad_norm":
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss": 1.
|
| 283 |
"step": 19500
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.805120566804879,
|
| 287 |
-
"grad_norm":
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss": 1.
|
| 290 |
"step": 20000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.825248580975001,
|
| 294 |
-
"grad_norm":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 20500
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.845376595145123,
|
| 301 |
-
"grad_norm":
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 21000
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.865504609315245,
|
| 308 |
-
"grad_norm":
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss": 1.
|
| 311 |
"step": 21500
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.885632623485367,
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss": 1.
|
| 318 |
"step": 22000
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.9057606376554889,
|
| 322 |
-
"grad_norm":
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss": 1.
|
| 325 |
"step": 22500
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.9258886518256109,
|
| 329 |
-
"grad_norm": 11.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 23000
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.9460166659957329,
|
| 336 |
-
"grad_norm": 10.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss": 1.
|
| 339 |
"step": 23500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.9661446801658549,
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss": 1.
|
| 346 |
"step": 24000
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.9862726943359769,
|
| 350 |
-
"grad_norm": 9.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss": 1.
|
| 353 |
"step": 24500
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.0064007085060989,
|
| 357 |
-
"grad_norm":
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss": 1.
|
| 360 |
"step": 25000
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.0265287226762208,
|
| 364 |
-
"grad_norm":
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss": 1.
|
| 367 |
"step": 25500
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.0466567368463426,
|
| 371 |
-
"grad_norm":
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss":
|
| 374 |
"step": 26000
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.0667847510164647,
|
| 378 |
-
"grad_norm": 11.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss":
|
| 381 |
"step": 26500
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.0869127651865866,
|
| 385 |
-
"grad_norm":
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss": 1.
|
| 388 |
"step": 27000
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.1070407793567087,
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss":
|
| 395 |
"step": 27500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.1271687935268306,
|
| 399 |
-
"grad_norm":
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss":
|
| 402 |
"step": 28000
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.1472968076969525,
|
| 406 |
-
"grad_norm":
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss":
|
| 409 |
"step": 28500
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.1674248218670746,
|
| 413 |
-
"grad_norm":
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss":
|
| 416 |
"step": 29000
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.1875528360371965,
|
| 420 |
-
"grad_norm":
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss":
|
| 423 |
"step": 29500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.2076808502073186,
|
| 427 |
-
"grad_norm":
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss":
|
| 430 |
"step": 30000
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.2278088643774405,
|
| 434 |
-
"grad_norm":
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss":
|
| 437 |
"step": 30500
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.2479368785475624,
|
| 441 |
-
"grad_norm":
|
| 442 |
-
"learning_rate":
|
| 443 |
-
"loss":
|
| 444 |
"step": 31000
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.2680648927176845,
|
| 448 |
-
"grad_norm":
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss":
|
| 451 |
"step": 31500
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.2881929068878064,
|
| 455 |
-
"grad_norm": 10.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss":
|
| 458 |
"step": 32000
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.3083209210579283,
|
| 462 |
-
"grad_norm":
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss":
|
| 465 |
"step": 32500
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.3284489352280504,
|
| 469 |
-
"grad_norm":
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss":
|
| 472 |
"step": 33000
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.3485769493981723,
|
| 476 |
-
"grad_norm":
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss":
|
| 479 |
"step": 33500
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.3687049635682944,
|
| 483 |
-
"grad_norm":
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss":
|
| 486 |
"step": 34000
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.3888329777384163,
|
| 490 |
-
"grad_norm":
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss":
|
| 493 |
"step": 34500
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.4089609919085384,
|
| 497 |
-
"grad_norm":
|
| 498 |
-
"learning_rate":
|
| 499 |
-
"loss":
|
| 500 |
"step": 35000
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.4290890060786603,
|
| 504 |
-
"grad_norm": 13.
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss":
|
| 507 |
"step": 35500
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.4492170202487822,
|
| 511 |
-
"grad_norm":
|
| 512 |
-
"learning_rate":
|
| 513 |
-
"loss":
|
| 514 |
"step": 36000
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.4693450344189043,
|
| 518 |
-
"grad_norm":
|
| 519 |
-
"learning_rate":
|
| 520 |
-
"loss":
|
| 521 |
"step": 36500
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.4894730485890262,
|
| 525 |
-
"grad_norm":
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss":
|
| 528 |
"step": 37000
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.509601062759148,
|
| 532 |
-
"grad_norm":
|
| 533 |
-
"learning_rate": 3.
|
| 534 |
-
"loss":
|
| 535 |
"step": 37500
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.5297290769292702,
|
| 539 |
-
"grad_norm":
|
| 540 |
-
"learning_rate": 3.
|
| 541 |
-
"loss": 0.
|
| 542 |
"step": 38000
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.5498570910993923,
|
| 546 |
-
"grad_norm": 8.
|
| 547 |
-
"learning_rate": 3.
|
| 548 |
-
"loss": 0.
|
| 549 |
"step": 38500
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.569985105269514,
|
| 553 |
-
"grad_norm":
|
| 554 |
-
"learning_rate": 3.
|
| 555 |
-
"loss": 0.
|
| 556 |
"step": 39000
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.590113119439636,
|
| 560 |
-
"grad_norm":
|
| 561 |
-
"learning_rate": 3.
|
| 562 |
-
"loss": 0.
|
| 563 |
"step": 39500
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.6102411336097582,
|
| 567 |
-
"grad_norm":
|
| 568 |
-
"learning_rate": 3.
|
| 569 |
-
"loss": 0.
|
| 570 |
"step": 40000
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 1.63036914777988,
|
| 574 |
-
"grad_norm":
|
| 575 |
-
"learning_rate": 3.
|
| 576 |
-
"loss": 0.
|
| 577 |
"step": 40500
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 1.650497161950002,
|
| 581 |
-
"grad_norm":
|
| 582 |
-
"learning_rate": 3.
|
| 583 |
-
"loss": 0.
|
| 584 |
"step": 41000
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 1.670625176120124,
|
| 588 |
-
"grad_norm":
|
| 589 |
-
"learning_rate": 3.
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 41500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 1.690753190290246,
|
| 595 |
-
"grad_norm":
|
| 596 |
-
"learning_rate": 3.
|
| 597 |
-
"loss": 0.
|
| 598 |
"step": 42000
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 1.7108812044603678,
|
| 602 |
-
"grad_norm":
|
| 603 |
-
"learning_rate": 3.
|
| 604 |
-
"loss": 0.
|
| 605 |
"step": 42500
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 1.73100921863049,
|
| 609 |
-
"grad_norm":
|
| 610 |
-
"learning_rate": 3.
|
| 611 |
-
"loss": 0.
|
| 612 |
"step": 43000
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 1.7511372328006118,
|
| 616 |
-
"grad_norm": 10.
|
| 617 |
-
"learning_rate": 3.
|
| 618 |
-
"loss": 0.
|
| 619 |
"step": 43500
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 1.7712652469707337,
|
| 623 |
-
"grad_norm":
|
| 624 |
-
"learning_rate": 3.
|
| 625 |
-
"loss": 0.
|
| 626 |
"step": 44000
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 1.7913932611408558,
|
| 630 |
-
"grad_norm": 7.
|
| 631 |
-
"learning_rate": 3.
|
| 632 |
-
"loss": 0.
|
| 633 |
"step": 44500
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 1.811521275310978,
|
| 637 |
-
"grad_norm": 11.
|
| 638 |
-
"learning_rate": 3.
|
| 639 |
-
"loss": 0.
|
| 640 |
"step": 45000
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 1.8316492894810998,
|
| 644 |
-
"grad_norm":
|
| 645 |
-
"learning_rate": 3.
|
| 646 |
-
"loss": 0.
|
| 647 |
"step": 45500
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 1.8517773036512217,
|
| 651 |
-
"grad_norm":
|
| 652 |
-
"learning_rate": 3.
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 46000
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 1.8719053178213438,
|
| 658 |
-
"grad_norm":
|
| 659 |
-
"learning_rate": 3.
|
| 660 |
-
"loss": 0.
|
| 661 |
"step": 46500
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 1.8920333319914657,
|
| 665 |
-
"grad_norm":
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss": 0.
|
| 668 |
"step": 47000
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 1.9121613461615876,
|
| 672 |
-
"grad_norm":
|
| 673 |
-
"learning_rate":
|
| 674 |
-
"loss": 0.
|
| 675 |
"step": 47500
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 1.9322893603317097,
|
| 679 |
-
"grad_norm":
|
| 680 |
-
"learning_rate":
|
| 681 |
-
"loss": 0.
|
| 682 |
"step": 48000
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 1.9524173745018316,
|
| 686 |
-
"grad_norm":
|
| 687 |
-
"learning_rate":
|
| 688 |
-
"loss": 0.
|
| 689 |
"step": 48500
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 1.9725453886719535,
|
| 693 |
-
"grad_norm":
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss": 0.
|
| 696 |
"step": 49000
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 1.9926734028420756,
|
| 700 |
-
"grad_norm":
|
| 701 |
-
"learning_rate":
|
| 702 |
-
"loss": 0.
|
| 703 |
"step": 49500
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 2.0128014170121977,
|
| 707 |
-
"grad_norm":
|
| 708 |
-
"learning_rate": 2.
|
| 709 |
-
"loss": 0.
|
| 710 |
"step": 50000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 2.0329294311823194,
|
| 714 |
-
"grad_norm":
|
| 715 |
-
"learning_rate": 2.
|
| 716 |
-
"loss": 0.
|
| 717 |
"step": 50500
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 2.0530574453524415,
|
| 721 |
-
"grad_norm":
|
| 722 |
-
"learning_rate": 2.
|
| 723 |
-
"loss": 0.
|
| 724 |
"step": 51000
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 2.0731854595225636,
|
| 728 |
-
"grad_norm": 8.
|
| 729 |
-
"learning_rate": 2.
|
| 730 |
-
"loss": 0.
|
| 731 |
"step": 51500
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 2.0933134736926853,
|
| 735 |
-
"grad_norm":
|
| 736 |
-
"learning_rate": 2.
|
| 737 |
-
"loss": 0.
|
| 738 |
"step": 52000
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 2.1134414878628074,
|
| 742 |
-
"grad_norm":
|
| 743 |
-
"learning_rate": 2.
|
| 744 |
-
"loss": 0.
|
| 745 |
"step": 52500
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 2.1335695020329295,
|
| 749 |
-
"grad_norm":
|
| 750 |
-
"learning_rate": 2.
|
| 751 |
-
"loss": 0.
|
| 752 |
"step": 53000
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 2.1536975162030516,
|
| 756 |
-
"grad_norm":
|
| 757 |
-
"learning_rate": 2.
|
| 758 |
-
"loss": 0.
|
| 759 |
"step": 53500
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 2.1738255303731733,
|
| 763 |
-
"grad_norm":
|
| 764 |
-
"learning_rate": 2.
|
| 765 |
-
"loss": 0.
|
| 766 |
"step": 54000
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 2.1939535445432954,
|
| 770 |
-
"grad_norm":
|
| 771 |
-
"learning_rate": 2.
|
| 772 |
-
"loss": 0.
