Training in progress, step 150, checkpoint
Browse files
last-checkpoint/adapter_config.json
CHANGED
|
@@ -20,10 +20,10 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"
|
| 24 |
"o_proj",
|
| 25 |
-
"
|
| 26 |
-
"
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
+
"v_proj",
|
| 24 |
"o_proj",
|
| 25 |
+
"k_proj",
|
| 26 |
+
"q_proj"
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
| 29 |
"use_dora": false,
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6832520
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d22559dc9f3539c34081db29203f0c0625e53a718119282ae5e9cbbbbfea04ad
|
| 3 |
size 6832520
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13739450
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:065d4ede74e3982d45eabd75bcf3791c1732814b3c2687bdef88b71c498b41db
|
| 3 |
size 13739450
|
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:ba79f79f4200644bcde298b8ba358af98910b10cc152e720addca023a2e47a37
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1256
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6186f1e9836e20fd270e1ab773c83f1027d92e426fd1d0a8c7816f8a9115c5fd
|
| 3 |
size 1256
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1594 +1,125 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "./output/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 150,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01277139208173691,
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
"learning_rate": 4.666666666666666e-06,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.02554278416347382,
|
| 20 |
-
"grad_norm": 2.
|
| 21 |
"learning_rate": 9.333333333333333e-06,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.038314176245210725,
|
| 27 |
-
"grad_norm": 2.
|
| 28 |
"learning_rate": 1.4e-05,
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.05108556832694764,
|
| 34 |
-
"grad_norm": 1.
|
| 35 |
"learning_rate": 1.8666666666666665e-05,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.06385696040868455,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
"learning_rate": 2.333333333333333e-05,
|
| 43 |
"loss": 1.9779,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.07662835249042145,
|
| 48 |
-
"grad_norm": 2.
|
| 49 |
"learning_rate": 2.8e-05,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.08939974457215837,
|
| 55 |
-
"grad_norm": 1.
|
| 56 |
"learning_rate": 3.266666666666666e-05,
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.10217113665389528,
|
| 62 |
-
"grad_norm": 2.
|
| 63 |
"learning_rate": 3.733333333333333e-05,
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.11494252873563218,
|
| 69 |
-
"grad_norm": 2.
|
| 70 |
"learning_rate": 4.2e-05,
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.1277139208173691,
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
"learning_rate": 4.666666666666666e-05,
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.140485312899106,
|
| 83 |
-
"grad_norm": 1.
|
| 84 |
"learning_rate": 5.1333333333333325e-05,
|
| 85 |
"loss": 1.9233,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.1532567049808429,
|
| 90 |
-
"grad_norm": 2.
|
| 91 |
"learning_rate": 5.6e-05,
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.16602809706257982,
|
| 97 |
-
"grad_norm": 2.
|
| 98 |
"learning_rate": 6.0666666666666666e-05,
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.17879948914431673,
|
| 104 |
-
"grad_norm": 2.
|
| 105 |
"learning_rate": 6.533333333333333e-05,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.19157088122605365,
|
| 111 |
-
"grad_norm": 2.
|
| 112 |
"learning_rate": 7e-05,
|
| 113 |
"loss": 1.9405,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.19157088122605365,
|
| 118 |
-
"eval_loss": 1.
|
| 119 |
-
"eval_runtime":
|
| 120 |
-
"eval_samples_per_second":
|
| 121 |
-
"eval_steps_per_second":
|
| 122 |
"step": 150
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.20434227330779056,
|
| 126 |
-
"grad_norm": 2.221877336502075,
|
| 127 |
-
"learning_rate": 6.999926573504895e-05,
|
| 128 |
-
"loss": 1.881,
|
| 129 |
-
"step": 160
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.21711366538952745,
|
| 133 |
-
"grad_norm": 2.2773871421813965,
|
| 134 |
-
"learning_rate": 6.999706297100412e-05,
|
| 135 |
-
"loss": 1.9073,
|
| 136 |
-
"step": 170
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.22988505747126436,
|
| 140 |
-
"grad_norm": 2.2678256034851074,
|
| 141 |
-
"learning_rate": 6.999339180028904e-05,
|
| 142 |
-
"loss": 1.8844,
|
| 143 |
-
"step": 180
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.24265644955300128,
|
| 147 |
-
"grad_norm": 2.0379583835601807,
|
| 148 |
-
"learning_rate": 6.99882523769387e-05,
|
| 149 |
-
"loss": 1.9496,
|
| 150 |
-
"step": 190
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.2554278416347382,
|
| 154 |
-
"grad_norm": 2.1946399211883545,
|
| 155 |
-
"learning_rate": 6.998164491659302e-05,
|
| 156 |
-
"loss": 1.9193,
|
| 157 |
-
"step": 200
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.2681992337164751,
|
| 161 |
-
"grad_norm": 2.218657970428467,
|
| 162 |
-
"learning_rate": 6.997356969648778e-05,
|
| 163 |
-
"loss": 1.834,
|
| 164 |
-
"step": 210
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.280970625798212,
|
| 168 |
-
"grad_norm": 2.1720638275146484,
|
| 169 |
-
"learning_rate": 6.996402705544307e-05,
|
| 170 |
-
"loss": 1.9135,
|
| 171 |
-
"step": 220
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.2937420178799489,
|
| 175 |
-
"grad_norm": 2.029083251953125,
|
| 176 |
-
"learning_rate": 6.995301739384896e-05,
|
| 177 |
-
"loss": 1.9661,
|
| 178 |
-
"step": 230
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.3065134099616858,
|
| 182 |
-
"grad_norm": 1.9260584115982056,
|
| 183 |
-
"learning_rate": 6.994054117364885e-05,
|
| 184 |
-
"loss": 1.9046,
|
| 185 |
-
"step": 240
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.31928480204342274,
|
| 189 |
-
"grad_norm": 2.0648281574249268,
|
| 190 |
-
"learning_rate": 6.992659891831991e-05,
|
| 191 |
-
"loss": 1.9672,
|
| 192 |
-
"step": 250
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.33205619412515963,
|
| 196 |
-
"grad_norm": 1.964780330657959,
|
| 197 |
-
"learning_rate": 6.991119121285126e-05,
|
| 198 |
-
"loss": 1.9125,
|
| 199 |
-
"step": 260
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.3448275862068966,
|
| 203 |
-
"grad_norm": 2.132140636444092,
|
| 204 |
-
"learning_rate": 6.989431870371936e-05,
|
| 205 |
-
"loss": 1.9624,
|
| 206 |
-
"step": 270
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.35759897828863346,
|
| 210 |
-
"grad_norm": 2.0243074893951416,
|
| 211 |
-
"learning_rate": 6.987598209886091e-05,
|
| 212 |
-
"loss": 1.9409,
|
| 213 |
-
"step": 280
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.37037037037037035,
|
| 217 |
-
"grad_norm": 1.9534668922424316,
|
| 218 |
-
