Training in progress, step 500, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/ar_diffusion_info.bin +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scaler.pt +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +33 -1925
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 36730224
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:56e8ec13f5b750bff43bf95c8f426d5c76de12c9614c601c59257c731d90aad9
|
| 3 |
size 36730224
|
last-checkpoint/ar_diffusion_info.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1736
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcf5fd030c264f8615af2d6bd5dd8f07c7ac6052929ca71266901b45d38e573c
|
| 3 |
size 1736
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 73588346
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed69c86e6661b1f68a99aa8b123937781810ccc7aacd7d2f7ad96f9fb90dd267
|
| 3 |
size 73588346
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa10217fea7d7bdff8c6b7812212a3b0fcc34c0411d6adbf4710eb28a9f096e8
|
| 3 |
size 14244
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6d2c7b185157ea92508fe4deec82c591b54fb96c18b69b1ba12fb4fa0a5f624
|
| 3 |
size 988
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:65d8086826e16c0b4479d43a31b6b223fba2daaa33e58ffbb91a05247a535912
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1991 +1,99 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric":
|
| 4 |
-
"best_model_checkpoint": "./ar-diffusion-checkpoints-progressive-attention/checkpoint-
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 250,
|
| 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": 0.003845858010922237,
|
| 14 |
-
"grad_norm":
|
| 15 |
"learning_rate": 1.84e-05,
|
| 16 |
-
"loss": 12.
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.007691716021844474,
|
| 21 |
-
"grad_norm": 5.
|
| 22 |
"learning_rate": 3.8e-05,
|
| 23 |
-
"loss": 5.
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.01153757403276671,
|
| 28 |
-
"grad_norm": 7.
|
| 29 |
"learning_rate": 5.8e-05,
|
| 30 |
-
"loss": 4.
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.015383432043688947,
|
| 35 |
-
"grad_norm": 4.
|
| 36 |
"learning_rate": 7.800000000000001e-05,
|
| 37 |
-
"loss": 3.
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.019229290054611183,
|
| 42 |
-
"grad_norm": 4.
|
| 43 |
"learning_rate": 9.8e-05,
|
| 44 |
-
"loss": 3.
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.019229290054611183,
|
| 49 |
-
"eval_loss": 3.
|
| 50 |
-
"eval_runtime":
|
| 51 |
-
"eval_samples_per_second":
|
| 52 |
-
"eval_steps_per_second":
|
| 53 |
"step": 250
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.02307514806553342,
|
| 57 |
-
"grad_norm":
|
| 58 |
"learning_rate": 0.000118,
|
| 59 |
-
"loss": 3.
|
| 60 |
"step": 300
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.02692100607645566,
|
| 64 |
-
"grad_norm": 4.
|
| 65 |
"learning_rate": 0.000138,
|
| 66 |
-
"loss": 3.
|
| 67 |
"step": 350
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 0.030766864087377895,
|
| 71 |
-
"grad_norm": 5.
|
| 72 |
"learning_rate": 0.00015800000000000002,
|
| 73 |
-
"loss": 3.
|
| 74 |
"step": 400
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.03461272209830013,
|
| 78 |
-
"grad_norm": 3.
|
| 79 |
"learning_rate": 0.00017800000000000002,
|
| 80 |
-
"loss": 3.
|
| 81 |
"step": 450
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.038458580109222366,
|
| 85 |
-
"grad_norm": 4.
|
| 86 |
"learning_rate": 0.00019800000000000002,
|
| 87 |
-
"loss": 3.
|
| 88 |
"step": 500
|
| 89 |
},
|
| 90 |
{
|
| 91 |
"epoch": 0.038458580109222366,
|
| 92 |
-
"eval_loss": 3.
|
| 93 |
-
"eval_runtime": 17.
|
| 94 |
-
"eval_samples_per_second":
|
| 95 |
-
"eval_steps_per_second":
|
| 96 |
"step": 500
|
| 97 |
-
},
|
| 98 |
-
{
|
| 99 |
-
"epoch": 0.0423044381201446,
|
| 100 |
-
"grad_norm": 3.885101079940796,
|
| 101 |
-
"learning_rate": 0.0001997662519803652,
|
| 102 |
-
"loss": 3.3463,
|
| 103 |
-
"step": 550
|
| 104 |
-
},
|
| 105 |
-
{
|
| 106 |
-
"epoch": 0.04615029613106684,
|
| 107 |
-
"grad_norm": 5.273513317108154,
|
| 108 |
-
"learning_rate": 0.0001995065319585487,
|
| 109 |
-
"loss": 3.252,
|
| 110 |
-
"step": 600
|
| 111 |
-
},
|
| 112 |
-
{
|
| 113 |
-
"epoch": 0.04999615414198908,
|
| 114 |
-
"grad_norm": 3.779219627380371,
|
| 115 |
-
"learning_rate": 0.0001992468119367322,
|
| 116 |
-
"loss": 3.1188,
|
| 117 |
-
"step": 650
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.05384201215291132,
|
| 121 |
-
"grad_norm": 3.2600204944610596,
|
| 122 |
-
"learning_rate": 0.00019898709191491573,
|
| 123 |
-
"loss": 2.9879,
|
| 124 |
-
"step": 700
|
| 125 |
-
},
|
| 126 |
-
{
|
| 127 |
-
"epoch": 0.05768787016383355,
|
| 128 |
-
"grad_norm": 7.561593055725098,
|
| 129 |
-
"learning_rate": 0.00019872737189309924,
|
| 130 |
-
"loss": 3.2548,
|
| 131 |
-
"step": 750
|
| 132 |
-
},
|
| 133 |
-
{
|
| 134 |
-
"epoch": 0.05768787016383355,
|
| 135 |
-
"eval_loss": 3.1286704540252686,
|
| 136 |
-
"eval_runtime": 17.5722,
|
| 137 |
-
"eval_samples_per_second": 56.908,
|
| 138 |
-
"eval_steps_per_second": 14.227,
|
| 139 |
-
"step": 750
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 0.06153372817475579,
|
| 143 |
-
"grad_norm": 3.786135673522949,
|
| 144 |
-
"learning_rate": 0.00019846765187128275,
|
| 145 |
-
"loss": 3.1959,
|
| 146 |
-
"step": 800
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.06537958618567802,
|
| 150 |
-
"grad_norm": 4.230893611907959,
|
| 151 |
-
"learning_rate": 0.0001982079318494663,
|
| 152 |
-
"loss": 3.1453,
|
| 153 |
-
"step": 850
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.06922544419660026,
|
| 157 |
-
"grad_norm": 2.7769834995269775,
|
| 158 |
-
"learning_rate": 0.0001979482118276498,
|
| 159 |
-
"loss": 3.1789,
|
| 160 |
-
"step": 900
|
| 161 |
-
},
|
| 162 |
-
{
|
| 163 |
-
"epoch": 0.0730713022075225,
|
| 164 |
-
"grad_norm": 1.746205449104309,
|
| 165 |
-
"learning_rate": 0.00019768849180583334,
|
| 166 |
-
"loss": 2.9569,
|
| 167 |
-
"step": 950
|
| 168 |
-
},
|
| 169 |
-
{
|
| 170 |
-
"epoch": 0.07691716021844473,
|
| 171 |
-
"grad_norm": 3.5990424156188965,
|
| 172 |
-
"learning_rate": 0.00019742877178401683,
|
| 173 |
-
"loss": 3.0739,
|
| 174 |
-
"step": 1000
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 0.07691716021844473,
|
| 178 |
-
"eval_loss": 3.0572922229766846,
|
| 179 |
-
"eval_runtime": 17.4776,
|
| 180 |
-
"eval_samples_per_second": 57.216,
|
| 181 |
-
"eval_steps_per_second": 14.304,
|
| 182 |
-
"step": 1000
|
| 183 |
-
},
|
| 184 |
-
{
|
| 185 |
-
"epoch": 0.08076301822936698,
|
| 186 |
-
"grad_norm": 2.302044153213501,
|
| 187 |
-
"learning_rate": 0.00019716905176220034,
|
| 188 |
-
"loss": 2.9218,
|
| 189 |
-
"step": 1050
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.0846088762402892,
|
| 193 |
-
"grad_norm": 2.9099605083465576,
|
| 194 |
-
"learning_rate": 0.00019690933174038388,
|
| 195 |
-
"loss": 3.027,
|
| 196 |
-
"step": 1100
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 0.08845473425121145,
|
| 200 |
-
"grad_norm": 2.307961940765381,
|
| 201 |
-
"learning_rate": 0.0001966496117185674,
|
| 202 |
-
"loss": 2.9845,
|
| 203 |
-
"step": 1150
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.09230059226213368,
|
| 207 |
-
"grad_norm": 4.332247734069824,
|
| 208 |
-
"learning_rate": 0.0001963898916967509,
|
| 209 |
-
"loss": 2.9047,
|
| 210 |
-
"step": 1200
|
| 211 |
-
},
|
| 212 |
-
{
|
| 213 |
-
"epoch": 0.09614645027305592,
|
| 214 |
-
"grad_norm": 3.9856581687927246,
|
| 215 |
-
"learning_rate": 0.00019613017167493445,
|
| 216 |
-
"loss": 3.0003,
|
| 217 |
-
"step": 1250
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"epoch": 0.09614645027305592,
|
| 221 |
-
"eval_loss": 2.8999271392822266,
|
| 222 |
-
"eval_runtime": 17.4674,
|
| 223 |
-
"eval_samples_per_second": 57.25,
|
| 224 |
-
"eval_steps_per_second": 14.312,
|
| 225 |
-
"step": 1250
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.09999230828397816,
|
| 229 |
-
"grad_norm": 8.007057189941406,
|
| 230 |
-
"learning_rate": 0.00019587045165311796,
|
| 231 |
-
"loss": 2.9034,
|
| 232 |
-
"step": 1300
|
| 233 |
-
},
|
| 234 |
-
{
|
| 235 |
-
"epoch": 0.10383816629490039,
|
| 236 |
-
"grad_norm": 2.3496313095092773,
|
| 237 |
-
"learning_rate": 0.00019561073163130145,
|
| 238 |
-
"loss": 2.8742,
|
| 239 |
-
"step": 1350
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"epoch": 0.10768402430582263,
|
| 243 |
-
"grad_norm": 2.3679521083831787,
|
| 244 |
-
"learning_rate": 0.000195351011609485,
|
| 245 |
-
"loss": 2.8189,
|
| 246 |
-
"step": 1400
|
| 247 |
-
},
|
| 248 |
-
{
|
| 249 |