|
| 773 |
"step": 54500
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 2.2140815587134175,
|
| 777 |
-
"grad_norm":
|
| 778 |
-
"learning_rate": 2.
|
| 779 |
-
"loss": 0.
|
| 780 |
"step": 55000
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 2.234209572883539,
|
| 784 |
-
"grad_norm":
|
| 785 |
-
"learning_rate": 2.
|
| 786 |
-
"loss": 0.
|
| 787 |
"step": 55500
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 2.2543375870536613,
|
| 791 |
-
"grad_norm":
|
| 792 |
-
"learning_rate":
|
| 793 |
-
"loss": 0.
|
| 794 |
"step": 56000
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 2.2744656012237834,
|
| 798 |
-
"grad_norm": 7.
|
| 799 |
-
"learning_rate":
|
| 800 |
-
"loss": 0.
|
| 801 |
"step": 56500
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 2.294593615393905,
|
| 805 |
-
"grad_norm":
|
| 806 |
-
"learning_rate":
|
| 807 |
-
"loss": 0.
|
| 808 |
"step": 57000
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 2.314721629564027,
|
| 812 |
-
"grad_norm":
|
| 813 |
-
"learning_rate":
|
| 814 |
-
"loss": 0.
|
| 815 |
"step": 57500
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 2.3348496437341493,
|
| 819 |
-
"grad_norm":
|
| 820 |
-
"learning_rate":
|
| 821 |
-
"loss": 0.
|
| 822 |
"step": 58000
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 2.3549776579042714,
|
| 826 |
-
"grad_norm":
|
| 827 |
-
"learning_rate":
|
| 828 |
-
"loss": 0.
|
| 829 |
"step": 58500
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 2.375105672074393,
|
| 833 |
-
"grad_norm":
|
| 834 |
-
"learning_rate":
|
| 835 |
-
"loss": 0.
|
| 836 |
"step": 59000
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 2.395233686244515,
|
| 840 |
-
"grad_norm":
|
| 841 |
-
"learning_rate":
|
| 842 |
-
"loss": 0.
|
| 843 |
"step": 59500
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 2.4153617004146373,
|
| 847 |
-
"grad_norm":
|
| 848 |
-
"learning_rate":
|
| 849 |
-
"loss": 0.
|
| 850 |
"step": 60000
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 2.435489714584759,
|
| 854 |
-
"grad_norm":
|
| 855 |
-
"learning_rate":
|
| 856 |
-
"loss": 0.
|
| 857 |
"step": 60500
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 2.455617728754881,
|
| 861 |
-
"grad_norm":
|
| 862 |
-
"learning_rate":
|
| 863 |
-
"loss": 0.
|
| 864 |
"step": 61000
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 2.475745742925003,
|
| 868 |
-
"grad_norm":
|
| 869 |
-
"learning_rate":
|
| 870 |
-
"loss": 0.
|
| 871 |
"step": 61500
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 2.495873757095125,
|
| 875 |
-
"grad_norm":
|
| 876 |
-
"learning_rate":
|
| 877 |
-
"loss": 0.
|
| 878 |
"step": 62000
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 2.516001771265247,
|
| 882 |
-
"grad_norm":
|
| 883 |
-
"learning_rate":
|
| 884 |
-
"loss": 0.
|
| 885 |
"step": 62500
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 2.536129785435369,
|
| 889 |
-
"grad_norm":
|
| 890 |
-
"learning_rate":
|
| 891 |
-
"loss": 0.
|
| 892 |
"step": 63000
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 2.5562577996054907,
|
| 896 |
-
"grad_norm":
|
| 897 |
-
"learning_rate":
|
| 898 |
-
"loss": 0.
|
| 899 |
"step": 63500
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 2.576385813775613,
|
| 903 |
-
"grad_norm":
|
| 904 |
-
"learning_rate":
|
| 905 |
-
"loss": 0.
|
| 906 |
"step": 64000
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 2.596513827945735,
|
| 910 |
-
"grad_norm":
|
| 911 |
-
"learning_rate":
|
| 912 |
-
"loss": 0.
|
| 913 |
"step": 64500
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 2.6166418421158566,
|
| 917 |
-
"grad_norm":
|
| 918 |
-
"learning_rate":
|
| 919 |
-
"loss": 0.
|
| 920 |
"step": 65000
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 2.6367698562859787,
|
| 924 |
-
"grad_norm":
|
| 925 |
-
"learning_rate":
|
| 926 |
-
"loss": 0.
|
| 927 |
"step": 65500
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 2.656897870456101,
|
| 931 |
-
"grad_norm":
|
| 932 |
-
"learning_rate":
|
| 933 |
-
"loss": 0.
|
| 934 |
"step": 66000
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 2.677025884626223,
|
| 938 |
-
"grad_norm":
|
| 939 |
-
"learning_rate":
|
| 940 |
-
"loss": 0.
|
| 941 |
"step": 66500
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 2.6971538987963446,
|
| 945 |
-
"grad_norm":
|
| 946 |
-
"learning_rate":
|
| 947 |
-
"loss": 0.
|
| 948 |
"step": 67000
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 2.7172819129664667,
|
| 952 |
-
"grad_norm":
|
| 953 |
-
"learning_rate":
|
| 954 |
-
"loss": 0.
|
| 955 |
"step": 67500
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 2.737409927136589,
|
| 959 |
-
"grad_norm": 10.
|
| 960 |
-
"learning_rate":
|
| 961 |
-
"loss": 0.
|
| 962 |
"step": 68000
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 2.757537941306711,
|
| 966 |
-
"grad_norm":
|
| 967 |
-
"learning_rate":
|
| 968 |
-
"loss": 0.
|
| 969 |
"step": 68500
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 2.7776659554768326,
|
| 973 |
-
"grad_norm":
|
| 974 |
-
"learning_rate":
|
| 975 |
-
"loss": 0.
|
| 976 |
"step": 69000
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 2.7977939696469547,
|
| 980 |
-
"grad_norm":
|
| 981 |
-
"learning_rate":
|
| 982 |
-
"loss": 0.
|
| 983 |
"step": 69500
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 2.817921983817077,
|
| 987 |
-
"grad_norm": 10.
|
| 988 |
-
"learning_rate":
|
| 989 |
-
"loss": 0.
|
| 990 |
"step": 70000
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 2.8380499979871985,
|
| 994 |
-
"grad_norm":
|
| 995 |
-
"learning_rate":
|
| 996 |
-
"loss": 0.
|
| 997 |
"step": 70500
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 2.8581780121573206,
|
| 1001 |
-
"grad_norm":
|
| 1002 |
-
"learning_rate":
|
| 1003 |
-
"loss": 0.
|
| 1004 |
"step": 71000
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 2.8783060263274427,
|
| 1008 |
-
"grad_norm":
|
| 1009 |
-
"learning_rate":
|
| 1010 |
-
"loss": 0.
|
| 1011 |
"step": 71500
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 2.8984340404975644,
|
| 1015 |
-
"grad_norm":
|
| 1016 |
-
"learning_rate": 2.
|
| 1017 |
-
"loss": 0.
|
| 1018 |
"step": 72000
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 2.9185620546676865,
|
| 1022 |
-
"grad_norm":
|
| 1023 |
-
"learning_rate": 2.
|
| 1024 |
-
"loss": 0.
|
| 1025 |
"step": 72500
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 2.9386900688378086,
|
| 1029 |
-
"grad_norm":
|
| 1030 |
-
"learning_rate":
|
| 1031 |
-
"loss": 0.
|
| 1032 |
"step": 73000
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 2.9588180830079303,
|
| 1036 |
-
"grad_norm":
|
| 1037 |
-
"learning_rate":
|
| 1038 |
-
"loss": 0.
|
| 1039 |
"step": 73500
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 2.9789460971780524,
|
| 1043 |
-
"grad_norm":
|
| 1044 |
-
"learning_rate":
|
| 1045 |
-
"loss": 0.
|
| 1046 |
"step": 74000
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 2.9990741113481745,
|
| 1050 |
-
"grad_norm":
|
| 1051 |
-
"learning_rate": 2.
|
| 1052 |
-
"loss": 0.
|
| 1053 |
"step": 74500
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
-
"epoch": 3.