"learning_rate": 6.985618216764314e-05,
|
| 219 |
-
"loss": 1.9589,
|
| 220 |
-
"step": 290
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.3831417624521073,
|
| 224 |
-
"grad_norm": 1.9132697582244873,
|
| 225 |
-
"learning_rate": 6.983491974083148e-05,
|
| 226 |
-
"loss": 1.9467,
|
| 227 |
-
"step": 300
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.3831417624521073,
|
| 231 |
-
"eval_loss": 1.962647795677185,
|
| 232 |
-
"eval_runtime": 24.5626,
|
| 233 |
-
"eval_samples_per_second": 20.356,
|
| 234 |
-
"eval_steps_per_second": 20.356,
|
| 235 |
-
"step": 300
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.3959131545338442,
|
| 239 |
-
"grad_norm": 2.107532262802124,
|
| 240 |
-
"learning_rate": 6.981219571055479e-05,
|
| 241 |
-
"loss": 1.8887,
|
| 242 |
-
"step": 310
|
| 243 |
-
},
|
| 244 |
-
{
|
| 245 |
-
"epoch": 0.4086845466155811,
|
| 246 |
-
"grad_norm": 1.9190975427627563,
|
| 247 |
-
"learning_rate": 6.978801103026786e-05,
|
| 248 |
-
"loss": 1.9321,
|
| 249 |
-
"step": 320
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.421455938697318,
|
| 253 |
-
"grad_norm": 2.313232421875,
|
| 254 |
-
"learning_rate": 6.976236671471145e-05,
|
| 255 |
-
"loss": 1.9673,
|
| 256 |
-
"step": 330
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 0.4342273307790549,
|
| 260 |
-
"grad_norm": 2.028118848800659,
|
| 261 |
-
"learning_rate": 6.973526383986968e-05,
|
| 262 |
-
"loss": 1.8951,
|
| 263 |
-
"step": 340
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 0.44699872286079184,
|
| 267 |
-
"grad_norm": 1.9303714036941528,
|
| 268 |
-
"learning_rate": 6.970670354292487e-05,
|
| 269 |
-
"loss": 1.888,
|
| 270 |
-
"step": 350
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"epoch": 0.45977011494252873,
|
| 274 |
-
"grad_norm": 1.9469075202941895,
|
| 275 |
-
"learning_rate": 6.967668702220992e-05,
|
| 276 |
-
"loss": 1.9534,
|
| 277 |
-
"step": 360
|
| 278 |
-
},
|
| 279 |
-
{
|
| 280 |
-
"epoch": 0.4725415070242657,
|
| 281 |
-
"grad_norm": 1.9907073974609375,
|
| 282 |
-
"learning_rate": 6.964521553715788e-05,
|
| 283 |
-
"loss": 1.928,
|
| 284 |
-
"step": 370
|
| 285 |
-
},
|
| 286 |
-
{
|
| 287 |
-
"epoch": 0.48531289910600256,
|
| 288 |
-
"grad_norm": 1.9961628913879395,
|
| 289 |
-
"learning_rate": 6.961229040824927e-05,
|
| 290 |
-
"loss": 1.8961,
|
| 291 |
-
"step": 380
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.49808429118773945,
|
| 295 |
-
"grad_norm": 2.0501065254211426,
|
| 296 |
-
"learning_rate": 6.957791301695653e-05,
|
| 297 |
-
"loss": 1.8465,
|
| 298 |
-
"step": 390
|
| 299 |
-
},
|
| 300 |
-
{
|
| 301 |
-
"epoch": 0.5108556832694764,
|
| 302 |
-
"grad_norm": 2.014704704284668,
|
| 303 |
-
"learning_rate": 6.954208480568616e-05,
|
| 304 |
-
"loss": 1.9278,
|
| 305 |
-
"step": 400
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 0.5236270753512133,
|
| 309 |
-
"grad_norm": 1.8024693727493286,
|
| 310 |
-
"learning_rate": 6.950480727771816e-05,
|
| 311 |
-
"loss": 1.945,
|
| 312 |
-
"step": 410
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"epoch": 0.5363984674329502,
|
| 316 |
-
"grad_norm": 2.1424338817596436,
|
| 317 |
-
"learning_rate": 6.946608199714291e-05,
|
| 318 |
-
"loss": 1.9117,
|
| 319 |
-
"step": 420
|
| 320 |
-
},
|
| 321 |
-
{
|
| 322 |
-
"epoch": 0.5491698595146871,
|
| 323 |
-
"grad_norm": 1.9119850397109985,
|
| 324 |
-
"learning_rate": 6.942591058879567e-05,
|
| 325 |
-
"loss": 1.9086,
|
| 326 |
-
"step": 430
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 0.561941251596424,
|
| 330 |
-
"grad_norm": 1.866670846939087,
|
| 331 |
-
"learning_rate": 6.938429473818823e-05,
|
| 332 |
-
"loss": 1.91,
|
| 333 |
-
"step": 440
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 0.5747126436781609,
|
| 337 |
-
"grad_norm": 1.9712364673614502,
|
| 338 |
-
"learning_rate": 6.934123619143835e-05,
|
| 339 |
-
"loss": 1.899,
|
| 340 |
-
"step": 450
|
| 341 |
-
},
|
| 342 |
-
{
|
| 343 |
-
"epoch": 0.5747126436781609,
|
| 344 |
-
"eval_loss": 1.9519777297973633,
|
| 345 |
-
"eval_runtime": 24.1832,
|
| 346 |
-
"eval_samples_per_second": 20.676,
|
| 347 |
-
"eval_steps_per_second": 20.676,
|
| 348 |
-
"step": 450
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 0.5874840357598978,
|
| 352 |
-
"grad_norm": 1.8985211849212646,
|
| 353 |
-
"learning_rate": 6.929673675519645e-05,
|
| 354 |
-
"loss": 1.8811,
|
| 355 |
-
"step": 460
|
| 356 |
-
},
|
| 357 |
-
{
|
| 358 |
-
"epoch": 0.6002554278416348,
|
| 359 |
-
"grad_norm": 1.8869976997375488,
|
| 360 |
-
"learning_rate": 6.92507982965697e-05,
|
| 361 |
-
"loss": 1.8547,
|
| 362 |
-
"step": 470
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 0.6130268199233716,
|
| 366 |
-
"grad_norm": 2.027930974960327,
|
| 367 |
-
"learning_rate": 6.920342274304384e-05,
|
| 368 |
-
"loss": 1.8605,
|
| 369 |
-
"step": 480
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.6257982120051085,
|
| 373 |
-
"grad_norm": 1.832667350769043,
|
| 374 |
-
"learning_rate": 6.915461208240223e-05,
|
| 375 |
-
"loss": 1.8656,
|
| 376 |
-
"step": 490
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 0.6385696040868455,
|
| 380 |
-
"grad_norm": 1.9933980703353882,
|
| 381 |
-
"learning_rate": 6.910436836264241e-05,
|
| 382 |
-
"loss": 1.9675,
|
| 383 |
-
"step": 500
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.6513409961685823,
|
| 387 |
-
"grad_norm": 1.8492072820663452,
|
| 388 |
-
"learning_rate": 6.905269369189023e-05,
|
| 389 |
-
"loss": 1.8948,
|
| 390 |
-
"step": 510
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 0.6641123882503193,
|
| 394 |
-
"grad_norm": 2.021852970123291,
|
| 395 |
-
"learning_rate": 6.899959023831139e-05,
|
| 396 |
-
"loss": 1.8755,
|
| 397 |
-
"step": 520
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 0.6768837803320562,
|
| 401 |
-
"grad_norm": 1.7901625633239746,
|
| 402 |
-
"learning_rate": 6.894506023002045e-05,
|
| 403 |
-
"loss": 1.8934,
|
| 404 |
-
"step": 530
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.6896551724137931,
|
| 408 |
-
"grad_norm": 1.9580366611480713,
|
| 409 |
-
"learning_rate": 6.888910595498735e-05,
|
| 410 |
-
"loss": 1.9229,
|
| 411 |
-
"step": 540
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.70242656449553,
|
| 415 |
-
"grad_norm": 1.8306083679199219,
|
| 416 |
-
"learning_rate": 6.883172976094139e-05,
|
| 417 |
-
"loss": 1.9469,
|
| 418 |
-
"step": 550
|
| 419 |
-
},
|
| 420 |
-
{
|
| 421 |
-
"epoch": 0.7151979565772669,
|
| 422 |
-
"grad_norm": 1.8871126174926758,
|
| 423 |
-
"learning_rate": 6.877293405527276e-05,
|
| 424 |
-
"loss": 1.914,
|
| 425 |
-
"step": 560
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 0.7279693486590039,
|
| 429 |
-
"grad_norm": 1.7971593141555786,
|
| 430 |
-
"learning_rate": 6.871272130493153e-05,
|
| 431 |
-
"loss": 1.8504,
|
| 432 |
-
"step": 570
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 0.7407407407407407,
