Initial upload of Vocalino_0.1_alpha
Browse files- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state_0.pth +1 -1
- rng_state_1.pth +1 -1
- rng_state_2.pth +1 -1
- rng_state_3.pth +1 -1
- rng_state_4.pth +1 -1
- rng_state_5.pth +1 -1
- rng_state_6.pth +1 -1
- rng_state_7.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +345 -1955
- training_args.bin +1 -1
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4991037968
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a71b6adcef4539fb5482f20abe9b1538f393f31a56c1eb1e15c2658ef980b217
|
| 3 |
size 4991037968
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1610725592
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf8f47e69590cc9d5682340b5b2db04c088225dfa08d29f5289ea537a20414d4
|
| 3 |
size 1610725592
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13203690391
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7c9e30cc6ba4e695dc260982dc19384db615b96f806cd7c31f0cb7c1aee3136
|
| 3 |
size 13203690391
|
rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8f92cf63e0989759370d24108b469c492c12202403f036015307ce49f12cedc
|
| 3 |
size 16389
|
rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ed40a0a4e9f365d2c6cc004d97e6705894eba46c8be4c160c1455bc3062dee1
|
| 3 |
size 16389
|
rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d688b304d19c260b5cfa471535ed51d7e1d60b3a0d0159dfd1a04b87904a9f42
|
| 3 |
size 16389
|
rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9967425ebcaee80d9b518fa0244d52f739b1b983d87cda71d5fede0c073e9d3b
|
| 3 |
size 16389
|
rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:469900fd39c667ffbd49c3c407c0ba317a1e9f5f9339a99b5d38423b7d0ce6d4
|
| 3 |
size 16389
|
rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:124688471ff2a6e80f2fcefedbf741fb18d08dd539d5bd07a52e81be545142a5
|
| 3 |
size 16389
|
rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e69f1ced9f992a72c948698e5eb06088610788988cdb2fdbdd624e064319d60
|
| 3 |
size 16389
|
rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a376268a55d6ee10c371c06aa952334c4c6a1af9ea2d71b1951a57367a0c6722
|
| 3 |
size 16389
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:88dc02118a954c9e13d2edebcae6232e190a54b8a4ac26352359138b0034c512
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,2433 +2,823 @@
|
|
| 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": 0.008694707097054668,
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
-
"learning_rate":
|
| 16 |
-
"loss": 4.
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.017389414194109335,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss": 4.
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.026084121291164004,
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss": 4.
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.03477882838821867,
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 4.
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.04347353548527334,
|
| 42 |
-
"grad_norm": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 4.
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.05216824258232801,
|
| 49 |
-
"grad_norm": 0.
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 4.
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.060862949679382675,
|
| 56 |
-
"grad_norm": 0.
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 4.
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.06955765677643734,
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 4.
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.078252363873492,
|
| 70 |
-
"grad_norm": 0.
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 4.
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.08694707097054669,
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 4.
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.09564177806760135,
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 4.
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.10433648516465602,
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss":
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.11303119226171068,
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss":
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.12172589935876535,
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss":
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.13042060645582002,
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss":
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.13911531355287468,
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss":
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.14781002064992935,
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss":
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.156504727746984,
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss":
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.16519943484403868,
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss":
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.17389414194109337,
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss":
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.18258884903814804,
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss":
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.1912835561352027,
|
| 161 |
-
"grad_norm": 0.
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss":
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.19997826323225737,
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss":
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.20867297032931204,
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss":
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.2173676774263667,
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss":
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.22606238452342137,
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss":
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.23475709162047603,
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss":
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.2434517987175307,
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss":
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.25214650581458536,
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss":
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.26084121291164003,
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss":
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.2695359200086947,
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss":
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.27823062710574936,
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss":
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.286925334202804,
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss":
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.2956200412998587,
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss":
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.30431474839691336,
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss":
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.313009455493968,
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss":
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.3217041625910227,
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss":
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.33039886968807736,
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss":
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.3390935767851321,
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss":
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.34778828388218674,
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.3564829909792414,
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss":
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.3651776980762961,
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss":
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.37387240517335074,
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss":
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.3825671122704054,
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss":
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.3912618193674601,
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss":
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.39995652646451474,
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss":
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.4086512335615694,
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss":
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.41734594065862407,
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss":
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.42604064775567874,
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss":
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.4347353548527334,
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss":
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.44343006194978807,
|
| 364 |
-
"grad_norm": 0.
|
| 365 |
-
"learning_rate":
|
| 366 |
-
"loss":
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.45212476904684273,
|
| 371 |
-
"grad_norm": 0.
|
| 372 |
-
"learning_rate":
|
| 373 |
-
"loss":
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.4608194761438974,
|
| 378 |
-
"grad_norm": 0.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss":
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.46951418324095207,
|
| 385 |
-
"grad_norm": 0.
|
| 386 |
-
"learning_rate":
|
| 387 |
-
"loss":
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.47820889033800673,
|
| 392 |
-
"grad_norm": 0.
|
| 393 |
-
"learning_rate":
|
| 394 |
-
"loss":
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.4869035974350614,
|
| 399 |
-
"grad_norm": 0.
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss":
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.49559830453211606,
|
| 406 |
-
"grad_norm": 0.
|
| 407 |
-
"learning_rate":
|
| 408 |
-
"loss":
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.5042930116291707,
|
| 413 |
-
"grad_norm": 0.
|
| 414 |
-
"learning_rate":
|
| 415 |
-
"loss":
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.5129877187262254,
|
| 420 |
-
"grad_norm": 0.
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss":
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.5216824258232801,
|
| 427 |
"grad_norm": 0.03662109375,
|
| 428 |
-
"learning_rate":
|
| 429 |
-
"loss":
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.5303771329203347,
|
| 434 |
-
"grad_norm": 0.
|
| 435 |
-
"learning_rate":
|
| 436 |
-
"loss":
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.5390718400173894,
|
| 441 |
-
"grad_norm": 0.
|
| 442 |
-
"learning_rate":
|
| 443 |
-
"loss":
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.5477665471144441,
|
| 448 |
-
"grad_norm": 0.
|
| 449 |
-
"learning_rate":
|
| 450 |
-
"loss":
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.5564612542114987,
|
| 455 |
-
"grad_norm": 0.
|
| 456 |
-
"learning_rate":
|
| 457 |
-
"loss":
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.5651559613085534,
|
| 462 |
-
"grad_norm": 0.
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss":
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.573850668405608,
|
| 469 |
-
"grad_norm": 0.
|
| 470 |
-
"learning_rate":
|
| 471 |
-
"loss":
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.5825453755026627,
|
| 476 |
-
"grad_norm": 0.
|
| 477 |
-
"learning_rate":
|
| 478 |
-
"loss":
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.5912400825997174,
|
| 483 |
-
"grad_norm": 0.
|
| 484 |
-
"learning_rate":
|
| 485 |
-
"loss":
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.599934789696772,
|
| 490 |
-
"grad_norm": 0.
|
| 491 |
-
"learning_rate":
|
| 492 |
-
"loss":
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.6086294967938267,
|
| 497 |
-
"grad_norm": 0.
|
| 498 |
-
"learning_rate": 3.
|
| 499 |
-
"loss":
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.6173242038908814,
|
| 504 |
-
"grad_norm": 0.
|
| 505 |
-
"learning_rate": 3.
|
| 506 |
-
"loss":
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.626018910987936,
|
| 511 |
-
"grad_norm": 0.
|
| 512 |
-
"learning_rate": 3.
|
| 513 |
-
"loss":
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.6347136180849907,
|
| 518 |
-
"grad_norm": 0.
|
| 519 |
-
"learning_rate": 3.
|
| 520 |
-
"loss":
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.6434083251820454,
|
| 525 |
-
"grad_norm": 0.
|
| 526 |
-
"learning_rate": 3.
|
| 527 |
-
"loss":
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.6521030322791,
|
| 532 |
-
"grad_norm": 0.
|
| 533 |
-
"learning_rate": 3.
|
| 534 |
-
"loss":
|
| 535 |
"step": 750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.6607977393761547,
|
| 539 |
"grad_norm": 0.037109375,
|
| 540 |
-
"learning_rate": 3.
|
| 541 |
-
"loss":
|
| 542 |
"step": 760
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.6694924464732094,
|
| 546 |
-
"grad_norm": 0.
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss":
|
| 549 |
"step": 770
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.6781871535702642,
|
| 553 |
-
"grad_norm": 0.
|
| 554 |
-
"learning_rate":
|
| 555 |
-
"loss":
|
| 556 |
"step": 780
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.6868818606673188,
|
| 560 |
-
"grad_norm": 0.
|
| 561 |
-
"learning_rate":
|
| 562 |
-
"loss":
|
| 563 |
"step": 790
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.6955765677643735,
|
| 567 |
-
"grad_norm": 0.
|
| 568 |
-
"learning_rate":
|
| 569 |
-
"loss":
|
| 570 |
"step": 800
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.7042712748614282,
|
| 574 |
"grad_norm": 0.035400390625,
|
| 575 |
-
"learning_rate":
|
| 576 |
-
"loss":
|
| 577 |
"step": 810
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.7129659819584828,
|
| 581 |
-
"grad_norm": 0.
|
| 582 |
-
"learning_rate":
|
| 583 |
-
"loss":
|
| 584 |
"step": 820
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.7216606890555375,
|
| 588 |
-
"grad_norm": 0.
|
| 589 |
-
"learning_rate":
|
| 590 |
-
"loss":
|
| 591 |
"step": 830
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.7303553961525922,
|
| 595 |
-
"grad_norm": 0.
|
| 596 |
-
"learning_rate":
|
| 597 |
-
"loss":
|
| 598 |
"step": 840
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.7390501032496468,
|
| 602 |
-
"grad_norm": 0.
|
| 603 |
-
"learning_rate":
|
| 604 |
-
"loss":
|
| 605 |
"step": 850
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.7477448103467015,
|
| 609 |
-
"grad_norm": 0.
|
| 610 |
-
"learning_rate":
|
| 611 |
-
"loss":
|
| 612 |
"step": 860
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.7564395174437561,
|
| 616 |
"grad_norm": 0.038330078125,
|
| 617 |
-
"learning_rate":
|
| 618 |
-
"loss":
|
| 619 |
"step": 870
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.7651342245408108,
|
| 623 |
-
"grad_norm": 0.
|
| 624 |
-
"learning_rate":
|
| 625 |
-
"loss":
|
| 626 |
"step": 880
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.7738289316378655,
|
| 630 |
"grad_norm": 0.036865234375,
|
| 631 |
-
"learning_rate":
|
| 632 |
-
"loss":
|
| 633 |
"step": 890
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.7825236387349201,
|
| 637 |
-
"grad_norm": 0.
|
| 638 |
-
"learning_rate":
|
| 639 |
-
"loss":
|
| 640 |
"step": 900
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.7912183458319748,
|
| 644 |
-
"grad_norm": 0.
|
| 645 |
-
"learning_rate":
|
| 646 |
-
"loss":
|
| 647 |
"step": 910
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.7999130529290295,
|
| 651 |
-
"grad_norm": 0.
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss":
|
| 654 |
"step": 920
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.8086077600260841,
|
| 658 |
-
"grad_norm": 0.
|
| 659 |
-
"learning_rate":
|
| 660 |
-
"loss":
|
| 661 |
"step": 930
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.8173024671231388,
|
| 665 |
-
"grad_norm": 0.
|
| 666 |
-
"learning_rate":
|
| 667 |
-
"loss":
|
| 668 |
"step": 940
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.8259971742201935,
|
| 672 |
-
"grad_norm": 0.
|
| 673 |
-
"learning_rate":
|
| 674 |
-
"loss":
|
| 675 |
"step": 950
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.8346918813172481,
|
| 679 |
-
"grad_norm": 0.
|
| 680 |
-
"learning_rate":
|
| 681 |
-
"loss":
|
| 682 |
"step": 960
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.8433865884143028,
|
| 686 |
-
"grad_norm": 0.
|
| 687 |
-
"learning_rate":
|
| 688 |
-
"loss":
|
| 689 |
"step": 970
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.8520812955113575,
|
| 693 |
-
"grad_norm": 0.
|
| 694 |
-
"learning_rate":
|
| 695 |
-
"loss":
|
| 696 |
"step": 980
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.8607760026084121,
|
| 700 |
-
"grad_norm": 0.
|
| 701 |
-
"learning_rate":
|
| 702 |
-
"loss":
|
| 703 |
"step": 990
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.8694707097054668,
|
| 707 |
-
"grad_norm": 0.
|
| 708 |
-
"learning_rate":
|
| 709 |
-
"loss":
|
| 710 |
"step": 1000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.8781654168025215,
|
| 714 |
-
"grad_norm": 0.
|
| 715 |
-
"learning_rate":
|
| 716 |
-
"loss":
|
| 717 |
"step": 1010
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.8868601238995761,
|
| 721 |
-
"grad_norm": 0.
|
| 722 |
-
"learning_rate": 3.
|
| 723 |
-
"loss":
|
| 724 |
"step": 1020
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.8955548309966308,
|
| 728 |
-
"grad_norm": 0.
|
| 729 |
-
"learning_rate": 3.
|
| 730 |
-
"loss":
|
| 731 |
"step": 1030
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.9042495380936855,
|
| 735 |
-
"grad_norm": 0.
|
| 736 |
-
"learning_rate":
|
| 737 |
-
"loss":
|
| 738 |
"step": 1040
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.9129442451907401,
|
| 742 |
-
"grad_norm": 0.
|
| 743 |
-
"learning_rate":
|
| 744 |
-
"loss":
|
| 745 |
"step": 1050
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.9216389522877948,
|
| 749 |
-
"grad_norm": 0.
|
| 750 |
-
"learning_rate":
|
| 751 |
-
"loss":
|
| 752 |
"step": 1060
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.9303336593848495,
|
| 756 |
-
"grad_norm": 0.
|
| 757 |
-
"learning_rate":
|
| 758 |
-
"loss":
|
| 759 |
"step": 1070
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.9390283664819041,
|
| 763 |
-
"grad_norm": 0.
|
| 764 |
-
"learning_rate":
|
| 765 |
-
"loss":
|
| 766 |
"step": 1080
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.9477230735789588,
|
| 770 |
-
"grad_norm": 0.
|
| 771 |
-
"learning_rate":
|
| 772 |
-
"loss":
|
| 773 |
"step": 1090
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.9564177806760135,
|
| 777 |
-
"grad_norm": 0.
|
| 778 |
-
"learning_rate":
|
| 779 |
-
"loss":
|
| 780 |
"step": 1100
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.9651124877730681,
|
| 784 |
-
"grad_norm": 0.
|
| 785 |
-
"learning_rate": 3.
|
| 786 |
-
"loss":
|
| 787 |
"step": 1110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.9738071948701228,
|
| 791 |
-
"grad_norm": 0.
|
| 792 |
-
"learning_rate":
|
| 793 |
-
"loss":
|
| 794 |
"step": 1120
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.9825019019671775,
|
| 798 |
-
"grad_norm": 0.
|
| 799 |
-
"learning_rate":
|
| 800 |
-
"loss":
|
| 801 |
"step": 1130
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.9911966090642321,
|
| 805 |
-
"grad_norm": 0.
|
| 806 |
-
"learning_rate": 3.