|
| 1057 |
-
"
|
| 1058 |
-
"
|
| 1059 |
-
"
|
| 1060 |
-
"
|
| 1061 |
-
|
| 1062 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1063 |
}
|
| 1064 |
],
|
| 1065 |
"logging_steps": 500,
|
| 1066 |
-
"max_steps":
|
| 1067 |
"num_input_tokens_seen": 0,
|
| 1068 |
-
"num_train_epochs":
|
| 1069 |
"save_steps": 500,
|
| 1070 |
"stateful_callbacks": {
|
| 1071 |
"TrainerControl": {
|
|
@@ -1079,7 +1772,7 @@
|
|
| 1079 |
"attributes": {}
|
| 1080 |
}
|
| 1081 |
},
|
| 1082 |
-
"total_flos":
|
| 1083 |
"train_batch_size": 16,
|
| 1084 |
"trial_name": null,
|
| 1085 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 124205,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.020128014170121975,
|
| 14 |
+
"grad_norm": 4.4243292808532715,
|
| 15 |
+
"learning_rate": 4.9799122418582184e-05,
|
| 16 |
+
"loss": 3.5581,
|
| 17 |
"step": 500
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.04025602834024395,
|
| 21 |
+
"grad_norm": 5.467681884765625,
|
| 22 |
+
"learning_rate": 4.9597842276880965e-05,
|
| 23 |
+
"loss": 3.0931,
|
| 24 |
"step": 1000
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.06038404251036593,
|
| 28 |
+
"grad_norm": 6.596652507781982,
|
| 29 |
+
"learning_rate": 4.9396562135179747e-05,
|
| 30 |
+
"loss": 2.9698,
|
| 31 |
"step": 1500
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.0805120566804879,
|
| 35 |
+
"grad_norm": 6.438045501708984,
|
| 36 |
+
"learning_rate": 4.919528199347853e-05,
|
| 37 |
+
"loss": 2.8506,
|
| 38 |
"step": 2000
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.10064007085060987,
|
| 42 |
+
"grad_norm": 9.303359031677246,
|
| 43 |
+
"learning_rate": 4.89940018517773e-05,
|
| 44 |
+
"loss": 2.6189,
|
| 45 |
"step": 2500
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.12076808502073186,
|
| 49 |
+
"grad_norm": 10.118370056152344,
|
| 50 |
+
"learning_rate": 4.8792721710076084e-05,
|
| 51 |
+
"loss": 2.4486,
|
| 52 |
"step": 3000
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.14089609919085383,
|
| 56 |
+
"grad_norm": 9.903593063354492,
|
| 57 |
+
"learning_rate": 4.8591441568374865e-05,
|
| 58 |
+
"loss": 2.3446,
|
| 59 |
"step": 3500
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.1610241133609758,
|
| 63 |
+
"grad_norm": 14.505561828613281,
|
| 64 |
+
"learning_rate": 4.8390161426673646e-05,
|
| 65 |
+
"loss": 2.3041,
|
| 66 |
"step": 4000
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.18115212753109777,
|
| 70 |
+
"grad_norm": 8.208205223083496,
|
| 71 |
+
"learning_rate": 4.818888128497243e-05,
|
| 72 |
+
"loss": 2.221,
|
| 73 |
"step": 4500
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.20128014170121974,
|
| 77 |
+
"grad_norm": 9.10877799987793,
|
| 78 |
+
"learning_rate": 4.79876011432712e-05,
|
| 79 |
+
"loss": 2.2055,
|
| 80 |
"step": 5000
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.22140815587134174,
|
| 84 |
+
"grad_norm": 10.197419166564941,
|
| 85 |
+
"learning_rate": 4.778632100156999e-05,
|
| 86 |
+
"loss": 2.0875,
|
| 87 |
"step": 5500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.24153617004146372,
|
| 91 |
+
"grad_norm": 13.130525588989258,
|
| 92 |
+
"learning_rate": 4.758504085986877e-05,
|
| 93 |
+
"loss": 2.0619,
|
| 94 |
"step": 6000
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.26166418421158566,
|
| 98 |
+
"grad_norm": 10.468463897705078,
|
| 99 |
+
"learning_rate": 4.7383760718167546e-05,
|
| 100 |
+
"loss": 2.0394,
|
| 101 |
"step": 6500
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.28179219838170766,
|
| 105 |
+
"grad_norm": 10.524393081665039,
|
| 106 |
+
"learning_rate": 4.718248057646633e-05,
|
| 107 |
+
"loss": 1.9868,
|
| 108 |
"step": 7000
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.30192021255182966,
|
| 112 |
+
"grad_norm": 12.109511375427246,
|
| 113 |
+
"learning_rate": 4.698120043476511e-05,
|
| 114 |
+
"loss": 1.9657,
|
| 115 |
"step": 7500
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.3220482267219516,
|
| 119 |
+
"grad_norm": 9.086125373840332,
|
| 120 |
+
"learning_rate": 4.677992029306389e-05,
|
| 121 |
+
"loss": 1.9197,
|
| 122 |
"step": 8000
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.3421762408920736,
|
| 126 |
+
"grad_norm": 11.7044038772583,
|
| 127 |
+
"learning_rate": 4.657864015136267e-05,
|
| 128 |
+
"loss": 1.9035,
|
| 129 |
"step": 8500
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.36230425506219555,
|
| 133 |
+
"grad_norm": 10.403022766113281,
|
| 134 |
+
"learning_rate": 4.6377360009661446e-05,
|
| 135 |
+
"loss": 1.8933,
|
| 136 |
"step": 9000
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.38243226923231755,
|
| 140 |
+
"grad_norm": 11.45753002166748,
|
| 141 |
+
"learning_rate": 4.6176079867960234e-05,
|
| 142 |
+
"loss": 1.8203,
|
| 143 |
"step": 9500
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.4025602834024395,
|
| 147 |
+
"grad_norm": 15.168684959411621,
|
| 148 |
+
"learning_rate": 4.597479972625901e-05,
|
| 149 |
+
"loss": 1.8252,
|
| 150 |
"step": 10000
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.4226882975725615,
|
| 154 |
+
"grad_norm": 11.098946571350098,
|
| 155 |
+
"learning_rate": 4.577351958455779e-05,
|
| 156 |
+
"loss": 1.7927,
|
| 157 |
"step": 10500
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.4428163117426835,
|
| 161 |
+
"grad_norm": 14.646119117736816,
|
| 162 |
+
"learning_rate": 4.557223944285657e-05,
|
| 163 |
+
"loss": 1.7588,
|
| 164 |
"step": 11000
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.46294432591280543,
|
| 168 |
+
"grad_norm": 9.974624633789062,
|
| 169 |
+
"learning_rate": 4.537095930115535e-05,
|
| 170 |
+
"loss": 1.744,
|
| 171 |
"step": 11500
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.48307234008292743,
|
| 175 |
+
"grad_norm": 12.911872863769531,
|
| 176 |
+
"learning_rate": 4.516967915945413e-05,
|
| 177 |
+
"loss": 1.6844,
|
| 178 |
"step": 12000
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.5032003542530494,
|
| 182 |
+
"grad_norm": 12.667566299438477,
|
| 183 |
+
"learning_rate": 4.496839901775291e-05,
|
| 184 |
+
"loss": 1.6536,
|
| 185 |
"step": 12500
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.5233283684231713,
|
| 189 |
+
"grad_norm": 16.291105270385742,
|
| 190 |
+
"learning_rate": 4.476711887605169e-05,
|
| 191 |
+
"loss": 1.6482,
|
| 192 |
"step": 13000
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.5434563825932933,
|
| 196 |
+
"grad_norm": 13.714164733886719,
|
| 197 |
+
"learning_rate": 4.456583873435047e-05,
|
| 198 |
+
"loss": 1.6148,
|
| 199 |
"step": 13500
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.5635843967634153,
|
| 203 |
+
"grad_norm": 14.619283676147461,
|
| 204 |
+
"learning_rate": 4.436455859264925e-05,
|
| 205 |
+
"loss": 1.6384,
|
| 206 |
"step": 14000
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.5837124109335373,
|
| 210 |
+
"grad_norm": 11.949310302734375,
|
| 211 |
+
"learning_rate": 4.416327845094803e-05,
|
| 212 |
+
"loss": 1.5816,
|
| 213 |
"step": 14500
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.6038404251036593,
|
| 217 |
+
"grad_norm": 12.366539001464844,
|
| 218 |
+
"learning_rate": 4.396199830924681e-05,
|
| 219 |
+
"loss": 1.5615,
|
| 220 |
"step": 15000
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.6239684392737812,
|
| 224 |
+
"grad_norm": 9.16579818725586,
|
| 225 |
+
"learning_rate": 4.3760718167545596e-05,
|
| 226 |
+
"loss": 1.5674,
|
| 227 |
"step": 15500
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.6440964534439032,
|
| 231 |
+
"grad_norm": 11.544962882995605,
|
| 232 |
+
"learning_rate": 4.355943802584437e-05,
|
| 233 |
+
"loss": 1.5268,
|
| 234 |
"step": 16000
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.6642244676140252,
|
| 238 |
+
"grad_norm": 15.489496231079102,
|
| 239 |
+
"learning_rate": 4.335815788414315e-05,
|
| 240 |
+
"loss": 1.4859,
|
| 241 |
"step": 16500
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.6843524817841472,
|
| 245 |
+
"grad_norm": 12.722689628601074,
|
| 246 |
+
"learning_rate": 4.315687774244193e-05,
|
| 247 |
+
"loss": 1.4752,
|
| 248 |
"step": 17000
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.7044804959542692,
|
| 252 |
+
"grad_norm": 14.4386568069458,
|
| 253 |
+
"learning_rate": 4.2955597600740714e-05,
|
| 254 |
+
"loss": 1.4646,
|
| 255 |
"step": 17500
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.7246085101243911,
|
| 259 |
+
"grad_norm": 9.312575340270996,
|
| 260 |
+
"learning_rate": 4.2754317459039495e-05,
|
| 261 |
+
"loss": 1.481,
|
| 262 |
"step": 18000
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.7447365242945131,
|
| 266 |
+
"grad_norm": 13.926103591918945,
|
| 267 |
+
"learning_rate": 4.2553037317338277e-05,
|
| 268 |
+
"loss": 1.4305,
|
| 269 |
"step": 18500
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.7648645384646351,
|
| 273 |
+
"grad_norm": 9.286456108093262,
|
| 274 |
+
"learning_rate": 4.235175717563705e-05,
|
| 275 |
+
"loss": 1.419,
|
| 276 |
"step": 19000
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.7849925526347571,
|
| 280 |
+
"grad_norm": 18.389366149902344,
|
| 281 |
+
"learning_rate": 4.215047703393583e-05,
|
| 282 |
+
"loss": 1.3926,
|
| 283 |
"step": 19500
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.805120566804879,
|
| 287 |
+
"grad_norm": 21.281620025634766,
|
| 288 |
+
"learning_rate": 4.1949196892234614e-05,
|
| 289 |
+
"loss": 1.393,
|
| 290 |
"step": 20000
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.825248580975001,