-
"epoch": 0.11152988231674486,
|
| 250 |
-
"grad_norm": 5.371161937713623,
|
| 251 |
-
"learning_rate": 0.0001950912915876685,
|
| 252 |
-
"loss": 2.826,
|
| 253 |
-
"step": 1450
|
| 254 |
-
},
|
| 255 |
-
{
|
| 256 |
-
"epoch": 0.1153757403276671,
|
| 257 |
-
"grad_norm": 4.442603588104248,
|
| 258 |
-
"learning_rate": 0.000194831571565852,
|
| 259 |
-
"loss": 2.9645,
|
| 260 |
-
"step": 1500
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"epoch": 0.1153757403276671,
|
| 264 |
-
"eval_loss": 2.864973783493042,
|
| 265 |
-
"eval_runtime": 17.4467,
|
| 266 |
-
"eval_samples_per_second": 57.317,
|
| 267 |
-
"eval_steps_per_second": 14.329,
|
| 268 |
-
"step": 1500
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.11922159833858934,
|
| 272 |
-
"grad_norm": 3.153311252593994,
|
| 273 |
-
"learning_rate": 0.00019457185154403555,
|
| 274 |
-
"loss": 2.8506,
|
| 275 |
-
"step": 1550
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.12306745634951158,
|
| 279 |
-
"grad_norm": 4.94291877746582,
|
| 280 |
-
"learning_rate": 0.00019431213152221907,
|
| 281 |
-
"loss": 2.9362,
|
| 282 |
-
"step": 1600
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.12691331436043382,
|
| 286 |
-
"grad_norm": 3.0833139419555664,
|
| 287 |
-
"learning_rate": 0.00019405241150040255,
|
| 288 |
-
"loss": 2.7926,
|
| 289 |
-
"step": 1650
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.13075917237135604,
|
| 293 |
-
"grad_norm": 4.330313205718994,
|
| 294 |
-
"learning_rate": 0.00019379788587902243,
|
| 295 |
-
"loss": 3.0154,
|
| 296 |
-
"step": 1700
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.13460503038227828,
|
| 300 |
-
"grad_norm": 3.247084379196167,
|
| 301 |
-
"learning_rate": 0.00019353816585720594,
|
| 302 |
-
"loss": 2.8993,
|
| 303 |
-
"step": 1750
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.13460503038227828,
|
| 307 |
-
"eval_loss": 2.797525644302368,
|
| 308 |
-
"eval_runtime": 17.3938,
|
| 309 |
-
"eval_samples_per_second": 57.492,
|
| 310 |
-
"eval_steps_per_second": 14.373,
|
| 311 |
-
"step": 1750
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.13845088839320052,
|
| 315 |
-
"grad_norm": 1.2497533559799194,
|
| 316 |
-
"learning_rate": 0.00019327844583538946,
|
| 317 |
-
"loss": 2.8769,
|
| 318 |
-
"step": 1800
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.14229674640412276,
|
| 322 |
-
"grad_norm": 3.2852976322174072,
|
| 323 |
-
"learning_rate": 0.000193018725813573,
|
| 324 |
-
"loss": 2.7027,
|
| 325 |
-
"step": 1850
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.146142604415045,
|
| 329 |
-
"grad_norm": 3.199697732925415,
|
| 330 |
-
"learning_rate": 0.0001927590057917565,
|
| 331 |
-
"loss": 2.8646,
|
| 332 |
-
"step": 1900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.14998846242596722,
|
| 336 |
-
"grad_norm": 2.30946946144104,
|
| 337 |
-
"learning_rate": 0.00019249928576994002,
|
| 338 |
-
"loss": 2.7485,
|
| 339 |
-
"step": 1950
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.15383432043688947,
|
| 343 |
-
"grad_norm": 2.7489562034606934,
|
| 344 |
-
"learning_rate": 0.00019223956574812353,
|
| 345 |
-
"loss": 2.8288,
|
| 346 |
-
"step": 2000
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.15383432043688947,
|
| 350 |
-
"eval_loss": 2.732482671737671,
|
| 351 |
-
"eval_runtime": 17.4426,
|
| 352 |
-
"eval_samples_per_second": 57.331,
|
| 353 |
-
"eval_steps_per_second": 14.333,
|
| 354 |
-
"step": 2000
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 0.1576801784478117,
|
| 358 |
-
"grad_norm": 2.094149351119995,
|
| 359 |
-
"learning_rate": 0.00019197984572630705,
|
| 360 |
-
"loss": 2.8181,
|
| 361 |
-
"step": 2050
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 0.16152603645873395,
|
| 365 |
-
"grad_norm": 2.880585193634033,
|
| 366 |
-
"learning_rate": 0.00019172012570449056,
|
| 367 |
-
"loss": 2.7495,
|
| 368 |
-
"step": 2100
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 0.16537189446965617,
|
| 372 |
-
"grad_norm": 3.0462090969085693,
|
| 373 |
-
"learning_rate": 0.0001914604056826741,
|
| 374 |
-
"loss": 2.6552,
|
| 375 |
-
"step": 2150
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 0.1692177524805784,
|
| 379 |
-
"grad_norm": 2.8088278770446777,
|
| 380 |
-
"learning_rate": 0.0001912006856608576,
|
| 381 |
-
"loss": 2.6424,
|
| 382 |
-
"step": 2200
|
| 383 |
-
},
|
| 384 |
-
{
|
| 385 |
-
"epoch": 0.17306361049150065,
|
| 386 |
-
"grad_norm": 2.3041441440582275,
|
| 387 |
-
"learning_rate": 0.00019094096563904112,
|
| 388 |
-
"loss": 2.6002,
|
| 389 |
-
"step": 2250
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 0.17306361049150065,
|
| 393 |
-
"eval_loss": 2.6283011436462402,
|
| 394 |
-
"eval_runtime": 17.4951,
|
| 395 |
-
"eval_samples_per_second": 57.159,
|
| 396 |
-
"eval_steps_per_second": 14.29,
|
| 397 |
-
"step": 2250
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 0.1769094685024229,
|
| 401 |
-
"grad_norm": 2.4782445430755615,
|
| 402 |
-
"learning_rate": 0.00019068124561722464,
|
| 403 |
-
"loss": 2.7373,
|
| 404 |
-
"step": 2300
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.18075532651334514,
|
| 408 |
-
"grad_norm": 2.799548864364624,
|
| 409 |
-
"learning_rate": 0.00019042152559540815,
|
| 410 |
-
"loss": 2.6528,
|
| 411 |
-
"step": 2350
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.18460118452426735,
|
| 415 |
-
"grad_norm": 3.158393144607544,
|
| 416 |
-
"learning_rate": 0.00019016180557359166,
|
| 417 |
-
"loss": 2.6495,
|
| 418 |
-
"step": 2400
|
| 419 |
-
},
|
| 420 |
-
{
|
| 421 |
-
"epoch": 0.1884470425351896,
|
| 422 |
-
"grad_norm": 2.4797089099884033,
|
| 423 |
-
"learning_rate": 0.0001899020855517752,
|
| 424 |
-
"loss": 2.6815,
|
| 425 |
-
"step": 2450
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 0.19229290054611184,
|
| 429 |
-
"grad_norm": 2.5662102699279785,
|
| 430 |
-
"learning_rate": 0.00018964236552995872,
|
| 431 |
-
"loss": 2.6435,
|
| 432 |
-
"step": 2500
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 0.19229290054611184,
|
| 436 |
-
"eval_loss": 2.6090738773345947,
|
| 437 |
-
"eval_runtime": 17.3946,
|
| 438 |
-
"eval_samples_per_second": 57.489,
|
| 439 |
-
"eval_steps_per_second": 14.372,
|
| 440 |
-
"step": 2500
|
| 441 |
-
},
|
| 442 |
-
{
|
| 443 |
-
"epoch": 0.19613875855703408,
|
| 444 |
-
"grad_norm": 2.0812065601348877,
|
| 445 |
-
"learning_rate": 0.00018938264550814223,
|
| 446 |
-
"loss": 2.4975,
|
| 447 |
-
"step": 2550
|
| 448 |
-
},
|
| 449 |
-
{
|
| 450 |
-
"epoch": 0.19998461656795632,
|
| 451 |
-
"grad_norm": 3.9404165744781494,
|
| 452 |
-
"learning_rate": 0.00018912292548632574,
|
| 453 |
-
"loss": 2.6395,
|
| 454 |
-
"step": 2600
|
| 455 |
-
},
|
| 456 |
-
{
|
| 457 |
-
"epoch": 0.20383047457887854,
|
| 458 |
-
"grad_norm": 2.122506856918335,
|
| 459 |
-
"learning_rate": 0.00018886320546450925,
|
| 460 |
-
"loss": 2.59,
|
| 461 |
-
"step": 2650
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 0.20767633258980078,
|
| 465 |
-
"grad_norm": 4.2281293869018555,
|
| 466 |
-
"learning_rate": 0.0001886034854426928,
|
| 467 |
-
"loss": 2.5563,
|
| 468 |
-
"step": 2700
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.21152219060072303,
|
| 472 |
-
"grad_norm": 2.492363691329956,
|
| 473 |
-
"learning_rate": 0.0001883437654208763,
|
| 474 |
-
"loss": 2.4601,
|
| 475 |
-
"step": 2750
|
| 476 |
-
},
|
| 477 |
-
{
|
| 478 |
-
"epoch": 0.21152219060072303,
|
| 479 |
-
"eval_loss": 2.576280355453491,
|
| 480 |
-
"eval_runtime": 17.2831,
|
| 481 |
-
"eval_samples_per_second": 57.86,
|
| 482 |
-
"eval_steps_per_second": 14.465,
|
| 483 |
-
"step": 2750
|
| 484 |
-
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 0.21536804861164527,
|
| 487 |
-
"grad_norm": 3.6676604747772217,
|
| 488 |
-
"learning_rate": 0.00018808404539905982,
|
| 489 |
-
"loss": 2.4723,
|
| 490 |
-
"step": 2800
|
| 491 |
-
},
|
| 492 |
-
{
|
| 493 |
-
"epoch": 0.21921390662256748,
|
| 494 |
-
"grad_norm": 2.855665922164917,
|
| 495 |
-
"learning_rate": 0.00018782432537724336,
|
| 496 |
-
"loss": 2.5224,
|
| 497 |
-
"step": 2850
|
| 498 |
-
},
|
| 499 |
-
{
|
| 500 |
-
"epoch": 0.22305976463348973,
|
| 501 |
-
"grad_norm": 2.6359572410583496,
|
| 502 |
-
"learning_rate": 0.00018756460535542685,
|
| 503 |
-
"loss": 2.6181,
|
| 504 |
-
"step": 2900
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 0.22690562264441197,
|
| 508 |
-
"grad_norm": 2.5210607051849365,
|
| 509 |
-
"learning_rate": 0.00018730488533361036,
|
| 510 |
-
"loss": 2.4691,
|
| 511 |
-
"step": 2950
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 0.2307514806553342,