|
| 436 |
-
"grad_norm": 1.9786813259124756,
|
| 437 |
-
"learning_rate": 6.86510940363241e-05,
|
| 438 |
-
"loss": 1.876,
|
| 439 |
-
"step": 580
|
| 440 |
-
},
|
| 441 |
-
{
|
| 442 |
-
"epoch": 0.7535121328224776,
|
| 443 |
-
"grad_norm": 1.9042707681655884,
|
| 444 |
-
"learning_rate": 6.858805483520723e-05,
|
| 445 |
-
"loss": 1.8948,
|
| 446 |
-
"step": 590
|
| 447 |
-
},
|
| 448 |
-
{
|
| 449 |
-
"epoch": 0.7662835249042146,
|
| 450 |
-
"grad_norm": 2.191734552383423,
|
| 451 |
-
"learning_rate": 6.852360634657953e-05,
|
| 452 |
-
"loss": 1.9266,
|
| 453 |
-
"step": 600
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 0.7662835249042146,
|
| 457 |
-
"eval_loss": 1.9498045444488525,
|
| 458 |
-
"eval_runtime": 24.164,
|
| 459 |
-
"eval_samples_per_second": 20.692,
|
| 460 |
-
"eval_steps_per_second": 20.692,
|
| 461 |
-
"step": 600
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 0.7790549169859514,
|
| 465 |
-
"grad_norm": 1.8927336931228638,
|
| 466 |
-
"learning_rate": 6.845775127457056e-05,
|
| 467 |
-
"loss": 1.8992,
|
| 468 |
-
"step": 610
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.7918263090676884,
|
| 472 |
-
"grad_norm": 2.0473830699920654,
|
| 473 |
-
"learning_rate": 6.839049238232719e-05,
|
| 474 |
-
"loss": 1.8489,
|
| 475 |
-
"step": 620
|
| 476 |
-
},
|
| 477 |
-
{
|
| 478 |
-
"epoch": 0.8045977011494253,
|
| 479 |
-
"grad_norm": 1.8225151300430298,
|
| 480 |
-
"learning_rate": 6.832183249189786e-05,
|
| 481 |
-
"loss": 1.923,
|
| 482 |
-
"step": 630
|
| 483 |
-
},
|
| 484 |
-
{
|
| 485 |
-
"epoch": 0.8173690932311622,
|
| 486 |
-
"grad_norm": 1.8365392684936523,
|
| 487 |
-
"learning_rate": 6.825177448411404e-05,
|
| 488 |
-
"loss": 1.8782,
|
| 489 |
-
"step": 640
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.8301404853128991,
|
| 493 |
-
"grad_norm": 1.8688030242919922,
|
| 494 |
-
"learning_rate": 6.818032129846945e-05,
|
| 495 |
-
"loss": 1.9063,
|
| 496 |
-
"step": 650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 0.842911877394636,
|
| 500 |
-
"grad_norm": 1.7445991039276123,
|
| 501 |
-
"learning_rate": 6.810747593299666e-05,
|
| 502 |
-
"loss": 1.8548,
|
| 503 |
-
"step": 660
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.855683269476373,
|
| 507 |
-
"grad_norm": 1.7961305379867554,
|
| 508 |
-
"learning_rate": 6.803324144414127e-05,
|
| 509 |
-
"loss": 1.8504,
|
| 510 |
-
"step": 670
|
| 511 |
-
},
|
| 512 |
-
{
|
| 513 |
-
"epoch": 0.8684546615581098,
|
| 514 |
-
"grad_norm": 1.8798601627349854,
|
| 515 |
-
"learning_rate": 6.795762094663379e-05,
|
| 516 |
-
"loss": 1.9029,
|
| 517 |
-
"step": 680
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 0.8812260536398467,
|
| 521 |
-
"grad_norm": 1.9716835021972656,
|
| 522 |
-
"learning_rate": 6.788061761335882e-05,
|
| 523 |
-
"loss": 1.9092,
|
| 524 |
-
"step": 690
|
| 525 |
-
},
|
| 526 |
-
{
|
| 527 |
-
"epoch": 0.8939974457215837,
|
| 528 |
-
"grad_norm": 1.8877177238464355,
|
| 529 |
-
"learning_rate": 6.780223467522203e-05,
|
| 530 |
-
"loss": 1.9268,
|
| 531 |
-
"step": 700
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 0.9067688378033205,
|
| 535 |
-
"grad_norm": 2.017631769180298,
|
| 536 |
-
"learning_rate": 6.772247542101449e-05,
|
| 537 |
-
"loss": 1.9685,
|
| 538 |
-
"step": 710
|
| 539 |
-
},
|
| 540 |
-
{
|
| 541 |
-
"epoch": 0.9195402298850575,
|
| 542 |
-
"grad_norm": 1.9911872148513794,
|
| 543 |
-
"learning_rate": 6.764134319727477e-05,
|
| 544 |
-
"loss": 1.8774,
|
| 545 |
-
"step": 720
|
| 546 |
-
},
|
| 547 |
-
{
|
| 548 |
-
"epoch": 0.9323116219667944,
|
| 549 |
-
"grad_norm": 1.7452633380889893,
|
| 550 |
-
"learning_rate": 6.755884140814851e-05,
|
| 551 |
-
"loss": 1.8693,
|
| 552 |
-
"step": 730
|
| 553 |
-
},
|
| 554 |
-
{
|
| 555 |
-
"epoch": 0.9450830140485313,
|
| 556 |
-
"grad_norm": 1.9967455863952637,
|
| 557 |
-
"learning_rate": 6.747497351524552e-05,
|
| 558 |
-
"loss": 1.897,
|
| 559 |
-
"step": 740
|
| 560 |
-
},
|
| 561 |
-
{
|
| 562 |
-
"epoch": 0.9578544061302682,
|
| 563 |
-
"grad_norm": 1.7866498231887817,
|
| 564 |
-
"learning_rate": 6.738974303749464e-05,
|
| 565 |
-
"loss": 1.8744,
|
| 566 |
-
"step": 750
|
| 567 |
-
},
|
| 568 |
-
{
|
| 569 |
-
"epoch": 0.9578544061302682,
|
| 570 |
-
"eval_loss": 1.9411674737930298,
|
| 571 |
-
"eval_runtime": 25.0752,
|
| 572 |
-
"eval_samples_per_second": 19.94,
|
| 573 |
-
"eval_steps_per_second": 19.94,
|
| 574 |
-
"step": 750
|
| 575 |
-
},
|
| 576 |
-
{
|
| 577 |
-
"epoch": 0.9706257982120051,
|
| 578 |
-
"grad_norm": 1.8407279253005981,
|
| 579 |
-
"learning_rate": 6.7303153550996e-05,
|
| 580 |
-
"loss": 1.8936,
|
| 581 |
-
"step": 760
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 0.9833971902937421,
|
| 585 |
-
"grad_norm": 2.055628538131714,
|
| 586 |
-
"learning_rate": 6.721520868887103e-05,
|
| 587 |
-
"loss": 1.8761,
|
| 588 |
-
"step": 770
|
| 589 |
-
},
|
| 590 |
-
{
|
| 591 |
-
"epoch": 0.9961685823754789,
|
| 592 |
-
"grad_norm": 1.8534340858459473,
|
| 593 |
-
"learning_rate": 6.712591214111003e-05,
|
| 594 |
-
"loss": 1.8828,
|
| 595 |
-
"step": 780
|
| 596 |
-
},
|
| 597 |
-
{
|
| 598 |
-
"epoch": 1.0089399744572158,
|
| 599 |
-
"grad_norm": 1.8435447216033936,
|
| 600 |
-
"learning_rate": 6.703526765441728e-05,
|
| 601 |
-
"loss": 1.8907,
|
| 602 |
-
"step": 790
|
| 603 |
-
},
|
| 604 |
-
{
|
| 605 |
-
"epoch": 1.0217113665389528,
|
| 606 |
-
"grad_norm": 1.971137285232544,
|
| 607 |
-
"learning_rate": 6.69432790320539e-05,
|
| 608 |
-
"loss": 1.8402,
|
| 609 |
-
"step": 800
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 1.0344827586206897,
|
| 613 |
-
"grad_norm": 2.0040283203125,
|
| 614 |
-
"learning_rate": 6.684995013367826e-05,
|
| 615 |
-
"loss": 1.9047,
|
| 616 |
-
"step": 810
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 1.0472541507024267,
|
| 620 |
-
"grad_norm": 1.9755958318710327,
|
| 621 |
-
"learning_rate": 6.675528487518401e-05,
|
| 622 |
-
"loss": 1.8602,
|
| 623 |
-
"step": 820
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 1.0600255427841634,
|
| 627 |
-
"grad_norm": 1.8699226379394531,
|
| 628 |
-
"learning_rate": 6.665928722853581e-05,
|
| 629 |
-
"loss": 1.8777,
|
| 630 |
-
"step": 830
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 1.0727969348659003,
|
| 634 |
-
"grad_norm": 2.011084794998169,
|
| 635 |
-
"learning_rate": 6.656196122160264e-05,
|
| 636 |
-
"loss": 1.8186,
|
| 637 |
-
"step": 840
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 1.0855683269476373,
|
| 641 |
-
"grad_norm": 1.9715994596481323,
|
| 642 |
-
"learning_rate": 6.646331093798883e-05,
|
| 643 |
-
"loss": 1.8738,
|
| 644 |
-
"step": 850
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 1.0983397190293742,
|
| 648 |
-
"grad_norm": 1.997855305671692,
|
| 649 |
-
"learning_rate": 6.63633405168627e-05,
|
| 650 |
-
"loss": 1.8658,
|
| 651 |
-
"step": 860