|
| 807 |
-
"loss":
|
| 808 |
"step": 1140
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.9998913161612868,
|
| 812 |
-
"grad_norm": 0.0341796875,
|
| 813 |
-
"learning_rate": 3.1092895461166534e-05,
|
| 814 |
-
"loss": 3.9182,
|
| 815 |
-
"step": 1150
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 1.0078252363873492,
|
| 819 |
-
"grad_norm": 0.03515625,
|
| 820 |
-
"learning_rate": 3.093648604393756e-05,
|
| 821 |
-
"loss": 3.8922,
|
| 822 |
-
"step": 1160
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 1.0165199434844039,
|
| 826 |
-
"grad_norm": 0.03515625,
|
| 827 |
-
"learning_rate": 3.077911654705843e-05,
|
| 828 |
-
"loss": 3.8674,
|
| 829 |
-
"step": 1170
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 1.0252146505814586,
|
| 833 |
-
"grad_norm": 0.03515625,
|
| 834 |
-
"learning_rate": 3.062080078550145e-05,
|
| 835 |
-
"loss": 3.8918,
|
| 836 |
-
"step": 1180
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 1.0339093576785132,
|
| 840 |
-
"grad_norm": 0.034423828125,
|
| 841 |
-
"learning_rate": 3.0461552657308462e-05,
|
| 842 |
-
"loss": 3.8896,
|
| 843 |
-
"step": 1190
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 1.0426040647755679,
|
| 847 |
-
"grad_norm": 0.0341796875,
|
| 848 |
-
"learning_rate": 3.0301386142370883e-05,
|
| 849 |
-
"loss": 3.8919,
|
| 850 |
-
"step": 1200
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 1.0512987718726226,
|
| 854 |
-
"grad_norm": 0.036376953125,
|
| 855 |
-
"learning_rate": 3.0140315301202386e-05,
|
| 856 |
-
"loss": 3.8883,
|
| 857 |
-
"step": 1210
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 1.0599934789696772,
|
| 861 |
-
"grad_norm": 0.038330078125,
|
| 862 |
-
"learning_rate": 2.9978354273704604e-05,
|
| 863 |
-
"loss": 3.8872,
|
| 864 |
-
"step": 1220
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 1.0686881860667319,
|
| 868 |
-
"grad_norm": 0.03515625,
|
| 869 |
-
"learning_rate": 2.981551727792582e-05,
|
| 870 |
-
"loss": 3.8726,
|
| 871 |
-
"step": 1230
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 1.0773828931637865,
|
| 875 |
-
"grad_norm": 0.03369140625,
|
| 876 |
-
"learning_rate": 2.9651818608812807e-05,
|
| 877 |
-
"loss": 3.9143,
|
| 878 |
-
"step": 1240
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 1.0860776002608412,
|
| 882 |
-
"grad_norm": 0.033447265625,
|
| 883 |
-
"learning_rate": 2.948727263695594e-05,
|
| 884 |
-
"loss": 3.8643,
|
| 885 |
-
"step": 1250
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 1.0947723073578959,
|
| 889 |
-
"grad_norm": 0.035888671875,
|
| 890 |
-
"learning_rate": 2.9321893807327613e-05,
|
| 891 |
-
"loss": 3.8943,
|
| 892 |
-
"step": 1260
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 1.1034670144549505,
|
| 896 |
-
"grad_norm": 0.03515625,
|
| 897 |
-
"learning_rate": 2.9155696638014196e-05,
|
| 898 |
-
"loss": 3.893,
|
| 899 |
-
"step": 1270
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 1.1121617215520052,
|
| 903 |
-
"grad_norm": 0.036865234375,
|
| 904 |
-
"learning_rate": 2.8988695718941503e-05,
|
| 905 |
-
"loss": 3.8933,
|
| 906 |
-
"step": 1280
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 1.1208564286490599,
|
| 910 |
-
"grad_norm": 0.035400390625,
|
| 911 |
-
"learning_rate": 2.8820905710594033e-05,
|
| 912 |
-
"loss": 3.8639,
|
| 913 |
-
"step": 1290
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 1.1295511357461145,
|
| 917 |
-
"grad_norm": 0.034912109375,
|
| 918 |
-
"learning_rate": 2.8652341342727924e-05,
|
| 919 |
-
"loss": 3.9006,
|
| 920 |
-
"step": 1300
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 1.1382458428431692,
|
| 924 |
-
"grad_norm": 0.035400390625,
|
| 925 |
-
"learning_rate": 2.84830174130779e-05,
|
| 926 |
-
"loss": 3.8713,
|
| 927 |
-
"step": 1310
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 1.1469405499402239,
|
| 931 |
-
"grad_norm": 0.0341796875,
|
| 932 |
-
"learning_rate": 2.8312948786058226e-05,
|
| 933 |
-
"loss": 3.8979,
|
| 934 |
-
"step": 1320
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 1.1556352570372785,
|
| 938 |
-
"grad_norm": 0.0341796875,
|
| 939 |
-
"learning_rate": 2.81421503914578e-05,
|
| 940 |
-
"loss": 3.8656,
|
| 941 |
-
"step": 1330
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 1.1643299641343332,
|
| 945 |
-
"grad_norm": 0.0361328125,
|
| 946 |
-
"learning_rate": 2.7970637223129523e-05,
|
| 947 |
-
"loss": 3.8645,
|
| 948 |
-
"step": 1340
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 1.1730246712313879,
|
| 952 |
-
"grad_norm": 0.03564453125,
|
| 953 |
-
"learning_rate": 2.7798424337674018e-05,
|
| 954 |
-
"loss": 3.8833,
|
| 955 |
-
"step": 1350
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 1.1817193783284425,
|
| 959 |
-
"grad_norm": 0.033935546875,
|
| 960 |
-
"learning_rate": 2.762552685311788e-05,
|
| 961 |
-
"loss": 3.8935,
|
| 962 |
-
"step": 1360
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 1.1904140854254972,
|
| 966 |
-
"grad_norm": 0.035888671875,
|
| 967 |
-
"learning_rate": 2.7451959947586487e-05,
|
| 968 |
-
"loss": 3.8626,
|
| 969 |
-
"step": 1370
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 1.1991087925225519,
|
| 973 |
"grad_norm": 0.035400390625,
|
| 974 |
-
"learning_rate":
|
| 975 |
-
"loss":
|
| 976 |
-
"step":
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 1.2078034996196065,
|
| 980 |
-
"grad_norm": 0.033203125,
|
| 981 |
-
"learning_rate": 2.7102878878593718e-05,
|
| 982 |
-
"loss": 3.8976,
|
| 983 |
-
"step": 1390
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 1.2164982067166612,
|
| 987 |
-
"grad_norm": 0.03662109375,
|
| 988 |
-
"learning_rate": 2.6927395359859466e-05,
|
| 989 |
-
"loss": 3.8638,
|
| 990 |
-
"step": 1400
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 1.2251929138137159,
|
| 994 |
-
"grad_norm": 0.036376953125,
|
| 995 |
-
"learning_rate": 2.6751303706914024e-05,
|
| 996 |
-
"loss": 3.8755,
|
| 997 |
-
"step": 1410
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 1.2338876209107705,
|
| 1001 |
-
"grad_norm": 0.035888671875,
|
| 1002 |
-
"learning_rate": 2.6574619378288756e-05,
|
| 1003 |
-
"loss": 3.8552,
|
| 1004 |
-
"step": 1420
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 1.2425823280078252,
|
| 1008 |
-
"grad_norm": 0.03564453125,
|
| 1009 |
-
"learning_rate": 2.639735788454417e-05,
|
| 1010 |
-
"loss": 3.8723,
|
| 1011 |
-
"step": 1430
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 1.2512770351048799,
|
| 1015 |
-
"grad_norm": 0.03662109375,
|
| 1016 |
-
"learning_rate": 2.621953478690826e-05,
|
| 1017 |
-
"loss": 3.8719,
|
| 1018 |
-
"step": 1440
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 1.2599717422019345,
|
| 1022 |
-
"grad_norm": 0.03564453125,
|
| 1023 |
-
"learning_rate": 2.6041165695910474e-05,
|
| 1024 |
-
"loss": 3.8929,
|
| 1025 |
-
"step": 1450
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 1.2686664492989892,
|
| 1029 |
-
"grad_norm": 0.0341796875,
|
| 1030 |
-
"learning_rate": 2.586226627001128e-05,
|
| 1031 |
-
"loss": 3.8787,
|
| 1032 |
-
"step": 1460
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 1.2773611563960439,
|
| 1036 |
-
"grad_norm": 0.0361328125,
|
| 1037 |
-
"learning_rate": 2.5682852214227595e-05,
|
| 1038 |
-
"loss": 3.8795,
|
| 1039 |
-
"step": 1470
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 1.2860558634930985,
|
| 1043 |
-
"grad_norm": 0.036376953125,
|
| 1044 |
-
"learning_rate": 2.5502939278754056e-05,
|
| 1045 |
-
"loss": 3.8857,
|
| 1046 |
-
"step": 1480
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 1.2947505705901532,
|
| 1050 |
-
"grad_norm": 0.033935546875,
|
| 1051 |
-
"learning_rate": 2.532254325758039e-05,
|
| 1052 |
-
"loss": 3.8849,
|
| 1053 |
-
"step": 1490
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 1.3034452776872079,
|
| 1057 |
-
"grad_norm": 0.033935546875,
|
| 1058 |
-
"learning_rate": 2.5141679987104874e-05,
|
| 1059 |
-
"loss": 3.8941,
|
| 1060 |
-
"step": 1500
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 1.3121399847842625,
|
| 1064 |
-
"grad_norm": 0.033935546875,
|
| 1065 |
-
"learning_rate": 2.496036534474416e-05,
|
| 1066 |
-
"loss": 3.9017,
|
| 1067 |
-
"step": 1510
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 1.3208346918813172,
|
| 1071 |
-
"grad_norm": 0.037841796875,
|
| 1072 |
-
"learning_rate": 2.4778615247539393e-05,
|
| 1073 |
-
"loss": 3.9013,
|
| 1074 |
-
"step": 1520
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 1.3295293989783719,
|
| 1078 |
-
"grad_norm": 0.03564453125,
|
| 1079 |
-
"learning_rate": 2.459644565075893e-05,
|
| 1080 |
-
"loss": 3.8722,
|
| 1081 |
-
"step": 1530
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 1.3382241060754265,
|
| 1085 |
-
"grad_norm": 0.03466796875,
|
| 1086 |
-
"learning_rate": 2.441387254649767e-05,
|
| 1087 |
-
"loss": 3.875,
|
| 1088 |
-
"step": 1540
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 1.3469188131724812,
|
| 1092 |
-
"grad_norm": 0.03466796875,
|
| 1093 |
-
"learning_rate": 2.4230911962273174e-05,
|
| 1094 |
-
"loss": 3.8724,
|
| 1095 |
-
"step": 1550
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 1.3556135202695359,
|
| 1099 |
-
"grad_norm": 0.0361328125,
|
| 1100 |
-
"learning_rate": 2.404757995961865e-05,
|
| 1101 |
-
"loss": 3.8942,
|
| 1102 |
-
"step": 1560
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 1.3643082273665905,
|
| 1106 |
-
"grad_norm": 0.035888671875,
|
| 1107 |
-
"learning_rate": 2.3863892632672946e-05,
|
| 1108 |
-
"loss": 3.8788,
|
| 1109 |
-
"step": 1570
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 1.3730029344636452,
|
| 1113 |
-
"grad_norm": 0.036865234375,
|
| 1114 |
-
"learning_rate": 2.3679866106767717e-05,
|
| 1115 |
-
"loss": 3.8805,
|
| 1116 |
-
"step": 1580
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 1.3816976415606999,
|
| 1120 |
-
"grad_norm": 0.035400390625,
|
| 1121 |
-
"learning_rate": 2.3495516537011817e-05,
|
| 1122 |
-
"loss": 3.8724,
|
| 1123 |
-
"step": 1590
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 1.3903923486577545,
|
| 1127 |
-
"grad_norm": 0.0341796875,
|
| 1128 |
-
"learning_rate": 2.331086010687312e-05,
|
| 1129 |
-
"loss": 3.8605,
|
| 1130 |
-
"step": 1600
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 1.3990870557548092,
|
| 1134 |
-
"grad_norm": 0.035888671875,
|
| 1135 |
-
"learning_rate": 2.3125913026757795e-05,
|
| 1136 |
-
"loss": 3.8631,
|
| 1137 |
-
"step": 1610
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 1.4077817628518638,
|
| 1141 |
-
"grad_norm": 0.033935546875,
|
| 1142 |
-
"learning_rate": 2.2940691532587254e-05,
|
| 1143 |
-
"loss": 3.8697,
|
| 1144 |
-
"step": 1620
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 1.4164764699489185,
|
| 1148 |
-
"grad_norm": 0.0361328125,
|
| 1149 |
-
"learning_rate": 2.275521188437286e-05,
|
| 1150 |
-
"loss": 3.8924,
|
| 1151 |
-
"step": 1630
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 1.4251711770459732,
|
| 1155 |