|
| 294 |
+
"grad_norm": 13.243730545043945,
|
| 295 |
+
"learning_rate": 4.1747916750533395e-05,
|
| 296 |
+
"loss": 1.3645,
|
| 297 |
"step": 20500
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.845376595145123,
|
| 301 |
+
"grad_norm": 16.452898025512695,
|
| 302 |
+
"learning_rate": 4.1546636608832176e-05,
|
| 303 |
+
"loss": 1.3608,
|
| 304 |
"step": 21000
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.865504609315245,
|
| 308 |
+
"grad_norm": 13.110088348388672,
|
| 309 |
+
"learning_rate": 4.134535646713095e-05,
|
| 310 |
+
"loss": 1.3747,
|
| 311 |
"step": 21500
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.885632623485367,
|
| 315 |
+
"grad_norm": 11.320960998535156,
|
| 316 |
+
"learning_rate": 4.114407632542974e-05,
|
| 317 |
+
"loss": 1.3154,
|
| 318 |
"step": 22000
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.9057606376554889,
|
| 322 |
+
"grad_norm": 11.911636352539062,
|
| 323 |
+
"learning_rate": 4.094279618372851e-05,
|
| 324 |
+
"loss": 1.3131,
|
| 325 |
"step": 22500
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.9258886518256109,
|
| 329 |
+
"grad_norm": 11.672017097473145,
|
| 330 |
+
"learning_rate": 4.0741516042027295e-05,
|
| 331 |
+
"loss": 1.3185,
|
| 332 |
"step": 23000
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.9460166659957329,
|
| 336 |
+
"grad_norm": 10.78576374053955,
|
| 337 |
+
"learning_rate": 4.0540235900326076e-05,
|
| 338 |
+
"loss": 1.2969,
|
| 339 |
"step": 23500
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.9661446801658549,
|
| 343 |
+
"grad_norm": 10.77676773071289,
|
| 344 |
+
"learning_rate": 4.033895575862486e-05,
|
| 345 |
+
"loss": 1.2918,
|
| 346 |
"step": 24000
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.9862726943359769,
|
| 350 |
+
"grad_norm": 9.955876350402832,
|
| 351 |
+
"learning_rate": 4.013767561692364e-05,
|
| 352 |
+
"loss": 1.2866,
|
| 353 |
"step": 24500
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 1.0064007085060989,
|
| 357 |
+
"grad_norm": 9.662604331970215,
|
| 358 |
+
"learning_rate": 3.993639547522241e-05,
|
| 359 |
+
"loss": 1.2468,
|
| 360 |
"step": 25000
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 1.0265287226762208,
|
| 364 |
+
"grad_norm": 17.390344619750977,
|
| 365 |
+
"learning_rate": 3.9735115333521194e-05,
|
| 366 |
+
"loss": 1.2174,
|
| 367 |
"step": 25500
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 1.0466567368463426,
|
| 371 |
+
"grad_norm": 10.578307151794434,
|
| 372 |
+
"learning_rate": 3.9533835191819976e-05,
|
| 373 |
+
"loss": 1.1876,
|
| 374 |
"step": 26000
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 1.0667847510164647,
|
| 378 |
+
"grad_norm": 11.745862007141113,
|
| 379 |
+
"learning_rate": 3.933255505011876e-05,
|
| 380 |
+
"loss": 1.1913,
|
| 381 |
"step": 26500
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 1.0869127651865866,
|
| 385 |
+
"grad_norm": 16.16767120361328,
|
| 386 |
+
"learning_rate": 3.913127490841754e-05,
|
| 387 |
+
"loss": 1.2254,
|
| 388 |
"step": 27000
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 1.1070407793567087,
|
| 392 |
+
"grad_norm": 11.383201599121094,
|
| 393 |
+
"learning_rate": 3.892999476671631e-05,
|
| 394 |
+
"loss": 1.1772,
|
| 395 |
"step": 27500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.1271687935268306,
|
| 399 |
+
"grad_norm": 9.412652015686035,
|
| 400 |
+
"learning_rate": 3.87287146250151e-05,
|
| 401 |
+
"loss": 1.1518,
|
| 402 |
"step": 28000
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 1.1472968076969525,
|
| 406 |
+
"grad_norm": 18.917278289794922,
|
| 407 |
+
"learning_rate": 3.852743448331388e-05,
|
| 408 |
+
"loss": 1.1634,
|
| 409 |
"step": 28500
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 1.1674248218670746,
|
| 413 |
+
"grad_norm": 24.050168991088867,
|
| 414 |
+
"learning_rate": 3.8326154341612657e-05,
|
| 415 |
+
"loss": 1.1303,
|
| 416 |
"step": 29000
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 1.1875528360371965,
|
| 420 |
+
"grad_norm": 16.491037368774414,
|
| 421 |
+
"learning_rate": 3.812487419991144e-05,
|
| 422 |
+
"loss": 1.1527,
|
| 423 |
"step": 29500
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 1.2076808502073186,
|
| 427 |
+
"grad_norm": 8.464524269104004,
|
| 428 |
+
"learning_rate": 3.792359405821022e-05,
|
| 429 |
+
"loss": 1.097,
|
| 430 |
"step": 30000
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 1.2278088643774405,
|
| 434 |
+
"grad_norm": 25.68097686767578,
|
| 435 |
+
"learning_rate": 3.7722313916509e-05,
|
| 436 |
+
"loss": 1.13,
|
| 437 |
"step": 30500
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 1.2479368785475624,
|
| 441 |
+
"grad_norm": 11.21618938446045,
|
| 442 |
+
"learning_rate": 3.752103377480778e-05,
|
| 443 |
+
"loss": 1.089,
|
| 444 |
"step": 31000
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 1.2680648927176845,
|
| 448 |
+
"grad_norm": 14.719616889953613,
|
| 449 |
+
"learning_rate": 3.7319753633106556e-05,
|
| 450 |
+
"loss": 1.0869,
|
| 451 |
"step": 31500
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 1.2881929068878064,
|
| 455 |
+
"grad_norm": 10.524362564086914,
|
| 456 |
+
"learning_rate": 3.7118473491405344e-05,
|
| 457 |
+
"loss": 1.0621,
|
| 458 |
"step": 32000
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 1.3083209210579283,
|
| 462 |
+
"grad_norm": 17.096012115478516,
|
| 463 |
+
"learning_rate": 3.691719334970412e-05,
|
| 464 |
+
"loss": 1.0583,
|
| 465 |
"step": 32500
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 1.3284489352280504,
|
| 469 |
+
"grad_norm": 8.337162017822266,
|
| 470 |
+
"learning_rate": 3.67159132080029e-05,
|
| 471 |
+
"loss": 1.062,
|
| 472 |
"step": 33000
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 1.3485769493981723,
|
| 476 |
+
"grad_norm": 16.806882858276367,
|
| 477 |
+
"learning_rate": 3.651463306630168e-05,
|
| 478 |
+
"loss": 1.0498,
|
| 479 |
"step": 33500
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 1.3687049635682944,
|
| 483 |
+
"grad_norm": 14.466264724731445,
|
| 484 |
+
"learning_rate": 3.631335292460046e-05,
|
| 485 |
+
"loss": 1.0378,
|
| 486 |
"step": 34000
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 1.3888329777384163,
|
| 490 |
+
"grad_norm": 10.003057479858398,
|
| 491 |
+
"learning_rate": 3.6112072782899244e-05,
|
| 492 |
+
"loss": 1.0587,
|
| 493 |
"step": 34500
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 1.4089609919085384,
|
| 497 |
+
"grad_norm": 14.94166374206543,
|
| 498 |
+
"learning_rate": 3.591079264119802e-05,
|
| 499 |
+
"loss": 1.0299,
|
| 500 |
"step": 35000
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 1.4290890060786603,
|
| 504 |
+
"grad_norm": 13.608207702636719,
|
| 505 |
+
"learning_rate": 3.57095124994968e-05,
|
| 506 |
+
"loss": 1.0059,
|
| 507 |
"step": 35500
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 1.4492170202487822,
|
| 511 |
+
"grad_norm": 15.849994659423828,
|
| 512 |
+
"learning_rate": 3.550823235779558e-05,
|
| 513 |
+
"loss": 1.0209,
|
| 514 |
"step": 36000
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 1.4693450344189043,
|
| 518 |
+
"grad_norm": 17.382606506347656,
|
| 519 |
+
"learning_rate": 3.530695221609436e-05,
|
| 520 |
+
"loss": 1.0337,
|
| 521 |
"step": 36500
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 1.4894730485890262,
|
| 525 |
+
"grad_norm": 16.811100006103516,
|
| 526 |
+
"learning_rate": 3.5105672074393144e-05,
|
| 527 |
+
"loss": 1.0181,
|
| 528 |
"step": 37000
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 1.509601062759148,
|
| 532 |
+
"grad_norm": 11.390883445739746,
|
| 533 |
+
"learning_rate": 3.490439193269192e-05,
|
| 534 |
+
"loss": 1.0099,
|
| 535 |
"step": 37500
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 1.5297290769292702,
|
| 539 |
+
"grad_norm": 9.782214164733887,
|
| 540 |
+
"learning_rate": 3.47031117909907e-05,
|
| 541 |
+
"loss": 0.9937,
|
| 542 |
"step": 38000
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 1.5498570910993923,
|
| 546 |
+
"grad_norm": 8.597905158996582,
|
| 547 |
+
"learning_rate": 3.450183164928949e-05,
|
| 548 |
+
"loss": 0.9506,
|
| 549 |
"step": 38500
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 1.569985105269514,
|
| 553 |
+
"grad_norm": 14.413061141967773,
|
| 554 |
+
"learning_rate": 3.430055150758826e-05,
|
| 555 |
+
"loss": 0.9617,
|
| 556 |
"step": 39000
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 1.590113119439636,
|
| 560 |
+
"grad_norm": 11.166732788085938,
|
| 561 |
+
"learning_rate": 3.409927136588704e-05,
|
| 562 |
+
"loss": 0.9621,
|
| 563 |
"step": 39500
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 1.6102411336097582,
|
| 567 |
+
"grad_norm": 14.102859497070312,
|
| 568 |
+
"learning_rate": 3.389799122418582e-05,
|
| 569 |
+
"loss": 0.9586,
|
| 570 |
"step": 40000
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 1.63036914777988,
|
| 574 |
+
"grad_norm": 15.011527061462402,
|
| 575 |
+
"learning_rate": 3.3696711082484606e-05,
|
| 576 |
+
"loss": 0.9449,
|
| 577 |
"step": 40500
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 1.650497161950002,
|
| 581 |
+
"grad_norm": 14.565163612365723,
|
| 582 |
+
"learning_rate": 3.349543094078339e-05,
|
| 583 |
+
"loss": 0.9437,
|
| 584 |
"step": 41000
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 1.670625176120124,
|
| 588 |
+
"grad_norm": 6.743287086486816,
|
| 589 |
+
"learning_rate": 3.329415079908216e-05,
|
| 590 |
+
"loss": 0.9548,