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 1.1111111111111112,
|
| 655 |
-
"grad_norm": 1.9899482727050781,
|
| 656 |
-
"learning_rate": 6.626205415278291e-05,
|
| 657 |
-
"loss": 1.8366,
|
| 658 |
-
"step": 870
|
| 659 |
-
},
|
| 660 |
-
{
|
| 661 |
-
"epoch": 1.123882503192848,
|
| 662 |
-
"grad_norm": 1.9865031242370605,
|
| 663 |
-
"learning_rate": 6.615945609552244e-05,
|
| 664 |
-
"loss": 1.8194,
|
| 665 |
-
"step": 880
|
| 666 |
-
},
|
| 667 |
-
{
|
| 668 |
-
"epoch": 1.136653895274585,
|
| 669 |
-
"grad_norm": 1.8949304819107056,
|
| 670 |
-
"learning_rate": 6.605555064989027e-05,
|
| 671 |
-
"loss": 1.8091,
|
| 672 |
-
"step": 890
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 1.1494252873563218,
|
| 676 |
-
"grad_norm": 2.125781774520874,
|
| 677 |
-
"learning_rate": 6.595034217555082e-05,
|
| 678 |
-
"loss": 1.8225,
|
| 679 |
-
"step": 900
|
| 680 |
-
},
|
| 681 |
-
{
|
| 682 |
-
"epoch": 1.1494252873563218,
|
| 683 |
-
"eval_loss": 1.9406205415725708,
|
| 684 |
-
"eval_runtime": 24.2668,
|
| 685 |
-
"eval_samples_per_second": 20.604,
|
| 686 |
-
"eval_steps_per_second": 20.604,
|
| 687 |
-
"step": 900
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 1.1621966794380587,
|
| 691 |
-
"grad_norm": 1.8366427421569824,
|
| 692 |
-
"learning_rate": 6.584383508684096e-05,
|
| 693 |
-
"loss": 1.8386,
|
| 694 |
-
"step": 910
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 1.1749680715197957,
|
| 698 |
-
"grad_norm": 2.075599431991577,
|
| 699 |
-
"learning_rate": 6.57360338525848e-05,
|
| 700 |
-
"loss": 1.8182,
|
| 701 |
-
"step": 920
|
| 702 |
-
},
|
| 703 |
-
{
|
| 704 |
-
"epoch": 1.1877394636015326,
|
| 705 |
-
"grad_norm": 1.9955580234527588,
|
| 706 |
-
"learning_rate": 6.562694299590624e-05,
|
| 707 |
-
"loss": 1.8186,
|
| 708 |
-
"step": 930
|
| 709 |
-
},
|
| 710 |
-
{
|
| 711 |
-
"epoch": 1.2005108556832695,
|
| 712 |
-
"grad_norm": 2.012458086013794,
|
| 713 |
-
"learning_rate": 6.551656709403914e-05,
|
| 714 |
-
"loss": 1.8484,
|
| 715 |
-
"step": 940
|
| 716 |
-
},
|
| 717 |
-
{
|
| 718 |
-
"epoch": 1.2132822477650063,
|
| 719 |
-
"grad_norm": 2.052921772003174,
|
| 720 |
-
"learning_rate": 6.540491077813528e-05,
|
| 721 |
-
"loss": 1.8583,
|
| 722 |
-
"step": 950
|
| 723 |
-
},
|
| 724 |
-
{
|
| 725 |
-
"epoch": 1.2260536398467432,
|
| 726 |
-
"grad_norm": 1.8998031616210938,
|
| 727 |
-
"learning_rate": 6.529197873307006e-05,
|
| 728 |
-
"loss": 1.7882,
|
| 729 |
-
"step": 960
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 1.2388250319284801,
|
| 733 |
-
"grad_norm": 2.0680484771728516,
|
| 734 |
-
"learning_rate": 6.517777569724587e-05,
|
| 735 |
-
"loss": 1.8675,
|
| 736 |
-
"step": 970
|
| 737 |
-
},
|
| 738 |
-
{
|
| 739 |
-
"epoch": 1.251596424010217,
|
| 740 |
-
"grad_norm": 2.0891168117523193,
|
| 741 |
-
"learning_rate": 6.506230646239343e-05,
|
| 742 |
-
"loss": 1.8287,
|
| 743 |
-
"step": 980
|
| 744 |
-
},
|
| 745 |
-
{
|
| 746 |
-
"epoch": 1.264367816091954,
|
| 747 |
-
"grad_norm": 1.9214017391204834,
|
| 748 |
-
"learning_rate": 6.494557587337054e-05,
|
| 749 |
-
"loss": 1.81,
|
| 750 |
-
"step": 990
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 1.277139208173691,
|
| 754 |
-
"grad_norm": 2.0630176067352295,
|
| 755 |
-
"learning_rate": 6.482758882795892e-05,
|
| 756 |
-
"loss": 1.8507,
|
| 757 |
-
"step": 1000
|
| 758 |
-
},
|
| 759 |
-
{
|
| 760 |
-
"epoch": 1.289910600255428,
|
| 761 |
-
"grad_norm": 2.1162936687469482,
|
| 762 |
-
"learning_rate": 6.47083502766587e-05,
|
| 763 |
-
"loss": 1.9005,
|
| 764 |
-
"step": 1010
|
| 765 |
-
},
|
| 766 |
-
{
|
| 767 |
-
"epoch": 1.3026819923371646,
|
| 768 |
-
"grad_norm": 2.150144100189209,
|
| 769 |
-
"learning_rate": 6.458786522248068e-05,
|
| 770 |
-
"loss": 1.8002,
|
| 771 |
-
"step": 1020
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 1.3154533844189016,
|
| 775 |
-
"grad_norm": 1.9569624662399292,
|
| 776 |
-
"learning_rate": 6.446613872073644e-05,
|
| 777 |
-
"loss": 1.8005,
|
| 778 |
-
"step": 1030
|
| 779 |
-
},
|
| 780 |
-
{
|
| 781 |
-
"epoch": 1.3282247765006385,
|
| 782 |
-
"grad_norm": 2.029466390609741,
|
| 783 |
-
"learning_rate": 6.434317587882618e-05,
|
| 784 |
-
"loss": 1.8143,
|
| 785 |
-
"step": 1040
|
| 786 |
-
},
|
| 787 |
-
{
|
| 788 |
-
"epoch": 1.3409961685823755,
|
| 789 |
-
"grad_norm": 2.1835060119628906,
|
| 790 |
-
"learning_rate": 6.421898185602448e-05,
|
| 791 |
-
"loss": 1.9125,
|
| 792 |
-
"step": 1050
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 1.3409961685823755,
|
| 796 |
-
"eval_loss": 1.9381868839263916,
|
| 797 |
-
"eval_runtime": 24.2287,
|
| 798 |
-
"eval_samples_per_second": 20.637,
|
| 799 |
-
"eval_steps_per_second": 20.637,
|
| 800 |
-
"step": 1050
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 1.3537675606641124,
|
| 804 |
-
"grad_norm": 2.052374839782715,
|
| 805 |
-
"learning_rate": 6.409356186326383e-05,
|
| 806 |
-
"loss": 1.8531,
|
| 807 |
-
"step": 1060
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 1.3665389527458494,
|
| 811 |
-
"grad_norm": 2.042969226837158,
|
| 812 |
-
"learning_rate": 6.396692116291589e-05,
|
| 813 |
-
"loss": 1.8755,
|
| 814 |
-
"step": 1070
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 1.3793103448275863,
|
| 818 |
-
"grad_norm": 1.9522734880447388,
|
| 819 |
-
"learning_rate": 6.383906506857083e-05,
|
| 820 |
-
"loss": 1.8719,
|
| 821 |
-
"step": 1080
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 1.392081736909323,
|
| 825 |
-
"grad_norm": 1.8483445644378662,
|
| 826 |
-
"learning_rate": 6.37099989448143e-05,
|
| 827 |
-
"loss": 1.8538,
|
| 828 |
-
"step": 1090
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 1.40485312899106,
|
| 832 |
-
"grad_norm": 2.2309532165527344,
|
| 833 |
-
"learning_rate": 6.357972820700237e-05,
|
| 834 |
-
"loss": 1.8851,
|
| 835 |
-
"step": 1100
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 1.417624521072797,
|
| 839 |
-
"grad_norm": 1.9797334671020508,
|
| 840 |
-
"learning_rate": 6.344825832103429e-05,
|
| 841 |
-
"loss": 1.8704,
|
| 842 |
-
"step": 1110
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 1.4303959131545338,
|
| 846 |
-
"grad_norm": 2.023789167404175,
|
| 847 |
-
"learning_rate": 6.331559480312315e-05,
|
| 848 |
-
"loss": 1.8692,
|
| 849 |
-
"step": 1120
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 1.4431673052362708,
|
| 853 |
-
"grad_norm": 1.9136630296707153,
|
| 854 |
-
"learning_rate": 6.318174321956448e-05,
|
| 855 |
-
"loss": 1.8521,
|
| 856 |
-
"step": 1130
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 1.4559386973180077,
|
| 860 |
-
"grad_norm": 2.0023677349090576,
|
| 861 |
-
"learning_rate": 6.304670918650265e-05,
|
| 862 |
-
"loss": 1.8565,
|
| 863 |
-
"step": 1140
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 1.4687100893997447,
|
| 867 |
-
"grad_norm": 2.022937536239624,
|
| 868 |
-