-
"grad_norm": 0.03515625,
|
| 1156 |
-
"learning_rate": 2.25694903647885e-05,
|
| 1157 |
-
"loss": 3.8735,
|
| 1158 |
-
"step": 1640
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 1.433865884143028,
|
| 1162 |
-
"grad_norm": 0.036865234375,
|
| 1163 |
-
"learning_rate": 2.2383543277741203e-05,
|
| 1164 |
-
"loss": 3.8695,
|
| 1165 |
-
"step": 1650
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 1.4425605912400825,
|
| 1169 |
-
"grad_norm": 0.0341796875,
|
| 1170 |
-
"learning_rate": 2.219738694693982e-05,
|
| 1171 |
-
"loss": 3.8616,
|
| 1172 |
-
"step": 1660
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 1.4512552983371374,
|
| 1176 |
-
"grad_norm": 0.034912109375,
|
| 1177 |
-
"learning_rate": 2.2011037714462097e-05,
|
| 1178 |
-
"loss": 3.8746,
|
| 1179 |
-
"step": 1670
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 1.4599500054341918,
|
| 1183 |
-
"grad_norm": 0.03564453125,
|
| 1184 |
-
"learning_rate": 2.1824511939319973e-05,
|
| 1185 |
-
"loss": 3.9004,
|
| 1186 |
-
"step": 1680
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 1.4686447125312467,
|
| 1190 |
-
"grad_norm": 0.036376953125,
|
| 1191 |
-
"learning_rate": 2.1637825996023532e-05,
|
| 1192 |
-
"loss": 3.8916,
|
| 1193 |
-
"step": 1690
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 1.4773394196283012,
|
| 1197 |
-
"grad_norm": 0.03515625,
|
| 1198 |
-
"learning_rate": 2.1450996273143505e-05,
|
| 1199 |
-
"loss": 3.8843,
|
| 1200 |
-
"step": 1700
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 1.486034126725356,
|
| 1204 |
-
"grad_norm": 0.035400390625,
|
| 1205 |
-
"learning_rate": 2.1264039171872595e-05,
|
| 1206 |
-
"loss": 3.8682,
|
| 1207 |
-
"step": 1710
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 1.4947288338224105,
|
| 1211 |
-
"grad_norm": 0.03515625,
|
| 1212 |
-
"learning_rate": 2.1076971104585627e-05,
|
| 1213 |
-
"loss": 3.8677,
|
| 1214 |
-
"step": 1720
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 1.5034235409194654,
|
| 1218 |
-
"grad_norm": 0.035400390625,
|
| 1219 |
-
"learning_rate": 2.0889808493398806e-05,
|
| 1220 |
-
"loss": 3.8765,
|
| 1221 |
-
"step": 1730
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 1.5121182480165198,
|
| 1225 |
-
"grad_norm": 0.03662109375,
|
| 1226 |
-
"learning_rate": 2.0702567768728038e-05,
|
| 1227 |
-
"loss": 3.8712,
|
| 1228 |
-
"step": 1740
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 1.5208129551135747,
|
| 1232 |
-
"grad_norm": 0.03564453125,
|
| 1233 |
-
"learning_rate": 2.051526536784656e-05,
|
| 1234 |
-
"loss": 3.8496,
|
| 1235 |
-
"step": 1750
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 1.5295076622106292,
|
| 1239 |
-
"grad_norm": 0.03564453125,
|
| 1240 |
-
"learning_rate": 2.0327917733441992e-05,
|
| 1241 |
-
"loss": 3.8995,
|
| 1242 |
-
"step": 1760
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 1.538202369307684,
|
| 1246 |
-
"grad_norm": 0.035400390625,
|
| 1247 |
-
"learning_rate": 2.0140541312172827e-05,
|
| 1248 |
-
"loss": 3.8507,
|
| 1249 |
-
"step": 1770
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 1.5468970764047385,
|
| 1253 |
-
"grad_norm": 0.03662109375,
|
| 1254 |
-
"learning_rate": 1.99531525532247e-05,
|
| 1255 |
-
"loss": 3.868,
|
| 1256 |
-
"step": 1780
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 1.5555917835017934,
|
| 1260 |
-
"grad_norm": 0.0361328125,
|
| 1261 |
-
"learning_rate": 1.9765767906866317e-05,
|
| 1262 |
-
"loss": 3.8813,
|
| 1263 |
-
"step": 1790
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 1.5642864905988478,
|
| 1267 |
-
"grad_norm": 0.03515625,
|
| 1268 |
-
"learning_rate": 1.9578403823005347e-05,
|
| 1269 |
-
"loss": 3.8748,
|
| 1270 |
-
"step": 1800
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 1.5729811976959027,
|
| 1274 |
-
"grad_norm": 0.035888671875,
|
| 1275 |
-
"learning_rate": 1.9391076749744358e-05,
|
| 1276 |
-
"loss": 3.8824,
|
| 1277 |
-
"step": 1810
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 1.5816759047929572,
|
| 1281 |
-
"grad_norm": 0.03564453125,
|
| 1282 |
-
"learning_rate": 1.9203803131936856e-05,
|
| 1283 |
-
"loss": 3.8608,
|
| 1284 |
-
"step": 1820
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 1.590370611890012,
|
| 1288 |
-
"grad_norm": 0.035888671875,
|
| 1289 |
-
"learning_rate": 1.901659940974368e-05,
|
| 1290 |
-
"loss": 3.881,
|
| 1291 |
-
"step": 1830
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 1.5990653189870665,
|
| 1295 |
-
"grad_norm": 0.03515625,
|
| 1296 |
-
"learning_rate": 1.8829482017189728e-05,
|
| 1297 |
-
"loss": 3.872,
|
| 1298 |
-
"step": 1840
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 1.6077600260841214,
|
| 1302 |
-
"grad_norm": 0.036376953125,
|
| 1303 |
-
"learning_rate": 1.864246738072132e-05,
|
| 1304 |
-
"loss": 3.8695,
|
| 1305 |
-
"step": 1850
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 1.6164547331811758,
|
| 1309 |
-
"grad_norm": 0.03515625,
|
| 1310 |
-
"learning_rate": 1.8455571917764116e-05,
|
| 1311 |
-
"loss": 3.8826,
|
| 1312 |
-
"step": 1860
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 1.6251494402782307,
|
| 1316 |
-
"grad_norm": 0.035888671875,
|
| 1317 |
-
"learning_rate": 1.8268812035281944e-05,
|
| 1318 |
-
"loss": 3.8739,
|
| 1319 |
-
"step": 1870
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 1.6338441473752852,
|
| 1323 |
-
"grad_norm": 0.03515625,
|
| 1324 |
-
"learning_rate": 1.8082204128336403e-05,
|
| 1325 |
-
"loss": 3.8691,
|
| 1326 |
-
"step": 1880
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 1.64253885447234,
|
| 1330 |
-
"grad_norm": 0.03466796875,
|
| 1331 |
-
"learning_rate": 1.789576457864768e-05,
|
| 1332 |
-
"loss": 3.8715,
|
| 1333 |
-
"step": 1890
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 1.6512335615693945,
|
| 1337 |
-
"grad_norm": 0.03466796875,
|
| 1338 |
-
"learning_rate": 1.770950975315638e-05,
|
| 1339 |
-
"loss": 3.8714,
|
| 1340 |
-
"step": 1900
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 1.6599282686664494,
|
| 1344 |
-
"grad_norm": 0.03759765625,
|
| 1345 |
-
"learning_rate": 1.7523456002586773e-05,
|
| 1346 |
-
"loss": 3.8646,
|
| 1347 |
-
"step": 1910
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 1.6686229757635038,
|
| 1351 |
-
"grad_norm": 0.033935546875,
|
| 1352 |
-
"learning_rate": 1.733761966001138e-05,
|
| 1353 |
-
"loss": 3.8885,
|
| 1354 |
-
"step": 1920
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 1.6773176828605587,
|
| 1358 |
-
"grad_norm": 0.033935546875,
|
| 1359 |
-
"learning_rate": 1.7152017039417174e-05,
|
| 1360 |
-
"loss": 3.881,
|
| 1361 |
-
"step": 1930
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 1.6860123899576132,
|
| 1365 |
-
"grad_norm": 0.0361328125,
|
| 1366 |
-
"learning_rate": 1.69666644342734e-05,
|
| 1367 |
-
"loss": 3.8763,
|
| 1368 |
-
"step": 1940
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 1.694707097054668,
|
| 1372 |
-
"grad_norm": 0.036865234375,
|
| 1373 |
-
"learning_rate": 1.678157811610124e-05,
|
| 1374 |
-
"loss": 3.9071,
|
| 1375 |
-
"step": 1950
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 1.7034018041517225,
|
| 1379 |
-
"grad_norm": 0.03564453125,
|
| 1380 |
-
"learning_rate": 1.6596774333045368e-05,
|
| 1381 |
-
"loss": 3.8678,
|
| 1382 |
-
"step": 1960
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 1.7120965112487774,
|
| 1386 |
-
"grad_norm": 0.034912109375,
|
| 1387 |
-
"learning_rate": 1.641226930844762e-05,
|
| 1388 |
-
"loss": 3.8607,
|
| 1389 |
-
"step": 1970
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 1.7207912183458318,
|
| 1393 |
-
"grad_norm": 0.035400390625,
|
| 1394 |
-
"learning_rate": 1.622807923942274e-05,
|
| 1395 |
-
"loss": 3.8679,
|
| 1396 |
-
"step": 1980
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 1.7294859254428867,
|
| 1400 |
-
"grad_norm": 0.03662109375,
|
| 1401 |
-
"learning_rate": 1.6044220295436533e-05,
|
| 1402 |
-
"loss": 3.8905,
|
| 1403 |
-
"step": 1990
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 1.7381806325399412,
|
| 1407 |
-
"grad_norm": 0.037841796875,
|
| 1408 |
-
"learning_rate": 1.586070861688636e-05,
|
| 1409 |
-
"loss": 3.8872,
|
| 1410 |
-
"step": 2000
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 1.746875339636996,
|
| 1414 |
-
"grad_norm": 0.03564453125,
|
| 1415 |
-
"learning_rate": 1.5677560313684267e-05,
|
| 1416 |
-
"loss": 3.8655,
|
| 1417 |
-
"step": 2010
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 1.7555700467340505,
|
| 1421 |
-
"grad_norm": 0.0341796875,
|
| 1422 |
-
"learning_rate": 1.5494791463842707e-05,
|
| 1423 |
-
"loss": 3.8652,
|
| 1424 |
-
"step": 2020
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 1.7642647538311054,
|
| 1428 |
-
"grad_norm": 0.0341796875,
|
| 1429 |
-
"learning_rate": 1.531241811206313e-05,
|
| 1430 |
-
"loss": 3.8559,
|
| 1431 |
-
"step": 2030
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 1.77295946092816,
|
| 1435 |
-
"grad_norm": 0.035400390625,
|
| 1436 |
-
"learning_rate": 1.5130456268327439e-05,
|
| 1437 |
-
"loss": 3.9012,
|
| 1438 |
-
"step": 2040
|
| 1439 |
-
},
|
| 1440 |
-
{
|
| 1441 |
-
"epoch": 1.7816541680252147,
|
| 1442 |
-
"grad_norm": 0.03369140625,
|
| 1443 |
-
"learning_rate": 1.4948921906492566e-05,
|
| 1444 |
-
"loss": 3.8698,
|
| 1445 |
-
"step": 2050
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 1.7903488751222694,
|
| 1449 |
-
"grad_norm": 0.03466796875,
|
| 1450 |
-
"learning_rate": 1.4767830962888128e-05,
|
| 1451 |
-
"loss": 3.8875,
|
| 1452 |
-
"step": 2060
|
| 1453 |
-
},
|
| 1454 |
-
{
|
| 1455 |
-
"epoch": 1.799043582219324,
|
| 1456 |
-
"grad_norm": 0.03564453125,
|
| 1457 |
-
"learning_rate": 1.4587199334917485e-05,
|
| 1458 |
-
"loss": 3.8823,
|
| 1459 |
-
"step": 2070
|
| 1460 |
-
},
|
| 1461 |
-
{
|
| 1462 |
-
"epoch": 1.8077382893163787,
|
| 1463 |
-
"grad_norm": 0.034423828125,
|
| 1464 |
-
"learning_rate": 1.4407042879662086e-05,
|
| 1465 |
-
"loss": 3.8793,
|
| 1466 |
-
"step": 2080
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 1.8164329964134334,
|
| 1470 |
-
"grad_norm": 0.035888671875,
|
| 1471 |
-
"learning_rate": 1.4227377412489505e-05,
|
| 1472 |
-
"loss": 3.8899,
|
| 1473 |
-
"step": 2090
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 1.825127703510488,
|
| 1477 |
-
"grad_norm": 0.03515625,
|
| 1478 |
-
"learning_rate": 1.4048218705664983e-05,
|
| 1479 |
-
"loss": 3.866,
|
| 1480 |
-
"step": 2100
|
| 1481 |
-
},
|
| 1482 |
-
{
|
| 1483 |
-
"epoch": 1.8338224106075427,
|
| 1484 |
-
"grad_norm": 0.035400390625,
|
| 1485 |
-
"learning_rate": 1.3869582486966891e-05,
|
| 1486 |
-
"loss": 3.888,
|
| 1487 |
-
"step": 2110
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 1.8425171177045974,
|
| 1491 |
-
"grad_norm": 0.03515625,
|
| 1492 |
-
"learning_rate": 1.3691484438306e-05,
|
| 1493 |
-
"loss": 3.8963,
|
| 1494 |
-
"step": 2120
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 1.851211824801652,
|
| 1498 |
-