|
| 515 |
-
"grad_norm": 3.2655303478240967,
|
| 516 |
-
"learning_rate": 0.0001870451653117939,
|
| 517 |
-
"loss": 2.5791,
|
| 518 |
-
"step": 3000
|
| 519 |
-
},
|
| 520 |
-
{
|
| 521 |
-
"epoch": 0.2307514806553342,
|
| 522 |
-
"eval_loss": 2.5004756450653076,
|
| 523 |
-
"eval_runtime": 17.3413,
|
| 524 |
-
"eval_samples_per_second": 57.666,
|
| 525 |
-
"eval_steps_per_second": 14.416,
|
| 526 |
-
"step": 3000
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 0.23459733866625646,
|
| 530 |
-
"grad_norm": 3.1443817615509033,
|
| 531 |
-
"learning_rate": 0.0001867854452899774,
|
| 532 |
-
"loss": 2.5689,
|
| 533 |
-
"step": 3050
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.23844319667717867,
|
| 537 |
-
"grad_norm": 2.394235610961914,
|
| 538 |
-
"learning_rate": 0.00018652572526816092,
|
| 539 |
-
"loss": 2.4802,
|
| 540 |
-
"step": 3100
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"epoch": 0.2422890546881009,
|
| 544 |
-
"grad_norm": 3.216654062271118,
|
| 545 |
-
"learning_rate": 0.00018626600524634446,
|
| 546 |
-
"loss": 2.5239,
|
| 547 |
-
"step": 3150
|
| 548 |
-
},
|
| 549 |
-
{
|
| 550 |
-
"epoch": 0.24613491269902316,
|
| 551 |
-
"grad_norm": 2.438185691833496,
|
| 552 |
-
"learning_rate": 0.00018600628522452798,
|
| 553 |
-
"loss": 2.6504,
|
| 554 |
-
"step": 3200
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"epoch": 0.2499807707099454,
|
| 558 |
-
"grad_norm": 3.8275232315063477,
|
| 559 |
-
"learning_rate": 0.00018574656520271146,
|
| 560 |
-
"loss": 2.4848,
|
| 561 |
-
"step": 3250
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"epoch": 0.2499807707099454,
|
| 565 |
-
"eval_loss": 2.4598114490509033,
|
| 566 |
-
"eval_runtime": 17.3676,
|
| 567 |
-
"eval_samples_per_second": 57.578,
|
| 568 |
-
"eval_steps_per_second": 14.395,
|
| 569 |
-
"step": 3250
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.25382662872086764,
|
| 573 |
-
"grad_norm": 2.3598976135253906,
|
| 574 |
-
"learning_rate": 0.000185486845180895,
|
| 575 |
-
"loss": 2.3763,
|
| 576 |
-
"step": 3300
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.2576724867317899,
|
| 580 |
-
"grad_norm": 2.642561912536621,
|
| 581 |
-
"learning_rate": 0.00018522712515907852,
|
| 582 |
-
"loss": 2.5,
|
| 583 |
-
"step": 3350
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.26151834474271207,
|
| 587 |
-
"grad_norm": 1.7602814435958862,
|
| 588 |
-
"learning_rate": 0.00018496740513726206,
|
| 589 |
-
"loss": 2.4706,
|
| 590 |
-
"step": 3400
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.2653642027536343,
|
| 594 |
-
"grad_norm": 1.542140245437622,
|
| 595 |
-
"learning_rate": 0.00018470768511544557,
|
| 596 |
-
"loss": 2.4107,
|
| 597 |
-
"step": 3450
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.26921006076455656,
|
| 601 |
-
"grad_norm": 1.7666794061660767,
|
| 602 |
-
"learning_rate": 0.00018444796509362908,
|
| 603 |
-
"loss": 2.4018,
|
| 604 |
-
"step": 3500
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.26921006076455656,
|
| 608 |
-
"eval_loss": 2.4329559803009033,
|
| 609 |
-
"eval_runtime": 17.3412,
|
| 610 |
-
"eval_samples_per_second": 57.666,
|
| 611 |
-
"eval_steps_per_second": 14.417,
|
| 612 |
-
"step": 3500
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.2730559187754788,
|
| 616 |
-
"grad_norm": 2.434197425842285,
|
| 617 |
-
"learning_rate": 0.0001841882450718126,
|
| 618 |
-
"loss": 2.4402,
|
| 619 |
-
"step": 3550
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.27690177678640104,
|
| 623 |
-
"grad_norm": 2.393425226211548,
|
| 624 |
-
"learning_rate": 0.0001839285250499961,
|
| 625 |
-
"loss": 2.4952,
|
| 626 |
-
"step": 3600
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.2807476347973233,
|
| 630 |
-
"grad_norm": 2.2749860286712646,
|
| 631 |
-
"learning_rate": 0.00018366880502817962,
|
| 632 |
-
"loss": 2.4638,
|
| 633 |
-
"step": 3650
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.28459349280824553,
|
| 637 |
-
"grad_norm": 1.988981008529663,
|
| 638 |
-
"learning_rate": 0.00018340908500636316,
|
| 639 |
-
"loss": 2.3414,
|
| 640 |
-
"step": 3700
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.2884393508191678,
|
| 644 |
-
"grad_norm": 2.1251513957977295,
|
| 645 |
-
"learning_rate": 0.00018314936498454667,
|
| 646 |
-
"loss": 2.4272,
|
| 647 |
-
"step": 3750
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.2884393508191678,
|
| 651 |
-
"eval_loss": 2.4050426483154297,
|
| 652 |
-
"eval_runtime": 17.2358,
|
| 653 |
-
"eval_samples_per_second": 58.019,
|
| 654 |
-
"eval_steps_per_second": 14.505,
|
| 655 |
-
"step": 3750
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 0.29228520883009,
|
| 659 |
-
"grad_norm": 2.7036380767822266,
|
| 660 |
-
"learning_rate": 0.00018288964496273019,
|
| 661 |
-
"loss": 2.3573,
|
| 662 |
-
"step": 3800
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 0.2961310668410122,
|
| 666 |
-
"grad_norm": 2.6849310398101807,
|
| 667 |
-
"learning_rate": 0.0001826299249409137,
|
| 668 |
-
"loss": 2.3911,
|
| 669 |
-
"step": 3850
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.29997692485193445,
|
| 673 |
-
"grad_norm": 3.012932777404785,
|
| 674 |
-
"learning_rate": 0.0001823702049190972,
|
| 675 |
-
"loss": 2.451,
|
| 676 |
-
"step": 3900
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 0.3038227828628567,
|
| 680 |
-
"grad_norm": 2.8903510570526123,
|
| 681 |
-
"learning_rate": 0.00018211048489728072,
|
| 682 |
-
"loss": 2.3306,
|
| 683 |
-
"step": 3950
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 0.30766864087377893,
|
| 687 |
-
"grad_norm": 1.4737602472305298,
|
| 688 |
-
"learning_rate": 0.00018185076487546426,
|
| 689 |
-
"loss": 2.3145,
|
| 690 |
-
"step": 4000
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 0.30766864087377893,
|
| 694 |
-
"eval_loss": 2.364640712738037,
|
| 695 |
-
"eval_runtime": 17.2167,
|
| 696 |
-
"eval_samples_per_second": 58.083,
|
| 697 |
-
"eval_steps_per_second": 14.521,
|
| 698 |
-
"step": 4000
|
| 699 |
-
},
|
| 700 |
-
{
|
| 701 |
-
"epoch": 0.3115144988847012,
|
| 702 |
-
"grad_norm": 1.781186819076538,
|
| 703 |
-
"learning_rate": 0.00018159104485364778,
|
| 704 |
-
"loss": 2.3993,
|
| 705 |
-
"step": 4050
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 0.3153603568956234,
|
| 709 |
-
"grad_norm": 2.6136422157287598,
|
| 710 |
-
"learning_rate": 0.0001813313248318313,
|
| 711 |
-
"loss": 2.3217,
|
| 712 |
-
"step": 4100
|
| 713 |
-
},
|
| 714 |
-
{
|
| 715 |
-
"epoch": 0.31920621490654566,
|
| 716 |
-
"grad_norm": 2.6466028690338135,
|
| 717 |
-
"learning_rate": 0.00018107160481001483,
|
| 718 |
-
"loss": 2.369,
|
| 719 |
-
"step": 4150
|
| 720 |
-
},
|
| 721 |
-
{
|
| 722 |
-
"epoch": 0.3230520729174679,
|
| 723 |
-
"grad_norm": 2.442426919937134,
|
| 724 |
-
"learning_rate": 0.00018081188478819832,
|
| 725 |
-
"loss": 2.3943,
|
| 726 |
-
"step": 4200
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 0.32689793092839015,
|
| 730 |
-
"grad_norm": 5.356844902038574,
|
| 731 |
-
"learning_rate": 0.00018055216476638186,
|
| 732 |
-
"loss": 2.39,
|
| 733 |
-
"step": 4250
|
| 734 |
-
},
|
| 735 |
-
{
|
| 736 |
-
"epoch": 0.32689793092839015,
|
| 737 |
-
"eval_loss": 2.3410093784332275,
|
| 738 |
-
"eval_runtime": 17.2633,
|
| 739 |
-
"eval_samples_per_second": 57.926,
|
| 740 |
-
"eval_steps_per_second": 14.482,
|
| 741 |
-
"step": 4250
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 0.33074378893931233,
|
| 745 |
-
"grad_norm": 1.8614141941070557,
|
| 746 |
-
"learning_rate": 0.00018029244474456537,
|
| 747 |
-
"loss": 2.3852,
|
| 748 |
-
"step": 4300
|
| 749 |
-
},
|
| 750 |
-
{
|
| 751 |
-
"epoch": 0.3345896469502346,
|
| 752 |
-
"grad_norm": 1.7005974054336548,
|
| 753 |
-
"learning_rate": 0.00018003272472274888,
|
| 754 |
-
"loss": 2.2891,
|
| 755 |
-
"step": 4350
|
| 756 |
-
},
|
| 757 |
-
{
|
| 758 |
-
"epoch": 0.3384355049611568,
|
| 759 |
-
"grad_norm": 2.0309245586395264,
|
| 760 |
-
"learning_rate": 0.00017977300470093242,
|
| 761 |
-
"loss": 2.3093,
|
| 762 |
-
"step": 4400
|
| 763 |
-
},
|
| 764 |
-
{
|
| 765 |
-
"epoch": 0.34228136297207906,
|
| 766 |
-
"grad_norm": 2.6562678813934326,
|
| 767 |
-
"learning_rate": 0.00017951328467911593,
|
| 768 |
-
"loss": 2.2637,
|
| 769 |
-
"step": 4450
|
| 770 |
-
},
|
| 771 |
-
{
|
| 772 |
-
"epoch": 0.3461272209830013,
|
| 773 |
-
"grad_norm": 1.352026343345642,
|
| 774 |
-
"learning_rate": 0.00017925356465729945,
|
| 775 |
-
"loss": 2.2263,
|
| 776 |
-
"step": 4500
|
| 777 |
-
},
|
| 778 |
-
{
|
| 779 |
-
"epoch": 0.3461272209830013,
|
| 780 |
-