"learning_rate": 6.291049836969522e-05,
|
| 869 |
-
"loss": 1.8222,
|
| 870 |
-
"step": 1150
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 1.4814814814814814,
|
| 874 |
-
"grad_norm": 2.031189203262329,
|
| 875 |
-
"learning_rate": 6.277311648427529e-05,
|
| 876 |
-
"loss": 1.8933,
|
| 877 |
-
"step": 1160
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 1.4942528735632183,
|
| 881 |
-
"grad_norm": 2.021033763885498,
|
| 882 |
-
"learning_rate": 6.26345692945116e-05,
|
| 883 |
-
"loss": 1.8851,
|
| 884 |
-
"step": 1170
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 1.5070242656449553,
|
| 888 |
-
"grad_norm": 1.9933768510818481,
|
| 889 |
-
"learning_rate": 6.249486261356676e-05,
|
| 890 |
-
"loss": 1.8925,
|
| 891 |
-
"step": 1180
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 1.5197956577266922,
|
| 895 |
-
"grad_norm": 2.0429482460021973,
|
| 896 |
-
"learning_rate": 6.23540023032533e-05,
|
| 897 |
-
"loss": 1.8536,
|
| 898 |
-
"step": 1190
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 1.5325670498084292,
|
| 902 |
-
"grad_norm": 1.950465440750122,
|
| 903 |
-
"learning_rate": 6.221199427378772e-05,
|
| 904 |
-
"loss": 1.9116,
|
| 905 |
-
"step": 1200
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 1.5325670498084292,
|
| 909 |
-
"eval_loss": 1.9389454126358032,
|
| 910 |
-
"eval_runtime": 24.2278,
|
| 911 |
-
"eval_samples_per_second": 20.637,
|
| 912 |
-
"eval_steps_per_second": 20.637,
|
| 913 |
-
"step": 1200
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 1.545338441890166,
|
| 917 |
-
"grad_norm": 1.9224931001663208,
|
| 918 |
-
"learning_rate": 6.206884448354253e-05,
|
| 919 |
-
"loss": 1.8073,
|
| 920 |
-
"step": 1210
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 1.558109833971903,
|
| 924 |
-
"grad_norm": 1.9591902494430542,
|
| 925 |
-
"learning_rate": 6.192455893879623e-05,
|
| 926 |
-
"loss": 1.8249,
|
| 927 |
-
"step": 1220
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 1.5708812260536398,
|
| 931 |
-
"grad_norm": 1.9398179054260254,
|
| 932 |
-
"learning_rate": 6.177914369348129e-05,
|
| 933 |
-
"loss": 1.8701,
|
| 934 |
-
"step": 1230
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 1.5836526181353767,
|
| 938 |
-
"grad_norm": 1.8603562116622925,
|
| 939 |
-
"learning_rate": 6.163260484893019e-05,
|
| 940 |
-
"loss": 1.8398,
|
| 941 |
-
"step": 1240
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 1.5964240102171137,
|
| 945 |
-
"grad_norm": 2.005242347717285,
|
| 946 |
-
"learning_rate": 6.148494855361933e-05,
|
| 947 |
-
"loss": 1.8626,
|
| 948 |
-
"step": 1250
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 1.6091954022988506,
|
| 952 |
-
"grad_norm": 1.8576045036315918,
|
| 953 |
-
"learning_rate": 6.133618100291116e-05,
|
| 954 |
-
"loss": 1.8573,
|
| 955 |
-
"step": 1260
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 1.6219667943805876,
|
| 959 |
-
"grad_norm": 2.0463547706604004,
|
| 960 |
-
"learning_rate": 6.118630843879414e-05,
|
| 961 |
-
"loss": 1.8162,
|
| 962 |
-
"step": 1270
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 1.6347381864623243,
|
| 966 |
-
"grad_norm": 1.9812066555023193,
|
| 967 |
-
"learning_rate": 6.10353371496209e-05,
|
| 968 |
-
"loss": 1.8147,
|
| 969 |
-
"step": 1280
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 1.6475095785440614,
|
| 973 |
-
"grad_norm": 2.0018861293792725,
|
| 974 |
-
"learning_rate": 6.088327346984437e-05,
|
| 975 |
-
"loss": 1.7751,
|
| 976 |
-
"step": 1290
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 1.6602809706257982,
|
| 980 |
-
"grad_norm": 2.017937183380127,
|
| 981 |
-
"learning_rate": 6.073012377975201e-05,
|
| 982 |
-
"loss": 1.8877,
|
| 983 |
-
"step": 1300
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 1.673052362707535,
|
| 987 |
-
"grad_norm": 1.9787315130233765,
|
| 988 |
-
"learning_rate": 6.057589450519807e-05,
|
| 989 |
-
"loss": 1.8283,
|
| 990 |
-
"step": 1310
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 1.685823754789272,
|
| 994 |
-
"grad_norm": 2.075465679168701,
|
| 995 |
-
"learning_rate": 6.042059211733404e-05,
|
| 996 |
-
"loss": 1.8498,
|
| 997 |
-
"step": 1320
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 1.698595146871009,
|
| 1001 |
-
"grad_norm": 2.0676026344299316,
|
| 1002 |
-
"learning_rate": 6.026422313233707e-05,
|
| 1003 |
-
"loss": 1.8781,
|
| 1004 |
-
"step": 1330
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 1.711366538952746,
|
| 1008 |
-
"grad_norm": 2.158276319503784,
|
| 1009 |
-
"learning_rate": 6.010679411113658e-05,
|
| 1010 |
-
"loss": 1.8436,
|
| 1011 |
-
"step": 1340
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 1.7241379310344827,
|
| 1015 |
-
"grad_norm": 2.1169486045837402,
|
| 1016 |
-
"learning_rate": 5.9948311659139024e-05,
|
| 1017 |
-
"loss": 1.8629,
|
| 1018 |
-
"step": 1350
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 1.7241379310344827,
|
| 1022 |
-
"eval_loss": 1.9343986511230469,
|
| 1023 |
-
"eval_runtime": 24.1802,
|
| 1024 |
-
"eval_samples_per_second": 20.678,
|
| 1025 |
-
"eval_steps_per_second": 20.678,
|
| 1026 |
-
"step": 1350
|
| 1027 |
-
},
|
| 1028 |
-
{
|
| 1029 |
-
"epoch": 1.7369093231162198,
|
| 1030 |
-
"grad_norm": 1.857434868812561,
|
| 1031 |
-
"learning_rate": 5.978878242595065e-05,
|
| 1032 |
-
"loss": 1.8505,
|
| 1033 |
-
"step": 1360
|
| 1034 |
-
},
|
| 1035 |
-
{
|
| 1036 |
-
"epoch": 1.7496807151979565,
|
| 1037 |
-
"grad_norm": 2.055877923965454,
|
| 1038 |
-
"learning_rate": 5.962821310509861e-05,
|
| 1039 |
-
"loss": 1.8437,
|
| 1040 |
-
"step": 1370
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 1.7624521072796935,
|
| 1044 |
-
"grad_norm": 1.8985681533813477,
|
| 1045 |
-
"learning_rate": 5.9466610433750005e-05,
|
| 1046 |
-
"loss": 1.8312,
|
| 1047 |
-
"step": 1380
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 1.7752234993614304,
|
| 1051 |
-
"grad_norm": 1.991992712020874,
|
| 1052 |
-
"learning_rate": 5.9303981192429265e-05,
|
| 1053 |
-
"loss": 1.9209,
|
| 1054 |
-
"step": 1390
|
| 1055 |
-
},
|
| 1056 |
-
{
|
| 1057 |
-
"epoch": 1.7879948914431671,
|
| 1058 |
-
"grad_norm": 1.845304012298584,
|
| 1059 |
-
"learning_rate": 5.914033220473365e-05,
|
| 1060 |
-
"loss": 1.85,
|
| 1061 |
-
"step": 1400
|
| 1062 |
-
},
|
| 1063 |
-
{
|
| 1064 |
-
"epoch": 1.8007662835249043,
|
| 1065 |
-
"grad_norm": 1.8313244581222534,
|
| 1066 |
-
"learning_rate": 5.897567033704691e-05,
|
| 1067 |
-
"loss": 1.8897,
|
| 1068 |
-
"step": 1410
|
| 1069 |
-
},
|
| 1070 |
-
{
|
| 1071 |
-
"epoch": 1.813537675606641,
|
| 1072 |
-
"grad_norm": 1.8554408550262451,
|
| 1073 |
-
"learning_rate": 5.881000249825124e-05,
|
| 1074 |
-
"loss": 1.8012,
|
| 1075 |
-
"step": 1420
|
| 1076 |
-
},
|
| 1077 |
-
{
|
| 1078 |
-
"epoch": 1.8263090676883782,
|
| 1079 |
-
"grad_norm": 2.053835868835449,
|
| 1080 |
-
"learning_rate": 5.8643335639437366e-05,
