Training in progress, step 150, checkpoint
Browse files- last-checkpoint/adapter_config.json +3 -5
- last-checkpoint/adapter_model.safetensors +2 -2
- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/tokenizer.json +2 -2
- last-checkpoint/trainer_state.json +54 -1862
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/adapter_config.json
CHANGED
|
@@ -14,17 +14,15 @@
|
|
| 14 |
"lora_dropout": 0.1,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
-
"modules_to_save": [
|
| 18 |
-
"lm_head"
|
| 19 |
-
],
|
| 20 |
"peft_type": "LORA",
|
| 21 |
"r": 8,
|
| 22 |
"rank_pattern": {},
|
| 23 |
"revision": null,
|
| 24 |
"target_modules": [
|
| 25 |
-
"q_proj",
|
| 26 |
-
"o_proj",
|
| 27 |
"k_proj",
|
|
|
|
|
|
|
| 28 |
"v_proj"
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 14 |
"lora_dropout": 0.1,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
+
"modules_to_save": [],
|
|
|
|
|
|
|
| 18 |
"peft_type": "LORA",
|
| 19 |
"r": 8,
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"k_proj",
|
| 24 |
+
"o_proj",
|
| 25 |
+
"q_proj",
|
| 26 |
"v_proj"
|
| 27 |
],
|
| 28 |
"task_type": "CAUSAL_LM",
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:58a6dac7b8788cdf686e0493d47f426ab4e6370365936cf0127fc02865db7e27
|
| 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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7192f33aec7dae1d9eb5f8b448f192cc1be6d1e1d1fce1464dfc671d4cd435f0
|
| 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/tokenizer.json
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52716f60c3ad328509fa37cdded9a2f1196ecae463f5480f5d38c66a25e7a7dc
|
| 3 |
+
size 17210019
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1933 +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":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.02554278416347382,
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.038314176245210725,
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.05108556832694764,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.06385696040868455,
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.07662835249042145,
|
| 48 |
-
"grad_norm":
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.08939974457215837,
|
| 55 |
-
"grad_norm":
|
| 56 |
-
"learning_rate":
|
| 57 |
-
"loss": 1.
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.10217113665389528,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 1.
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.11494252873563218,
|
| 69 |
-
"grad_norm":
|
| 70 |
-
"learning_rate":
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.1277139208173691,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate":
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.140485312899106,
|
| 83 |
-
"grad_norm":
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.1532567049808429,
|
| 90 |
-
"grad_norm": 2.
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.16602809706257982,
|
| 97 |
-
"grad_norm":
|
| 98 |
-
"learning_rate":
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.17879948914431673,
|
| 104 |
-
"grad_norm":
|
| 105 |
-
"learning_rate":
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.19157088122605365,
|
| 111 |
-
"grad_norm":
|
| 112 |
-
"learning_rate":
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.19157088122605365,
|
| 118 |
-
"eval_loss": 1.
|
| 119 |
-
"eval_runtime": 24.
|
| 120 |
-
"eval_samples_per_second": 20.
|
| 121 |
-
"eval_steps_per_second": 20.