"eval_loss": 2.2719833850860596,
|
| 781 |
-
"eval_runtime": 17.4568,
|
| 782 |
-
"eval_samples_per_second": 57.284,
|
| 783 |
-
"eval_steps_per_second": 14.321,
|
| 784 |
-
"step": 4500
|
| 785 |
-
},
|
| 786 |
-
{
|
| 787 |
-
"epoch": 0.34997307899392355,
|
| 788 |
-
"grad_norm": 2.6207292079925537,
|
| 789 |
-
"learning_rate": 0.00017899384463548296,
|
| 790 |
-
"loss": 2.325,
|
| 791 |
-
"step": 4550
|
| 792 |
-
},
|
| 793 |
-
{
|
| 794 |
-
"epoch": 0.3538189370048458,
|
| 795 |
-
"grad_norm": 1.9380669593811035,
|
| 796 |
-
"learning_rate": 0.00017873412461366647,
|
| 797 |
-
"loss": 2.3208,
|
| 798 |
-
"step": 4600
|
| 799 |
-
},
|
| 800 |
-
{
|
| 801 |
-
"epoch": 0.35766479501576803,
|
| 802 |
-
"grad_norm": 2.2287373542785645,
|
| 803 |
-
"learning_rate": 0.00017847440459184999,
|
| 804 |
-
"loss": 2.3115,
|
| 805 |
-
"step": 4650
|
| 806 |
-
},
|
| 807 |
-
{
|
| 808 |
-
"epoch": 0.3615106530266903,
|
| 809 |
-
"grad_norm": 2.3176259994506836,
|
| 810 |
-
"learning_rate": 0.00017821468457003353,
|
| 811 |
-
"loss": 2.2203,
|
| 812 |
-
"step": 4700
|
| 813 |
-
},
|
| 814 |
-
{
|
| 815 |
-
"epoch": 0.36535651103761246,
|
| 816 |
-
"grad_norm": 1.9587980508804321,
|
| 817 |
-
"learning_rate": 0.00017795496454821704,
|
| 818 |
-
"loss": 2.2162,
|
| 819 |
-
"step": 4750
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 0.36535651103761246,
|
| 823 |
-
"eval_loss": 2.263962984085083,
|
| 824 |
-
"eval_runtime": 17.2653,
|
| 825 |
-
"eval_samples_per_second": 57.92,
|
| 826 |
-
"eval_steps_per_second": 14.48,
|
| 827 |
-
"step": 4750
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 0.3692023690485347,
|
| 831 |
-
"grad_norm": 1.8225115537643433,
|
| 832 |
-
"learning_rate": 0.00017769524452640055,
|
| 833 |
-
"loss": 2.2906,
|
| 834 |
-
"step": 4800
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 0.37304822705945695,
|
| 838 |
-
"grad_norm": 2.2642672061920166,
|
| 839 |
-
"learning_rate": 0.00017743552450458406,
|
| 840 |
-
"loss": 2.2819,
|
| 841 |
-
"step": 4850
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 0.3768940850703792,
|
| 845 |
-
"grad_norm": 2.1667556762695312,
|
| 846 |
-
"learning_rate": 0.00017717580448276758,
|
| 847 |
-
"loss": 2.2756,
|
| 848 |
-
"step": 4900
|
| 849 |
-
},
|
| 850 |
-
{
|
| 851 |
-
"epoch": 0.38073994308130144,
|
| 852 |
-
"grad_norm": 2.112743854522705,
|
| 853 |
-
"learning_rate": 0.0001769160844609511,
|
| 854 |
-
"loss": 2.2583,
|
| 855 |
-
"step": 4950
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 0.3845858010922237,
|
| 859 |
-
"grad_norm": 3.304570436477661,
|
| 860 |
-
"learning_rate": 0.00017665636443913463,
|
| 861 |
-
"loss": 2.2407,
|
| 862 |
-
"step": 5000
|
| 863 |
-
},
|
| 864 |
-
{
|
| 865 |
-
"epoch": 0.3845858010922237,
|
| 866 |
-
"eval_loss": 2.229123830795288,
|
| 867 |
-
"eval_runtime": 17.1378,
|
| 868 |
-
"eval_samples_per_second": 58.351,
|
| 869 |
-
"eval_steps_per_second": 14.588,
|
| 870 |
-
"step": 5000
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 0.3884316591031459,
|
| 874 |
-
"grad_norm": 1.6103578805923462,
|
| 875 |
-
"learning_rate": 0.00017639664441731814,
|
| 876 |
-
"loss": 2.3529,
|
| 877 |
-
"step": 5050
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 0.39227751711406816,
|
| 881 |
-
"grad_norm": 1.2587641477584839,
|
| 882 |
-
"learning_rate": 0.00017613692439550165,
|
| 883 |
-
"loss": 2.2583,
|
| 884 |
-
"step": 5100
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 0.3961233751249904,
|
| 888 |
-
"grad_norm": 3.0143070220947266,
|
| 889 |
-
"learning_rate": 0.00017587720437368517,
|
| 890 |
-
"loss": 2.1529,
|
| 891 |
-
"step": 5150
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 0.39996923313591265,
|
| 895 |
-
"grad_norm": 2.145265579223633,
|
| 896 |
-
"learning_rate": 0.00017561748435186868,
|
| 897 |
-
"loss": 2.2835,
|
| 898 |
-
"step": 5200
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 0.40381509114683484,
|
| 902 |
-
"grad_norm": 1.7174979448318481,
|
| 903 |
-
"learning_rate": 0.00017535776433005222,
|
| 904 |
-
"loss": 2.2701,
|
| 905 |
-
"step": 5250
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 0.40381509114683484,
|
| 909 |
-
"eval_loss": 2.215972661972046,
|
| 910 |
-
"eval_runtime": 17.0477,
|
| 911 |
-
"eval_samples_per_second": 58.659,
|
| 912 |
-
"eval_steps_per_second": 14.665,
|
| 913 |
-
"step": 5250
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 0.4076609491577571,
|
| 917 |
-
"grad_norm": 1.8635681867599487,
|
| 918 |
-
"learning_rate": 0.00017509804430823573,
|
| 919 |
-
"loss": 2.1989,
|
| 920 |
-
"step": 5300
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.4115068071686793,
|
| 924 |
-
"grad_norm": 4.2349700927734375,
|
| 925 |
-
"learning_rate": 0.00017483832428641925,
|
| 926 |
-
"loss": 2.2203,
|
| 927 |
-
"step": 5350
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.41535266517960157,
|
| 931 |
-
"grad_norm": 2.0099525451660156,
|
| 932 |
-
"learning_rate": 0.00017457860426460279,
|
| 933 |
-
"loss": 2.2021,
|
| 934 |
-
"step": 5400
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 0.4191985231905238,
|
| 938 |
-
"grad_norm": 2.215400457382202,
|
| 939 |
-
"learning_rate": 0.00017431888424278627,
|
| 940 |
-
"loss": 2.1583,
|
| 941 |
-
"step": 5450
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 0.42304438120144605,
|
| 945 |
-
"grad_norm": 2.2177133560180664,
|
| 946 |
-
"learning_rate": 0.00017405916422096978,
|
| 947 |
-
"loss": 2.2502,
|
| 948 |
-
"step": 5500
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 0.42304438120144605,
|
| 952 |
-
"eval_loss": 2.16481876373291,
|
| 953 |
-
"eval_runtime": 17.3392,
|
| 954 |
-
"eval_samples_per_second": 57.673,
|
| 955 |
-
"eval_steps_per_second": 14.418,
|
| 956 |
-
"step": 5500
|
| 957 |
-
},
|
| 958 |
-
{
|
| 959 |
-
"epoch": 0.4268902392123683,
|
| 960 |
-
"grad_norm": 2.4246537685394287,
|
| 961 |
-
"learning_rate": 0.00017379944419915332,
|
| 962 |
-
"loss": 2.1506,
|
| 963 |
-
"step": 5550
|
| 964 |
-
},
|
| 965 |
-
{
|
| 966 |
-
"epoch": 0.43073609722329054,
|
| 967 |
-
"grad_norm": 2.5737595558166504,
|
| 968 |
-
"learning_rate": 0.00017353972417733684,
|
| 969 |
-
"loss": 2.1155,
|
| 970 |
-
"step": 5600
|
| 971 |
-
},
|
| 972 |
-
{
|
| 973 |
-
"epoch": 0.4345819552342128,
|
| 974 |
-
"grad_norm": 1.9406490325927734,
|
| 975 |
-
"learning_rate": 0.00017328000415552035,
|
| 976 |
-
"loss": 2.1716,
|
| 977 |
-
"step": 5650
|
| 978 |
-
},
|
| 979 |
-
{
|
| 980 |
-
"epoch": 0.43842781324513497,
|
| 981 |
-
"grad_norm": 1.8545359373092651,
|
| 982 |
-
"learning_rate": 0.0001730202841337039,
|
| 983 |
-
"loss": 2.3107,
|
| 984 |
-
"step": 5700
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 0.4422736712560572,
|
| 988 |
-
"grad_norm": 1.4506388902664185,
|
| 989 |
-
"learning_rate": 0.0001727605641118874,
|
| 990 |
-
"loss": 2.1839,
|
| 991 |
-
"step": 5750
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 0.4422736712560572,
|
| 995 |
-
"eval_loss": 2.1573686599731445,
|
| 996 |
-
"eval_runtime": 17.1406,
|
| 997 |
-
"eval_samples_per_second": 58.341,
|
| 998 |
-
"eval_steps_per_second": 14.585,
|
| 999 |
-
"step": 5750
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.44611952926697945,
|
| 1003 |
-
"grad_norm": 2.4454505443573,
|
| 1004 |
-
"learning_rate": 0.0001725008440900709,
|
| 1005 |
-
"loss": 2.2399,
|
| 1006 |
-
"step": 5800
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 0.4499653872779017,
|
| 1010 |
-
"grad_norm": 1.7306182384490967,
|
| 1011 |
-
"learning_rate": 0.00017224112406825443,
|
| 1012 |
-
"loss": 2.2319,
|
| 1013 |
-
"step": 5850
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 0.45381124528882394,
|
| 1017 |
-
"grad_norm": 1.8138234615325928,
|
| 1018 |
-
"learning_rate": 0.00017198140404643794,
|
| 1019 |
-
"loss": 2.1173,
|
| 1020 |
-
"step": 5900
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 0.4576571032997462,
|
| 1024 |
-
"grad_norm": 1.6768412590026855,
|
| 1025 |
-
"learning_rate": 0.00017172168402462148,
|
| 1026 |
-
"loss": 2.2145,
|
| 1027 |
-
"step": 5950
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 0.4615029613106684,
|
| 1031 |
-
"grad_norm": 2.1484451293945312,
|
| 1032 |
-
"learning_rate": 0.000171461964002805,
|
| 1033 |
-
"loss": 2.171,
|
| 1034 |
-
"step": 6000
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 0.4615029613106684,
|
| 1038 |
-
"eval_loss": 2.1329145431518555,
|
| 1039 |
-
"eval_runtime": 17.1772,
|
| 1040 |
-
"eval_samples_per_second": 58.217,
|
| 1041 |
-
"eval_steps_per_second": 14.554,
|
| 1042 |
-
"step": 6000
|
| 1043 |
-
},
|
| 1044 |
-
{
|
| 1045 |