|
| 591 |
"step": 41500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 1.690753190290246,
|
| 595 |
+
"grad_norm": 8.881048202514648,
|
| 596 |
+
"learning_rate": 3.309287065738094e-05,
|
| 597 |
+
"loss": 0.941,
|
| 598 |
"step": 42000
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 1.7108812044603678,
|
| 602 |
+
"grad_norm": 24.312618255615234,
|
| 603 |
+
"learning_rate": 3.2891590515679724e-05,
|
| 604 |
+
"loss": 0.9096,
|
| 605 |
"step": 42500
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 1.73100921863049,
|
| 609 |
+
"grad_norm": 12.95536994934082,
|
| 610 |
+
"learning_rate": 3.2690310373978506e-05,
|
| 611 |
+
"loss": 0.9403,
|
| 612 |
"step": 43000
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 1.7511372328006118,
|
| 616 |
+
"grad_norm": 10.41937255859375,
|
| 617 |
+
"learning_rate": 3.248903023227729e-05,
|
| 618 |
+
"loss": 0.9075,
|
| 619 |
"step": 43500
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 1.7712652469707337,
|
| 623 |
+
"grad_norm": 7.3060383796691895,
|
| 624 |
+
"learning_rate": 3.228775009057606e-05,
|
| 625 |
+
"loss": 0.9169,
|
| 626 |
"step": 44000
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 1.7913932611408558,
|
| 630 |
+
"grad_norm": 7.2622222900390625,
|
| 631 |
+
"learning_rate": 3.208646994887485e-05,
|
| 632 |
+
"loss": 0.9088,
|
| 633 |
"step": 44500
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 1.811521275310978,
|
| 637 |
+
"grad_norm": 11.950494766235352,
|
| 638 |
+
"learning_rate": 3.1885189807173624e-05,
|
| 639 |
+
"loss": 0.8966,
|
| 640 |
"step": 45000
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 1.8316492894810998,
|
| 644 |
+
"grad_norm": 10.778806686401367,
|
| 645 |
+
"learning_rate": 3.1683909665472405e-05,
|
| 646 |
+
"loss": 0.8911,
|
| 647 |
"step": 45500
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 1.8517773036512217,
|
| 651 |
+
"grad_norm": 19.629791259765625,
|
| 652 |
+
"learning_rate": 3.148262952377119e-05,
|
| 653 |
+
"loss": 0.9132,
|
| 654 |
"step": 46000
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 1.8719053178213438,
|
| 658 |
+
"grad_norm": 6.642657279968262,
|
| 659 |
+
"learning_rate": 3.128134938206997e-05,
|
| 660 |
+
"loss": 0.9081,
|
| 661 |
"step": 46500
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 1.8920333319914657,
|
| 665 |
+
"grad_norm": 13.492182731628418,
|
| 666 |
+
"learning_rate": 3.108006924036875e-05,
|
| 667 |
+
"loss": 0.8847,
|
| 668 |
"step": 47000
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 1.9121613461615876,
|
| 672 |
+
"grad_norm": 15.45358943939209,
|
| 673 |
+
"learning_rate": 3.0878789098667524e-05,
|
| 674 |
+
"loss": 0.8931,
|
| 675 |
"step": 47500
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 1.9322893603317097,
|
| 679 |
+
"grad_norm": 9.835533142089844,
|
| 680 |
+
"learning_rate": 3.0677508956966305e-05,
|
| 681 |
+
"loss": 0.8618,
|
| 682 |
"step": 48000
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 1.9524173745018316,
|
| 686 |
+
"grad_norm": 20.065549850463867,
|
| 687 |
+
"learning_rate": 3.047622881526509e-05,
|
| 688 |
+
"loss": 0.8708,
|
| 689 |
"step": 48500
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 1.9725453886719535,
|
| 693 |
+
"grad_norm": 10.010771751403809,
|
| 694 |
+
"learning_rate": 3.0274948673563868e-05,
|
| 695 |
+
"loss": 0.8453,
|
| 696 |
"step": 49000
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 1.9926734028420756,
|
| 700 |
+
"grad_norm": 9.881052017211914,
|
| 701 |
+
"learning_rate": 3.007366853186265e-05,
|
| 702 |
+
"loss": 0.8517,
|
| 703 |
"step": 49500
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 2.0128014170121977,
|
| 707 |
+
"grad_norm": 14.831236839294434,
|
| 708 |
+
"learning_rate": 2.9872388390161427e-05,
|
| 709 |
+
"loss": 0.8147,
|
| 710 |
"step": 50000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 2.0329294311823194,
|
| 714 |
+
"grad_norm": 4.119289398193359,
|
| 715 |
+
"learning_rate": 2.967110824846021e-05,
|
| 716 |
+
"loss": 0.8103,
|
| 717 |
"step": 50500
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 2.0530574453524415,
|
| 721 |
+
"grad_norm": 14.809292793273926,
|
| 722 |
+
"learning_rate": 2.946982810675899e-05,
|
| 723 |
+
"loss": 0.8039,
|
| 724 |
"step": 51000
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 2.0731854595225636,
|
| 728 |
+
"grad_norm": 8.020957946777344,
|
| 729 |
+
"learning_rate": 2.9268547965057767e-05,
|
| 730 |
+
"loss": 0.7863,
|
| 731 |
"step": 51500
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 2.0933134736926853,
|
| 735 |
+
"grad_norm": 13.140082359313965,
|
| 736 |
+
"learning_rate": 2.906726782335655e-05,
|
| 737 |
+
"loss": 0.8164,
|
| 738 |
"step": 52000
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 2.1134414878628074,
|
| 742 |
+
"grad_norm": 12.682424545288086,
|
| 743 |
+
"learning_rate": 2.8865987681655326e-05,
|
| 744 |
+
"loss": 0.8106,
|
| 745 |
"step": 52500
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 2.1335695020329295,
|
| 749 |
+
"grad_norm": 6.823521137237549,
|
| 750 |
+
"learning_rate": 2.866470753995411e-05,
|
| 751 |
+
"loss": 0.7787,
|
| 752 |
"step": 53000
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 2.1536975162030516,
|
| 756 |
+
"grad_norm": 18.312480926513672,
|
| 757 |
+
"learning_rate": 2.846342739825289e-05,
|
| 758 |
+
"loss": 0.7644,
|
| 759 |
"step": 53500
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 2.1738255303731733,
|
| 763 |
+
"grad_norm": 16.279190063476562,
|
| 764 |
+
"learning_rate": 2.826214725655167e-05,
|
| 765 |
+
"loss": 0.7722,
|
| 766 |
"step": 54000
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 2.1939535445432954,
|
| 770 |
+
"grad_norm": 14.282658576965332,
|
| 771 |
+
"learning_rate": 2.8060867114850448e-05,
|
| 772 |
+
"loss": 0.7714,
|
| 773 |
"step": 54500
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 2.2140815587134175,
|
| 777 |
+
"grad_norm": 15.930342674255371,
|
| 778 |
+
"learning_rate": 2.7859586973149233e-05,
|
| 779 |
+
"loss": 0.7626,
|
| 780 |
"step": 55000
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 2.234209572883539,
|
| 784 |
+
"grad_norm": 14.36133098602295,
|
| 785 |
+
"learning_rate": 2.765830683144801e-05,
|
| 786 |
+
"loss": 0.7489,
|
| 787 |
"step": 55500
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 2.2543375870536613,
|
| 791 |
+
"grad_norm": 12.644648551940918,
|
| 792 |
+
"learning_rate": 2.745702668974679e-05,
|
| 793 |
+
"loss": 0.7686,
|
| 794 |
"step": 56000
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 2.2744656012237834,
|
| 798 |
+
"grad_norm": 7.564332485198975,
|
| 799 |
+
"learning_rate": 2.725574654804557e-05,
|
| 800 |
+
"loss": 0.8139,
|
| 801 |
"step": 56500
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 2.294593615393905,
|
| 805 |
+
"grad_norm": 9.281700134277344,
|
| 806 |
+
"learning_rate": 2.7054466406344355e-05,
|
| 807 |
+
"loss": 0.7678,
|
| 808 |
"step": 57000
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 2.314721629564027,
|
| 812 |
+
"grad_norm": 16.21125030517578,
|
| 813 |
+
"learning_rate": 2.6853186264643133e-05,
|
| 814 |
+
"loss": 0.7511,
|
| 815 |
"step": 57500
|
| 816 |
},
|
| 817 |
{
|
| 818 |
"epoch": 2.3348496437341493,
|
| 819 |
+
"grad_norm": 15.86844253540039,
|
| 820 |
+
"learning_rate": 2.665190612294191e-05,
|
| 821 |
+
"loss": 0.7478,
|
| 822 |
"step": 58000
|
| 823 |
},
|
| 824 |
{
|
| 825 |
"epoch": 2.3549776579042714,
|
| 826 |
+
"grad_norm": 8.65857982635498,
|
| 827 |
+
"learning_rate": 2.645062598124069e-05,
|
| 828 |
+
"loss": 0.7501,
|
| 829 |
"step": 58500
|
| 830 |
},
|
| 831 |
{
|
| 832 |
"epoch": 2.375105672074393,
|
| 833 |
+
"grad_norm": 13.931543350219727,
|
| 834 |
+
"learning_rate": 2.6249345839539473e-05,
|
| 835 |
+
"loss": 0.7437,
|
| 836 |
"step": 59000
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 2.395233686244515,
|
| 840 |
+
"grad_norm": 29.372133255004883,
|
| 841 |
+
"learning_rate": 2.6048065697838254e-05,
|
| 842 |
+
"loss": 0.7437,
|
| 843 |
"step": 59500
|
| 844 |
},
|
| 845 |
{
|
| 846 |
"epoch": 2.4153617004146373,
|
| 847 |
+
"grad_norm": 21.69481658935547,
|
| 848 |
+
"learning_rate": 2.5846785556137032e-05,
|
| 849 |
+
"loss": 0.7422,
|
| 850 |
"step": 60000
|
| 851 |
},
|
| 852 |
{
|
| 853 |
"epoch": 2.435489714584759,
|
| 854 |
+
"grad_norm": 9.028382301330566,
|
| 855 |
+
"learning_rate": 2.564550541443581e-05,
|
| 856 |
+
"loss": 0.715,
|
| 857 |
"step": 60500
|
| 858 |
},
|
| 859 |
{
|
| 860 |
"epoch": 2.455617728754881,
|
| 861 |
+
"grad_norm": 10.219104766845703,
|
| 862 |
+
"learning_rate": 2.5444225272734595e-05,
|
| 863 |
+
"loss": 0.765,
|
| 864 |
"step": 61000
|
| 865 |
},
|
| 866 |
{
|
| 867 |
"epoch": 2.475745742925003,
|
| 868 |
+
"grad_norm": 15.746648788452148,
|
| 869 |
+
"learning_rate": 2.5242945131033373e-05,
|
| 870 |
+
"loss": 0.7294,
|
| 871 |
"step": 61500
|
| 872 |
},
|
| 873 |
{
|
| 874 |
"epoch": 2.495873757095125,
|
| 875 |
+
"grad_norm": 11.065299034118652,
|
| 876 |
+
"learning_rate": 2.5041664989332154e-05,
|
| 877 |
+
"loss": 0.7318,
|
| 878 |
"step": 62000
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 2.516001771265247,
|
| 882 |
+
"grad_norm": 9.226717948913574,
|
| 883 |
+
"learning_rate": 2.4840384847630935e-05,
|
| 884 |
+
"loss": 0.7297,
|
| 885 |
"step": 62500
|
| 886 |
},
|
| 887 |
{
|
| 888 |
"epoch": 2.536129785435369,
|
| 889 |
+
"grad_norm": 7.417213439941406,
|
| 890 |
+
"learning_rate": 2.4639104705929713e-05,
|