|
| 122 |
"step": 150
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.20434227330779056,
|
| 126 |
-
"grad_norm": 3.3137006759643555,
|
| 127 |
-
"learning_rate": 2.24e-05,
|
| 128 |
-
"loss": 1.8722,
|
| 129 |
-
"step": 160
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.21711366538952745,
|
| 133 |
-
"grad_norm": 3.511589288711548,
|
| 134 |
-
"learning_rate": 2.38e-05,
|
| 135 |
-
"loss": 1.897,
|
| 136 |
-
"step": 170
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.22988505747126436,
|
| 140 |
-
"grad_norm": 3.4723284244537354,
|
| 141 |
-
"learning_rate": 2.5199999999999996e-05,
|
| 142 |
-
"loss": 1.8827,
|
| 143 |
-
"step": 180
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.24265644955300128,
|
| 147 |
-
"grad_norm": 3.296243667602539,
|
| 148 |
-
"learning_rate": 2.66e-05,
|
| 149 |
-
"loss": 1.9298,
|
| 150 |
-
"step": 190
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.2554278416347382,
|
| 154 |
-
"grad_norm": 3.188915491104126,
|
| 155 |
-
"learning_rate": 2.8e-05,
|
| 156 |
-
"loss": 1.9129,
|
| 157 |
-
"step": 200
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.2681992337164751,
|
| 161 |
-
"grad_norm": 3.2900846004486084,
|
| 162 |
-
"learning_rate": 2.9399999999999996e-05,
|
| 163 |
-
"loss": 1.8282,
|
| 164 |
-
"step": 210
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.280970625798212,
|
| 168 |
-
"grad_norm": 3.3161473274230957,
|
| 169 |
-
"learning_rate": 3.0799999999999996e-05,
|
| 170 |
-
"loss": 1.9102,
|
| 171 |
-
"step": 220
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.2937420178799489,
|
| 175 |
-
"grad_norm": 3.259005546569824,
|
| 176 |
-
"learning_rate": 3.22e-05,
|
| 177 |
-
"loss": 1.948,
|
| 178 |
-
"step": 230
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.3065134099616858,
|
| 182 |
-
"grad_norm": 3.051255941390991,
|
| 183 |
-
"learning_rate": 3.36e-05,
|
| 184 |
-
"loss": 1.8954,
|
| 185 |
-
"step": 240
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.31928480204342274,
|
| 189 |
-
"grad_norm": 3.3548150062561035,
|
| 190 |
-
"learning_rate": 3.5e-05,
|
| 191 |
-
"loss": 1.9701,
|
| 192 |
-
"step": 250
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.33205619412515963,
|
| 196 |
-
"grad_norm": 2.9780640602111816,
|
| 197 |
-
"learning_rate": 3.64e-05,
|
| 198 |
-
"loss": 1.908,
|
| 199 |
-
"step": 260
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.3448275862068966,
|
| 203 |
-
"grad_norm": 3.386885404586792,
|
| 204 |
-
"learning_rate": 3.78e-05,
|
| 205 |
-
"loss": 1.9511,
|
| 206 |
-
"step": 270
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.35759897828863346,
|
| 210 |
-
"grad_norm": 3.263582229614258,
|
| 211 |
-
"learning_rate": 3.92e-05,
|
| 212 |
-
"loss": 1.9289,
|
| 213 |
-
"step": 280
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.37037037037037035,
|
| 217 |
-
"grad_norm": 3.1607635021209717,