-
"epoch": 0.46534881932159067,
|
| 1046 |
-
"grad_norm": 2.217759609222412,
|
| 1047 |
-
"learning_rate": 0.0001712022439809885,
|
| 1048 |
-
"loss": 2.1679,
|
| 1049 |
-
"step": 6050
|
| 1050 |
-
},
|
| 1051 |
-
{
|
| 1052 |
-
"epoch": 0.4691946773325129,
|
| 1053 |
-
"grad_norm": 1.4925270080566406,
|
| 1054 |
-
"learning_rate": 0.00017094252395917202,
|
| 1055 |
-
"loss": 2.1634,
|
| 1056 |
-
"step": 6100
|
| 1057 |
-
},
|
| 1058 |
-
{
|
| 1059 |
-
"epoch": 0.4730405353434351,
|
| 1060 |
-
"grad_norm": 1.500253438949585,
|
| 1061 |
-
"learning_rate": 0.00017068280393735553,
|
| 1062 |
-
"loss": 2.1947,
|
| 1063 |
-
"step": 6150
|
| 1064 |
-
},
|
| 1065 |
-
{
|
| 1066 |
-
"epoch": 0.47688639335435734,
|
| 1067 |
-
"grad_norm": 1.7444700002670288,
|
| 1068 |
-
"learning_rate": 0.00017042308391553905,
|
| 1069 |
-
"loss": 2.1864,
|
| 1070 |
-
"step": 6200
|
| 1071 |
-
},
|
| 1072 |
-
{
|
| 1073 |
-
"epoch": 0.4807322513652796,
|
| 1074 |
-
"grad_norm": 2.21281099319458,
|
| 1075 |
-
"learning_rate": 0.00017016336389372259,
|
| 1076 |
-
"loss": 2.1346,
|
| 1077 |
-
"step": 6250
|
| 1078 |
-
},
|
| 1079 |
-
{
|
| 1080 |
-
"epoch": 0.4807322513652796,
|
| 1081 |
-
"eval_loss": 2.1062209606170654,
|
| 1082 |
-
"eval_runtime": 17.175,
|
| 1083 |
-
"eval_samples_per_second": 58.224,
|
| 1084 |
-
"eval_steps_per_second": 14.556,
|
| 1085 |
-
"step": 6250
|
| 1086 |
-
},
|
| 1087 |
-
{
|
| 1088 |
-
"epoch": 0.4845781093762018,
|
| 1089 |
-
"grad_norm": 2.066366672515869,
|
| 1090 |
-
"learning_rate": 0.0001699036438719061,
|
| 1091 |
-
"loss": 2.1034,
|
| 1092 |
-
"step": 6300
|
| 1093 |
-
},
|
| 1094 |
-
{
|
| 1095 |
-
"epoch": 0.48842396738712407,
|
| 1096 |
-
"grad_norm": 2.236504554748535,
|
| 1097 |
-
"learning_rate": 0.0001696439238500896,
|
| 1098 |
-
"loss": 2.1293,
|
| 1099 |
-
"step": 6350
|
| 1100 |
-
},
|
| 1101 |
-
{
|
| 1102 |
-
"epoch": 0.4922698253980463,
|
| 1103 |
-
"grad_norm": 2.368986129760742,
|
| 1104 |
-
"learning_rate": 0.00016938420382827312,
|
| 1105 |
-
"loss": 2.2147,
|
| 1106 |
-
"step": 6400
|
| 1107 |
-
},
|
| 1108 |
-
{
|
| 1109 |
-
"epoch": 0.49611568340896856,
|
| 1110 |
-
"grad_norm": 3.3609116077423096,
|
| 1111 |
-
"learning_rate": 0.00016912448380645664,
|
| 1112 |
-
"loss": 2.1395,
|
| 1113 |
-
"step": 6450
|
| 1114 |
-
},
|
| 1115 |
-
{
|
| 1116 |
-
"epoch": 0.4999615414198908,
|
| 1117 |
-
"grad_norm": 1.7160570621490479,
|
| 1118 |
-
"learning_rate": 0.00016886476378464015,
|
| 1119 |
-
"loss": 2.0777,
|
| 1120 |
-
"step": 6500
|
| 1121 |
-
},
|
| 1122 |
-
{
|
| 1123 |
-
"epoch": 0.4999615414198908,
|
| 1124 |
-
"eval_loss": 2.0895941257476807,
|
| 1125 |
-
"eval_runtime": 17.0411,
|
| 1126 |
-
"eval_samples_per_second": 58.682,
|
| 1127 |
-
"eval_steps_per_second": 14.67,
|
| 1128 |
-
"step": 6500
|
| 1129 |
-
},
|
| 1130 |
-
{
|
| 1131 |
-
"epoch": 0.503807399430813,
|
| 1132 |
-
"grad_norm": 2.2851459980010986,
|
| 1133 |
-
"learning_rate": 0.0001686050437628237,
|
| 1134 |
-
"loss": 2.155,
|
| 1135 |
-
"step": 6550
|
| 1136 |
-
},
|
| 1137 |
-
{
|
| 1138 |
-
"epoch": 0.5076532574417353,
|
| 1139 |
-
"grad_norm": 3.6158559322357178,
|
| 1140 |
-
"learning_rate": 0.0001683453237410072,
|
| 1141 |
-
"loss": 2.1105,
|
| 1142 |
-
"step": 6600
|
| 1143 |
-
},
|
| 1144 |
-
{
|
| 1145 |
-
"epoch": 0.5114991154526575,
|
| 1146 |
-
"grad_norm": 1.608955979347229,
|
| 1147 |
-
"learning_rate": 0.00016808560371919072,
|
| 1148 |
-
"loss": 2.0908,
|
| 1149 |
-
"step": 6650
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 0.5153449734635798,
|
| 1153 |
-
"grad_norm": 2.7433218955993652,
|
| 1154 |
-
"learning_rate": 0.00016782588369737426,
|
| 1155 |
-
"loss": 2.04,
|
| 1156 |
-
"step": 6700
|
| 1157 |
-
},
|
| 1158 |
-
{
|
| 1159 |
-
"epoch": 0.519190831474502,
|
| 1160 |
-
"grad_norm": 3.286970376968384,
|
| 1161 |
-
"learning_rate": 0.00016756616367555774,
|
| 1162 |
-
"loss": 2.1608,
|
| 1163 |
-
"step": 6750
|
| 1164 |
-
},
|
| 1165 |
-
{
|
| 1166 |
-
"epoch": 0.519190831474502,
|
| 1167 |
-
"eval_loss": 2.0548338890075684,
|
| 1168 |
-
"eval_runtime": 17.3257,
|
| 1169 |
-
"eval_samples_per_second": 57.718,
|
| 1170 |
-
"eval_steps_per_second": 14.429,
|
| 1171 |
-
"step": 6750
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 0.5230366894854241,
|
| 1175 |
-
"grad_norm": 1.7461440563201904,
|
| 1176 |
-
"learning_rate": 0.00016730644365374128,
|
| 1177 |
-
"loss": 2.0562,
|
| 1178 |
-
"step": 6800
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 0.5268825474963464,
|
| 1182 |
-
"grad_norm": 3.4528918266296387,
|
| 1183 |
-
"learning_rate": 0.0001670467236319248,
|
| 1184 |
-
"loss": 2.0983,
|
| 1185 |
-
"step": 6850
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 0.5307284055072686,
|
| 1189 |
-
"grad_norm": 2.23818302154541,
|
| 1190 |
-
"learning_rate": 0.0001667870036101083,
|
| 1191 |
-
"loss": 2.0187,
|
| 1192 |
-
"step": 6900
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 0.5345742635181909,
|
| 1196 |
-
"grad_norm": 1.7558523416519165,
|
| 1197 |
-
"learning_rate": 0.00016652728358829185,
|
| 1198 |
-
"loss": 2.0996,
|
| 1199 |
-
"step": 6950
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 0.5384201215291131,
|
| 1203 |
-
"grad_norm": 1.8939173221588135,
|
| 1204 |
-
"learning_rate": 0.00016626756356647536,
|
| 1205 |
-
"loss": 2.0499,
|
| 1206 |
-
"step": 7000
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 0.5384201215291131,
|
| 1210 |
-
"eval_loss": 2.0413780212402344,
|
| 1211 |
-
"eval_runtime": 17.1018,
|
| 1212 |
-
"eval_samples_per_second": 58.473,
|
| 1213 |
-
"eval_steps_per_second": 14.618,
|
| 1214 |
-
"step": 7000
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.5422659795400354,
|
| 1218 |
-
"grad_norm": 2.8039333820343018,
|
| 1219 |
-
"learning_rate": 0.00016600784354465885,
|
| 1220 |
-
"loss": 2.0606,
|
| 1221 |
-
"step": 7050
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 0.5461118375509576,
|
| 1225 |
-
"grad_norm": 1.4562283754348755,
|
| 1226 |
-
"learning_rate": 0.00016574812352284239,
|
| 1227 |
-
"loss": 2.1611,
|
| 1228 |
-
"step": 7100
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 0.5499576955618799,
|
| 1232 |
-
"grad_norm": 1.4812073707580566,
|
| 1233 |
-
"learning_rate": 0.0001654884035010259,
|
| 1234 |
-
"loss": 2.125,
|
| 1235 |
-
"step": 7150
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 0.5538035535728021,
|
| 1239 |
-
"grad_norm": 2.279209613800049,
|
| 1240 |
-
"learning_rate": 0.0001652286834792094,
|
| 1241 |
-
"loss": 2.0865,
|
| 1242 |
-
"step": 7200
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 0.5576494115837243,
|
| 1246 |
-
"grad_norm": 3.1694416999816895,
|
| 1247 |
-
"learning_rate": 0.00016496896345739295,
|
| 1248 |
-
"loss": 2.1783,
|
| 1249 |
-
"step": 7250
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 0.5576494115837243,
|
| 1253 |
-
"eval_loss": 2.0270566940307617,
|
| 1254 |
-
"eval_runtime": 16.9999,
|
| 1255 |
-
"eval_samples_per_second": 58.824,
|
| 1256 |
-
"eval_steps_per_second": 14.706,
|
| 1257 |
-
"step": 7250
|
| 1258 |
-
},
|
| 1259 |
-
{
|
| 1260 |
-
"epoch": 0.5614952695946466,
|
| 1261 |
-
"grad_norm": 2.7421817779541016,
|
| 1262 |
-
"learning_rate": 0.00016470924343557646,
|
| 1263 |
-
"loss": 2.0691,
|
| 1264 |
-
"step": 7300
|
| 1265 |
-
},
|
| 1266 |
-
{
|
| 1267 |
-
"epoch": 0.5653411276055688,
|
| 1268 |
-
"grad_norm": 1.6490452289581299,
|
| 1269 |
-
"learning_rate": 0.00016444952341375998,
|
| 1270 |
-
"loss": 2.0467,
|
| 1271 |
-
"step": 7350
|
| 1272 |
-
},
|
| 1273 |
-
{
|
| 1274 |
-
"epoch": 0.5691869856164911,
|
| 1275 |
-
"grad_norm": 2.1511409282684326,
|
| 1276 |
-
"learning_rate": 0.0001641898033919435,
|
| 1277 |
-
"loss": 2.0369,
|
| 1278 |
-
"step": 7400
|
| 1279 |
-
},
|
| 1280 |
-
{
|
| 1281 |
-
"epoch": 0.5730328436274132,
|
| 1282 |
-
"grad_norm": 1.9185343980789185,
|
| 1283 |
-
"learning_rate": 0.000163930083370127,
|
| 1284 |
-
"loss": 2.1182,
|
| 1285 |
-
"step": 7450
|
| 1286 |
-
},
|
| 1287 |
-
{
|
| 1288 |
-
"epoch": 0.5768787016383355,
|
| 1289 |
-
"grad_norm": 1.5140857696533203,
|
| 1290 |
-
"learning_rate": 0.00016367036334831052,
|
| 1291 |
-
"loss": 2.1145,
|
| 1292 |
-
"step": 7500
|
| 1293 |
-
},
|
| 1294 |
-
{
|
| 1295 |
-
"epoch": 0.5768787016383355,
|
| 1296 |
-
"eval_loss": 2.016103744506836,
|
| 1297 |
-
"eval_runtime": 17.1749,
|
| 1298 |
-
"eval_samples_per_second": 58.225,
|
| 1299 |
-
"eval_steps_per_second": 14.556,
|
| 1300 |
-
"step": 7500
|
| 1301 |
-
},
|
| 1302 |
-
{
|
| 1303 |
-
"epoch": 0.5807245596492577,
|
| 1304 |
-
"grad_norm": 1.6893503665924072,
|
| 1305 |
-