"grad_norm": 0.036376953125,
|
| 1499 |
-
"learning_rate": 1.3513940194348844e-05,
|
| 1500 |
-
"loss": 3.8881,
|
| 1501 |
-
"step": 2130
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 1.8599065318987067,
|
| 1505 |
-
"grad_norm": 0.036376953125,
|
| 1506 |
-
"learning_rate": 1.333696534114517e-05,
|
| 1507 |
-
"loss": 3.8843,
|
| 1508 |
-
"step": 2140
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 1.8686012389957614,
|
| 1512 |
-
"grad_norm": 0.03564453125,
|
| 1513 |
-
"learning_rate": 1.3160575414759716e-05,
|
| 1514 |
-
"loss": 3.8938,
|
| 1515 |
-
"step": 2150
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 1.877295946092816,
|
| 1519 |
-
"grad_norm": 0.035888671875,
|
| 1520 |
-
"learning_rate": 1.298478589990833e-05,
|
| 1521 |
-
"loss": 3.8784,
|
| 1522 |
-
"step": 2160
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 1.8859906531898707,
|
| 1526 |
-
"grad_norm": 0.035400390625,
|
| 1527 |
-
"learning_rate": 1.2809612228598635e-05,
|
| 1528 |
-
"loss": 3.8603,
|
| 1529 |
-
"step": 2170
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 1.8946853602869254,
|
| 1533 |
-
"grad_norm": 0.034912109375,
|
| 1534 |
-
"learning_rate": 1.2635069778775276e-05,
|
| 1535 |
-
"loss": 3.8714,
|
| 1536 |
-
"step": 2180
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 1.90338006738398,
|
| 1540 |
-
"grad_norm": 0.035400390625,
|
| 1541 |
-
"learning_rate": 1.2461173872969963e-05,
|
| 1542 |
-
"loss": 3.8824,
|
| 1543 |
-
"step": 2190
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 1.9120747744810347,
|
| 1547 |
-
"grad_norm": 0.03466796875,
|
| 1548 |
-
"learning_rate": 1.2287939776956334e-05,
|
| 1549 |
-
"loss": 3.8788,
|
| 1550 |
-
"step": 2200
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 1.9207694815780894,
|
| 1554 |
-
"grad_norm": 0.036376953125,
|
| 1555 |
-
"learning_rate": 1.211538269840985e-05,
|
| 1556 |
-
"loss": 3.8888,
|
| 1557 |
-
"step": 2210
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 1.929464188675144,
|
| 1561 |
-
"grad_norm": 0.03369140625,
|
| 1562 |
-
"learning_rate": 1.1943517785572714e-05,
|
| 1563 |
-
"loss": 3.8742,
|
| 1564 |
-
"step": 2220
|
| 1565 |
-
},
|
| 1566 |
-
{
|
| 1567 |
-
"epoch": 1.9381588957721987,
|
| 1568 |
-
"grad_norm": 0.03662109375,
|
| 1569 |
-
"learning_rate": 1.1772360125924102e-05,
|
| 1570 |
-
"loss": 3.9104,
|
| 1571 |
-
"step": 2230
|
| 1572 |
-
},
|
| 1573 |
-
{
|
| 1574 |
-
"epoch": 1.9468536028692534,
|
| 1575 |
-
"grad_norm": 0.03564453125,
|
| 1576 |
-
"learning_rate": 1.1601924744855655e-05,
|
| 1577 |
-
"loss": 3.87,
|
| 1578 |
-
"step": 2240
|
| 1579 |
-
},
|
| 1580 |
-
{
|
| 1581 |
-
"epoch": 1.955548309966308,
|
| 1582 |
-
"grad_norm": 0.035400390625,
|
| 1583 |
-
"learning_rate": 1.1432226604352463e-05,
|
| 1584 |
-
"loss": 3.8769,
|
| 1585 |
-
"step": 2250
|
| 1586 |
-
},
|
| 1587 |
-
{
|
| 1588 |
-
"epoch": 1.9642430170633627,
|
| 1589 |
-
"grad_norm": 0.03564453125,
|
| 1590 |
-
"learning_rate": 1.1263280601679567e-05,
|
| 1591 |
-
"loss": 3.8673,
|
| 1592 |
-
"step": 2260
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 1.9729377241604173,
|
| 1596 |
-
"grad_norm": 0.03515625,
|
| 1597 |
-
"learning_rate": 1.1095101568074216e-05,
|
| 1598 |
-
"loss": 3.8609,
|
| 1599 |
-
"step": 2270
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 1.981632431257472,
|
| 1603 |
-
"grad_norm": 0.03466796875,
|
| 1604 |
-
"learning_rate": 1.092770426744386e-05,
|
| 1605 |
-
"loss": 3.8815,
|
| 1606 |
-
"step": 2280
|
| 1607 |
-
},
|
| 1608 |
-
{
|
| 1609 |
-
"epoch": 1.9903271383545267,
|
| 1610 |
-
"grad_norm": 0.035888671875,
|
| 1611 |
-
"learning_rate": 1.0761103395070074e-05,
|
| 1612 |
-
"loss": 3.8784,
|
| 1613 |
-
"step": 2290
|
| 1614 |
-
},
|
| 1615 |
-
{
|
| 1616 |
-
"epoch": 1.9990218454515813,
|
| 1617 |
-
"grad_norm": 0.035400390625,
|
| 1618 |
-
"learning_rate": 1.059531357631849e-05,
|
| 1619 |
-
"loss": 3.8979,
|
| 1620 |
-
"step": 2300
|
| 1621 |
-
},
|
| 1622 |
-
{
|
| 1623 |
-
"epoch": 2.0069557656776436,
|
| 1624 |
-
"grad_norm": 0.035400390625,
|
| 1625 |
-
"learning_rate": 1.0430349365354922e-05,
|
| 1626 |
-
"loss": 3.8533,
|
| 1627 |
-
"step": 2310
|
| 1628 |
-
},
|
| 1629 |
-
{
|
| 1630 |
-
"epoch": 2.0156504727746984,
|
| 1631 |
-
"grad_norm": 0.03369140625,
|
| 1632 |
-
"learning_rate": 1.0266225243867662e-05,
|
| 1633 |
-
"loss": 3.8844,
|
| 1634 |
-
"step": 2320
|
| 1635 |
-
},
|
| 1636 |
-
{
|
| 1637 |
-
"epoch": 2.024345179871753,
|
| 1638 |
-
"grad_norm": 0.03515625,
|
| 1639 |
-
"learning_rate": 1.0102955619796208e-05,
|
| 1640 |
-
"loss": 3.873,
|
| 1641 |
-
"step": 2330
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 2.0330398869688078,
|
| 1645 |
-
"grad_norm": 0.03466796875,
|
| 1646 |
-
"learning_rate": 9.9405548260664e-06,
|
| 1647 |
-
"loss": 3.8934,
|
| 1648 |
-
"step": 2340
|
| 1649 |
-
},
|
| 1650 |
-
{
|
| 1651 |
-
"epoch": 2.041734594065862,
|
| 1652 |
-
"grad_norm": 0.03515625,
|
| 1653 |
-
"learning_rate": 9.779037119332223e-06,
|
| 1654 |
-
"loss": 3.8802,
|
| 1655 |
-
"step": 2350
|
| 1656 |
-
},
|
| 1657 |
-
{
|
| 1658 |
-
"epoch": 2.050429301162917,
|
| 1659 |
-
"grad_norm": 0.036865234375,
|
| 1660 |
-
"learning_rate": 9.618416678724205e-06,
|
| 1661 |
-
"loss": 3.889,
|
| 1662 |
-
"step": 2360
|
| 1663 |
-
},
|
| 1664 |
-
{
|
| 1665 |
-
"epoch": 2.0591240082599715,
|
| 1666 |
-
"grad_norm": 0.0361328125,
|
| 1667 |
-
"learning_rate": 9.458707604604764e-06,
|
| 1668 |
-
"loss": 3.8743,
|
| 1669 |
-
"step": 2370
|
| 1670 |
-
},
|
| 1671 |
-
{
|
| 1672 |
-
"epoch": 2.0678187153570264,
|
| 1673 |
-
"grad_norm": 0.035400390625,
|
| 1674 |
-
"learning_rate": 9.299923917330265e-06,
|
| 1675 |
-
"loss": 3.8706,
|
| 1676 |
-
"step": 2380
|
| 1677 |
-
},
|
| 1678 |
-
{
|
| 1679 |
-
"epoch": 2.076513422454081,
|
| 1680 |
-
"grad_norm": 0.03466796875,
|
| 1681 |
-
"learning_rate": 9.14207955602032e-06,
|
| 1682 |
-
"loss": 3.8572,
|
| 1683 |
-
"step": 2390
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 2.0852081295511358,
|
| 1687 |
-
"grad_norm": 0.03369140625,
|
| 1688 |
-
"learning_rate": 8.985188377334044e-06,
|
| 1689 |
-
"loss": 3.8926,
|
| 1690 |
-
"step": 2400
|
| 1691 |
-
},
|
| 1692 |
-
{
|
| 1693 |
-
"epoch": 2.09390283664819,
|
| 1694 |
-
"grad_norm": 0.03515625,
|
| 1695 |
-
"learning_rate": 8.829264154253673e-06,
|
| 1696 |
-
"loss": 3.8978,
|
| 1697 |
-
"step": 2410
|
| 1698 |
-
},
|
| 1699 |
-
{
|
| 1700 |
-
"epoch": 2.102597543745245,
|
| 1701 |
-
"grad_norm": 0.0341796875,
|
| 1702 |
-
"learning_rate": 8.674320574875456e-06,
|
| 1703 |
-
"loss": 3.8852,
|
| 1704 |
-
"step": 2420
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 2.1112922508422995,
|
| 1708 |
-
"grad_norm": 0.03564453125,
|
| 1709 |
-
"learning_rate": 8.520371241208019e-06,
|
| 1710 |
-
"loss": 3.8864,
|
| 1711 |
-
"step": 2430
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 2.1199869579393544,
|
| 1715 |
-
"grad_norm": 0.035400390625,
|
| 1716 |
-
"learning_rate": 8.367429667978275e-06,
|
| 1717 |
-
"loss": 3.8833,
|
| 1718 |
-
"step": 2440
|
| 1719 |
-
},
|
| 1720 |
-
{
|
| 1721 |
-
"epoch": 2.1286816650364093,
|
| 1722 |
-
"grad_norm": 0.036865234375,
|
| 1723 |
-
"learning_rate": 8.215509281445043e-06,
|
| 1724 |
-
"loss": 3.8611,
|
| 1725 |
-
"step": 2450
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 2.1373763721334638,
|
| 1729 |
-
"grad_norm": 0.03564453125,
|
| 1730 |
-
"learning_rate": 8.06462341822037e-06,
|
| 1731 |
-
"loss": 3.8443,
|
| 1732 |
-
"step": 2460
|
| 1733 |
-
},
|
| 1734 |
-
{
|
| 1735 |
-
"epoch": 2.146071079230518,
|
| 1736 |
-
"grad_norm": 0.035400390625,
|
| 1737 |
-
"learning_rate": 7.914785324098775e-06,
|
| 1738 |
-
"loss": 3.8879,
|
| 1739 |
-
"step": 2470
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 2.154765786327573,
|
| 1743 |
-
"grad_norm": 0.034423828125,
|
| 1744 |
-
"learning_rate": 7.7660081528944e-06,
|
| 1745 |
-
"loss": 3.8596,
|
| 1746 |
-
"step": 2480
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 2.163460493424628,
|
| 1750 |
-
"grad_norm": 0.03515625,
|
| 1751 |
-
"learning_rate": 7.618304965286334e-06,
|
| 1752 |
-
"loss": 3.8491,
|
| 1753 |
-
"step": 2490
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 2.1721552005216824,
|
| 1757 |
-
"grad_norm": 0.037109375,
|
| 1758 |
-
"learning_rate": 7.471688727672022e-06,
|
| 1759 |
-
"loss": 3.901,
|
| 1760 |
-
"step": 2500
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 2.180849907618737,
|
| 1764 |
-
"grad_norm": 0.0390625,
|
| 1765 |
-
"learning_rate": 7.326172311028996e-06,
|
| 1766 |
-
"loss": 3.865,
|
| 1767 |
-
"step": 2510
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 2.1895446147157918,
|
| 1771 |
-
"grad_norm": 0.03515625,
|
| 1772 |
-
"learning_rate": 7.181768489784957e-06,
|
| 1773 |
-
"loss": 3.8544,
|
| 1774 |
-
"step": 2520
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 2.1982393218128466,
|
| 1778 |
-
"grad_norm": 0.035400390625,
|
| 1779 |
-
"learning_rate": 7.038489940696375e-06,
|
| 1780 |
-
"loss": 3.8794,
|
| 1781 |
-
"step": 2530
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 2.206934028909901,
|
| 1785 |
-
"grad_norm": 0.03662109375,
|
| 1786 |
-
"learning_rate": 6.8963492417356224e-06,
|
| 1787 |
-
"loss": 3.863,
|
| 1788 |
-
"step": 2540
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 2.2156287360069555,
|
| 1792 |
-
"grad_norm": 0.03515625,
|
| 1793 |
-
"learning_rate": 6.755358870986797e-06,
|
| 1794 |
-
"loss": 3.8727,
|
| 1795 |
-
"step": 2550
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 2.2243234431040104,
|
| 1799 |
-
"grad_norm": 0.03515625,
|
| 1800 |
-
"learning_rate": 6.615531205550288e-06,
|
| 1801 |
-
"loss": 3.8702,
|
| 1802 |
-
"step": 2560
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 2.2330181502010653,
|
| 1806 |
-
"grad_norm": 0.0361328125,
|
| 1807 |
-
"learning_rate": 6.476878520456278e-06,
|
| 1808 |
-
"loss": 3.8891,
|
| 1809 |
-
"step": 2570
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 2.2417128572981198,
|
| 1813 |
-
"grad_norm": 0.0341796875,
|
| 1814 |
-
"learning_rate": 6.339412987587088e-06,
|
| 1815 |
-
"loss": 3.8554,
|
| 1816 |
-
"step": 2580
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 2.250407564395174,
|
| 1820 |
-
"grad_norm": 0.034423828125,
|
| 1821 |
-
"learning_rate": 6.203146674608742e-06,
|
| 1822 |
-
"loss": 3.8889,
|
| 1823 |
-
"step": 2590
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 2.259102271492229,
|
| 1827 |
-
"grad_norm": 0.03466796875,
|
| 1828 |
-
"learning_rate": 6.068091543911472e-06,
|
| 1829 |
-
"loss": 3.875,
|
| 1830 |
-
"step": 2600
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 2.267796978589284,
|
| 1834 |
-
"grad_norm": 0.033203125,
|
| 1835 |
-
"learning_rate": 5.9342594515596716e-06,
|
| 1836 |
-
"loss": 3.8745,
|
| 1837 |
-
"step": 2610
|
| 1838 |
-
},
|
| 1839 |
-
{
|
| 1840 |
-
"epoch": 2.2764916856863384,
|
| 1841 |
-
"grad_norm": 0.035888671875,
|
| 1842 |
-
"learning_rate": 5.8016621462510305e-06,