| 891 |
+
"loss": 0.7458,
|
| 892 |
"step": 63000
|
| 893 |
},
|
| 894 |
{
|
| 895 |
"epoch": 2.5562577996054907,
|
| 896 |
+
"grad_norm": 7.675484657287598,
|
| 897 |
+
"learning_rate": 2.4437824564228495e-05,
|
| 898 |
+
"loss": 0.7408,
|
| 899 |
"step": 63500
|
| 900 |
},
|
| 901 |
{
|
| 902 |
"epoch": 2.576385813775613,
|
| 903 |
+
"grad_norm": 12.612874984741211,
|
| 904 |
+
"learning_rate": 2.4236544422527276e-05,
|
| 905 |
+
"loss": 0.7425,
|
| 906 |
"step": 64000
|
| 907 |
},
|
| 908 |
{
|
| 909 |
"epoch": 2.596513827945735,
|
| 910 |
+
"grad_norm": 19.371625900268555,
|
| 911 |
+
"learning_rate": 2.4035264280826057e-05,
|
| 912 |
+
"loss": 0.7373,
|
| 913 |
"step": 64500
|
| 914 |
},
|
| 915 |
{
|
| 916 |
"epoch": 2.6166418421158566,
|
| 917 |
+
"grad_norm": 6.016005516052246,
|
| 918 |
+
"learning_rate": 2.3833984139124835e-05,
|
| 919 |
+
"loss": 0.7204,
|
| 920 |
"step": 65000
|
| 921 |
},
|
| 922 |
{
|
| 923 |
"epoch": 2.6367698562859787,
|
| 924 |
+
"grad_norm": 11.826494216918945,
|
| 925 |
+
"learning_rate": 2.3632703997423616e-05,
|
| 926 |
+
"loss": 0.7598,
|
| 927 |
"step": 65500
|
| 928 |
},
|
| 929 |
{
|
| 930 |
"epoch": 2.656897870456101,
|
| 931 |
+
"grad_norm": 17.242576599121094,
|
| 932 |
+
"learning_rate": 2.3431423855722394e-05,
|
| 933 |
+
"loss": 0.7009,
|
| 934 |
"step": 66000
|
| 935 |
},
|
| 936 |
{
|
| 937 |
"epoch": 2.677025884626223,
|
| 938 |
+
"grad_norm": 13.259429931640625,
|
| 939 |
+
"learning_rate": 2.3230143714021176e-05,
|
| 940 |
+
"loss": 0.7172,
|
| 941 |
"step": 66500
|
| 942 |
},
|
| 943 |
{
|
| 944 |
"epoch": 2.6971538987963446,
|
| 945 |
+
"grad_norm": 10.00273609161377,
|
| 946 |
+
"learning_rate": 2.3028863572319957e-05,
|
| 947 |
+
"loss": 0.7105,
|
| 948 |
"step": 67000
|
| 949 |
},
|
| 950 |
{
|
| 951 |
"epoch": 2.7172819129664667,
|
| 952 |
+
"grad_norm": 18.454111099243164,
|
| 953 |
+
"learning_rate": 2.2827583430618738e-05,
|
| 954 |
+
"loss": 0.7271,
|
| 955 |
"step": 67500
|
| 956 |
},
|
| 957 |
{
|
| 958 |
"epoch": 2.737409927136589,
|
| 959 |
+
"grad_norm": 10.182483673095703,
|
| 960 |
+
"learning_rate": 2.2626303288917516e-05,
|
| 961 |
+
"loss": 0.6922,
|
| 962 |
"step": 68000
|
| 963 |
},
|
| 964 |
{
|
| 965 |
"epoch": 2.757537941306711,
|
| 966 |
+
"grad_norm": 12.108529090881348,
|
| 967 |
+
"learning_rate": 2.2425023147216297e-05,
|
| 968 |
+
"loss": 0.7037,
|
| 969 |
"step": 68500
|
| 970 |
},
|
| 971 |
{
|
| 972 |
"epoch": 2.7776659554768326,
|
| 973 |
+
"grad_norm": 10.792649269104004,
|
| 974 |
+
"learning_rate": 2.222374300551508e-05,
|
| 975 |
+
"loss": 0.7148,
|
| 976 |
"step": 69000
|
| 977 |
},
|
| 978 |
{
|
| 979 |
"epoch": 2.7977939696469547,
|
| 980 |
+
"grad_norm": 9.332015037536621,
|
| 981 |
+
"learning_rate": 2.2022462863813856e-05,
|
| 982 |
+
"loss": 0.7035,
|
| 983 |
"step": 69500
|
| 984 |
},
|
| 985 |
{
|
| 986 |
"epoch": 2.817921983817077,
|
| 987 |
+
"grad_norm": 10.509349822998047,
|
| 988 |
+
"learning_rate": 2.1821182722112638e-05,
|
| 989 |
+
"loss": 0.6964,
|
| 990 |
"step": 70000
|
| 991 |
},
|
| 992 |
{
|
| 993 |
"epoch": 2.8380499979871985,
|
| 994 |
+
"grad_norm": 15.044132232666016,
|
| 995 |
+
"learning_rate": 2.1619902580411416e-05,
|
| 996 |
+
"loss": 0.6966,
|
| 997 |
"step": 70500
|
| 998 |
},
|
| 999 |
{
|
| 1000 |
"epoch": 2.8581780121573206,
|
| 1001 |
+
"grad_norm": 16.552404403686523,
|
| 1002 |
+
"learning_rate": 2.1418622438710197e-05,
|
| 1003 |
+
"loss": 0.6888,
|
| 1004 |
"step": 71000
|
| 1005 |
},
|
| 1006 |
{
|
| 1007 |
"epoch": 2.8783060263274427,
|
| 1008 |
+
"grad_norm": 12.294528007507324,
|
| 1009 |
+
"learning_rate": 2.1217342297008978e-05,
|
| 1010 |
+
"loss": 0.72,
|
| 1011 |
"step": 71500
|
| 1012 |
},
|
| 1013 |
{
|
| 1014 |
"epoch": 2.8984340404975644,
|
| 1015 |
+
"grad_norm": 8.884589195251465,
|
| 1016 |
+
"learning_rate": 2.101606215530776e-05,
|
| 1017 |
+
"loss": 0.7135,
|
| 1018 |
"step": 72000
|
| 1019 |
},
|
| 1020 |
{
|
| 1021 |
"epoch": 2.9185620546676865,
|
| 1022 |
+
"grad_norm": 16.005447387695312,
|
| 1023 |
+
"learning_rate": 2.0814782013606537e-05,
|
| 1024 |
+
"loss": 0.6937,
|
| 1025 |
"step": 72500
|
| 1026 |
},
|
| 1027 |
{
|
| 1028 |
"epoch": 2.9386900688378086,
|
| 1029 |
+
"grad_norm": 8.800543785095215,
|
| 1030 |
+
"learning_rate": 2.061350187190532e-05,
|
| 1031 |
+
"loss": 0.719,
|
| 1032 |
"step": 73000
|
| 1033 |
},
|
| 1034 |
{
|
| 1035 |
"epoch": 2.9588180830079303,
|
| 1036 |
+
"grad_norm": 11.393016815185547,
|
| 1037 |
+
"learning_rate": 2.0412221730204097e-05,
|
| 1038 |
+
"loss": 0.696,
|
| 1039 |
"step": 73500
|
| 1040 |
},
|
| 1041 |
{
|
| 1042 |
"epoch": 2.9789460971780524,
|
| 1043 |
+
"grad_norm": 11.305217742919922,
|
| 1044 |
+
"learning_rate": 2.021094158850288e-05,
|
| 1045 |
+
"loss": 0.6821,
|
| 1046 |
"step": 74000
|
| 1047 |
},
|
| 1048 |
{
|
| 1049 |
"epoch": 2.9990741113481745,
|
| 1050 |
+
"grad_norm": 10.046639442443848,
|
| 1051 |
+
"learning_rate": 2.000966144680166e-05,
|
| 1052 |
+
"loss": 0.6915,
|
| 1053 |
"step": 74500
|
| 1054 |
},
|
| 1055 |
{
|
| 1056 |
+
"epoch": 3.019202125518296,
|
| 1057 |
+
"grad_norm": 24.034393310546875,
|
| 1058 |
+
"learning_rate": 1.980838130510044e-05,
|
| 1059 |
+
"loss": 0.6306,
|
| 1060 |
+
"step": 75000
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 3.0393301396884183,
|
| 1064 |
+
"grad_norm": 8.862112998962402,
|
| 1065 |
+
"learning_rate": 1.960710116339922e-05,
|
| 1066 |
+
"loss": 0.6263,
|
| 1067 |
+
"step": 75500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 3.0594581538585404,
|
| 1071 |
+
"grad_norm": 17.781280517578125,
|
| 1072 |
+
"learning_rate": 1.9405821021698e-05,
|
| 1073 |
+
"loss": 0.6333,
|
| 1074 |
+
"step": 76000
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 3.0795861680286625,
|
| 1078 |
+
"grad_norm": 18.18588638305664,
|
| 1079 |
+
"learning_rate": 1.920454087999678e-05,
|
| 1080 |
+
"loss": 0.6316,
|
| 1081 |
+
"step": 76500
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 3.099714182198784,
|
| 1085 |
+
"grad_norm": 8.854592323303223,
|
| 1086 |
+
"learning_rate": 1.9003260738295562e-05,
|
| 1087 |
+
"loss": 0.6429,
|
| 1088 |
+
"step": 77000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 3.1198421963689063,
|
| 1092 |
+
"grad_norm": 9.773287773132324,
|
| 1093 |
+
"learning_rate": 1.880198059659434e-05,
|
| 1094 |
+
"loss": 0.6495,
|
| 1095 |
+
"step": 77500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 3.1399702105390284,
|
| 1099 |
+
"grad_norm": 3.36272931098938,
|
| 1100 |
+
"learning_rate": 1.860070045489312e-05,
|
| 1101 |
+
"loss": 0.6475,
|
| 1102 |
+
"step": 78000
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 3.16009822470915,
|
| 1106 |
+
"grad_norm": 12.333951950073242,
|
| 1107 |
+
"learning_rate": 1.83994203131919e-05,
|
| 1108 |
+
"loss": 0.6533,
|
| 1109 |
+
"step": 78500
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 3.180226238879272,
|
| 1113 |
+
"grad_norm": 14.517784118652344,
|
| 1114 |
+
"learning_rate": 1.8198140171490684e-05,
|
| 1115 |
+
"loss": 0.628,
|
| 1116 |
+
"step": 79000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 3.2003542530493942,
|
| 1120 |
+
"grad_norm": 11.176767349243164,
|
| 1121 |
+
"learning_rate": 1.7996860029789462e-05,
|
| 1122 |
+
"loss": 0.6377,
|
| 1123 |
+
"step": 79500
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 3.220482267219516,
|
| 1127 |
+
"grad_norm": 10.889203071594238,
|
| 1128 |
+
"learning_rate": 1.7795579888088243e-05,
|
| 1129 |
+
"loss": 0.6295,
|
| 1130 |
+
"step": 80000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 3.240610281389638,
|
| 1134 |
+
"grad_norm": 7.140369415283203,
|
| 1135 |
+
"learning_rate": 1.759429974638702e-05,
|
| 1136 |
+
"loss": 0.6287,
|
| 1137 |
+
"step": 80500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 3.26073829555976,
|
| 1141 |
+
"grad_norm": 16.54709243774414,
|
| 1142 |
+
"learning_rate": 1.7393019604685802e-05,
|
| 1143 |
+
"loss": 0.6416,
|
| 1144 |
+
"step": 81000
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 3.2808663097298822,
|
| 1148 |
+
"grad_norm": 13.592228889465332,
|
| 1149 |
+
"learning_rate": 1.7191739462984584e-05,
|
| 1150 |
+
"loss": 0.6234,
|
| 1151 |
+
"step": 81500
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 3.300994323900004,
|
| 1155 |
+
"grad_norm": 27.86805534362793,
|
| 1156 |
+
"learning_rate": 1.6990459321283365e-05,
|
| 1157 |
+
"loss": 0.6606,
|
| 1158 |
+
"step": 82000
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 3.321122338070126,
|
| 1162 |
+
"grad_norm": 15.816032409667969,
|
| 1163 |
+
"learning_rate": 1.6789179179582143e-05,
|
| 1164 |
+
"loss": 0.5969,
|
| 1165 |
+
"step": 82500
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 3.341250352240248,
|
| 1169 |
+
"grad_norm": 20.494857788085938,
|
| 1170 |
+
"learning_rate": 1.6587899037880924e-05,
|
| 1171 |
+
"loss": 0.6167,
|
| 1172 |
+
"step": 83000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 3.36137836641037,
|
| 1176 |
+
"grad_norm": 18.471277236938477,
|
| 1177 |
+
"learning_rate": 1.6386618896179702e-05,
|
| 1178 |
+
"loss": 0.6191,
|
| 1179 |
+
"step": 83500
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 3.381506380580492,
|
| 1183 |
+
"grad_norm": 11.292787551879883,
|
| 1184 |
+
"learning_rate": 1.6185338754478483e-05,
|
| 1185 |
+
"loss": 0.6448,
|
| 1186 |
+
"step": 84000
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 3.401634394750614,
|
| 1190 |
+
"grad_norm": 11.072359085083008,
|
| 1191 |
+