|
| 1081 |
-
"loss": 1.8708,
|
| 1082 |
-
"step": 1430
|
| 1083 |
-
},
|
| 1084 |
-
{
|
| 1085 |
-
"epoch": 1.839080459770115,
|
| 1086 |
-
"grad_norm": 1.9372299909591675,
|
| 1087 |
-
"learning_rate": 5.847567675361288e-05,
|
| 1088 |
-
"loss": 1.8338,
|
| 1089 |
-
"step": 1440
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 1.8518518518518519,
|
| 1093 |
-
"grad_norm": 2.2738914489746094,
|
| 1094 |
-
"learning_rate": 5.830703287540883e-05,
|
| 1095 |
-
"loss": 1.8748,
|
| 1096 |
-
"step": 1450
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 1.8646232439335888,
|
| 1100 |
-
"grad_norm": 2.297995090484619,
|
| 1101 |
-
"learning_rate": 5.813741108078461e-05,
|
| 1102 |
-
"loss": 1.8276,
|
| 1103 |
-
"step": 1460
|
| 1104 |
-
},
|
| 1105 |
-
{
|
| 1106 |
-
"epoch": 1.8773946360153255,
|
| 1107 |
-
"grad_norm": 1.985796332359314,
|
| 1108 |
-
"learning_rate": 5.796681848673098e-05,
|
| 1109 |
-
"loss": 1.9204,
|
| 1110 |
-
"step": 1470
|
| 1111 |
-
},
|
| 1112 |
-
{
|
| 1113 |
-
"epoch": 1.8901660280970627,
|
| 1114 |
-
"grad_norm": 1.8658971786499023,
|
| 1115 |
-
"learning_rate": 5.779526225097153e-05,
|
| 1116 |
-
"loss": 1.8205,
|
| 1117 |
-
"step": 1480
|
| 1118 |
-
},
|
| 1119 |
-
{
|
| 1120 |
-
"epoch": 1.9029374201787994,
|
| 1121 |
-
"grad_norm": 1.9643296003341675,
|
| 1122 |
-
"learning_rate": 5.762274957166233e-05,
|
| 1123 |
-
"loss": 1.8421,
|
| 1124 |
-
"step": 1490
|
| 1125 |
-
},
|
| 1126 |
-
{
|
| 1127 |
-
"epoch": 1.9157088122605364,
|
| 1128 |
-
"grad_norm": 1.924149751663208,
|
| 1129 |
-
"learning_rate": 5.7449287687089895e-05,
|
| 1130 |
-
"loss": 1.8696,
|
| 1131 |
-
"step": 1500
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 1.9157088122605364,
|
| 1135 |
-
"eval_loss": 1.9337016344070435,
|
| 1136 |
-
"eval_runtime": 24.181,
|
| 1137 |
-
"eval_samples_per_second": 20.677,
|
| 1138 |
-
"eval_steps_per_second": 20.677,
|
| 1139 |
-
"step": 1500
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 1.9284802043422733,
|
| 1143 |
-
"grad_norm": 1.9967809915542603,
|
| 1144 |
-
"learning_rate": 5.727488387536748e-05,
|
| 1145 |
-
"loss": 1.8602,
|
| 1146 |
-
"step": 1510
|
| 1147 |
-
},
|
| 1148 |
-
{
|
| 1149 |
-
"epoch": 1.9412515964240102,
|
| 1150 |
-
"grad_norm": 1.9925228357315063,
|
| 1151 |
-
"learning_rate": 5.709954545412975e-05,
|
| 1152 |
-
"loss": 1.8775,
|
| 1153 |
-
"step": 1520
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 1.9540229885057472,
|
| 1157 |
-
"grad_norm": 2.0943105220794678,
|
| 1158 |
-
"learning_rate": 5.692327978022566e-05,
|
| 1159 |
-
"loss": 1.8434,
|
| 1160 |
-
"step": 1530
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 1.966794380587484,
|
| 1164 |
-
"grad_norm": 1.9929866790771484,
|
| 1165 |
-
"learning_rate": 5.67460942494099e-05,
|
| 1166 |
-
"loss": 1.8754,
|
| 1167 |
-
"step": 1540
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 1.979565772669221,
|
| 1171 |
-
"grad_norm": 1.9730305671691895,
|
| 1172 |
-
"learning_rate": 5.656799629603245e-05,
|
| 1173 |
-
"loss": 1.8585,
|
| 1174 |
-
"step": 1550
|
| 1175 |
-
},
|
| 1176 |
-
{
|
| 1177 |
-
"epoch": 1.9923371647509578,
|
| 1178 |
-
"grad_norm": 2.1760172843933105,
|
| 1179 |
-
"learning_rate": 5.638899339272675e-05,
|
| 1180 |
-
"loss": 1.8677,
|
| 1181 |
-
"step": 1560
|
| 1182 |
-
},
|
| 1183 |
-
{
|
| 1184 |
-
"epoch": 2.005108556832695,
|
| 1185 |
-
"grad_norm": 1.958113193511963,
|
| 1186 |
-
"learning_rate": 5.6209093050096115e-05,
|
| 1187 |
-
"loss": 1.7717,
|
| 1188 |
-
"step": 1570
|
| 1189 |
-
},
|
| 1190 |
-
{
|
| 1191 |
-
"epoch": 2.0178799489144317,
|
| 1192 |
-
"grad_norm": 1.8135682344436646,
|
| 1193 |
-
"learning_rate": 5.6028302816398624e-05,
|
| 1194 |
-
"loss": 1.7952,
|
| 1195 |
-
"step": 1580
|
| 1196 |
-
},
|
| 1197 |
-
{
|
| 1198 |
-
"epoch": 2.0306513409961684,
|
| 1199 |
-
"grad_norm": 1.9992247819900513,
|
| 1200 |
-
"learning_rate": 5.5846630277230384e-05,
|
| 1201 |
-
"loss": 1.8559,
|
| 1202 |
-
"step": 1590
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 2.0434227330779056,
|
| 1206 |
-
"grad_norm": 1.9503588676452637,
|
| 1207 |
-
"learning_rate": 5.566408305520729e-05,
|
| 1208 |
-
"loss": 1.8136,
|
| 1209 |
-
"step": 1600
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 2.0561941251596423,
|
| 1213 |
-
"grad_norm": 2.0600390434265137,
|
| 1214 |
-
"learning_rate": 5.5480668809645175e-05,
|
| 1215 |
-
"loss": 1.8126,
|
| 1216 |
-
"step": 1610
|
| 1217 |
-
},
|
| 1218 |
-
{
|
| 1219 |
-
"epoch": 2.0689655172413794,
|
| 1220 |
-
"grad_norm": 1.9783591032028198,
|
| 1221 |
-
"learning_rate": 5.5296395236238445e-05,
|
| 1222 |
-
"loss": 1.8198,
|
| 1223 |
-
"step": 1620
|
| 1224 |
-
},
|
| 1225 |
-
{
|
| 1226 |
-
"epoch": 2.081736909323116,
|
| 1227 |
-
"grad_norm": 2.09946346282959,
|
| 1228 |
-
"learning_rate": 5.5111270066737163e-05,
|
| 1229 |
-
"loss": 1.7505,
|
| 1230 |
-
"step": 1630
|
| 1231 |
-
},
|
| 1232 |
-
{
|
| 1233 |
-
"epoch": 2.0945083014048533,
|
| 1234 |
-
"grad_norm": 1.9546903371810913,
|
| 1235 |
-
"learning_rate": 5.492530106862269e-05,
|
| 1236 |
-
"loss": 1.8198,
|
| 1237 |
-
"step": 1640
|
| 1238 |
-
},
|
| 1239 |
-
{
|
| 1240 |
-
"epoch": 2.10727969348659,
|
| 1241 |
-
"grad_norm": 2.0820565223693848,
|
| 1242 |
-
"learning_rate": 5.4738496044781724e-05,
|
| 1243 |
-
"loss": 1.8473,
|
| 1244 |
-
"step": 1650
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 2.10727969348659,
|
| 1248 |
-
"eval_loss": 1.9384523630142212,
|
| 1249 |
-
"eval_runtime": 24.2888,
|
| 1250 |
-
"eval_samples_per_second": 20.586,
|
| 1251 |
-
"eval_steps_per_second": 20.586,
|
| 1252 |
-
"step": 1650
|
| 1253 |
-
},
|
| 1254 |
-
{
|
| 1255 |
-
"epoch": 2.1200510855683268,
|
| 1256 |
-
"grad_norm": 2.078505516052246,
|
| 1257 |
-
"learning_rate": 5.4550862833178926e-05,
|
| 1258 |
-
"loss": 1.8142,
|
| 1259 |
-
"step": 1660
|
| 1260 |
-
},
|
| 1261 |
-
{
|
| 1262 |
-
"epoch": 2.132822477650064,
|
| 1263 |
-
"grad_norm": 2.0138731002807617,
|
| 1264 |
-
"learning_rate": 5.436240930652807e-05,
|
| 1265 |
-
"loss": 1.8039,
|
| 1266 |
-
"step": 1670
|
| 1267 |
-
},
|
| 1268 |
-
{
|
| 1269 |
-
"epoch": 2.1455938697318007,
|
| 1270 |
-
"grad_norm": 1.958559513092041,
|
| 1271 |
-
"learning_rate": 5.41731433719617e-05,
|
| 1272 |
-
"loss": 1.8039,
|
| 1273 |
-
"step": 1680
|
| 1274 |
-
},
|
| 1275 |
-
{
|
| 1276 |
-
"epoch": 2.158365261813538,
|
| 1277 |
-
"grad_norm": 2.14679217338562,
|
| 1278 |
-
"learning_rate": 5.398307297069936e-05,
|
| 1279 |
-
"loss": 1.7873,
|
| 1280 |
-
"step": 1690
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 2.1711366538952745,
|
| 1284 |
-
"grad_norm": 2.2213780879974365,
|
| 1285 |
-
"learning_rate": 5.379220607771443e-05,
|
| 1286 |
-
"loss": 1.7886,
|
| 1287 |
-
"step": 1700
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 2.1839080459770113,
|
| 1291 |
-
"grad_norm": 1.9779000282287598,