|
| 1843 |
-
"loss": 3.9004,
|
| 1844 |
-
"step": 2620
|
| 1845 |
-
},
|
| 1846 |
-
{
|
| 1847 |
-
"epoch": 2.285186392783393,
|
| 1848 |
-
"grad_norm": 0.033935546875,
|
| 1849 |
-
"learning_rate": 5.670311268285183e-06,
|
| 1850 |
-
"loss": 3.8669,
|
| 1851 |
-
"step": 2630
|
| 1852 |
-
},
|
| 1853 |
-
{
|
| 1854 |
-
"epoch": 2.2938810998804477,
|
| 1855 |
-
"grad_norm": 0.0361328125,
|
| 1856 |
-
"learning_rate": 5.540218348541837e-06,
|
| 1857 |
-
"loss": 3.8654,
|
| 1858 |
-
"step": 2640
|
| 1859 |
-
},
|
| 1860 |
-
{
|
| 1861 |
-
"epoch": 2.3025758069775026,
|
| 1862 |
-
"grad_norm": 0.0341796875,
|
| 1863 |
-
"learning_rate": 5.411394807468513e-06,
|
| 1864 |
-
"loss": 3.871,
|
| 1865 |
-
"step": 2650
|
| 1866 |
-
},
|
| 1867 |
-
{
|
| 1868 |
-
"epoch": 2.311270514074557,
|
| 1869 |
-
"grad_norm": 0.03515625,
|
| 1870 |
-
"learning_rate": 5.283851954077961e-06,
|
| 1871 |
-
"loss": 3.8598,
|
| 1872 |
-
"step": 2660
|
| 1873 |
-
},
|
| 1874 |
-
{
|
| 1875 |
-
"epoch": 2.319965221171612,
|
| 1876 |
-
"grad_norm": 0.035400390625,
|
| 1877 |
-
"learning_rate": 5.157600984955413e-06,
|
| 1878 |
-
"loss": 3.8763,
|
| 1879 |
-
"step": 2670
|
| 1880 |
-
},
|
| 1881 |
-
{
|
| 1882 |
-
"epoch": 2.3286599282686664,
|
| 1883 |
-
"grad_norm": 0.034912109375,
|
| 1884 |
-
"learning_rate": 5.032652983275645e-06,
|
| 1885 |
-
"loss": 3.8473,
|
| 1886 |
-
"step": 2680
|
| 1887 |
-
},
|
| 1888 |
-
{
|
| 1889 |
-
"epoch": 2.3373546353657213,
|
| 1890 |
-
"grad_norm": 0.03662109375,
|
| 1891 |
-
"learning_rate": 4.909018917830035e-06,
|
| 1892 |
-
"loss": 3.861,
|
| 1893 |
-
"step": 2690
|
| 1894 |
-
},
|
| 1895 |
-
{
|
| 1896 |
-
"epoch": 2.3460493424627757,
|
| 1897 |
-
"grad_norm": 0.03564453125,
|
| 1898 |
-
"learning_rate": 4.786709642063618e-06,
|
| 1899 |
-
"loss": 3.8676,
|
| 1900 |
-
"step": 2700
|
| 1901 |
-
},
|
| 1902 |
-
{
|
| 1903 |
-
"epoch": 2.3547440495598306,
|
| 1904 |
-
"grad_norm": 0.035888671875,
|
| 1905 |
-
"learning_rate": 4.665735893122338e-06,
|
| 1906 |
-
"loss": 3.858,
|
| 1907 |
-
"step": 2710
|
| 1908 |
-
},
|
| 1909 |
-
{
|
| 1910 |
-
"epoch": 2.363438756656885,
|
| 1911 |
-
"grad_norm": 0.033935546875,
|
| 1912 |
-
"learning_rate": 4.546108290910438e-06,
|
| 1913 |
-
"loss": 3.8869,
|
| 1914 |
-
"step": 2720
|
| 1915 |
-
},
|
| 1916 |
-
{
|
| 1917 |
-
"epoch": 2.37213346375394,
|
| 1918 |
-
"grad_norm": 0.036376953125,
|
| 1919 |
-
"learning_rate": 4.427837337158187e-06,
|
| 1920 |
-
"loss": 3.882,
|
| 1921 |
-
"step": 2730
|
| 1922 |
-
},
|
| 1923 |
-
{
|
| 1924 |
-
"epoch": 2.3808281708509944,
|
| 1925 |
-
"grad_norm": 0.0341796875,
|
| 1926 |
-
"learning_rate": 4.310933414499949e-06,
|
| 1927 |
-
"loss": 3.8742,
|
| 1928 |
-
"step": 2740
|
| 1929 |
-
},
|
| 1930 |
-
{
|
| 1931 |
-
"epoch": 2.3895228779480493,
|
| 1932 |
-
"grad_norm": 0.035888671875,
|
| 1933 |
-
"learning_rate": 4.1954067855627476e-06,
|
| 1934 |
-
"loss": 3.858,
|
| 1935 |
-
"step": 2750
|
| 1936 |
-
},
|
| 1937 |
-
{
|
| 1938 |
-
"epoch": 2.3982175850451037,
|
| 1939 |
-
"grad_norm": 0.035888671875,
|
| 1940 |
-
"learning_rate": 4.0812675920653236e-06,
|
| 1941 |
-
"loss": 3.8501,
|
| 1942 |
-
"step": 2760
|
| 1943 |
-
},
|
| 1944 |
-
{
|
| 1945 |
-
"epoch": 2.4069122921421586,
|
| 1946 |
-
"grad_norm": 0.03662109375,
|
| 1947 |
-
"learning_rate": 3.968525853927841e-06,
|
| 1948 |
-
"loss": 3.8677,
|
| 1949 |
-
"step": 2770
|
| 1950 |
-
},
|
| 1951 |
-
{
|
| 1952 |
-
"epoch": 2.415606999239213,
|
| 1953 |
-
"grad_norm": 0.035400390625,
|
| 1954 |
-
"learning_rate": 3.857191468392241e-06,
|
| 1955 |
-
"loss": 3.8632,
|
| 1956 |
-
"step": 2780
|
| 1957 |
-
},
|
| 1958 |
-
{
|
| 1959 |
-
"epoch": 2.424301706336268,
|
| 1960 |
-
"grad_norm": 0.03564453125,
|
| 1961 |
-
"learning_rate": 3.747274209153442e-06,
|
| 1962 |
-
"loss": 3.8909,
|
| 1963 |
-
"step": 2790
|
| 1964 |
-
},
|
| 1965 |
-
{
|
| 1966 |
-
"epoch": 2.4329964134333224,
|
| 1967 |
-
"grad_norm": 0.036376953125,
|
| 1968 |
-
"learning_rate": 3.6387837255012783e-06,
|
| 1969 |
-
"loss": 3.8596,
|
| 1970 |
-
"step": 2800
|
| 1971 |
-
},
|
| 1972 |
-
{
|
| 1973 |
-
"epoch": 2.4416911205303773,
|
| 1974 |
-
"grad_norm": 0.034423828125,
|
| 1975 |
-
"learning_rate": 3.5317295414734875e-06,
|
| 1976 |
-
"loss": 3.8785,
|
| 1977 |
-
"step": 2810
|
| 1978 |
-
},
|
| 1979 |
-
{
|
| 1980 |
-
"epoch": 2.4503858276274317,
|
| 1981 |
-
"grad_norm": 0.035400390625,
|
| 1982 |
-
"learning_rate": 3.426121055019556e-06,
|
| 1983 |
-
"loss": 3.876,
|
| 1984 |
-
"step": 2820
|
| 1985 |
-
},
|
| 1986 |
-
{
|
| 1987 |
-
"epoch": 2.4590805347244866,
|
| 1988 |
-
"grad_norm": 0.03466796875,
|
| 1989 |
-
"learning_rate": 3.321967537175759e-06,
|
| 1990 |
-
"loss": 3.8893,
|
| 1991 |
-
"step": 2830
|
| 1992 |
-
},
|
| 1993 |
-
{
|
| 1994 |
-
"epoch": 2.467775241821541,
|
| 1995 |
-
"grad_norm": 0.035400390625,
|
| 1996 |
-
"learning_rate": 3.21927813125124e-06,
|
| 1997 |
-
"loss": 3.8601,
|
| 1998 |
-
"step": 2840
|
| 1999 |
-
},
|
| 2000 |
-
{
|
| 2001 |
-
"epoch": 2.476469948918596,
|
| 2002 |
-
"grad_norm": 0.03515625,
|
| 2003 |
-
"learning_rate": 3.1180618520253893e-06,
|
| 2004 |
-
"loss": 3.8608,
|
| 2005 |
-
"step": 2850
|
| 2006 |
-
},
|
| 2007 |
-
{
|
| 2008 |
-
"epoch": 2.4851646560156504,
|
| 2009 |
-
"grad_norm": 0.03515625,
|
| 2010 |
-
"learning_rate": 3.0183275849564396e-06,
|
| 2011 |
-
"loss": 3.8894,
|
| 2012 |
-
"step": 2860
|
| 2013 |
-
},
|
| 2014 |
-
{
|
| 2015 |
-
"epoch": 2.4938593631127053,
|
| 2016 |
-
"grad_norm": 0.03515625,
|
| 2017 |
-
"learning_rate": 2.9200840854014466e-06,
|
| 2018 |
-
"loss": 3.8522,
|
| 2019 |
-
"step": 2870
|
| 2020 |
-
},
|
| 2021 |
-
{
|
| 2022 |
-
"epoch": 2.5025540702097597,
|
| 2023 |
-
"grad_norm": 0.03515625,
|
| 2024 |
-
"learning_rate": 2.8233399778476744e-06,
|
| 2025 |
-
"loss": 3.8801,
|
| 2026 |
-
"step": 2880
|
| 2027 |
-
},
|
| 2028 |
-
{
|
| 2029 |
-
"epoch": 2.5112487773068146,
|
| 2030 |
-
"grad_norm": 0.0361328125,
|
| 2031 |
-
"learning_rate": 2.7281037551555e-06,
|
| 2032 |
-
"loss": 3.8751,
|
| 2033 |
-
"step": 2890
|
| 2034 |
-
},
|
| 2035 |
-
{
|
| 2036 |
-
"epoch": 2.519943484403869,
|
| 2037 |
-
"grad_norm": 0.03564453125,
|
| 2038 |
-
"learning_rate": 2.6343837778128366e-06,
|
| 2039 |
-
"loss": 3.863,
|
| 2040 |
-
"step": 2900
|
| 2041 |
-
},
|
| 2042 |
-
{
|
| 2043 |
-
"epoch": 2.528638191500924,
|
| 2044 |
-
"grad_norm": 0.03515625,
|
| 2045 |
-
"learning_rate": 2.5421882732011937e-06,
|
| 2046 |
-
"loss": 3.8768,
|
| 2047 |
-
"step": 2910
|
| 2048 |
-
},
|
| 2049 |
-
{
|
| 2050 |
-
"epoch": 2.5373328985979784,
|
| 2051 |
-
"grad_norm": 0.0361328125,
|
| 2052 |
-
"learning_rate": 2.4515253348734192e-06,
|
| 2053 |
-
"loss": 3.911,
|
| 2054 |
-
"step": 2920
|
| 2055 |
-
},
|
| 2056 |
-
{
|
| 2057 |
-
"epoch": 2.5460276056950333,
|
| 2058 |
-
"grad_norm": 0.034912109375,
|
| 2059 |
-
"learning_rate": 2.362402921843201e-06,
|
| 2060 |
-
"loss": 3.8806,
|
| 2061 |
-
"step": 2930
|
| 2062 |
-
},
|
| 2063 |
-
{
|
| 2064 |
-
"epoch": 2.5547223127920877,
|
| 2065 |
-
"grad_norm": 0.034912109375,
|
| 2066 |
-
"learning_rate": 2.2748288578863664e-06,
|
| 2067 |
-
"loss": 3.8654,
|
| 2068 |
-
"step": 2940
|
| 2069 |
-
},
|
| 2070 |
-
{
|
| 2071 |
-
"epoch": 2.5634170198891426,
|
| 2072 |
-
"grad_norm": 0.035888671875,
|
| 2073 |
-
"learning_rate": 2.188810830854058e-06,
|
| 2074 |
-
"loss": 3.8697,
|
| 2075 |
-
"step": 2950
|
| 2076 |
-
},
|
| 2077 |
-
{
|
| 2078 |
-
"epoch": 2.572111726986197,
|
| 2079 |
-
"grad_norm": 0.034423828125,
|
| 2080 |
-
"learning_rate": 2.104356391997826e-06,
|
| 2081 |
-
"loss": 3.8804,
|
| 2082 |
-
"step": 2960
|
| 2083 |
-
},
|
| 2084 |
-
{
|
| 2085 |
-
"epoch": 2.580806434083252,
|
| 2086 |
-
"grad_norm": 0.035888671875,
|
| 2087 |
-
"learning_rate": 2.02147295530676e-06,
|
| 2088 |
-
"loss": 3.8637,
|
| 2089 |
-
"step": 2970
|
| 2090 |
-
},
|
| 2091 |
-
{
|
| 2092 |
-
"epoch": 2.5895011411803064,
|
| 2093 |
-
"grad_norm": 0.03564453125,
|
| 2094 |
-
"learning_rate": 1.940167796856598e-06,
|
| 2095 |
-
"loss": 3.8623,
|
| 2096 |
-
"step": 2980
|
| 2097 |
-
},
|
| 2098 |
-
{
|
| 2099 |
-
"epoch": 2.5981958482773613,
|
| 2100 |
-
"grad_norm": 0.03564453125,
|
| 2101 |
-
"learning_rate": 1.8604480541710357e-06,
|
| 2102 |
-
"loss": 3.8593,
|
| 2103 |
-
"step": 2990
|
| 2104 |
-
},
|
| 2105 |
-
{
|
| 2106 |
-
"epoch": 2.6068905553744157,
|
| 2107 |
-
"grad_norm": 0.0341796875,
|
| 2108 |
-
"learning_rate": 1.7823207255950792e-06,
|
| 2109 |
-
"loss": 3.8711,
|
| 2110 |
-
"step": 3000
|
| 2111 |
-
},
|
| 2112 |
-
{
|
| 2113 |
-
"epoch": 2.6155852624714706,
|
| 2114 |
-
"grad_norm": 0.035888671875,
|
| 2115 |
-
"learning_rate": 1.7057926696807437e-06,
|
| 2116 |
-
"loss": 3.8767,
|
| 2117 |
-
"step": 3010
|
| 2118 |
-
},
|
| 2119 |
-
{
|
| 2120 |
-
"epoch": 2.624279969568525,
|
| 2121 |
-
"grad_norm": 0.036376953125,
|
| 2122 |
-
"learning_rate": 1.6308706045849154e-06,
|
| 2123 |
-
"loss": 3.8813,
|
| 2124 |
-
"step": 3020
|
| 2125 |
-
},
|
| 2126 |
-
{
|
| 2127 |
-
"epoch": 2.63297467666558,
|
| 2128 |
-
"grad_norm": 0.034912109375,
|
| 2129 |
-
"learning_rate": 1.5575611074796326e-06,
|
| 2130 |
-
"loss": 3.8709,
|
| 2131 |
-
"step": 3030
|
| 2132 |
-
},
|
| 2133 |
-
{
|
| 2134 |
-
"epoch": 2.6416693837626344,
|
| 2135 |
-
"grad_norm": 0.0361328125,
|
| 2136 |
-
"learning_rate": 1.4858706139746449e-06,
|
| 2137 |
-
"loss": 3.8536,
|
| 2138 |
-
"step": 3040
|
| 2139 |
-
},
|
| 2140 |
-
{
|
| 2141 |
-
"epoch": 2.6503640908596893,
|
| 2142 |
-
"grad_norm": 0.03515625,
|
| 2143 |
-
"learning_rate": 1.415805417552496e-06,
|
| 2144 |
-
"loss": 3.8704,
|
| 2145 |
-
"step": 3050
|
| 2146 |
-
},
|
| 2147 |
-
{
|
| 2148 |
-
"epoch": 2.6590587979567437,
|
| 2149 |
-
"grad_norm": 0.03662109375,
|
| 2150 |
-
"learning_rate": 1.3473716690160022e-06,
|
| 2151 |
-
"loss": 3.8812,
|
| 2152 |
-
"step": 3060
|
| 2153 |
-
},
|
| 2154 |
-
{
|
| 2155 |
-
"epoch": 2.6677535050537986,
|
| 2156 |
-
"grad_norm": 0.03515625,
|
| 2157 |
-
"learning_rate": 1.2805753759483296e-06,
|
| 2158 |
-
"loss": 3.8587,
|
| 2159 |
-
"step": 3070
|
| 2160 |
-
},
|
| 2161 |
-
{
|
| 2162 |
-
"epoch": 2.676448212150853,
|
| 2163 |
-
"grad_norm": 0.03515625,
|
| 2164 |
-
"learning_rate": 1.2154224021855733e-06,
|
| 2165 |
-
"loss": 3.855,
|
| 2166 |
-
"step": 3080
|
| 2167 |
-
},
|
| 2168 |
-
{
|
| 2169 |
-
"epoch": 2.685142919247908,
|
| 2170 |
-
"grad_norm": 0.0361328125,
|
| 2171 |
-
"learning_rate": 1.1519184673020156e-06,
|
| 2172 |
-
"loss": 3.892,
|
| 2173 |
-
"step": 3090
|
| 2174 |
-
},
|
| 2175 |
-
{
|
| 2176 |
-
"epoch": 2.6938376263449624,
|
| 2177 |
-
"grad_norm": 0.032958984375,
|
| 2178 |
-
"learning_rate": 1.0900691461080016e-06,
|
| 2179 |
-
"loss": 3.8608,
|
| 2180 |
-
"step": 3100
|
| 2181 |
-
},
|
| 2182 |
-
{
|
| 2183 |
-
"epoch": 2.7025323334420173,
|
| 2184 |
-
"grad_norm": 0.0341796875,
|
| 2185 |
-
"learning_rate": 1.0298798681605592e-06,
|
| 2186 |
-
"loss": 3.8663,
|
| 2187 |
-
"step": 3110
|