"learning_rate": 0.00016341064332649406,
|
| 1306 |
-
"loss": 2.0415,
|
| 1307 |
-
"step": 7550
|
| 1308 |
-
},
|
| 1309 |
-
{
|
| 1310 |
-
"epoch": 0.58457041766018,
|
| 1311 |
-
"grad_norm": 2.099968433380127,
|
| 1312 |
-
"learning_rate": 0.00016315092330467757,
|
| 1313 |
-
"loss": 1.9203,
|
| 1314 |
-
"step": 7600
|
| 1315 |
-
},
|
| 1316 |
-
{
|
| 1317 |
-
"epoch": 0.5884162756711022,
|
| 1318 |
-
"grad_norm": 2.0659477710723877,
|
| 1319 |
-
"learning_rate": 0.00016289120328286108,
|
| 1320 |
-
"loss": 2.1188,
|
| 1321 |
-
"step": 7650
|
| 1322 |
-
},
|
| 1323 |
-
{
|
| 1324 |
-
"epoch": 0.5922621336820244,
|
| 1325 |
-
"grad_norm": 1.599091649055481,
|
| 1326 |
-
"learning_rate": 0.0001626314832610446,
|
| 1327 |
-
"loss": 1.9583,
|
| 1328 |
-
"step": 7700
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 0.5961079916929467,
|
| 1332 |
-
"grad_norm": 2.5325448513031006,
|
| 1333 |
-
"learning_rate": 0.0001623717632392281,
|
| 1334 |
-
"loss": 2.088,
|
| 1335 |
-
"step": 7750
|
| 1336 |
-
},
|
| 1337 |
-
{
|
| 1338 |
-
"epoch": 0.5961079916929467,
|
| 1339 |
-
"eval_loss": 1.9883803129196167,
|
| 1340 |
-
"eval_runtime": 17.4065,
|
| 1341 |
-
"eval_samples_per_second": 57.45,
|
| 1342 |
-
"eval_steps_per_second": 14.362,
|
| 1343 |
-
"step": 7750
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 0.5999538497038689,
|
| 1347 |
-
"grad_norm": 1.6605775356292725,
|
| 1348 |
-
"learning_rate": 0.00016211204321741165,
|
| 1349 |
-
"loss": 2.0767,
|
| 1350 |
-
"step": 7800
|
| 1351 |
-
},
|
| 1352 |
-
{
|
| 1353 |
-
"epoch": 0.6037997077147912,
|
| 1354 |
-
"grad_norm": 1.4154255390167236,
|
| 1355 |
-
"learning_rate": 0.00016185232319559516,
|
| 1356 |
-
"loss": 1.9914,
|
| 1357 |
-
"step": 7850
|
| 1358 |
-
},
|
| 1359 |
-
{
|
| 1360 |
-
"epoch": 0.6076455657257134,
|
| 1361 |
-
"grad_norm": 1.7516275644302368,
|
| 1362 |
-
"learning_rate": 0.00016159260317377867,
|
| 1363 |
-
"loss": 2.0046,
|
| 1364 |
-
"step": 7900
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 0.6114914237366357,
|
| 1368 |
-
"grad_norm": 1.6723463535308838,
|
| 1369 |
-
"learning_rate": 0.0001613328831519622,
|
| 1370 |
-
"loss": 2.0312,
|
| 1371 |
-
"step": 7950
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 0.6153372817475579,
|
| 1375 |
-
"grad_norm": 2.417466163635254,
|
| 1376 |
-
"learning_rate": 0.0001610731631301457,
|
| 1377 |
-
"loss": 2.1176,
|
| 1378 |
-
"step": 8000
|
| 1379 |
-
},
|
| 1380 |
-
{
|
| 1381 |
-
"epoch": 0.6153372817475579,
|
| 1382 |
-
"eval_loss": 1.9759701490402222,
|
| 1383 |
-
"eval_runtime": 17.031,
|
| 1384 |
-
"eval_samples_per_second": 58.717,
|
| 1385 |
-
"eval_steps_per_second": 14.679,
|
| 1386 |
-
"step": 8000
|
| 1387 |
-
},
|
| 1388 |
-
{
|
| 1389 |
-
"epoch": 0.6191831397584802,
|
| 1390 |
-
"grad_norm": 2.4189867973327637,
|
| 1391 |
-
"learning_rate": 0.0001608134431083292,
|
| 1392 |
-
"loss": 1.9562,
|
| 1393 |
-
"step": 8050
|
| 1394 |
-
},
|
| 1395 |
-
{
|
| 1396 |
-
"epoch": 0.6230289977694023,
|
| 1397 |
-
"grad_norm": 2.1052446365356445,
|
| 1398 |
-
"learning_rate": 0.00016055372308651275,
|
| 1399 |
-
"loss": 2.0396,
|
| 1400 |
-
"step": 8100
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 0.6268748557803245,
|
| 1404 |
-
"grad_norm": 2.675004482269287,
|
| 1405 |
-
"learning_rate": 0.00016029400306469626,
|
| 1406 |
-
"loss": 2.0172,
|
| 1407 |
-
"step": 8150
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 0.6307207137912468,
|
| 1411 |
-
"grad_norm": 3.2394967079162598,
|
| 1412 |
-
"learning_rate": 0.00016003947744331611,
|
| 1413 |
-
"loss": 2.0258,
|
| 1414 |
-
"step": 8200
|
| 1415 |
-
},
|
| 1416 |
-
{
|
| 1417 |
-
"epoch": 0.634566571802169,
|
| 1418 |
-
"grad_norm": 1.610350489616394,
|
| 1419 |
-
"learning_rate": 0.00015977975742149963,
|
| 1420 |
-
"loss": 1.9348,
|
| 1421 |
-
"step": 8250
|
| 1422 |
-
},
|
| 1423 |
-
{
|
| 1424 |
-
"epoch": 0.634566571802169,
|
| 1425 |
-
"eval_loss": 1.9506243467330933,
|
| 1426 |
-
"eval_runtime": 17.1568,
|
| 1427 |
-
"eval_samples_per_second": 58.286,
|
| 1428 |
-
"eval_steps_per_second": 14.571,
|
| 1429 |
-
"step": 8250
|
| 1430 |
-
},
|
| 1431 |
-
{
|
| 1432 |
-
"epoch": 0.6384124298130913,
|
| 1433 |
-
"grad_norm": 2.4209282398223877,
|
| 1434 |
-
"learning_rate": 0.00015952003739968314,
|
| 1435 |
-
"loss": 1.9787,
|
| 1436 |
-
"step": 8300
|
| 1437 |
-
},
|
| 1438 |
-
{
|
| 1439 |
-
"epoch": 0.6422582878240135,
|
| 1440 |
-
"grad_norm": 1.3403830528259277,
|
| 1441 |
-
"learning_rate": 0.00015926031737786665,
|
| 1442 |
-
"loss": 1.9947,
|
| 1443 |
-
"step": 8350
|
| 1444 |
-
},
|
| 1445 |
-
{
|
| 1446 |
-
"epoch": 0.6461041458349358,
|
| 1447 |
-
"grad_norm": 1.963592290878296,
|
| 1448 |
-
"learning_rate": 0.0001590005973560502,
|
| 1449 |
-
"loss": 1.9534,
|
| 1450 |
-
"step": 8400
|
| 1451 |
-
},
|
| 1452 |
-
{
|
| 1453 |
-
"epoch": 0.649950003845858,
|
| 1454 |
-
"grad_norm": 1.5136828422546387,
|
| 1455 |
-
"learning_rate": 0.0001587408773342337,
|
| 1456 |
-
"loss": 1.9946,
|
| 1457 |
-
"step": 8450
|
| 1458 |
-
},
|
| 1459 |
-
{
|
| 1460 |
-
"epoch": 0.6537958618567803,
|
| 1461 |
-
"grad_norm": 2.382208824157715,
|
| 1462 |
-
"learning_rate": 0.00015848115731241722,
|
| 1463 |
-
"loss": 2.1216,
|
| 1464 |
-
"step": 8500
|
| 1465 |
-
},
|
| 1466 |
-
{
|
| 1467 |
-
"epoch": 0.6537958618567803,
|
| 1468 |
-
"eval_loss": 1.9431382417678833,
|
| 1469 |
-
"eval_runtime": 17.244,
|
| 1470 |
-
"eval_samples_per_second": 57.991,
|
| 1471 |
-
"eval_steps_per_second": 14.498,
|
| 1472 |
-
"step": 8500
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.6576417198677025,
|
| 1476 |
-
"grad_norm": 1.2141226530075073,
|
| 1477 |
-
"learning_rate": 0.00015822143729060076,
|
| 1478 |
-
"loss": 1.9762,
|
| 1479 |
-
"step": 8550
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.6614875778786247,
|
| 1483 |
-
"grad_norm": 2.0325729846954346,
|
| 1484 |
-
"learning_rate": 0.00015796171726878427,
|
| 1485 |
-
"loss": 1.9871,
|
| 1486 |
-
"step": 8600
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 0.665333435889547,
|
| 1490 |
-
"grad_norm": 1.6352391242980957,
|
| 1491 |
-
"learning_rate": 0.00015770199724696776,
|
| 1492 |
-
"loss": 1.9289,
|
| 1493 |
-
"step": 8650
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.6691792939004692,
|
| 1497 |
-
"grad_norm": 1.546341896057129,
|
| 1498 |
-
"learning_rate": 0.0001574422772251513,
|
| 1499 |
-
"loss": 1.9499,
|
| 1500 |
-
"step": 8700
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.6730251519113915,
|
| 1504 |
-
"grad_norm": 2.183812141418457,
|
| 1505 |
-
"learning_rate": 0.0001571825572033348,
|
| 1506 |
-
"loss": 1.9727,
|
| 1507 |
-
"step": 8750
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.6730251519113915,
|
| 1511 |
-
"eval_loss": 1.9287711381912231,
|
| 1512 |
-
"eval_runtime": 17.0409,
|
| 1513 |
-
"eval_samples_per_second": 58.682,
|
| 1514 |
-
"eval_steps_per_second": 14.671,
|
| 1515 |
-
"step": 8750
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 0.6768710099223136,
|
| 1519 |
-
"grad_norm": 2.3215548992156982,
|
| 1520 |
-
"learning_rate": 0.00015692283718151832,
|
| 1521 |
-
"loss": 2.0059,
|
| 1522 |
-
"step": 8800
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 0.6807168679332359,
|
| 1526 |
-
"grad_norm": 1.9137969017028809,
|
| 1527 |
-
"learning_rate": 0.00015666311715970186,
|
| 1528 |
-
"loss": 1.9883,
|
| 1529 |
-
"step": 8850
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 0.6845627259441581,
|
| 1533 |
-
"grad_norm": 2.4092469215393066,
|
| 1534 |
-
"learning_rate": 0.00015640339713788538,
|
| 1535 |
-
"loss": 1.9236,
|
| 1536 |
-
"step": 8900
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 0.6884085839550804,
|
| 1540 |
-
"grad_norm": 2.448526620864868,
|
| 1541 |
-
"learning_rate": 0.0001561436771160689,
|
| 1542 |
-
"loss": 1.9178,
|
| 1543 |
-
"step": 8950
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 0.6922544419660026,
|
| 1547 |
-
"grad_norm": 1.0038529634475708,
|
| 1548 |
-
"learning_rate": 0.0001558839570942524,
|
| 1549 |
-
"loss": 1.8992,
|
| 1550 |
-
"step": 9000
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 0.6922544419660026,
|
| 1554 |
-
"eval_loss": 1.902275800704956,
|
| 1555 |
-
"eval_runtime": 16.9573,
|
| 1556 |
-
"eval_samples_per_second": 58.972,
|
| 1557 |
-
"eval_steps_per_second": 14.743,
|
| 1558 |
-
"step": 9000
|
| 1559 |
-
},
|
| 1560 |
-
{
|
| 1561 |
-
"epoch": 0.6961002999769248,
|
| 1562 |
-
"grad_norm": 3.6892011165618896,
|
| 1563 |
-
"learning_rate": 0.00015562423707243591,
|
| 1564 |
-
"loss": 1.9551,
|
| 1565 |
-
"step": 9050
|
| 1566 |