|
| 1292 |
-
"learning_rate": 5.3600550701399455e-05,
|
| 1293 |
-
"loss": 1.7599,
|
| 1294 |
-
"step": 1710
|
| 1295 |
-
},
|
| 1296 |
-
{
|
| 1297 |
-
"epoch": 2.1966794380587484,
|
| 1298 |
-
"grad_norm": 2.0890724658966064,
|
| 1299 |
-
"learning_rate": 5.340811488323019e-05,
|
| 1300 |
-
"loss": 1.7752,
|
| 1301 |
-
"step": 1720
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 2.209450830140485,
|
| 1305 |
-
"grad_norm": 2.2274386882781982,
|
| 1306 |
-
"learning_rate": 5.321490669742815e-05,
|
| 1307 |
-
"loss": 1.8219,
|
| 1308 |
-
"step": 1730
|
| 1309 |
-
},
|
| 1310 |
-
{
|
| 1311 |
-
"epoch": 2.2222222222222223,
|
| 1312 |
-
"grad_norm": 2.2681832313537598,
|
| 1313 |
-
"learning_rate": 5.302093425062187e-05,
|
| 1314 |
-
"loss": 1.7979,
|
| 1315 |
-
"step": 1740
|
| 1316 |
-
},
|
| 1317 |
-
{
|
| 1318 |
-
"epoch": 2.234993614303959,
|
| 1319 |
-
"grad_norm": 2.136756420135498,
|
| 1320 |
-
"learning_rate": 5.28262056815067e-05,
|
| 1321 |
-
"loss": 1.7548,
|
| 1322 |
-
"step": 1750
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 2.247765006385696,
|
| 1326 |
-
"grad_norm": 2.3203325271606445,
|
| 1327 |
-
"learning_rate": 5.263072916050342e-05,
|
| 1328 |
-
"loss": 1.8238,
|
| 1329 |
-
"step": 1760
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 2.260536398467433,
|
| 1333 |
-
"grad_norm": 2.19116473197937,
|
| 1334 |
-
"learning_rate": 5.243451288941535e-05,
|
| 1335 |
-
"loss": 1.8143,
|
| 1336 |
-
"step": 1770
|
| 1337 |
-
},
|
| 1338 |
-
{
|
| 1339 |
-
"epoch": 2.27330779054917,
|
| 1340 |
-
"grad_norm": 2.180558204650879,
|
| 1341 |
-
"learning_rate": 5.2237565101084235e-05,
|
| 1342 |
-
"loss": 1.7681,
|
| 1343 |
-
"step": 1780
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 2.286079182630907,
|
| 1347 |
-
"grad_norm": 2.287687063217163,
|
| 1348 |
-
"learning_rate": 5.203989405904483e-05,
|
| 1349 |
-
"loss": 1.7872,
|
| 1350 |
-
"step": 1790
|
| 1351 |
-
},
|
| 1352 |
-
{
|
| 1353 |
-
"epoch": 2.2988505747126435,
|
| 1354 |
-
"grad_norm": 2.288097620010376,
|
| 1355 |
-
"learning_rate": 5.1841508057178177e-05,
|
| 1356 |
-
"loss": 1.7939,
|
| 1357 |
-
"step": 1800
|
| 1358 |
-
},
|
| 1359 |
-
{
|
| 1360 |
-
"epoch": 2.2988505747126435,
|
| 1361 |
-
"eval_loss": 1.9397783279418945,
|
| 1362 |
-
"eval_runtime": 24.3943,
|
| 1363 |
-
"eval_samples_per_second": 20.497,
|
| 1364 |
-
"eval_steps_per_second": 20.497,
|
| 1365 |
-
"step": 1800
|
| 1366 |
-
},
|
| 1367 |
-
{
|
| 1368 |
-
"epoch": 2.3116219667943807,
|
| 1369 |
-
"grad_norm": 2.1380770206451416,
|
| 1370 |
-
"learning_rate": 1.728050268572606e-06,
|
| 1371 |
-
"loss": 1.7778,
|
| 1372 |
-
"step": 1810
|
| 1373 |
-
},
|
| 1374 |
-
{
|
| 1375 |
-
"epoch": 2.3243933588761174,
|
| 1376 |
-
"grad_norm": 2.165830135345459,
|
| 1377 |
-
"learning_rate": 3.456100537145212e-06,
|
| 1378 |
-
"loss": 1.7769,
|
| 1379 |
-
"step": 1820
|
| 1380 |
-
},
|
| 1381 |
-
{
|
| 1382 |
-
"epoch": 2.3371647509578546,
|
| 1383 |
-
"grad_norm": 2.399066209793091,
|
| 1384 |
-
"learning_rate": 5.184150805717818e-06,
|
| 1385 |
-
"loss": 1.7873,
|
| 1386 |
-
"step": 1830
|
| 1387 |
-
},
|
| 1388 |
-
{
|
| 1389 |
-
"epoch": 2.3499361430395913,
|
| 1390 |
-
"grad_norm": 2.234840154647827,
|
| 1391 |
-
"learning_rate": 6.912201074290424e-06,
|
| 1392 |
-
"loss": 1.7203,
|
| 1393 |
-
"step": 1840
|
| 1394 |
-
},
|
| 1395 |
-
{
|
| 1396 |
-
"epoch": 2.362707535121328,
|
| 1397 |
-
"grad_norm": 2.12174916267395,
|
| 1398 |
-
"learning_rate": 8.640251342863028e-06,
|
| 1399 |
-
"loss": 1.8158,
|
| 1400 |
-
"step": 1850
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 2.375478927203065,
|
| 1404 |
-
"grad_norm": 2.109421968460083,
|
| 1405 |
-
"learning_rate": 1.0368301611435635e-05,
|
| 1406 |
-
"loss": 1.7586,
|
| 1407 |
-
"step": 1860
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 2.388250319284802,
|
| 1411 |
-
"grad_norm": 2.1194908618927,
|
| 1412 |
-
"learning_rate": 1.209635188000824e-05,
|
| 1413 |
-
"loss": 1.7936,
|
| 1414 |
-
"step": 1870
|
| 1415 |
-
},
|
| 1416 |
-
{
|
| 1417 |
-
"epoch": 2.401021711366539,
|
| 1418 |
-
"grad_norm": 2.211038589477539,
|
| 1419 |
-
"learning_rate": 1.3824402148580848e-05,
|
| 1420 |
-
"loss": 1.7774,
|
| 1421 |
-
"step": 1880
|
| 1422 |
-
},
|
| 1423 |
-
{
|
| 1424 |
-
"epoch": 2.413793103448276,
|
| 1425 |
-
"grad_norm": 2.295914888381958,
|
| 1426 |
-
"learning_rate": 1.5552452417153453e-05,
|
| 1427 |
-
"loss": 1.817,
|
| 1428 |
-
"step": 1890
|
| 1429 |
-
},
|
| 1430 |
-
{
|
| 1431 |
-
"epoch": 2.4265644955300125,
|
| 1432 |
-
"grad_norm": 2.0694994926452637,
|
| 1433 |
-
"learning_rate": 1.7280502685726057e-05,
|
| 1434 |
-
"loss": 1.8039,
|
| 1435 |
-
"step": 1900
|
| 1436 |
-
},
|
| 1437 |
-
{
|
| 1438 |
-
"epoch": 2.4393358876117497,
|
| 1439 |
-
"grad_norm": 2.2239274978637695,
|
| 1440 |
-
"learning_rate": 1.9008552954298664e-05,
|
| 1441 |
-
"loss": 1.8505,
|
| 1442 |
-
"step": 1910
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 2.4521072796934864,
|
| 1446 |
-
"grad_norm": 2.0061118602752686,
|
| 1447 |
-
"learning_rate": 2.073660322287127e-05,
|
| 1448 |
-
"loss": 1.7951,
|
| 1449 |
-
"step": 1920
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 2.4648786717752236,
|
| 1453 |
-
"grad_norm": 2.198082447052002,
|
| 1454 |
-
"learning_rate": 2.2464653491443878e-05,
|
| 1455 |
-
"loss": 1.8012,
|
| 1456 |
-
"step": 1930
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 2.4776500638569603,
|
| 1460 |
-
"grad_norm": 1.9910808801651,
|
| 1461 |
-
"learning_rate": 2.419270376001648e-05,
|
| 1462 |
-
"loss": 1.695,
|
| 1463 |
-
"step": 1940
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 2.4904214559386975,
|
| 1467 |
-
"grad_norm": 2.2533199787139893,
|
| 1468 |
-
"learning_rate": 2.5920754028589088e-05,
|
| 1469 |
-
"loss": 1.7852,
|
| 1470 |
-
"step": 1950
|
| 1471 |
-
},
|
| 1472 |
-
{
|
| 1473 |
-
"epoch": 2.4904214559386975,
|
| 1474 |
-
"eval_loss": 1.936609148979187,
|
| 1475 |
-
"eval_runtime": 24.3942,
|
| 1476 |
-
"eval_samples_per_second": 20.497,
|
| 1477 |
-
"eval_steps_per_second": 20.497,
|
| 1478 |
-
"step": 1950
|
| 1479 |
-
},
|
| 1480 |
-
{
|
| 1481 |
-
"epoch": 2.503192848020434,
|
| 1482 |
-
"grad_norm": 2.3623099327087402,
|
| 1483 |
-
"learning_rate": 2.5920482132857834e-05,
|
| 1484 |
-
"loss": 1.8228,
|
| 1485 |
-
"step": 1960
|
| 1486 |
-
},
|
| 1487 |
-
{
|
| 1488 |
-
"epoch": 2.5159642401021713,
|
| 1489 |
-
"grad_norm": 2.2549827098846436,
|
| 1490 |
-
"learning_rate": 2.5919666457072272e-05,
|
| 1491 |
-
"loss": 1.7792,
|
| 1492 |
-
"step": 1970
|
| 1493 |
-
},
|
| 1494 |
-
{
|
| 1495 |
-
"epoch": 2.528735632183908,
|
| 1496 |
-
"grad_norm": 2.1171176433563232,
|
| 1497 |
-
"learning_rate": 2.5918307035456523e-05,
|
| 1498 |
-
"loss": 1.8581,
|
| 1499 |
-
"step": 1980
|
| 1500 |
-
},
|
| 1501 |
-
{
|
| 1502 |
-
"epoch": 2.541507024265645,
|
| 1503 |