| 2188 |
-
},
|
| 2189 |
-
{
|
| 2190 |
-
"epoch": 2.7112270405390717,
|
| 2191 |
-
"grad_norm": 0.034423828125,
|
| 2192 |
-
"learning_rate": 9.713559172867492e-07,
|
| 2193 |
-
"loss": 3.8662,
|
| 2194 |
-
"step": 3120
|
| 2195 |
-
},
|
| 2196 |
-
{
|
| 2197 |
-
"epoch": 2.7199217476361266,
|
| 2198 |
-
"grad_norm": 0.035400390625,
|
| 2199 |
-
"learning_rate": 9.145024311198214e-07,
|
| 2200 |
-
"loss": 3.8786,
|
| 2201 |
-
"step": 3130
|
| 2202 |
-
},
|
| 2203 |
-
{
|
| 2204 |
-
"epoch": 2.728616454733181,
|
| 2205 |
-
"grad_norm": 0.03662109375,
|
| 2206 |
-
"learning_rate": 8.593244006481805e-07,
|
| 2207 |
-
"loss": 3.8842,
|
| 2208 |
-
"step": 3140
|
| 2209 |
-
},
|
| 2210 |
-
{
|
| 2211 |
-
"epoch": 2.737311161830236,
|
| 2212 |
-
"grad_norm": 0.035400390625,
|
| 2213 |
-
"learning_rate": 8.058266697772654e-07,
|
| 2214 |
-
"loss": 3.8759,
|
| 2215 |
-
"step": 3150
|
| 2216 |
-
},
|
| 2217 |
-
{
|
| 2218 |
-
"epoch": 2.7460058689272904,
|
| 2219 |
-
"grad_norm": 0.036865234375,
|
| 2220 |
-
"learning_rate": 7.54013934904303e-07,
|
| 2221 |
-
"loss": 3.8871,
|
| 2222 |
-
"step": 3160
|
| 2223 |
-
},
|
| 2224 |
-
{
|
| 2225 |
-
"epoch": 2.7547005760243453,
|
| 2226 |
-
"grad_norm": 0.0361328125,
|
| 2227 |
-
"learning_rate": 7.038907445060328e-07,
|
| 2228 |
-
"loss": 3.8612,
|
| 2229 |
-
"step": 3170
|
| 2230 |
-
},
|
| 2231 |
-
{
|
| 2232 |
-
"epoch": 2.7633952831213997,
|
| 2233 |
-
"grad_norm": 0.03466796875,
|
| 2234 |
-
"learning_rate": 6.55461498739407e-07,
|
| 2235 |
-
"loss": 3.8801,
|
| 2236 |
-
"step": 3180
|
| 2237 |
-
},
|
| 2238 |
-
{
|
| 2239 |
-
"epoch": 2.7720899902184546,
|
| 2240 |
-
"grad_norm": 0.035400390625,
|
| 2241 |
-
"learning_rate": 6.087304490553148e-07,
|
| 2242 |
-
"loss": 3.8663,
|
| 2243 |
-
"step": 3190
|
| 2244 |
-
},
|
| 2245 |
-
{
|
| 2246 |
-
"epoch": 2.780784697315509,
|
| 2247 |
-
"grad_norm": 0.03759765625,
|
| 2248 |
-
"learning_rate": 5.637016978253607e-07,
|
| 2249 |
-
"loss": 3.8537,
|
| 2250 |
-
"step": 3200
|
| 2251 |
-
},
|
| 2252 |
-
{
|
| 2253 |
-
"epoch": 2.789479404412564,
|
| 2254 |
-
"grad_norm": 0.03564453125,
|
| 2255 |
-
"learning_rate": 5.203791979817441e-07,
|
| 2256 |
-
"loss": 3.8857,
|
| 2257 |
-
"step": 3210
|
| 2258 |
-
},
|
| 2259 |
-
{
|
| 2260 |
-
"epoch": 2.7981741115096184,
|
| 2261 |
-
"grad_norm": 0.034423828125,
|
| 2262 |
-
"learning_rate": 4.787667526702122e-07,
|
| 2263 |
-
"loss": 3.8555,
|
| 2264 |
-
"step": 3220
|
| 2265 |
-
},
|
| 2266 |
-
{
|
| 2267 |
-
"epoch": 2.8068688186066733,
|
| 2268 |
-
"grad_norm": 0.03662109375,
|
| 2269 |
-
"learning_rate": 4.3886801491622946e-07,
|
| 2270 |
-
"loss": 3.9065,
|
| 2271 |
-
"step": 3230
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 2.8155635257037277,
|
| 2275 |
-
"grad_norm": 0.03515625,
|
| 2276 |
-
"learning_rate": 4.0068648730426264e-07,
|
| 2277 |
-
"loss": 3.8765,
|
| 2278 |
-
"step": 3240
|
| 2279 |
-
},
|
| 2280 |
-
{
|
| 2281 |
-
"epoch": 2.8242582328007826,
|
| 2282 |
-
"grad_norm": 0.03466796875,
|
| 2283 |
-
"learning_rate": 3.642255216703161e-07,
|
| 2284 |
-
"loss": 3.8651,
|
| 2285 |
-
"step": 3250
|
| 2286 |
-
},
|
| 2287 |
-
{
|
| 2288 |
-
"epoch": 2.832952939897837,
|
| 2289 |
-
"grad_norm": 0.034912109375,
|
| 2290 |
-
"learning_rate": 3.2948831880767806e-07,
|
| 2291 |
-
"loss": 3.8236,
|
| 2292 |
-
"step": 3260
|
| 2293 |
-
},
|
| 2294 |
-
{
|
| 2295 |
-
"epoch": 2.841647646994892,
|
| 2296 |
-
"grad_norm": 0.037353515625,
|
| 2297 |
-
"learning_rate": 2.9647792818593646e-07,
|
| 2298 |
-
"loss": 3.8867,
|
| 2299 |
-
"step": 3270
|
| 2300 |
-
},
|
| 2301 |
-
{
|
| 2302 |
-
"epoch": 2.8503423540919464,
|
| 2303 |
-
"grad_norm": 0.0361328125,
|
| 2304 |
-
"learning_rate": 2.651972476832709e-07,
|
| 2305 |
-
"loss": 3.8663,
|
| 2306 |
-
"step": 3280
|
| 2307 |
-
},
|
| 2308 |
-
{
|
| 2309 |
-
"epoch": 2.8590370611890012,
|
| 2310 |
-
"grad_norm": 0.0361328125,
|
| 2311 |
-
"learning_rate": 2.3564902333205851e-07,
|
| 2312 |
-
"loss": 3.8985,
|
| 2313 |
-
"step": 3290
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 2.867731768286056,
|
| 2317 |
-
"grad_norm": 0.035888671875,
|
| 2318 |
-
"learning_rate": 2.078358490778154e-07,
|
| 2319 |
-
"loss": 3.8638,
|
| 2320 |
-
"step": 3300
|
| 2321 |
-
},
|
| 2322 |
-
{
|
| 2323 |
-
"epoch": 2.8764264753831106,
|
| 2324 |
-
"grad_norm": 0.036865234375,
|
| 2325 |
-
"learning_rate": 1.8176016655147454e-07,
|
| 2326 |
-
"loss": 3.8679,
|
| 2327 |
-
"step": 3310
|
| 2328 |
-
},
|
| 2329 |
-
{
|
| 2330 |
-
"epoch": 2.885121182480165,
|
| 2331 |
-
"grad_norm": 0.035400390625,
|
| 2332 |
-
"learning_rate": 1.5742426485503725e-07,
|
| 2333 |
-
"loss": 3.8655,
|
| 2334 |
-
"step": 3320
|
| 2335 |
-
},
|
| 2336 |
-
{
|
| 2337 |
-
"epoch": 2.89381588957722,
|
| 2338 |
-
"grad_norm": 0.03515625,
|
| 2339 |
-
"learning_rate": 1.3483028036063829e-07,
|
| 2340 |
-
"loss": 3.8847,
|
| 2341 |
-
"step": 3330
|
| 2342 |
-
},
|
| 2343 |
-
{
|
| 2344 |
-
"epoch": 2.902510596674275,
|
| 2345 |
-
"grad_norm": 0.034423828125,
|
| 2346 |
-
"learning_rate": 1.1398019652298254e-07,
|
| 2347 |
-
"loss": 3.8767,
|
| 2348 |
-
"step": 3340
|
| 2349 |
-
},
|
| 2350 |
-
{
|
| 2351 |
-
"epoch": 2.9112053037713292,
|
| 2352 |
-
"grad_norm": 0.03369140625,
|
| 2353 |
-
"learning_rate": 9.487584370523772e-08,
|
| 2354 |
-
"loss": 3.8769,
|
| 2355 |
-
"step": 3350
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 2.9199000108683837,
|
| 2359 |
-
"grad_norm": 0.034912109375,
|
| 2360 |
-
"learning_rate": 7.751889901833842e-08,
|
| 2361 |
-
"loss": 3.8587,
|
| 2362 |
-
"step": 3360
|
| 2363 |
-
},
|
| 2364 |
-
{
|
| 2365 |
-
"epoch": 2.9285947179654386,
|
| 2366 |
-
"grad_norm": 0.03564453125,
|
| 2367 |
-
"learning_rate": 6.191088617376828e-08,
|
| 2368 |
-
"loss": 3.8742,
|
| 2369 |
-
"step": 3370
|
| 2370 |
-
},
|
| 2371 |
-
{
|
| 2372 |
-
"epoch": 2.9372894250624935,
|
| 2373 |
-
"grad_norm": 0.03466796875,
|
| 2374 |
-
"learning_rate": 4.805317534980036e-08,
|
| 2375 |
-
"loss": 3.8757,
|
| 2376 |
-
"step": 3380
|
| 2377 |
-
},
|
| 2378 |
-
{
|
| 2379 |
-
"epoch": 2.945984132159548,
|
| 2380 |
-
"grad_norm": 0.03466796875,
|
| 2381 |
-
"learning_rate": 3.5946983071200037e-08,
|
| 2382 |
-
"loss": 3.8686,
|
| 2383 |
-
"step": 3390
|
| 2384 |
-
},
|
| 2385 |
-
{
|
| 2386 |
-
"epoch": 2.9546788392566024,
|
| 2387 |
-
"grad_norm": 0.03466796875,
|
| 2388 |
-
"learning_rate": 2.5593372102443724e-08,
|
| 2389 |
-
"loss": 3.8705,
|
| 2390 |
-
"step": 3400
|
| 2391 |
-
},
|
| 2392 |
-
{
|
| 2393 |
-
"epoch": 2.9633735463536572,
|
| 2394 |
-
"grad_norm": 0.034912109375,
|
| 2395 |
-
"learning_rate": 1.6993251354415763e-08,
|
| 2396 |
-
"loss": 3.8715,
|
| 2397 |
-
"step": 3410
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 2.972068253450712,
|
| 2401 |
-
"grad_norm": 0.034423828125,
|
| 2402 |
-
"learning_rate": 1.0147375804614446e-08,
|
| 2403 |
-
"loss": 3.8607,
|
| 2404 |
-
"step": 3420
|
| 2405 |
-
},
|
| 2406 |
-
{
|
| 2407 |
-
"epoch": 2.9807629605477666,
|
| 2408 |
-
"grad_norm": 0.03564453125,
|
| 2409 |
-
"learning_rate": 5.05634643088948e-09,
|
| 2410 |
-
"loss": 3.8656,
|
| 2411 |
-
"step": 3430
|
| 2412 |
-
},
|
| 2413 |
-
{
|
| 2414 |
-
"epoch": 2.989457667644821,
|
| 2415 |
-
"grad_norm": 0.035400390625,
|
| 2416 |
-
"learning_rate": 1.7206101586664248e-09,
|
| 2417 |
-
"loss": 3.8606,
|
| 2418 |
-
"step": 3440
|
| 2419 |
-
},
|
| 2420 |
-
{
|
| 2421 |
-
"epoch": 2.998152374741876,
|
| 2422 |
-
"grad_norm": 0.034912109375,
|
| 2423 |
-
"learning_rate": 1.4045982171806416e-10,
|
| 2424 |
-
"loss": 3.8691,
|
| 2425 |
-
"step": 3450
|
| 2426 |
}
|
| 2427 |
],
|
| 2428 |
"logging_steps": 10,
|
| 2429 |
-
"max_steps":
|
| 2430 |
"num_input_tokens_seen": 0,
|
| 2431 |
-
"num_train_epochs":
|
| 2432 |
"save_steps": 157,
|
| 2433 |
"stateful_callbacks": {
|
| 2434 |
"TrainerControl": {
|
|
@@ -2442,7 +832,7 @@
|
|
| 2442 |
"attributes": {}
|
| 2443 |
}
|
| 2444 |
},
|
| 2445 |
-
"total_flos":
|
| 2446 |
"train_batch_size": 2,
|
| 2447 |
"trial_name": null,
|
| 2448 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1151,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.008694707097054668,
|
| 14 |
+
"grad_norm": 0.1513671875,
|
| 15 |
+
"learning_rate": 9.000000000000001e-07,
|
| 16 |
+
"loss": 4.1809,
|
| 17 |
"step": 10
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.017389414194109335,
|
| 21 |
+
"grad_norm": 0.1630859375,
|
| 22 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 23 |
+
"loss": 4.214,
|
| 24 |
"step": 20
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.026084121291164004,
|
| 28 |
+
"grad_norm": 0.158203125,
|
| 29 |
+
"learning_rate": 2.9e-06,
|
| 30 |
+
"loss": 4.1983,
|
| 31 |
"step": 30
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.03477882838821867,
|
| 35 |
+
"grad_norm": 0.1376953125,
|
| 36 |
+
"learning_rate": 3.900000000000001e-06,
|
| 37 |
+
"loss": 4.2097,
|
| 38 |
"step": 40
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.04347353548527334,
|
| 42 |
+
"grad_norm": 0.1494140625,
|
| 43 |
+
"learning_rate": 4.9000000000000005e-06,
|
| 44 |
+
"loss": 4.1934,
|
| 45 |
"step": 50
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.05216824258232801,
|
| 49 |
+
"grad_norm": 0.130859375,
|
| 50 |
+
"learning_rate": 5.9e-06,
|
| 51 |
+
"loss": 4.1777,
|
| 52 |
"step": 60
|
| 53 |
},
|
| 54 |
{
|
| 55 |
"epoch": 0.060862949679382675,
|
| 56 |
+
"grad_norm": 0.09228515625,
|
| 57 |
+
"learning_rate": 6.9e-06,
|
| 58 |
+
"loss": 4.1902,
|
| 59 |
"step": 70
|
| 60 |
},
|
| 61 |
{
|
| 62 |
"epoch": 0.06955765677643734,
|
| 63 |
+
"grad_norm": 0.11181640625,
|
| 64 |
+
"learning_rate": 7.9e-06,
|
| 65 |
+
"loss": 4.1578,
|
| 66 |
"step": 80
|
| 67 |
},
|
| 68 |
{
|
| 69 |
"epoch": 0.078252363873492,
|
| 70 |
+
"grad_norm": 0.10546875,
|
| 71 |
+
"learning_rate": 8.900000000000001e-06,
|
| 72 |
+
"loss": 4.1555,
|
| 73 |
"step": 90
|
| 74 |
},
|
| 75 |
{
|
| 76 |
"epoch": 0.08694707097054669,
|
| 77 |
+
"grad_norm": 0.07421875,
|
| 78 |
+
"learning_rate": 9.9e-06,
|
| 79 |
+
"loss": 4.1578,
|
| 80 |
"step": 100
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.09564177806760135,
|
| 84 |
+
"grad_norm": 0.059326171875,
|
| 85 |
+
"learning_rate": 9.998190772635151e-06,
|
| 86 |
+
"loss": 4.1387,
|
| 87 |
"step": 110
|
| 88 |
},
|
| 89 |
{
|
| 90 |
"epoch": 0.10433648516465602,
|
| 91 |
+
"grad_norm": 0.05712890625,
|
| 92 |
+
"learning_rate": 9.991938334301789e-06,
|
| 93 |
+
"loss": 4.121,
|
| 94 |
"step": 120
|
| 95 |
},
|
| 96 |
{
|
| 97 |
"epoch": 0.11303119226171068,
|
| 98 |
+
"grad_norm": 0.04931640625,
|
| 99 |
+
"learning_rate": 9.981225933663634e-06,
|
| 100 |
+
"loss": 4.0872,
|
| 101 |
"step": 130
|
| 102 |
},
|
| 103 |
{
|
| 104 |
"epoch": 0.12172589935876535,
|
| 105 |
+
"grad_norm": 0.05029296875,
|
| 106 |
+
"learning_rate": 9.966063141532634e-06,
|
| 107 |
+
"loss": 4.1091,
|
| 108 |
"step": 140
|
| 109 |
},
|
| 110 |
{
|
| 111 |
"epoch": 0.13042060645582002,
|
| 112 |
+