"learning_rate": 1.5984058612777265e-05,
|
| 1192 |
+
"loss": 0.6058,
|
| 1193 |
+
"step": 84500
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 3.4217624089207357,
|
| 1197 |
+
"grad_norm": 9.903491973876953,
|
| 1198 |
+
"learning_rate": 1.5782778471076046e-05,
|
| 1199 |
+
"loss": 0.6332,
|
| 1200 |
+
"step": 85000
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 3.441890423090858,
|
| 1204 |
+
"grad_norm": 6.9370269775390625,
|
| 1205 |
+
"learning_rate": 1.5581498329374824e-05,
|
| 1206 |
+
"loss": 0.5974,
|
| 1207 |
+
"step": 85500
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 3.46201843726098,
|
| 1211 |
+
"grad_norm": 7.209818363189697,
|
| 1212 |
+
"learning_rate": 1.5380218187673605e-05,
|
| 1213 |
+
"loss": 0.6163,
|
| 1214 |
+
"step": 86000
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 3.4821464514311016,
|
| 1218 |
+
"grad_norm": 11.230353355407715,
|
| 1219 |
+
"learning_rate": 1.5178938045972385e-05,
|
| 1220 |
+
"loss": 0.6124,
|
| 1221 |
+
"step": 86500
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 3.5022744656012237,
|
| 1225 |
+
"grad_norm": 9.454720497131348,
|
| 1226 |
+
"learning_rate": 1.4977657904271164e-05,
|
| 1227 |
+
"loss": 0.6029,
|
| 1228 |
+
"step": 87000
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 3.522402479771346,
|
| 1232 |
+
"grad_norm": 17.45578384399414,
|
| 1233 |
+
"learning_rate": 1.4776377762569946e-05,
|
| 1234 |
+
"loss": 0.6309,
|
| 1235 |
+
"step": 87500
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 3.5425304939414675,
|
| 1239 |
+
"grad_norm": 8.138897895812988,
|
| 1240 |
+
"learning_rate": 1.4575097620868725e-05,
|
| 1241 |
+
"loss": 0.6207,
|
| 1242 |
+
"step": 88000
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 3.5626585081115896,
|
| 1246 |
+
"grad_norm": 9.664793968200684,
|
| 1247 |
+
"learning_rate": 1.4373817479167507e-05,
|
| 1248 |
+
"loss": 0.6041,
|
| 1249 |
+
"step": 88500
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 3.5827865222817117,
|
| 1253 |
+
"grad_norm": 13.512253761291504,
|
| 1254 |
+
"learning_rate": 1.4172537337466285e-05,
|
| 1255 |
+
"loss": 0.6214,
|
| 1256 |
+
"step": 89000
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 3.602914536451834,
|
| 1260 |
+
"grad_norm": 23.86090660095215,
|
| 1261 |
+
"learning_rate": 1.3971257195765067e-05,
|
| 1262 |
+
"loss": 0.6014,
|
| 1263 |
+
"step": 89500
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 3.623042550621956,
|
| 1267 |
+
"grad_norm": 5.167603492736816,
|
| 1268 |
+
"learning_rate": 1.3769977054063845e-05,
|
| 1269 |
+
"loss": 0.6022,
|
| 1270 |
+
"step": 90000
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 3.6431705647920776,
|
| 1274 |
+
"grad_norm": 9.329811096191406,
|
| 1275 |
+
"learning_rate": 1.3568696912362627e-05,
|
| 1276 |
+
"loss": 0.6133,
|
| 1277 |
+
"step": 90500
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 3.6632985789621997,
|
| 1281 |
+
"grad_norm": 4.294738292694092,
|
| 1282 |
+
"learning_rate": 1.3367416770661406e-05,
|
| 1283 |
+
"loss": 0.6168,
|
| 1284 |
+
"step": 91000
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 3.683426593132322,
|
| 1288 |
+
"grad_norm": 9.759385108947754,
|
| 1289 |
+
"learning_rate": 1.3166136628960188e-05,
|
| 1290 |
+
"loss": 0.6101,
|
| 1291 |
+
"step": 91500
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 3.7035546073024435,
|
| 1295 |
+
"grad_norm": 10.828690528869629,
|
| 1296 |
+
"learning_rate": 1.2964856487258967e-05,
|
| 1297 |
+
"loss": 0.5941,
|
| 1298 |
+
"step": 92000
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 3.7236826214725656,
|
| 1302 |
+
"grad_norm": 17.616483688354492,
|
| 1303 |
+
"learning_rate": 1.2763576345557748e-05,
|
| 1304 |
+
"loss": 0.5971,
|
| 1305 |
+
"step": 92500
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 3.7438106356426877,
|
| 1309 |
+
"grad_norm": 14.077727317810059,
|
| 1310 |
+
"learning_rate": 1.2562296203856528e-05,
|
| 1311 |
+
"loss": 0.6198,
|
| 1312 |
+
"step": 93000
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 3.7639386498128093,
|
| 1316 |
+
"grad_norm": 17.22579002380371,
|
| 1317 |
+
"learning_rate": 1.236101606215531e-05,
|
| 1318 |
+
"loss": 0.6149,
|
| 1319 |
+
"step": 93500
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 3.7840666639829315,
|
| 1323 |
+
"grad_norm": 12.013233184814453,
|
| 1324 |
+
"learning_rate": 1.2159735920454089e-05,
|
| 1325 |
+
"loss": 0.6023,
|
| 1326 |
+
"step": 94000
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 3.8041946781530536,
|
| 1330 |
+
"grad_norm": 6.404044151306152,
|
| 1331 |
+
"learning_rate": 1.195845577875287e-05,
|
| 1332 |
+
"loss": 0.5816,
|
| 1333 |
+
"step": 94500
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 3.8243226923231752,
|
| 1337 |
+
"grad_norm": 15.49010181427002,
|
| 1338 |
+
"learning_rate": 1.1757175637051648e-05,
|
| 1339 |
+
"loss": 0.5994,
|
| 1340 |
+
"step": 95000
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.8444507064932973,
|
| 1344 |
+
"grad_norm": 8.060582160949707,
|
| 1345 |
+
"learning_rate": 1.155589549535043e-05,
|
| 1346 |
+
"loss": 0.6321,
|
| 1347 |
+
"step": 95500
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 3.8645787206634195,
|
| 1351 |
+
"grad_norm": 11.259038925170898,
|
| 1352 |
+
"learning_rate": 1.1354615353649209e-05,
|
| 1353 |
+
"loss": 0.5903,
|
| 1354 |
+
"step": 96000
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 3.884706734833541,
|
| 1358 |
+
"grad_norm": 5.94709587097168,
|
| 1359 |
+
"learning_rate": 1.1153335211947989e-05,
|
| 1360 |
+
"loss": 0.6072,
|
| 1361 |
+
"step": 96500
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 3.9048347490036632,
|
| 1365 |
+
"grad_norm": 6.936955451965332,
|
| 1366 |
+
"learning_rate": 1.095205507024677e-05,
|
| 1367 |
+
"loss": 0.5928,
|
| 1368 |
+
"step": 97000
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.9249627631737853,
|
| 1372 |
+
"grad_norm": 5.972278594970703,
|
| 1373 |
+
"learning_rate": 1.075077492854555e-05,
|
| 1374 |
+
"loss": 0.5831,
|
| 1375 |
+
"step": 97500
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 3.945090777343907,
|
| 1379 |
+
"grad_norm": 15.202763557434082,
|
| 1380 |
+
"learning_rate": 1.054949478684433e-05,
|
| 1381 |
+
"loss": 0.599,
|
| 1382 |
+
"step": 98000
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 3.965218791514029,
|
| 1386 |
+
"grad_norm": 11.337836265563965,
|
| 1387 |
+
"learning_rate": 1.034821464514311e-05,
|
| 1388 |
+
"loss": 0.5819,
|
| 1389 |
+
"step": 98500
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 3.9853468056841512,
|
| 1393 |
+
"grad_norm": 14.106744766235352,
|
| 1394 |
+
"learning_rate": 1.014693450344189e-05,
|
| 1395 |
+
"loss": 0.5854,
|
| 1396 |
+
"step": 99000
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 4.005474819854273,
|
| 1400 |
+
"grad_norm": 12.05351734161377,
|
| 1401 |
+
"learning_rate": 9.945654361740671e-06,
|
| 1402 |
+
"loss": 0.5763,
|
| 1403 |
+
"step": 99500
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 4.0256028340243954,
|
| 1407 |
+
"grad_norm": 5.722725868225098,
|
| 1408 |
+
"learning_rate": 9.744374220039451e-06,
|
| 1409 |
+
"loss": 0.546,
|
| 1410 |
+
"step": 100000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 4.045730848194517,
|
| 1414 |
+
"grad_norm": 7.079346179962158,
|
| 1415 |
+
"learning_rate": 9.543094078338232e-06,
|
| 1416 |
+
"loss": 0.5487,
|
| 1417 |
+
"step": 100500
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 4.065858862364639,
|
| 1421 |
+
"grad_norm": 11.30566692352295,
|
| 1422 |
+
"learning_rate": 9.341813936637012e-06,
|
| 1423 |
+
"loss": 0.5549,
|
| 1424 |
+
"step": 101000
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 4.085986876534761,
|
| 1428 |
+
"grad_norm": 10.975486755371094,
|
| 1429 |
+
"learning_rate": 9.140533794935791e-06,
|
| 1430 |
+
"loss": 0.557,
|
| 1431 |
+
"step": 101500
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 4.106114890704883,
|
| 1435 |
+
"grad_norm": 14.5142183303833,
|
| 1436 |
+
"learning_rate": 8.939253653234573e-06,
|
| 1437 |
+
"loss": 0.5433,
|
| 1438 |
+
"step": 102000
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 4.126242904875005,
|
| 1442 |
+
"grad_norm": 8.246289253234863,
|
| 1443 |
+
"learning_rate": 8.737973511533352e-06,
|
| 1444 |
+
"loss": 0.5534,
|
| 1445 |
+
"step": 102500
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 4.146370919045127,
|
| 1449 |
+
"grad_norm": 16.871402740478516,
|
| 1450 |
+
"learning_rate": 8.536693369832134e-06,
|
| 1451 |
+
"loss": 0.5575,
|
| 1452 |
+
"step": 103000
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 4.166498933215249,
|
| 1456 |
+
"grad_norm": 19.952173233032227,
|
| 1457 |
+
"learning_rate": 8.335413228130913e-06,
|
| 1458 |
+
"loss": 0.5349,
|
| 1459 |
+
"step": 103500
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 4.186626947385371,
|
| 1463 |
+
"grad_norm": 15.664671897888184,
|
| 1464 |
+
"learning_rate": 8.134133086429693e-06,
|
| 1465 |
+
"loss": 0.5644,
|
| 1466 |
+
"step": 104000
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 4.206754961555493,
|
| 1470 |
+
"grad_norm": 14.062329292297363,
|
| 1471 |
+
"learning_rate": 7.932852944728474e-06,
|
| 1472 |
+
"loss": 0.555,
|
| 1473 |
+
"step": 104500
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 4.226882975725615,
|
| 1477 |
+
"grad_norm": 7.628203392028809,
|
| 1478 |
+
"learning_rate": 7.731572803027254e-06,
|
| 1479 |
+
"loss": 0.5267,
|
| 1480 |
+
"step": 105000