|
| 218 |
-
"learning_rate": 4.059999999999999e-05,
|
| 219 |
-
"loss": 1.9594,
|
| 220 |
-
"step": 290
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.3831417624521073,
|
| 224 |
-
"grad_norm": 3.176176071166992,
|
| 225 |
-
"learning_rate": 4.2e-05,
|
| 226 |
-
"loss": 1.9399,
|
| 227 |
-
"step": 300
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.3831417624521073,
|
| 231 |
-
"eval_loss": 1.9805161952972412,
|
| 232 |
-
"eval_runtime": 28.0922,
|
| 233 |
-
"eval_samples_per_second": 17.799,
|
| 234 |
-
"eval_steps_per_second": 17.799,
|
| 235 |
-
"step": 300
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.3959131545338442,
|
| 239 |
-
"grad_norm": 3.4425015449523926,
|
| 240 |
-
"learning_rate": 4.34e-05,
|
| 241 |
-
"loss": 1.8902,
|
| 242 |
-
"step": 310
|
| 243 |
-
},
|
| 244 |
-
{
|
| 245 |
-
"epoch": 0.4086845466155811,
|
| 246 |
-
"grad_norm": 3.13578724861145,
|
| 247 |
-
"learning_rate": 4.48e-05,
|
| 248 |
-
"loss": 1.9275,
|
| 249 |
-
"step": 320
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.421455938697318,
|
| 253 |
-
"grad_norm": 3.6584582328796387,
|
| 254 |
-
"learning_rate": 4.62e-05,
|
| 255 |
-
"loss": 1.9625,
|
| 256 |
-
"step": 330
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 0.4342273307790549,
|
| 260 |
-
"grad_norm": 3.0934386253356934,
|
| 261 |
-
"learning_rate": 4.76e-05,
|
| 262 |
-
"loss": 1.882,
|
| 263 |
-
"step": 340
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 0.44699872286079184,
|
| 267 |
-
"grad_norm": 3.1788082122802734,
|
| 268 |
-
"learning_rate": 4.899999999999999e-05,
|
| 269 |
-
"loss": 1.8866,
|
| 270 |
-
"step": 350
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"epoch": 0.45977011494252873,
|
| 274 |
-
"grad_norm": 3.1974220275878906,
|
| 275 |
-
"learning_rate": 5.039999999999999e-05,
|
| 276 |
-
"loss": 1.9472,
|
| 277 |
-
"step": 360
|
| 278 |
-
},
|
| 279 |
-
{
|
| 280 |
-
"epoch": 0.4725415070242657,
|
| 281 |
-
"grad_norm": 3.3801984786987305,
|
| 282 |
-
"learning_rate": 5.179999999999999e-05,
|
| 283 |
-
"loss": 1.9249,
|
| 284 |
-
"step": 370
|
| 285 |
-
},
|
| 286 |
-
{
|
| 287 |
-
"epoch": 0.48531289910600256,
|
| 288 |
-
"grad_norm": 3.156172037124634,
|
| 289 |
-
"learning_rate": 5.32e-05,
|
| 290 |
-
"loss": 1.9025,
|
| 291 |
-
"step": 380
|
| 292 |
-
},
|
| 293 |
-
{
|
| 294 |
-
"epoch": 0.49808429118773945,
|
| 295 |
-
"grad_norm": 3.3196117877960205,
|
| 296 |
-
"learning_rate": 5.46e-05,
|
| 297 |
-
"loss": 1.8319,
|
| 298 |
-
"step": 390
|
| 299 |
-
},
|
| 300 |
-
{
|
| 301 |
-
"epoch": 0.5108556832694764,
|
| 302 |
-
"grad_norm": 3.410414457321167,
|
| 303 |
-
"learning_rate": 5.6e-05,
|
| 304 |
-
"loss": 1.9329,
|
| 305 |
-
"step": 400
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 0.5236270753512133,
|
| 309 |
-
"grad_norm": 3.050872325897217,
|
| 310 |
-
"learning_rate": 5.739999999999999e-05,