-
"grad_norm": 2.118044137954712,
|
| 1504 |
-
"learning_rate": 2.5916403925049198e-05,
|
| 1505 |
-
"loss": 1.8363,
|
| 1506 |
-
"step": 1990
|
| 1507 |
-
},
|
| 1508 |
-
{
|
| 1509 |
-
"epoch": 2.554278416347382,
|
| 1510 |
-
"grad_norm": 2.404895782470703,
|
| 1511 |
-
"learning_rate": 2.5913957205700998e-05,
|
| 1512 |
-
"loss": 1.8638,
|
| 1513 |
-
"step": 2000
|
| 1514 |
-
},
|
| 1515 |
-
{
|
| 1516 |
-
"epoch": 2.5670498084291187,
|
| 1517 |
-
"grad_norm": 2.2589292526245117,
|
| 1518 |
-
"learning_rate": 2.591096698007136e-05,
|
| 1519 |
-
"loss": 1.817,
|
| 1520 |
-
"step": 2010
|
| 1521 |
-
},
|
| 1522 |
-
{
|
| 1523 |
-
"epoch": 2.579821200510856,
|
| 1524 |
-
"grad_norm": 2.308377981185913,
|
| 1525 |
-
"learning_rate": 2.5907433373624172e-05,
|
| 1526 |
-
"loss": 1.7985,
|
| 1527 |
-
"step": 2020
|
| 1528 |
-
},
|
| 1529 |
-
{
|
| 1530 |
-
"epoch": 2.5925925925925926,
|
| 1531 |
-
"grad_norm": 2.1969752311706543,
|
| 1532 |
-
"learning_rate": 2.5903356534622476e-05,
|
| 1533 |
-
"loss": 1.8013,
|
| 1534 |
-
"step": 2030
|
| 1535 |
-
},
|
| 1536 |
-
{
|
| 1537 |
-
"epoch": 2.6053639846743293,
|
| 1538 |
-
"grad_norm": 2.2254557609558105,
|
| 1539 |
-
"learning_rate": 2.5898736634122276e-05,
|
| 1540 |
-
"loss": 1.7312,
|
| 1541 |
-
"step": 2040
|
| 1542 |
-
},
|
| 1543 |
-
{
|
| 1544 |
-
"epoch": 2.6181353767560664,
|
| 1545 |
-
"grad_norm": 2.2725250720977783,
|
| 1546 |
-
"learning_rate": 2.5893573865965345e-05,
|
| 1547 |
-
"loss": 1.8751,
|
| 1548 |
-
"step": 2050
|
| 1549 |
-
},
|
| 1550 |
-
{
|
| 1551 |
-
"epoch": 2.630906768837803,
|
| 1552 |
-
"grad_norm": 2.1987719535827637,
|
| 1553 |
-
"learning_rate": 2.588786844677109e-05,
|
| 1554 |
-
"loss": 1.8096,
|
| 1555 |
-
"step": 2060
|
| 1556 |
-
},
|
| 1557 |
-
{
|
| 1558 |
-
"epoch": 2.6436781609195403,
|
| 1559 |
-
"grad_norm": 2.1684072017669678,
|
| 1560 |
-
"learning_rate": 2.588162061592748e-05,
|
| 1561 |
-
"loss": 1.738,
|
| 1562 |
-
"step": 2070
|
| 1563 |
-
},
|
| 1564 |
-
{
|
| 1565 |
-
"epoch": 2.656449553001277,
|
| 1566 |
-
"grad_norm": 2.2119176387786865,
|
| 1567 |
-
"learning_rate": 2.5874830635580974e-05,
|
| 1568 |
-
"loss": 1.7888,
|
| 1569 |
-
"step": 2080
|
| 1570 |
-
},
|
| 1571 |
-
{
|
| 1572 |
-
"epoch": 2.6692209450830138,
|
| 1573 |
-
"grad_norm": 2.2012903690338135,
|
| 1574 |
-
"learning_rate": 2.586749879062556e-05,
|
| 1575 |
-
"loss": 1.8119,
|
| 1576 |
-
"step": 2090
|
| 1577 |
-
},
|
| 1578 |
-
{
|
| 1579 |
-
"epoch": 2.681992337164751,
|
| 1580 |
-
"grad_norm": 2.18770694732666,
|
| 1581 |
-
"learning_rate": 2.5859625388690762e-05,
|
| 1582 |
-
"loss": 1.7867,
|
| 1583 |
-
"step": 2100
|
| 1584 |
-
},
|
| 1585 |
-
{
|
| 1586 |
-
"epoch": 2.681992337164751,
|
| 1587 |
-
"eval_loss": 1.9344319105148315,
|
| 1588 |
-
"eval_runtime": 24.3873,
|
| 1589 |
-
"eval_samples_per_second": 20.502,
|
| 1590 |
-
"eval_steps_per_second": 20.502,
|
| 1591 |
-
"step": 2100
|
| 1592 |
}
|
| 1593 |
],
|
| 1594 |
"logging_steps": 10,
|
|
@@ -1608,7 +139,7 @@
|
|
| 1608 |
"attributes": {}
|
| 1609 |
}
|
| 1610 |
},
|
| 1611 |
-
"total_flos":
|
| 1612 |
"train_batch_size": 16,
|
| 1613 |
"trial_name": null,
|
| 1614 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.9779324531555176,
|
| 3 |
+
"best_model_checkpoint": "./output/checkpoint-150",
|
| 4 |
+
"epoch": 0.19157088122605365,
|
| 5 |
"eval_steps": 150,
|
| 6 |
+
"global_step": 150,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.01277139208173691,
|
| 13 |
+
"grad_norm": 2.1240530014038086,
|
| 14 |
"learning_rate": 4.666666666666666e-06,
|
| 15 |
+
"loss": 1.9176,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.02554278416347382,
|
| 20 |
+
"grad_norm": 2.017596960067749,
|
| 21 |
"learning_rate": 9.333333333333333e-06,
|
| 22 |
+
"loss": 1.9414,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.038314176245210725,
|
| 27 |
+
"grad_norm": 2.077119827270508,
|
| 28 |
"learning_rate": 1.4e-05,
|
| 29 |
+
"loss": 1.9121,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.05108556832694764,
|
| 34 |
+
"grad_norm": 1.9572290182113647,
|
| 35 |
"learning_rate": 1.8666666666666665e-05,
|
| 36 |
+
"loss": 1.8867,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.06385696040868455,
|
| 41 |
+
"grad_norm": 2.5854249000549316,
|
| 42 |
"learning_rate": 2.333333333333333e-05,
|
| 43 |
"loss": 1.9779,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.07662835249042145,
|
| 48 |
+
"grad_norm": 2.080289840698242,
|
| 49 |
"learning_rate": 2.8e-05,
|
| 50 |
+
"loss": 1.9335,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.08939974457215837,
|
| 55 |
+
"grad_norm": 1.978735089302063,
|
| 56 |
"learning_rate": 3.266666666666666e-05,
|
| 57 |
+
"loss": 1.9452,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.10217113665389528,
|
| 62 |
+
"grad_norm": 2.1327736377716064,
|
| 63 |
"learning_rate": 3.733333333333333e-05,
|
| 64 |
+
"loss": 1.9443,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.11494252873563218,
|
| 69 |
+
"grad_norm": 2.057833194732666,
|
| 70 |
"learning_rate": 4.2e-05,
|
| 71 |
+
"loss": 1.9261,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.1277139208173691,
|
| 76 |
+
"grad_norm": 2.1436352729797363,
|
| 77 |
"learning_rate": 4.666666666666666e-05,
|
| 78 |
+
"loss": 1.9776,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.140485312899106,
|
| 83 |
+
"grad_norm": 1.885575294494629,
|
| 84 |
"learning_rate": 5.1333333333333325e-05,
|
| 85 |
"loss": 1.9233,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.1532567049808429,
|
| 90 |
+
"grad_norm": 2.046653985977173,
|
| 91 |
"learning_rate": 5.6e-05,
|
| 92 |
+
"loss": 1.9403,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.16602809706257982,
|
| 97 |
+
"grad_norm": 2.272224187850952,
|
| 98 |
"learning_rate": 6.0666666666666666e-05,
|
| 99 |
+
"loss": 1.9421,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.17879948914431673,
|
| 104 |
+
"grad_norm": 2.098900318145752,
|
| 105 |
"learning_rate": 6.533333333333333e-05,
|
| 106 |
+
"loss": 1.919,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.19157088122605365,
|
| 111 |
+
"grad_norm": 2.07694411277771,
|
| 112 |
"learning_rate": 7e-05,
|
| 113 |
"loss": 1.9405,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.19157088122605365,
|
| 118 |
+
"eval_loss": 1.9779324531555176,
|
| 119 |
+
"eval_runtime": 25.3379,
|
| 120 |
+
"eval_samples_per_second": 19.733,
|
| 121 |
+
"eval_steps_per_second": 19.733,
|
| 122 |
"step": 150
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
}
|
| 124 |
],
|
| 125 |
"logging_steps": 10,
|
|
|
|
| 139 |
"attributes": {}
|
| 140 |
}
|
| 141 |
},
|
| 142 |
+
"total_flos": 2084245000224768.0,
|
| 143 |
"train_batch_size": 16,
|
| 144 |
"trial_name": null,
|
| 145 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5d09d197d3b76d99965de547b76e3d4e9afb2509c824e5faadb664c5461d6450
|
| 3 |
size 5496
|