"grad_norm": 0.04541015625,
|
| 113 |
+
"learning_rate": 9.946463504847235e-06,
|
| 114 |
+
"loss": 4.0941,
|
| 115 |
"step": 150
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 0.13911531355287468,
|
| 119 |
+
"grad_norm": 0.041748046875,
|
| 120 |
+
"learning_rate": 9.9224445345691e-06,
|
| 121 |
+
"loss": 4.0769,
|
| 122 |
"step": 160
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.14781002064992935,
|
| 126 |
+
"grad_norm": 0.044189453125,
|
| 127 |
+
"learning_rate": 9.894027690038244e-06,
|
| 128 |
+
"loss": 4.0895,
|
| 129 |
"step": 170
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 0.156504727746984,
|
| 133 |
+
"grad_norm": 0.042724609375,
|
| 134 |
+
"learning_rate": 9.861238359800543e-06,
|
| 135 |
+
"loss": 4.0772,
|
| 136 |
"step": 180
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 0.16519943484403868,
|
| 140 |
+
"grad_norm": 0.043212890625,
|
| 141 |
+
"learning_rate": 9.824105838924784e-06,
|
| 142 |
+
"loss": 4.0843,
|
| 143 |
"step": 190
|
| 144 |
},
|
| 145 |
{
|
| 146 |
"epoch": 0.17389414194109337,
|
| 147 |
+
"grad_norm": 0.040283203125,
|
| 148 |
+
"learning_rate": 9.782663302829467e-06,
|
| 149 |
+
"loss": 4.0945,
|
| 150 |
"step": 200
|
| 151 |
},
|
| 152 |
{
|
| 153 |
"epoch": 0.18258884903814804,
|
| 154 |
+
"grad_norm": 0.0419921875,
|
| 155 |
+
"learning_rate": 9.736947777642809e-06,
|
| 156 |
+
"loss": 4.0721,
|
| 157 |
"step": 210
|
| 158 |
},
|
| 159 |
{
|
| 160 |
"epoch": 0.1912835561352027,
|
| 161 |
+
"grad_norm": 0.039794921875,
|
| 162 |
+
"learning_rate": 9.687000107122367e-06,
|
| 163 |
+
"loss": 4.0701,
|
| 164 |
"step": 220
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.19997826323225737,
|
| 168 |
+
"grad_norm": 0.041259765625,
|
| 169 |
+
"learning_rate": 9.632864916163886e-06,
|
| 170 |
+
"loss": 4.0938,
|
| 171 |
"step": 230
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.20867297032931204,
|
| 175 |
+
"grad_norm": 0.038330078125,
|
| 176 |
+
"learning_rate": 9.574590570931944e-06,
|
| 177 |
+
"loss": 4.0663,
|
| 178 |
"step": 240
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.2173676774263667,
|
| 182 |
+
"grad_norm": 0.03857421875,
|
| 183 |
+
"learning_rate": 9.512229135648023e-06,
|
| 184 |
+
"loss": 4.0691,
|
| 185 |
"step": 250
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.22606238452342137,
|
| 189 |
+
"grad_norm": 0.039306640625,
|
| 190 |
+
"learning_rate": 9.445836326074625e-06,
|
| 191 |
+
"loss": 4.0786,
|
| 192 |
"step": 260
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.23475709162047603,
|
| 196 |
+
"grad_norm": 0.039794921875,
|
| 197 |
+
"learning_rate": 9.37547145973696e-06,
|
| 198 |
+
"loss": 4.0746,
|
| 199 |
"step": 270
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.2434517987175307,
|
| 203 |
+
"grad_norm": 0.03955078125,
|
| 204 |
+
"learning_rate": 9.301197402926726e-06,
|
| 205 |
+
"loss": 4.0718,
|
| 206 |
"step": 280
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.25214650581458536,
|
| 210 |
+
"grad_norm": 0.041259765625,
|
| 211 |
+
"learning_rate": 9.223080514535277e-06,
|
| 212 |
+
"loss": 4.0605,
|
| 213 |
"step": 290
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.26084121291164003,
|
| 217 |
+
"grad_norm": 0.03955078125,
|
| 218 |
+
"learning_rate": 9.141190586766418e-06,
|
| 219 |
+
"loss": 4.0513,
|
| 220 |
"step": 300
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.2695359200086947,
|
| 224 |
+
"grad_norm": 0.041259765625,
|
| 225 |
+
"learning_rate": 9.055600782781738e-06,
|
| 226 |
+
"loss": 4.0529,
|
| 227 |
"step": 310
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.27823062710574936,
|
| 231 |
+
"grad_norm": 0.037841796875,
|
| 232 |
+
"learning_rate": 8.96638757133423e-06,
|
| 233 |
+
"loss": 4.0448,
|
| 234 |
"step": 320
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.286925334202804,
|
| 238 |
+
"grad_norm": 0.041015625,
|
| 239 |
+
"learning_rate": 8.873630658448586e-06,
|
| 240 |
+
"loss": 4.042,
|
| 241 |
"step": 330
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.2956200412998587,
|
| 245 |
+
"grad_norm": 0.03857421875,
|
| 246 |
+
"learning_rate": 8.777412916209214e-06,
|
| 247 |
+
"loss": 4.0662,
|
| 248 |
"step": 340
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.30431474839691336,
|
| 252 |
+
"grad_norm": 0.03857421875,
|
| 253 |
+
"learning_rate": 8.677820308719572e-06,
|
| 254 |
+
"loss": 4.065,
|
| 255 |
"step": 350
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.313009455493968,
|
| 259 |
+
"grad_norm": 0.0361328125,
|
| 260 |
+
"learning_rate": 8.574941815299012e-06,
|
| 261 |
+
"loss": 4.0534,
|
| 262 |
"step": 360
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.3217041625910227,
|
| 266 |
+
"grad_norm": 0.039306640625,
|
| 267 |
+
"learning_rate": 8.468869350985725e-06,
|
| 268 |
+
"loss": 4.0515,
|
| 269 |
"step": 370
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.33039886968807736,
|
| 273 |
+
"grad_norm": 0.037353515625,
|
| 274 |
+
"learning_rate": 8.359697684416805e-06,
|
| 275 |
+
"loss": 4.0615,
|
| 276 |
"step": 380
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.3390935767851321,
|
| 280 |
+
"grad_norm": 0.036865234375,
|
| 281 |
+
"learning_rate": 8.247524353158836e-06,
|
| 282 |
+
"loss": 4.038,
|
| 283 |
"step": 390
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.34778828388218674,
|
| 287 |
+
"grad_norm": 0.037109375,
|
| 288 |
+
"learning_rate": 8.132449576564603e-06,
|
| 289 |
+
"loss": 4.0643,
|
| 290 |
"step": 400
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.3564829909792414,
|
| 294 |
+
"grad_norm": 0.036376953125,
|
| 295 |
+
"learning_rate": 8.014576166233823e-06,
|
| 296 |
+
"loss": 4.0567,
|
| 297 |
"step": 410
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.3651776980762961,
|
| 301 |
+
"grad_norm": 0.038818359375,
|
| 302 |
+
"learning_rate": 7.894009434157873e-06,
|
| 303 |
+
"loss": 4.0569,
|
| 304 |
"step": 420
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.37387240517335074,
|
| 308 |
+
"grad_norm": 0.037841796875,
|
| 309 |
+
"learning_rate": 7.77085709863058e-06,
|
| 310 |
+
"loss": 4.0561,
|
| 311 |
"step": 430
|
| 312 |
},
|
| 313 |
{
|
| 314 |
"epoch": 0.3825671122704054,
|
| 315 |
+
"grad_norm": 0.035400390625,
|
| 316 |
+
"learning_rate": 7.645229188009153e-06,
|
| 317 |
+
"loss": 4.0437,
|
| 318 |
"step": 440
|
| 319 |
},
|
| 320 |
{
|
| 321 |
"epoch": 0.3912618193674601,
|
| 322 |
+
"grad_norm": 0.037109375,
|
| 323 |
+
"learning_rate": 7.517237942411213e-06,
|
| 324 |
+
"loss": 4.0419,
|
| 325 |
"step": 450
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 0.39995652646451474,
|
| 329 |
+
"grad_norm": 0.036376953125,
|
| 330 |
+
"learning_rate": 7.386997713435774e-06,
|
| 331 |
+
"loss": 4.0653,
|
| 332 |
"step": 460
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.4086512335615694,
|
| 336 |
+
"grad_norm": 0.035888671875,
|
| 337 |
+
"learning_rate": 7.254624861997754e-06,
|
| 338 |
+
"loss": 4.0415,
|
| 339 |
"step": 470
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 0.41734594065862407,
|
| 343 |
+
"grad_norm": 0.03662109375,
|
| 344 |
+
"learning_rate": 7.120237654367301e-06,
|
| 345 |
+
"loss": 4.0362,
|
| 346 |
"step": 480
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 0.42604064775567874,
|
| 350 |
+
"grad_norm": 0.038818359375,
|
| 351 |
+
"learning_rate": 6.983956156506798e-06,
|
| 352 |
+
"loss": 4.0543,
|
| 353 |
"step": 490
|
| 354 |
},
|
| 355 |
{
|
| 356 |
"epoch": 0.4347353548527334,
|
| 357 |
+
"grad_norm": 0.03759765625,
|
| 358 |
+
"learning_rate": 6.845902126799981e-06,
|
| 359 |
+
"loss": 4.0623,
|
| 360 |
"step": 500
|
| 361 |
},
|
| 362 |
{
|
| 363 |
"epoch": 0.44343006194978807,
|
| 364 |
+
"grad_norm": 0.035400390625,
|
| 365 |
+
"learning_rate": 6.706198907268986e-06,
|
| 366 |
+
"loss": 4.0398,
|
| 367 |
"step": 510
|
| 368 |
},
|
| 369 |
{
|
| 370 |
"epoch": 0.45212476904684273,
|
| 371 |
+
"grad_norm": 0.035888671875,
|
| 372 |
+
"learning_rate": 6.5649713133765115e-06,
|
| 373 |
+
"loss": 4.0346,
|
| 374 |
"step": 520
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.4608194761438974,
|
| 378 |
+
"grad_norm": 0.036865234375,
|
| 379 |
+
"learning_rate": 6.422345522511575e-06,
|
| 380 |
+
"loss": 4.0574,
|
| 381 |
"step": 530
|
| 382 |
},
|
| 383 |
{
|
| 384 |
"epoch": 0.46951418324095207,
|
| 385 |
+
"grad_norm": 0.036865234375,
|
| 386 |
+
"learning_rate": 6.2784489612584695e-06,
|
| 387 |
+
"loss": 4.0289,
|
| 388 |
"step": 540
|
| 389 |
},
|
| 390 |
{
|
| 391 |
"epoch": 0.47820889033800673,
|
| 392 |
+
"grad_norm": 0.03564453125,
|
| 393 |
+
"learning_rate": 6.133410191549658e-06,
|
| 394 |
+
"loss": 4.0427,
|
| 395 |
"step": 550
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 0.4869035974350614,
|
| 399 |
+
"grad_norm": 0.0361328125,
|
| 400 |
+
"learning_rate": 5.987358795804294e-06,
|
| 401 |
+
"loss": 4.0587,
|
| 402 |
"step": 560
|
| 403 |
},
|
| 404 |
{
|
| 405 |
"epoch": 0.49559830453211606,
|
| 406 |
+
"grad_norm": 0.036376953125,
|
| 407 |
+
"learning_rate": 5.840425261155022e-06,
|
| 408 |
+
"loss": 4.0301,
|
| 409 |
"step": 570
|
| 410 |
},
|
| 411 |
{
|
| 412 |
"epoch": 0.5042930116291707,
|
| 413 |
+
"grad_norm": 0.0380859375,
|
| 414 |
+
"learning_rate": 5.692740862866472e-06,
|
| 415 |
+
"loss": 4.0396,
|
| 416 |
"step": 580
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.5129877187262254,
|
| 420 |
+
"grad_norm": 0.03759765625,
|
| 421 |
+
"learning_rate": 5.544437547049608e-06,
|
| 422 |
+
"loss": 4.0401,
|
| 423 |
"step": 590
|
| 424 |
},
|
| 425 |
{
|
| 426 |
"epoch": 0.5216824258232801,
|
| 427 |
"grad_norm": 0.03662109375,
|
| 428 |
+
"learning_rate": 5.3956478127767155e-06,
|
| 429 |
+
"loss": 4.0432,
|
| 430 |
"step": 600
|
| 431 |
},
|
| 432 |
{
|
| 433 |
"epoch": 0.5303771329203347,
|
| 434 |
+
"grad_norm": 0.03662109375,
|
| 435 |
+
"learning_rate": 5.2465045937023704e-06,
|
| 436 |
+
"loss": 4.0351,
|
| 437 |
"step": 610
|
| 438 |
},
|
| 439 |
{
|
| 440 |
"epoch": 0.5390718400173894,
|
| 441 |
+
"grad_norm": 0.0361328125,
|
| 442 |
+
"learning_rate": 5.097141139296129e-06,
|
| 443 |
+
"loss": 4.0482,
|
| 444 |
"step": 620
|
| 445 |
},
|
| 446 |
{
|
| 447 |
"epoch": 0.5477665471144441,
|
| 448 |
+
"grad_norm": 0.035888671875,
|
| 449 |
+
"learning_rate": 4.947690895793049e-06,
|
| 450 |
+
"loss": 4.044,
|
| 451 |
"step": 630
|
| 452 |
},
|
| 453 |
{
|
| 454 |
"epoch": 0.5564612542114987,
|
| 455 |
+
"grad_norm": 0.0400390625,
|
| 456 |
+
"learning_rate": 4.7982873869684315e-06,
|
| 457 |
+
"loss": 4.0323,
|
| 458 |
"step": 640
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.5651559613085534,
|
| 462 |
+
"grad_norm": 0.034912109375,
|
| 463 |
+
"learning_rate": 4.649064094843274e-06,
|
| 464 |
+
"loss": 4.0172,
|
| 465 |
"step": 650
|
| 466 |
},
|
| 467 |
{
|
| 468 |
"epoch": 0.573850668405608,
|
| 469 |
+
"grad_norm": 0.03564453125,
|
| 470 |
+
"learning_rate": 4.500154340427037e-06,
|
| 471 |
+
"loss": 4.0479,
|
| 472 |