|
| 311 |
-
"loss": 1.936,
|
| 312 |
-
"step": 410
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"epoch": 0.5363984674329502,
|
| 316 |
-
"grad_norm": 3.5527184009552,
|
| 317 |
-
"learning_rate": 5.879999999999999e-05,
|
| 318 |
-
"loss": 1.9004,
|
| 319 |
-
"step": 420
|
| 320 |
-
},
|
| 321 |
-
{
|
| 322 |
-
"epoch": 0.5491698595146871,
|
| 323 |
-
"grad_norm": 3.099611282348633,
|
| 324 |
-
"learning_rate": 6.019999999999999e-05,
|
| 325 |
-
"loss": 1.9061,
|
| 326 |
-
"step": 430
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 0.561941251596424,
|
| 330 |
-
"grad_norm": 3.0915510654449463,
|
| 331 |
-
"learning_rate": 6.159999999999999e-05,
|
| 332 |
-
"loss": 1.9015,
|
| 333 |
-
"step": 440
|
| 334 |
-
},
|
| 335 |
-
{
|
| 336 |
-
"epoch": 0.5747126436781609,
|
| 337 |
-
"grad_norm": 3.2416725158691406,
|
| 338 |
-
"learning_rate": 6.3e-05,
|
| 339 |
-
"loss": 1.8938,
|
| 340 |
-
"step": 450
|
| 341 |
-
},
|
| 342 |
-
{
|
| 343 |
-
"epoch": 0.5747126436781609,
|
| 344 |
-
"eval_loss": 1.970125675201416,
|
| 345 |
-
"eval_runtime": 24.7662,
|
| 346 |
-
"eval_samples_per_second": 20.189,
|
| 347 |
-
"eval_steps_per_second": 20.189,
|
| 348 |
-
"step": 450
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 0.5874840357598978,
|
| 352 |
-
"grad_norm": 2.9722657203674316,
|
| 353 |
-
"learning_rate": 6.44e-05,
|
| 354 |
-
"loss": 1.869,
|
| 355 |
-
"step": 460
|
| 356 |
-
},
|
| 357 |
-
{
|
| 358 |
-
"epoch": 0.6002554278416348,
|
| 359 |
-
"grad_norm": 3.058877468109131,
|
| 360 |
-
"learning_rate": 6.579999999999999e-05,
|
| 361 |
-
"loss": 1.856,
|
| 362 |
-
"step": 470
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 0.6130268199233716,
|
| 366 |
-
"grad_norm": 3.3449816703796387,
|
| 367 |
-
"learning_rate": 6.72e-05,
|
| 368 |
-
"loss": 1.8573,
|
| 369 |
-
"step": 480
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.6257982120051085,
|
| 373 |
-
"grad_norm": 3.0998528003692627,
|
| 374 |
-
"learning_rate": 6.859999999999999e-05,
|
| 375 |
-
"loss": 1.8721,
|
| 376 |
-
"step": 490
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 0.6385696040868455,
|
| 380 |
-
"grad_norm": 3.2949531078338623,
|
| 381 |
-
"learning_rate": 7e-05,
|
| 382 |
-
"loss": 1.9653,
|
| 383 |
-
"step": 500
|
| 384 |
-
},
|
| 385 |
-
{
|
| 386 |
-
"epoch": 0.6513409961685823,
|
| 387 |
-
"grad_norm": 2.965726137161255,
|
| 388 |
-
"learning_rate": 6.99991470746888e-05,
|
| 389 |
-
"loss": 1.8915,
|
| 390 |
-
"step": 510
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 0.6641123882503193,
|
| 394 |
-
"grad_norm": 3.32828950881958,
|
| 395 |
-
"learning_rate": 6.999658834032565e-05,
|
| 396 |
-
"loss": 1.8694,
|
| 397 |
-
"step": 520
|
| 398 |
-
},
|
| 399 |
-
{
|
| 400 |
-
"epoch": 0.6768837803320562,
|
| 401 |
-
"grad_norm": 3.0084595680236816,
|
| 402 |
-
"learning_rate": 6.999232392161959e-05,