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 4.247010989895736,
|
| 1484 |
+
"grad_norm": 7.687813758850098,
|
| 1485 |
+
"learning_rate": 7.530292661326034e-06,
|
| 1486 |
+
"loss": 0.5657,
|
| 1487 |
+
"step": 105500
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 4.267139004065859,
|
| 1491 |
+
"grad_norm": 14.242276191711426,
|
| 1492 |
+
"learning_rate": 7.3290125196248145e-06,
|
| 1493 |
+
"loss": 0.5611,
|
| 1494 |
+
"step": 106000
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 4.287267018235981,
|
| 1498 |
+
"grad_norm": 21.96781349182129,
|
| 1499 |
+
"learning_rate": 7.127732377923595e-06,
|
| 1500 |
+
"loss": 0.5523,
|
| 1501 |
+
"step": 106500
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 4.307395032406103,
|
| 1505 |
+
"grad_norm": 6.637162685394287,
|
| 1506 |
+
"learning_rate": 6.9264522362223746e-06,
|
| 1507 |
+
"loss": 0.536,
|
| 1508 |
+
"step": 107000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 4.327523046576225,
|
| 1512 |
+
"grad_norm": 24.86383628845215,
|
| 1513 |
+
"learning_rate": 6.725172094521155e-06,
|
| 1514 |
+
"loss": 0.5733,
|
| 1515 |
+
"step": 107500
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 4.3476510607463466,
|
| 1519 |
+
"grad_norm": 13.378886222839355,
|
| 1520 |
+
"learning_rate": 6.5238919528199355e-06,
|
| 1521 |
+
"loss": 0.54,
|
| 1522 |
+
"step": 108000
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 4.367779074916469,
|
| 1526 |
+
"grad_norm": 10.155076026916504,
|
| 1527 |
+
"learning_rate": 6.322611811118716e-06,
|
| 1528 |
+
"loss": 0.5431,
|
| 1529 |
+
"step": 108500
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 4.387907089086591,
|
| 1533 |
+
"grad_norm": 13.7531156539917,
|
| 1534 |
+
"learning_rate": 6.1213316694174955e-06,
|
| 1535 |
+
"loss": 0.557,
|
| 1536 |
+
"step": 109000
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 4.408035103256712,
|
| 1540 |
+
"grad_norm": 10.07307243347168,
|
| 1541 |
+
"learning_rate": 5.920051527716276e-06,
|
| 1542 |
+
"loss": 0.5467,
|
| 1543 |
+
"step": 109500
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 4.428163117426835,
|
| 1547 |
+
"grad_norm": 16.847694396972656,
|
| 1548 |
+
"learning_rate": 5.7187713860150555e-06,
|
| 1549 |
+
"loss": 0.5532,
|
| 1550 |
+
"step": 110000
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 4.448291131596957,
|
| 1554 |
+
"grad_norm": 9.273799896240234,
|
| 1555 |
+
"learning_rate": 5.517491244313836e-06,
|
| 1556 |
+
"loss": 0.5486,
|
| 1557 |
+
"step": 110500
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 4.468419145767078,
|
| 1561 |
+
"grad_norm": 18.95438575744629,
|
| 1562 |
+
"learning_rate": 5.3162111026126164e-06,
|
| 1563 |
+
"loss": 0.56,
|
| 1564 |
+
"step": 111000
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 4.488547159937201,
|
| 1568 |
+
"grad_norm": 27.988101959228516,
|
| 1569 |
+
"learning_rate": 5.114930960911397e-06,
|
| 1570 |
+
"loss": 0.5379,
|
| 1571 |
+
"step": 111500
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 4.5086751741073225,
|
| 1575 |
+
"grad_norm": 10.201438903808594,
|
| 1576 |
+
"learning_rate": 4.913650819210177e-06,
|
| 1577 |
+
"loss": 0.5415,
|
| 1578 |
+
"step": 112000
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 4.528803188277444,
|
| 1582 |
+
"grad_norm": 16.212665557861328,
|
| 1583 |
+
"learning_rate": 4.712370677508957e-06,
|
| 1584 |
+
"loss": 0.5452,
|
| 1585 |
+
"step": 112500
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 4.548931202447567,
|
| 1589 |
+
"grad_norm": 12.399296760559082,
|
| 1590 |
+
"learning_rate": 4.511090535807737e-06,
|
| 1591 |
+
"loss": 0.5208,
|
| 1592 |
+
"step": 113000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 4.569059216617688,
|
| 1596 |
+
"grad_norm": 12.065649032592773,
|
| 1597 |
+
"learning_rate": 4.309810394106518e-06,
|
| 1598 |
+
"loss": 0.5511,
|
| 1599 |
+
"step": 113500
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 4.58918723078781,
|
| 1603 |
+
"grad_norm": 22.801958084106445,
|
| 1604 |
+
"learning_rate": 4.108530252405298e-06,
|
| 1605 |
+
"loss": 0.5317,
|
| 1606 |
+
"step": 114000
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 4.609315244957933,
|
| 1610 |
+
"grad_norm": 8.57504653930664,
|
| 1611 |
+
"learning_rate": 3.907250110704079e-06,
|
| 1612 |
+
"loss": 0.5404,
|
| 1613 |
+
"step": 114500
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 4.629443259128054,
|
| 1617 |
+
"grad_norm": 12.048576354980469,
|
| 1618 |
+
"learning_rate": 3.7059699690028587e-06,
|
| 1619 |
+
"loss": 0.5244,
|
| 1620 |
+
"step": 115000
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 4.649571273298177,
|
| 1624 |
+
"grad_norm": 16.698043823242188,
|
| 1625 |
+
"learning_rate": 3.5046898273016387e-06,
|
| 1626 |
+
"loss": 0.5113,
|
| 1627 |
+
"step": 115500
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.6696992874682985,
|
| 1631 |
+
"grad_norm": 11.211584091186523,
|
| 1632 |
+
"learning_rate": 3.303409685600419e-06,
|
| 1633 |
+
"loss": 0.5594,
|
| 1634 |
+
"step": 116000
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.68982730163842,
|
| 1638 |
+
"grad_norm": 18.095754623413086,
|
| 1639 |
+
"learning_rate": 3.102129543899199e-06,
|
| 1640 |
+
"loss": 0.5549,
|
| 1641 |
+
"step": 116500
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 4.709955315808543,
|
| 1645 |
+
"grad_norm": 9.149567604064941,
|
| 1646 |
+
"learning_rate": 2.9008494021979792e-06,
|
| 1647 |
+
"loss": 0.5275,
|
| 1648 |
+
"step": 117000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 4.730083329978664,
|
| 1652 |
+
"grad_norm": 10.984237670898438,
|
| 1653 |
+
"learning_rate": 2.6995692604967597e-06,
|
| 1654 |
+
"loss": 0.5421,
|
| 1655 |
+
"step": 117500
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 4.750211344148786,
|
| 1659 |
+
"grad_norm": 11.422547340393066,
|
| 1660 |
+
"learning_rate": 2.49828911879554e-06,
|
| 1661 |
+
"loss": 0.5421,
|
| 1662 |
+
"step": 118000
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 4.770339358318909,
|
| 1666 |
+
"grad_norm": 10.895167350769043,
|
| 1667 |
+
"learning_rate": 2.2970089770943197e-06,
|
| 1668 |
+
"loss": 0.5468,
|
| 1669 |
+
"step": 118500
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 4.79046737248903,
|
| 1673 |
+
"grad_norm": 8.268088340759277,
|
| 1674 |
+
"learning_rate": 2.0957288353931e-06,
|
| 1675 |
+
"loss": 0.5333,
|
| 1676 |
+
"step": 119000
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.810595386659152,
|
| 1680 |
+
"grad_norm": 8.368425369262695,
|
| 1681 |
+
"learning_rate": 1.8944486936918804e-06,
|
| 1682 |
+
"loss": 0.5498,
|
| 1683 |
+
"step": 119500
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 4.8307234008292745,
|
| 1687 |
+
"grad_norm": 13.171350479125977,
|
| 1688 |
+
"learning_rate": 1.6931685519906606e-06,
|
| 1689 |
+
"loss": 0.5483,
|
| 1690 |
+
"step": 120000
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 4.850851414999396,
|
| 1694 |
+
"grad_norm": 10.929677963256836,
|
| 1695 |
+
"learning_rate": 1.4918884102894408e-06,
|
| 1696 |
+
"loss": 0.5368,
|
| 1697 |
+
"step": 120500
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.870979429169518,
|
| 1701 |
+
"grad_norm": 6.101780891418457,
|
| 1702 |
+
"learning_rate": 1.2906082685882213e-06,
|
| 1703 |
+
"loss": 0.5545,
|
| 1704 |
+
"step": 121000
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.89110744333964,
|
| 1708 |
+
"grad_norm": 10.938297271728516,
|
| 1709 |
+
"learning_rate": 1.0893281268870013e-06,
|
| 1710 |
+
"loss": 0.5387,
|
| 1711 |
+
"step": 121500
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 4.911235457509762,
|
| 1715 |
+
"grad_norm": 6.434581756591797,
|
| 1716 |
+
"learning_rate": 8.880479851857815e-07,
|
| 1717 |
+
"loss": 0.5317,
|
| 1718 |
+
"step": 122000
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.931363471679884,
|
| 1722 |
+
"grad_norm": 10.044371604919434,
|
| 1723 |
+
"learning_rate": 6.867678434845619e-07,
|
| 1724 |
+
"loss": 0.5261,
|
| 1725 |
+
"step": 122500
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 4.951491485850006,
|
| 1729 |
+
"grad_norm": 14.088526725769043,
|
| 1730 |
+
"learning_rate": 4.85487701783342e-07,
|
| 1731 |
+
"loss": 0.5499,
|
| 1732 |
+
"step": 123000
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 4.971619500020128,
|
| 1736 |
+
"grad_norm": 11.805033683776855,
|
| 1737 |
+
"learning_rate": 2.842075600821223e-07,
|
| 1738 |
+
"loss": 0.556,
|
| 1739 |
+
"step": 123500
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.99174751419025,
|
| 1743 |
+
"grad_norm": 12.106754302978516,
|
| 1744 |
+
"learning_rate": 8.292741838090254e-08,
|
| 1745 |
+
"loss": 0.5345,
|
| 1746 |
+
"step": 124000
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 5.0,
|
| 1750 |
+
"step": 124205,
|
| 1751 |
+
"total_flos": 2.789913716232192e+16,
|
| 1752 |
+
"train_loss": 0.945329581354701,
|
| 1753 |
+
"train_runtime": 20900.3624,
|
| 1754 |
+
"train_samples_per_second": 95.083,
|
| 1755 |
+
"train_steps_per_second": 5.943
|
| 1756 |
}
|
| 1757 |
],
|
| 1758 |
"logging_steps": 500,
|
| 1759 |
+
"max_steps": 124205,
|
| 1760 |
"num_input_tokens_seen": 0,
|
| 1761 |
+
"num_train_epochs": 5,
|
| 1762 |
"save_steps": 500,
|
| 1763 |
"stateful_callbacks": {
|
| 1764 |
"TrainerControl": {
|
|
|
|
| 1772 |
"attributes": {}
|
| 1773 |
}
|
| 1774 |
},
|
| 1775 |
+
"total_flos": 2.789913716232192e+16,
|
| 1776 |
"train_batch_size": 16,
|
| 1777 |
"trial_name": null,
|
| 1778 |
"trial_params": null
|