"step": 660
|
| 473 |
},
|
| 474 |
{
|
| 475 |
"epoch": 0.5825453755026627,
|
| 476 |
+
"grad_norm": 0.036376953125,
|
| 477 |
+
"learning_rate": 4.351691164604247e-06,
|
| 478 |
+
"loss": 4.0412,
|
| 479 |
"step": 670
|
| 480 |
},
|
| 481 |
{
|
| 482 |
"epoch": 0.5912400825997174,
|
| 483 |
+
"grad_norm": 0.035888671875,
|
| 484 |
+
"learning_rate": 4.203807209271393e-06,
|
| 485 |
+
"loss": 4.0207,
|
| 486 |
"step": 680
|
| 487 |
},
|
| 488 |
{
|
| 489 |
"epoch": 0.599934789696772,
|
| 490 |
+
"grad_norm": 0.0361328125,
|
| 491 |
+
"learning_rate": 4.056634598830282e-06,
|
| 492 |
+
"loss": 4.0281,
|
| 493 |
"step": 690
|
| 494 |
},
|
| 495 |
{
|
| 496 |
"epoch": 0.6086294967938267,
|
| 497 |
+
"grad_norm": 0.03564453125,
|
| 498 |
+
"learning_rate": 3.910304822143734e-06,
|
| 499 |
+
"loss": 4.0252,
|
| 500 |
"step": 700
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.6173242038908814,
|
| 504 |
+
"grad_norm": 0.03466796875,
|
| 505 |
+
"learning_rate": 3.7649486150591115e-06,
|
| 506 |
+
"loss": 4.0236,
|
| 507 |
"step": 710
|
| 508 |
},
|
| 509 |
{
|
| 510 |
"epoch": 0.626018910987936,
|
| 511 |
+
"grad_norm": 0.0361328125,
|
| 512 |
+
"learning_rate": 3.6206958436045856e-06,
|
| 513 |
+
"loss": 4.0225,
|
| 514 |
"step": 720
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 0.6347136180849907,
|
| 518 |
+
"grad_norm": 0.036865234375,
|
| 519 |
+
"learning_rate": 3.4776753879625563e-06,
|
| 520 |
+
"loss": 4.0214,
|
| 521 |
"step": 730
|
| 522 |
},
|
| 523 |
{
|
| 524 |
"epoch": 0.6434083251820454,
|
| 525 |
+
"grad_norm": 0.0361328125,
|
| 526 |
+
"learning_rate": 3.3360150273238413e-06,
|
| 527 |
+
"loss": 4.0541,
|
| 528 |
"step": 740
|
| 529 |
},
|
| 530 |
{
|
| 531 |
"epoch": 0.6521030322791,
|
| 532 |
+
"grad_norm": 0.037109375,
|
| 533 |
+
"learning_rate": 3.1958413257255403e-06,
|
| 534 |
+
"loss": 4.0453,
|
| 535 |
"step": 750
|
| 536 |
},
|
| 537 |
{
|
| 538 |
"epoch": 0.6607977393761547,
|
| 539 |
"grad_norm": 0.037109375,
|
| 540 |
+
"learning_rate": 3.057279518974544e-06,
|
| 541 |
+
"loss": 4.0308,
|
| 542 |
"step": 760
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.6694924464732094,
|
| 546 |
+
"grad_norm": 0.03515625,
|
| 547 |
+
"learning_rate": 2.9204534027577387e-06,
|
| 548 |
+
"loss": 4.0492,
|
| 549 |
"step": 770
|
| 550 |
},
|
| 551 |
{
|
| 552 |
"epoch": 0.6781871535702642,
|
| 553 |
+
"grad_norm": 0.041748046875,
|
| 554 |
+
"learning_rate": 2.7854852220388617e-06,
|
| 555 |
+
"loss": 4.0639,
|
| 556 |
"step": 780
|
| 557 |
},
|
| 558 |
{
|
| 559 |
"epoch": 0.6868818606673188,
|
| 560 |
+
"grad_norm": 0.035888671875,
|
| 561 |
+
"learning_rate": 2.6524955618408093e-06,
|
| 562 |
+
"loss": 4.0283,
|
| 563 |
"step": 790
|
| 564 |
},
|
| 565 |
{
|
| 566 |
"epoch": 0.6955765677643735,
|
| 567 |
+
"grad_norm": 0.035400390625,
|
| 568 |
+
"learning_rate": 2.521603239511011e-06,
|
| 569 |
+
"loss": 4.0418,
|
| 570 |
"step": 800
|
| 571 |
},
|
| 572 |
{
|
| 573 |
"epoch": 0.7042712748614282,
|
| 574 |
"grad_norm": 0.035400390625,
|
| 575 |
+
"learning_rate": 2.3929251985660866e-06,
|
| 576 |
+
"loss": 4.0417,
|
| 577 |
"step": 810
|
| 578 |
},
|
| 579 |
{
|
| 580 |
"epoch": 0.7129659819584828,
|
| 581 |
+
"grad_norm": 0.03564453125,
|
| 582 |
+
"learning_rate": 2.2665764042106647e-06,
|
| 583 |
+
"loss": 4.0288,
|
| 584 |
"step": 820
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.7216606890555375,
|
| 588 |
+
"grad_norm": 0.035888671875,
|
| 589 |
+
"learning_rate": 2.142669740623661e-06,
|
| 590 |
+
"loss": 4.0501,
|
| 591 |
"step": 830
|
| 592 |
},
|
| 593 |
{
|
| 594 |
"epoch": 0.7303553961525922,
|
| 595 |
+
"grad_norm": 0.036376953125,
|
| 596 |
+
"learning_rate": 2.021315910103841e-06,
|
| 597 |
+
"loss": 4.0402,
|
| 598 |
"step": 840
|
| 599 |
},
|
| 600 |
{
|
| 601 |
"epoch": 0.7390501032496468,
|
| 602 |
+
"grad_norm": 0.03466796875,
|
| 603 |
+
"learning_rate": 1.9026233341647398e-06,
|
| 604 |
+
"loss": 4.0303,
|
| 605 |
"step": 850
|
| 606 |
},
|
| 607 |
{
|
| 608 |
"epoch": 0.7477448103467015,
|
| 609 |
+
"grad_norm": 0.036865234375,
|
| 610 |
+
"learning_rate": 1.786698056667297e-06,
|
| 611 |
+
"loss": 4.0461,
|
| 612 |
"step": 860
|
| 613 |
},
|
| 614 |
{
|
| 615 |
"epoch": 0.7564395174437561,
|
| 616 |
"grad_norm": 0.038330078125,
|
| 617 |
+
"learning_rate": 1.6736436490767793e-06,
|
| 618 |
+
"loss": 4.0543,
|
| 619 |
"step": 870
|
| 620 |
},
|
| 621 |
{
|
| 622 |
"epoch": 0.7651342245408108,
|
| 623 |
+
"grad_norm": 0.035888671875,
|
| 624 |
+
"learning_rate": 1.5635611179286203e-06,
|
| 625 |
+
"loss": 4.0414,
|
| 626 |
"step": 880
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.7738289316378655,
|
| 630 |
"grad_norm": 0.036865234375,
|
| 631 |
+
"learning_rate": 1.4565488145858497e-06,
|
| 632 |
+
"loss": 4.028,
|
| 633 |
"step": 890
|
| 634 |
},
|
| 635 |
{
|
| 636 |
"epoch": 0.7825236387349201,
|
| 637 |
+
"grad_norm": 0.035888671875,
|
| 638 |
+
"learning_rate": 1.3527023473687417e-06,
|
| 639 |
+
"loss": 4.0582,
|
| 640 |
"step": 900
|
| 641 |
},
|
| 642 |
{
|
| 643 |
"epoch": 0.7912183458319748,
|
| 644 |
+
"grad_norm": 0.037353515625,
|
| 645 |
+
"learning_rate": 1.2521144961351893e-06,
|
| 646 |
+
"loss": 4.0381,
|
| 647 |
"step": 910
|
| 648 |
},
|
| 649 |
{
|
| 650 |
"epoch": 0.7999130529290295,
|
| 651 |
+
"grad_norm": 0.0361328125,
|
| 652 |
+
"learning_rate": 1.154875129388126e-06,
|
| 653 |
+
"loss": 4.0082,
|
| 654 |
"step": 920
|
| 655 |
},
|
| 656 |
{
|
| 657 |
"epoch": 0.8086077600260841,
|
| 658 |
+
"grad_norm": 0.038330078125,
|
| 659 |
+
"learning_rate": 1.061071123984031e-06,
|
| 660 |
+
"loss": 4.0471,
|
| 661 |
"step": 930
|
| 662 |
},
|
| 663 |
{
|
| 664 |
"epoch": 0.8173024671231388,
|
| 665 |
+
"grad_norm": 0.035888671875,
|
| 666 |
+
"learning_rate": 9.707862875142898e-07,
|
| 667 |
+
"loss": 4.0324,
|
| 668 |
"step": 940
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.8259971742201935,
|
| 672 |
+
"grad_norm": 0.037841796875,
|
| 673 |
+
"learning_rate": 8.841012834287254e-07,
|
| 674 |
+
"loss": 4.0469,
|
| 675 |
"step": 950
|
| 676 |
},
|
| 677 |
{
|
| 678 |
"epoch": 0.8346918813172481,
|
| 679 |
+
"grad_norm": 0.035400390625,
|
| 680 |
+
"learning_rate": 8.010935589682134e-07,
|
| 681 |
+
"loss": 4.0267,
|
| 682 |
"step": 960
|
| 683 |
},
|
| 684 |
{
|
| 685 |
"epoch": 0.8433865884143028,
|
| 686 |
+
"grad_norm": 0.036376953125,
|
| 687 |
+
"learning_rate": 7.218372759707626e-07,
|
| 688 |
+
"loss": 4.0408,
|
| 689 |
"step": 970
|
| 690 |
},
|
| 691 |
{
|
| 692 |
"epoch": 0.8520812955113575,
|
| 693 |
+
"grad_norm": 0.035888671875,
|
| 694 |
+
"learning_rate": 6.464032446128837e-07,
|
| 695 |
+
"loss": 4.0647,
|
| 696 |
"step": 980
|
| 697 |
},
|
| 698 |
{
|
| 699 |
"epoch": 0.8607760026084121,
|
| 700 |
+
"grad_norm": 0.03466796875,
|
| 701 |
+
"learning_rate": 5.748588601454463e-07,
|
| 702 |
+
"loss": 4.0245,
|
| 703 |
"step": 990
|
| 704 |
},
|
| 705 |
{
|
| 706 |
"epoch": 0.8694707097054668,
|
| 707 |
+
"grad_norm": 0.03759765625,
|
| 708 |
+
"learning_rate": 5.072680426805332e-07,
|
| 709 |
+
"loss": 4.0502,
|
| 710 |
"step": 1000
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.8781654168025215,
|
| 714 |
+
"grad_norm": 0.037841796875,
|
| 715 |
+
"learning_rate": 4.436911800831084e-07,
|
| 716 |
+
"loss": 4.0439,
|
| 717 |
"step": 1010
|
| 718 |
},
|
| 719 |
{
|
| 720 |
"epoch": 0.8868601238995761,
|
| 721 |
+
"grad_norm": 0.036865234375,
|
| 722 |
+
"learning_rate": 3.841850740185088e-07,
|
| 723 |
+
"loss": 4.0625,
|
| 724 |
"step": 1020
|
| 725 |
},
|
| 726 |
{
|
| 727 |
"epoch": 0.8955548309966308,
|
| 728 |
+
"grad_norm": 0.03515625,
|
| 729 |
+
"learning_rate": 3.288028892039585e-07,
|
| 730 |
+
"loss": 4.0363,
|
| 731 |
"step": 1030
|
| 732 |
},
|
| 733 |
{
|
| 734 |
"epoch": 0.9042495380936855,
|
| 735 |
+
"grad_norm": 0.036865234375,
|
| 736 |
+
"learning_rate": 2.7759410590946446e-07,
|
| 737 |
+
"loss": 4.0429,
|
| 738 |
"step": 1040
|
| 739 |
},
|
| 740 |
{
|
| 741 |
"epoch": 0.9129442451907401,
|
| 742 |
+
"grad_norm": 0.0380859375,
|
| 743 |
+
"learning_rate": 2.306044757505055e-07,
|
| 744 |
+
"loss": 4.0474,
|
| 745 |
"step": 1050
|
| 746 |
},
|
| 747 |
{
|
| 748 |
"epoch": 0.9216389522877948,
|
| 749 |
+
"grad_norm": 0.037109375,
|
| 750 |
+
"learning_rate": 1.8787598081203285e-07,
|
| 751 |
+
"loss": 4.0502,
|
| 752 |
"step": 1060
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.9303336593848495,
|
| 756 |
+
"grad_norm": 0.036376953125,
|
| 757 |
+
"learning_rate": 1.4944679614029346e-07,
|
| 758 |
+
"loss": 4.038,
|
| 759 |
"step": 1070
|
| 760 |
},
|
| 761 |
{
|
| 762 |
"epoch": 0.9390283664819041,
|
| 763 |
+
"grad_norm": 0.0390625,
|
| 764 |
+
"learning_rate": 1.1535125563597927e-07,
|
| 765 |
+
"loss": 4.0239,
|
| 766 |
"step": 1080
|
| 767 |
},
|
| 768 |
{
|
| 769 |
"epoch": 0.9477230735789588,
|
| 770 |
+
"grad_norm": 0.036865234375,
|
| 771 |
+
"learning_rate": 8.561982137919478e-08,
|
| 772 |
+
"loss": 4.0352,
|
| 773 |
"step": 1090
|
| 774 |
},
|
| 775 |
{
|
| 776 |
"epoch": 0.9564177806760135,
|
| 777 |
+
"grad_norm": 0.03564453125,
|
| 778 |
+
"learning_rate": 6.02790564136263e-08,
|
| 779 |
+
"loss": 4.0371,
|
| 780 |
"step": 1100
|
| 781 |
},
|
| 782 |
{
|
| 783 |
"epoch": 0.9651124877730681,
|
| 784 |
+
"grad_norm": 0.03466796875,
|
| 785 |
+
"learning_rate": 3.935160101424251e-08,
|
| 786 |
+
"loss": 4.0525,
|
| 787 |
"step": 1110
|
| 788 |
},
|
| 789 |
{
|
| 790 |
"epoch": 0.9738071948701228,
|
| 791 |
+
"grad_norm": 0.034912109375,
|
| 792 |
+
"learning_rate": 2.285615245972983e-08,
|
| 793 |
+
"loss": 4.0236,
|
| 794 |
"step": 1120
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.9825019019671775,
|
| 798 |
+
"grad_norm": 0.035888671875,
|
| 799 |
+
"learning_rate": 1.0807448327723735e-08,
|
| 800 |
+
"loss": 4.0348,
|
| 801 |
"step": 1130
|
| 802 |
},
|
| 803 |
{
|
| 804 |
"epoch": 0.9911966090642321,
|
| 805 |
+
"grad_norm": 0.035400390625,
|
| 806 |
+
"learning_rate": 3.2162533277713125e-09,
|
| 807 |
+
"loss": 4.0388,
|
| 808 |
"step": 1140
|
| 809 |
},
|
| 810 |
{
|
| 811 |
"epoch": 0.9998913161612868,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 812 |
"grad_norm": 0.035400390625,
|
| 813 |
+
"learning_rate": 8.93496837878205e-11,
|
| 814 |
+
"loss": 4.0638,
|
| 815 |
+
"step": 1150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 816 |
}
|
| 817 |
],
|
| 818 |
"logging_steps": 10,
|
| 819 |
+
"max_steps": 1151,
|
| 820 |
"num_input_tokens_seen": 0,
|
| 821 |
+
"num_train_epochs": 1,
|
| 822 |
"save_steps": 157,
|
| 823 |
"stateful_callbacks": {
|
| 824 |
"TrainerControl": {
|
|
|
|
| 832 |
"attributes": {}
|
| 833 |
}
|
| 834 |
},
|
| 835 |
+
"total_flos": 5.234099929420399e+18,
|
| 836 |
"train_batch_size": 2,
|
| 837 |
"trial_name": null,
|
| 838 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5905
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:301d89f1700999978d1aab66fccdff9692ce49fe1df2946ff54d29752274abac
|
| 3 |
size 5905
|