|
| 403 |
-
"loss": 1.8969,
|
| 404 |
-
"step": 530
|
| 405 |
-
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 0.6896551724137931,
|
| 408 |
-
"grad_norm": 3.276498556137085,
|
| 409 |
-
"learning_rate": 6.99863540264124e-05,
|
| 410 |
-
"loss": 1.9219,
|
| 411 |
-
"step": 540
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 0.70242656449553,
|
| 415 |
-
"grad_norm": 3.205116033554077,
|
| 416 |
-
"learning_rate": 6.997867894566835e-05,
|
| 417 |
-
"loss": 1.9566,
|
| 418 |
-
"step": 550
|
| 419 |
-
},
|
| 420 |
-
{
|
| 421 |
-
"epoch": 0.7151979565772669,
|
| 422 |
-
"grad_norm": 3.1132071018218994,
|
| 423 |
-
"learning_rate": 6.996929905346004e-05,
|
| 424 |
-
"loss": 1.9083,
|
| 425 |
-
"step": 560
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 0.7279693486590039,
|
| 429 |
-
"grad_norm": 2.9555888175964355,
|
| 430 |
-
"learning_rate": 6.995821480695019e-05,
|
| 431 |
-
"loss": 1.8563,
|
| 432 |
-
"step": 570
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 0.7407407407407407,
|
| 436 |
-
"grad_norm": 3.2956736087799072,
|
| 437 |
-
"learning_rate": 6.994542674636933e-05,
|
| 438 |
-
"loss": 1.8805,
|
| 439 |
-
"step": 580
|
| 440 |
-
},
|
| 441 |
-
{
|
| 442 |
-
"epoch": 0.7535121328224776,
|
| 443 |
-
"grad_norm": 3.101565361022949,
|
| 444 |
-
"learning_rate": 6.99309354949895e-05,
|
| 445 |
-
"loss": 1.8884,
|
| 446 |
-
"step": 590
|
| 447 |
-
},
|
| 448 |
-
{
|
| 449 |
-
"epoch": 0.7662835249042146,
|
| 450 |
-
"grad_norm": 3.5506794452667236,
|
| 451 |
-
"learning_rate": 6.991474175909384e-05,
|
| 452 |
-
"loss": 1.9324,
|
| 453 |
-
"step": 600
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 0.7662835249042146,
|
| 457 |
-
"eval_loss": 1.9721698760986328,
|
| 458 |
-
"eval_runtime": 27.9596,
|
| 459 |
-
"eval_samples_per_second": 17.883,
|
| 460 |
-
"eval_steps_per_second": 17.883,
|
| 461 |
-
"step": 600
|
| 462 |
-
},
|
| 463 |
-
{
|
| 464 |
-
"epoch": 0.7790549169859514,
|
| 465 |
-
"grad_norm": 3.1406140327453613,
|
| 466 |
-
"learning_rate": 6.989684632794221e-05,
|
| 467 |
-
"loss": 1.8976,
|
| 468 |
-
"step": 610
|
| 469 |
-
},
|
| 470 |
-
{
|
| 471 |
-
"epoch": 0.7918263090676884,
|
| 472 |
-
"grad_norm": 3.364961624145508,
|
| 473 |
-
"learning_rate": 6.987725007373265e-05,
|
| 474 |
-
"loss": 1.8477,
|
| 475 |
-
"step": 620
|
| 476 |
-
},
|
| 477 |
-
{
|
| 478 |
-
"epoch": 0.8045977011494253,
|
| 479 |
-
"grad_norm": 2.9893288612365723,
|
| 480 |
-
"learning_rate": 6.985595395155896e-05,
|
| 481 |
-
"loss": 1.9148,
|
| 482 |
-
"step": 630
|
| 483 |
-
},
|
| 484 |
-
{
|
| 485 |
-
"epoch": 0.8173690932311622,
|
| 486 |
-
"grad_norm": 3.107212781906128,
|
| 487 |
-
"learning_rate": 6.983295899936408e-05,
|
| 488 |
-
"loss": 1.8768,
|
| 489 |
-
"step": 640
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.8301404853128991,
|
| 493 |
-
"grad_norm": 3.053358554840088,