-
},
|
| 1567 |
-
{
|
| 1568 |
-
"epoch": 0.6999461579878471,
|
| 1569 |
-
"grad_norm": 1.524671196937561,
|
| 1570 |
-
"learning_rate": 0.00015536451705061943,
|
| 1571 |
-
"loss": 1.9109,
|
| 1572 |
-
"step": 9100
|
| 1573 |
-
},
|
| 1574 |
-
{
|
| 1575 |
-
"epoch": 0.7037920159987693,
|
| 1576 |
-
"grad_norm": 1.5293575525283813,
|
| 1577 |
-
"learning_rate": 0.00015510479702880297,
|
| 1578 |
-
"loss": 1.8891,
|
| 1579 |
-
"step": 9150
|
| 1580 |
-
},
|
| 1581 |
-
{
|
| 1582 |
-
"epoch": 0.7076378740096916,
|
| 1583 |
-
"grad_norm": 2.321476697921753,
|
| 1584 |
-
"learning_rate": 0.00015484507700698648,
|
| 1585 |
-
"loss": 1.9021,
|
| 1586 |
-
"step": 9200
|
| 1587 |
-
},
|
| 1588 |
-
{
|
| 1589 |
-
"epoch": 0.7114837320206138,
|
| 1590 |
-
"grad_norm": 2.710942029953003,
|
| 1591 |
-
"learning_rate": 0.00015458535698517,
|
| 1592 |
-
"loss": 1.9414,
|
| 1593 |
-
"step": 9250
|
| 1594 |
-
},
|
| 1595 |
-
{
|
| 1596 |
-
"epoch": 0.7114837320206138,
|
| 1597 |
-
"eval_loss": 1.9107917547225952,
|
| 1598 |
-
"eval_runtime": 16.997,
|
| 1599 |
-
"eval_samples_per_second": 58.834,
|
| 1600 |
-
"eval_steps_per_second": 14.708,
|
| 1601 |
-
"step": 9250
|
| 1602 |
-
},
|
| 1603 |
-
{
|
| 1604 |
-
"epoch": 0.7153295900315361,
|
| 1605 |
-
"grad_norm": 1.9385954141616821,
|
| 1606 |
-
"learning_rate": 0.0001543256369633535,
|
| 1607 |
-
"loss": 1.9401,
|
| 1608 |
-
"step": 9300
|
| 1609 |
-
},
|
| 1610 |
-
{
|
| 1611 |
-
"epoch": 0.7191754480424583,
|
| 1612 |
-
"grad_norm": 2.589629650115967,
|
| 1613 |
-
"learning_rate": 0.00015406591694153702,
|
| 1614 |
-
"loss": 1.9054,
|
| 1615 |
-
"step": 9350
|
| 1616 |
-
},
|
| 1617 |
-
{
|
| 1618 |
-
"epoch": 0.7230213060533806,
|
| 1619 |
-
"grad_norm": 1.6431207656860352,
|
| 1620 |
-
"learning_rate": 0.00015380619691972056,
|
| 1621 |
-
"loss": 1.9324,
|
| 1622 |
-
"step": 9400
|
| 1623 |
-
},
|
| 1624 |
-
{
|
| 1625 |
-
"epoch": 0.7268671640643027,
|
| 1626 |
-
"grad_norm": 4.840892314910889,
|
| 1627 |
-
"learning_rate": 0.00015354647689790407,
|
| 1628 |
-
"loss": 1.993,
|
| 1629 |
-
"step": 9450
|
| 1630 |
-
},
|
| 1631 |
-
{
|
| 1632 |
-
"epoch": 0.7307130220752249,
|
| 1633 |
-
"grad_norm": 0.9328492879867554,
|
| 1634 |
-
"learning_rate": 0.00015328675687608758,
|
| 1635 |
-
"loss": 1.8637,
|
| 1636 |
-
"step": 9500
|
| 1637 |
-
},
|
| 1638 |
-
{
|
| 1639 |
-
"epoch": 0.7307130220752249,
|
| 1640 |
-
"eval_loss": 1.8725571632385254,
|
| 1641 |
-
"eval_runtime": 16.9918,
|
| 1642 |
-
"eval_samples_per_second": 58.852,
|
| 1643 |
-
"eval_steps_per_second": 14.713,
|
| 1644 |
-
"step": 9500
|
| 1645 |
-
},
|
| 1646 |
-
{
|
| 1647 |
-
"epoch": 0.7345588800861472,
|
| 1648 |
-
"grad_norm": 1.6844549179077148,
|
| 1649 |
-
"learning_rate": 0.00015302703685427112,
|
| 1650 |
-
"loss": 1.8262,
|
| 1651 |
-
"step": 9550
|
| 1652 |
-
},
|
| 1653 |
-
{
|
| 1654 |
-
"epoch": 0.7384047380970694,
|
| 1655 |
-
"grad_norm": 2.512157678604126,
|
| 1656 |
-
"learning_rate": 0.0001527673168324546,
|
| 1657 |
-
"loss": 1.9319,
|
| 1658 |
-
"step": 9600
|
| 1659 |
-
},
|
| 1660 |
-
{
|
| 1661 |
-
"epoch": 0.7422505961079917,
|
| 1662 |
-
"grad_norm": 1.3238016366958618,
|
| 1663 |
-
"learning_rate": 0.00015250759681063812,
|
| 1664 |
-
"loss": 1.9001,
|
| 1665 |
-
"step": 9650
|
| 1666 |
-
},
|
| 1667 |
-
{
|
| 1668 |
-
"epoch": 0.7460964541189139,
|
| 1669 |
-
"grad_norm": 1.83181631565094,
|
| 1670 |
-
"learning_rate": 0.00015224787678882166,
|
| 1671 |
-
"loss": 1.8324,
|
| 1672 |
-
"step": 9700
|
| 1673 |
-
},
|
| 1674 |
-
{
|
| 1675 |
-
"epoch": 0.7499423121298362,
|
| 1676 |
-
"grad_norm": 1.6106966733932495,
|
| 1677 |
-
"learning_rate": 0.00015198815676700518,
|
| 1678 |
-
"loss": 1.9535,
|
| 1679 |
-
"step": 9750
|
| 1680 |
-
},
|
| 1681 |
-
{
|
| 1682 |
-
"epoch": 0.7499423121298362,
|
| 1683 |
-
"eval_loss": 1.873831033706665,
|
| 1684 |
-
"eval_runtime": 17.0343,
|
| 1685 |
-
"eval_samples_per_second": 58.705,
|
| 1686 |
-
"eval_steps_per_second": 14.676,
|
| 1687 |
-
"step": 9750
|
| 1688 |
-
},
|
| 1689 |
-
{
|
| 1690 |
-
"epoch": 0.7537881701407584,
|
| 1691 |
-
"grad_norm": 2.3586697578430176,
|
| 1692 |
-
"learning_rate": 0.0001517284367451887,
|
| 1693 |
-
"loss": 1.858,
|
| 1694 |
-
"step": 9800
|
| 1695 |
-
},
|
| 1696 |
-
{
|
| 1697 |
-
"epoch": 0.7576340281516807,
|
| 1698 |
-
"grad_norm": 0.7499716877937317,
|
| 1699 |
-
"learning_rate": 0.00015146871672337223,
|
| 1700 |
-
"loss": 1.866,
|
| 1701 |
-
"step": 9850
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 0.7614798861626029,
|
| 1705 |
-
"grad_norm": 1.3573709726333618,
|
| 1706 |
-
"learning_rate": 0.00015120899670155574,
|
| 1707 |
-
"loss": 1.8636,
|
| 1708 |
-
"step": 9900
|
| 1709 |
-
},
|
| 1710 |
-
{
|
| 1711 |
-
"epoch": 0.7653257441735252,
|
| 1712 |
-
"grad_norm": 2.271859884262085,
|
| 1713 |
-
"learning_rate": 0.00015094927667973923,
|
| 1714 |
-
"loss": 1.8625,
|
| 1715 |
-
"step": 9950
|
| 1716 |
-
},
|
| 1717 |
-
{
|
| 1718 |
-
"epoch": 0.7691716021844474,
|
| 1719 |
-
"grad_norm": 1.8813310861587524,
|
| 1720 |
-
"learning_rate": 0.00015068955665792277,
|
| 1721 |
-
"loss": 1.9318,
|
| 1722 |
-
"step": 10000
|
| 1723 |
-
},
|
| 1724 |
-
{
|
| 1725 |
-
"epoch": 0.7691716021844474,
|
| 1726 |
-
"eval_loss": 1.8548085689544678,
|
| 1727 |
-
"eval_runtime": 17.0223,
|
| 1728 |
-
"eval_samples_per_second": 58.746,
|
| 1729 |
-
"eval_steps_per_second": 14.687,
|
| 1730 |
-
"step": 10000
|
| 1731 |
-
},
|
| 1732 |
-
{
|
| 1733 |
-
"epoch": 0.7730174601953695,
|
| 1734 |
-
"grad_norm": 1.4667410850524902,
|
| 1735 |
-
"learning_rate": 0.00015042983663610628,
|
| 1736 |
-
"loss": 1.9318,
|
| 1737 |
-
"step": 10050
|
| 1738 |
-
},
|
| 1739 |
-
{
|
| 1740 |
-
"epoch": 0.7768633182062918,
|
| 1741 |
-
"grad_norm": 2.0754499435424805,
|
| 1742 |
-
"learning_rate": 0.00015017011661428982,
|
| 1743 |
-
"loss": 1.753,
|
| 1744 |
-
"step": 10100
|
| 1745 |
-
},
|
| 1746 |
-
{
|
| 1747 |
-
"epoch": 0.780709176217214,
|
| 1748 |
-
"grad_norm": 1.634293794631958,
|
| 1749 |
-
"learning_rate": 0.00014991039659247333,
|
| 1750 |
-
"loss": 1.9328,
|
| 1751 |
-
"step": 10150
|
| 1752 |
-
},
|
| 1753 |
-
{
|
| 1754 |
-
"epoch": 0.7845550342281363,
|
| 1755 |
-
"grad_norm": 1.5003916025161743,
|
| 1756 |
-
"learning_rate": 0.00014965067657065684,
|
| 1757 |
-
"loss": 1.8237,
|
| 1758 |
-
"step": 10200
|
| 1759 |
-
},
|
| 1760 |
-
{
|
| 1761 |
-
"epoch": 0.7884008922390585,
|
| 1762 |
-
"grad_norm": 1.7433470487594604,
|
| 1763 |
-
"learning_rate": 0.00014939095654884036,
|
| 1764 |
-
"loss": 1.859,
|
| 1765 |
-
"step": 10250
|
| 1766 |
-
},
|
| 1767 |
-
{
|
| 1768 |
-
"epoch": 0.7884008922390585,
|
| 1769 |
-
"eval_loss": 1.842873215675354,
|
| 1770 |
-
"eval_runtime": 16.9242,
|
| 1771 |
-
"eval_samples_per_second": 59.087,
|
| 1772 |
-
"eval_steps_per_second": 14.772,
|
| 1773 |
-
"step": 10250
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 0.7922467502499808,
|
| 1777 |
-
"grad_norm": 1.8703033924102783,
|
| 1778 |
-
"learning_rate": 0.00014913123652702387,
|
| 1779 |
-
"loss": 1.9019,
|
| 1780 |
-
"step": 10300
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 0.796092608260903,
|
| 1784 |
-
"grad_norm": 1.1216288805007935,
|
| 1785 |
-
"learning_rate": 0.00014887151650520738,
|
| 1786 |
-
"loss": 1.9812,
|
| 1787 |
-
"step": 10350
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 0.7999384662718253,
|
| 1791 |
-
"grad_norm": 3.2229816913604736,
|
| 1792 |
-
"learning_rate": 0.00014861179648339092,
|
| 1793 |
-
"loss": 1.7986,
|
| 1794 |
-
"step": 10400
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 0.8037843242827475,
|
| 1798 |
-
"grad_norm": 2.366506814956665,
|
| 1799 |
-
"learning_rate": 0.00014835207646157444,
|
| 1800 |
-
"loss": 1.8995,
|
| 1801 |
-
"step": 10450
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 0.8076301822936697,
|
| 1805 |
-
"grad_norm": 2.0333054065704346,
|
| 1806 |
-
"learning_rate": 0.00014809235643975795,
|
| 1807 |
-
"loss": 1.916,
|
| 1808 |
-
"step": 10500
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 0.8076301822936697,
|
| 1812 |
-
"eval_loss": 1.8220328092575073,
|
| 1813 |
-
"eval_runtime": 17.063,
|
| 1814 |
-
"eval_samples_per_second": 58.606,
|
| 1815 |
-
"eval_steps_per_second": 14.652,
|
| 1816 |
-
"step": 10500
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 0.811476040304592,
|
| 1820 |
-
"grad_norm": 1.8630917072296143,
|
| 1821 |
-
"learning_rate": 0.00014783263641794146,
|
| 1822 |
-
"loss": 1.9692,
|
| 1823 |
-
"step": 10550
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 0.8153218983155142,
|
| 1827 |
-
"grad_norm": 1.744325876235962,
|
| 1828 |
-
"learning_rate": 0.00014757291639612497,
|
| 1829 |
-
"loss": 1.8893,
|
| 1830 |
-
"step": 10600
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 0.8191677563264365,
|
| 1834 |
-
"grad_norm": 1.8451564311981201,
|
| 1835 |
-
"learning_rate": 0.0001473131963743085,
|
| 1836 |
-
"loss": 1.7862,
|
| 1837 |
-
"step": 10650
|
| 1838 |
-
},
|
| 1839 |
-
{
|
| 1840 |
-
"epoch": 0.8230136143373586,
|
| 1841 |
-
"grad_norm": 1.9895585775375366,
|
| 1842 |
-
"learning_rate": 0.00014705347635249203,
|
| 1843 |
-
"loss": 1.9978,
|
| 1844 |
-
"step": 10700
|
| 1845 |
-
},
|
| 1846 |
-
{
|
| 1847 |
-
"epoch": 0.8268594723482809,
|
| 1848 |
-
"grad_norm": 2.574155330657959,
|
| 1849 |
-
"learning_rate": 0.00014679375633067554,
|
| 1850 |
-
"loss": 1.8705,
|
| 1851 |
-
"step": 10750
|
| 1852 |
-
},
|
| 1853 |
-
{
|
| 1854 |
-
"epoch": 0.8268594723482809,
|
| 1855 |
-
"eval_loss": 1.8234485387802124,
|
| 1856 |
-
"eval_runtime": 17.061,
|
| 1857 |
-
"eval_samples_per_second": 58.613,
|
| 1858 |
-
"eval_steps_per_second": 14.653,
|
| 1859 |
-
"step": 10750
|
| 1860 |
-
},
|
| 1861 |
-
{
|
| 1862 |
-
"epoch": 0.8307053303592031,
|
| 1863 |
-
"grad_norm": 1.347955584526062,
|
| 1864 |
-
"learning_rate": 0.00014653403630885905,
|
| 1865 |
-
"loss": 1.7948,
|
| 1866 |
-
"step": 10800
|
| 1867 |
-
},
|
| 1868 |
-
{
|
| 1869 |
-
"epoch": 0.8345511883701254,
|
| 1870 |
-
"grad_norm": 2.8867287635803223,
|
| 1871 |
-
"learning_rate": 0.00014627431628704257,
|
| 1872 |
-
"loss": 1.8986,
|
| 1873 |
-
"step": 10850
|
| 1874 |
-
},
|
| 1875 |
-
{
|
| 1876 |
-
"epoch": 0.8383970463810476,
|
| 1877 |
-
"grad_norm": 2.692473888397217,
|
| 1878 |
-
"learning_rate": 0.00014601459626522608,
|
| 1879 |
-
"loss": 1.9209,
|
| 1880 |
-
"step": 10900
|
| 1881 |
-
},
|
| 1882 |
-
{
|
| 1883 |
-
"epoch": 0.8422429043919698,
|
| 1884 |
-
"grad_norm": 1.7835667133331299,
|
| 1885 |
-
"learning_rate": 0.00014575487624340962,
|
| 1886 |
-
"loss": 1.8406,
|
| 1887 |
-
"step": 10950
|
| 1888 |
-
},
|
| 1889 |
-
{
|
| 1890 |
-
"epoch": 0.8460887624028921,
|
| 1891 |
-
"grad_norm": 1.8864330053329468,
|
| 1892 |
-
"learning_rate": 0.00014549515622159313,
|
| 1893 |
-
"loss": 1.8001,
|
| 1894 |
-
"step": 11000
|
| 1895 |
-
},
|
| 1896 |
-
{
|
| 1897 |
-
"epoch": 0.8460887624028921,
|
| 1898 |
-
"eval_loss": 1.807216763496399,
|
| 1899 |
-
"eval_runtime": 17.2188,
|
| 1900 |
-
"eval_samples_per_second": 58.076,
|
| 1901 |
-
"eval_steps_per_second": 14.519,
|
| 1902 |
-
"step": 11000
|
| 1903 |
-
},
|
| 1904 |
-
{
|
| 1905 |
-
"epoch": 0.8499346204138143,
|
| 1906 |
-
"grad_norm": 2.47654128074646,
|
| 1907 |
-
"learning_rate": 0.00014523543619977664,
|
| 1908 |
-
"loss": 1.8003,
|
| 1909 |
-
"step": 11050
|
| 1910 |
-
},
|
| 1911 |
-
{
|
| 1912 |
-
"epoch": 0.8537804784247366,
|
| 1913 |
-
"grad_norm": 1.6507407426834106,
|
| 1914 |
-
"learning_rate": 0.00014497571617796018,
|
| 1915 |
-
"loss": 1.8999,
|
| 1916 |
-
"step": 11100
|
| 1917 |
-
},
|
| 1918 |
-
{
|
| 1919 |
-
"epoch": 0.8576263364356588,
|
| 1920 |
-
"grad_norm": 1.4758163690567017,
|
| 1921 |
-
"learning_rate": 0.0001447159961561437,
|
| 1922 |
-
"loss": 1.837,
|
| 1923 |
-
"step": 11150
|
| 1924 |
-
},
|
| 1925 |
-
{
|
| 1926 |
-
"epoch": 0.8614721944465811,
|
| 1927 |
-
"grad_norm": 2.2486917972564697,
|
| 1928 |
-
"learning_rate": 0.00014445627613432718,
|
| 1929 |
-
"loss": 1.8501,
|
| 1930 |
-
"step": 11200
|
| 1931 |
-
},
|
| 1932 |
-
{
|
| 1933 |
-
"epoch": 0.8653180524575033,
|
| 1934 |
-
"grad_norm": 1.682785987854004,
|
| 1935 |
-
"learning_rate": 0.00014419655611251072,
|
| 1936 |
-
"loss": 1.8486,
|
| 1937 |
-
"step": 11250
|
| 1938 |
-
},
|
| 1939 |
-
{
|
| 1940 |
-
"epoch": 0.8653180524575033,
|
| 1941 |
-
"eval_loss": 1.8015695810317993,
|
| 1942 |
-
"eval_runtime": 16.9649,
|
| 1943 |
-
"eval_samples_per_second": 58.945,
|
| 1944 |
-
"eval_steps_per_second": 14.736,
|
| 1945 |
-
"step": 11250
|
| 1946 |
-
},
|
| 1947 |
-
{
|
| 1948 |
-
"epoch": 0.8691639104684256,
|
| 1949 |
-
"grad_norm": 1.687892198562622,
|
| 1950 |
-
"learning_rate": 0.00014393683609069424,
|
| 1951 |
-
"loss": 1.8196,
|
| 1952 |
-
"step": 11300
|
| 1953 |
-
},
|
| 1954 |
-
{
|
| 1955 |
-
"epoch": 0.8730097684793477,
|
| 1956 |
-
"grad_norm": 1.6149276494979858,
|
| 1957 |
-
"learning_rate": 0.00014367711606887775,
|
| 1958 |
-
"loss": 1.829,
|
| 1959 |
-
"step": 11350
|
| 1960 |
-
},
|
| 1961 |
-
{
|
| 1962 |
-
"epoch": 0.8768556264902699,
|
| 1963 |
-
"grad_norm": 1.2599520683288574,
|
| 1964 |
-
"learning_rate": 0.0001434173960470613,
|
| 1965 |
-
"loss": 1.8398,
|
| 1966 |
-
"step": 11400
|
| 1967 |
-
},
|
| 1968 |
-
{
|
| 1969 |
-
"epoch": 0.8807014845011922,
|
| 1970 |
-
"grad_norm": 2.5826971530914307,
|
| 1971 |
-
"learning_rate": 0.0001431576760252448,
|
| 1972 |
-
"loss": 1.8362,
|
| 1973 |
-
"step": 11450
|
| 1974 |
-
},
|
| 1975 |
-
{
|
| 1976 |
-
"epoch": 0.8845473425121144,
|
| 1977 |
-
"grad_norm": 1.9814509153366089,
|
| 1978 |
-
"learning_rate": 0.00014289795600342831,
|
| 1979 |
-
"loss": 1.7922,
|
| 1980 |
-
"step": 11500
|
| 1981 |
-
},
|
| 1982 |
-
{
|
| 1983 |
-
"epoch": 0.8845473425121144,
|
| 1984 |
-
"eval_loss": 1.792478322982788,
|
| 1985 |
-
"eval_runtime": 16.8513,
|
| 1986 |
-
"eval_samples_per_second": 59.343,
|
| 1987 |
-
"eval_steps_per_second": 14.836,
|
| 1988 |
-
"step": 11500
|
| 1989 |
}
|
| 1990 |
],
|
| 1991 |
"logging_steps": 50,
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 500,
|
| 3 |
+
"best_metric": 3.3119897842407227,
|
| 4 |
+
"best_model_checkpoint": "./ar-diffusion-checkpoints-progressive-attention/checkpoint-500",
|
| 5 |
+
"epoch": 0.038458580109222366,
|
| 6 |
"eval_steps": 250,
|
| 7 |
+
"global_step": 500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.003845858010922237,
|
| 14 |
+
"grad_norm": 13.142335891723633,
|
| 15 |
"learning_rate": 1.84e-05,
|
| 16 |
+
"loss": 12.0117,
|
| 17 |
"step": 50
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.007691716021844474,
|
| 21 |
+
"grad_norm": 5.153238296508789,
|
| 22 |
"learning_rate": 3.8e-05,
|
| 23 |
+
"loss": 5.8408,
|
| 24 |
"step": 100
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.01153757403276671,
|
| 28 |
+
"grad_norm": 7.471889019012451,
|
| 29 |
"learning_rate": 5.8e-05,
|
| 30 |
+
"loss": 4.5173,
|
| 31 |
"step": 150
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.015383432043688947,
|
| 35 |
+
"grad_norm": 4.756172180175781,
|
| 36 |
"learning_rate": 7.800000000000001e-05,
|
| 37 |
+
"loss": 3.8946,
|
| 38 |
"step": 200
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.019229290054611183,
|
| 42 |
+
"grad_norm": 4.0329108238220215,
|
| 43 |
"learning_rate": 9.8e-05,
|
| 44 |
+
"loss": 3.842,
|
| 45 |
"step": 250
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"epoch": 0.019229290054611183,
|
| 49 |
+
"eval_loss": 3.7754223346710205,
|
| 50 |
+
"eval_runtime": 18.0149,
|
| 51 |
+
"eval_samples_per_second": 55.51,
|
| 52 |
+
"eval_steps_per_second": 13.877,
|
| 53 |
"step": 250
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.02307514806553342,
|
| 57 |
+
"grad_norm": 7.410353660583496,
|
| 58 |
"learning_rate": 0.000118,
|
| 59 |
+
"loss": 3.5501,
|
| 60 |
"step": 300
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.02692100607645566,
|
| 64 |
+
"grad_norm": 4.085294246673584,
|
| 65 |
"learning_rate": 0.000138,
|
| 66 |
+
"loss": 3.4015,
|
| 67 |
"step": 350
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 0.030766864087377895,
|
| 71 |
+
"grad_norm": 5.78131103515625,
|
| 72 |
"learning_rate": 0.00015800000000000002,
|
| 73 |
+
"loss": 3.4988,
|
| 74 |
"step": 400
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.03461272209830013,
|
| 78 |
+
"grad_norm": 3.2009644508361816,
|
| 79 |
"learning_rate": 0.00017800000000000002,
|
| 80 |
+
"loss": 3.4145,
|
| 81 |
"step": 450
|
| 82 |
},
|
| 83 |
{
|
| 84 |
"epoch": 0.038458580109222366,
|
| 85 |
+
"grad_norm": 4.380056858062744,
|
| 86 |
"learning_rate": 0.00019800000000000002,
|
| 87 |
+
"loss": 3.3179,
|
| 88 |
"step": 500
|
| 89 |
},
|
| 90 |
{
|
| 91 |
"epoch": 0.038458580109222366,
|
| 92 |
+
"eval_loss": 3.3119897842407227,
|
| 93 |
+
"eval_runtime": 17.9539,
|
| 94 |
+
"eval_samples_per_second": 55.698,
|
| 95 |
+
"eval_steps_per_second": 13.925,
|
| 96 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 97 |
}
|
| 98 |
],
|
| 99 |
"logging_steps": 50,
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5432
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e44af272c49f1414968320d715761884d929c74fc3d44815afabe4a9422f97fb
|
| 3 |
size 5432
|