|
| 494 |
-
"learning_rate": 6.980826633788956e-05,
|
| 495 |
-
"loss": 1.9073,
|
| 496 |
-
"step": 650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 0.842911877394636,
|
| 500 |
-
"grad_norm": 2.9061360359191895,
|
| 501 |
-
"learning_rate": 6.978187717062086e-05,
|
| 502 |
-
"loss": 1.8477,
|
| 503 |
-
"step": 660
|
| 504 |
-
},
|
| 505 |
-
{
|
| 506 |
-
"epoch": 0.855683269476373,
|
| 507 |
-
"grad_norm": 2.9826648235321045,
|
| 508 |
-
"learning_rate": 6.975379278372878e-05,
|
| 509 |
-
"loss": 1.8508,
|
| 510 |
-
"step": 670
|
| 511 |
-
},
|
| 512 |
-
{
|
| 513 |
-
"epoch": 0.8684546615581098,
|
| 514 |
-
"grad_norm": 3.2406656742095947,
|
| 515 |
-
"learning_rate": 6.972401454600672e-05,
|
| 516 |
-
"loss": 1.9066,
|
| 517 |
-
"step": 680
|
| 518 |
-
},
|
| 519 |
-
{
|
| 520 |
-
"epoch": 0.8812260536398467,
|
| 521 |
-
"grad_norm": 3.206275463104248,
|
| 522 |
-
"learning_rate": 6.969254390880395e-05,
|
| 523 |
-
"loss": 1.9108,
|
| 524 |
-
"step": 690
|
| 525 |
-
},
|
| 526 |
-
{
|
| 527 |
-
"epoch": 0.8939974457215837,
|
| 528 |
-
"grad_norm": 3.2281596660614014,
|
| 529 |
-
"learning_rate": 6.965938240595496e-05,
|
| 530 |
-
"loss": 1.9246,
|
| 531 |
-
"step": 700
|
| 532 |
-
},
|
| 533 |
-
{
|
| 534 |
-
"epoch": 0.9067688378033205,
|
| 535 |
-
"grad_norm": 3.3281443119049072,
|
| 536 |
-
"learning_rate": 6.962453165370459e-05,
|
| 537 |
-
"loss": 1.9756,
|
| 538 |
-
"step": 710
|
| 539 |
-
},
|
| 540 |
-
{
|
| 541 |
-
"epoch": 0.9195402298850575,
|
| 542 |
-
"grad_norm": 3.3063316345214844,
|
| 543 |
-
"learning_rate": 6.958799335062934e-05,
|
| 544 |
-
"loss": 1.8813,
|
| 545 |
-
"step": 720
|
| 546 |
-
},
|
| 547 |
-
{
|
| 548 |
-
"epoch": 0.9323116219667944,
|
| 549 |
-
"grad_norm": 2.862231969833374,
|
| 550 |
-
"learning_rate": 6.954976927755458e-05,
|
| 551 |
-
"loss": 1.8586,
|
| 552 |
-
"step": 730
|
| 553 |
-
},
|
| 554 |
-
{
|
| 555 |
-
"epoch": 0.9450830140485313,
|
| 556 |
-
"grad_norm": 3.3289144039154053,
|
| 557 |
-
"learning_rate": 6.950986129746767e-05,
|
| 558 |
-
"loss": 1.8932,
|
| 559 |
-
"step": 740
|
| 560 |
-
},
|
| 561 |
-
{
|
| 562 |
-
"epoch": 0.9578544061302682,
|
| 563 |
-
"grad_norm": 2.8503053188323975,
|
| 564 |
-
"learning_rate": 6.946827135542728e-05,
|
| 565 |
-
"loss": 1.8778,
|
| 566 |
-
"step": 750
|
| 567 |
-
},
|
| 568 |
-
{
|
| 569 |
-
"epoch": 0.9578544061302682,
|
| 570 |
-
"eval_loss": 1.9627635478973389,
|
| 571 |
-
"eval_runtime": 28.7652,
|
| 572 |
-
"eval_samples_per_second": 17.382,
|
| 573 |
-
"eval_steps_per_second": 17.382,
|
| 574 |
-
"step": 750
|
| 575 |
-
},
|
| 576 |
-
{
|
| 577 |
-
"epoch": 0.9706257982120051,
|
| 578 |
-
"grad_norm": 3.0852253437042236,
|
| 579 |
-
"learning_rate": 6.94250014784685e-05,
|
| 580 |
-
"loss": 1.9039,
|
| 581 |
-
"step": 760
|
| 582 |
-
},
|
| 583 |
-
{
|
| 584 |
-
"epoch": 0.9833971902937421,
|
| 585 |
-
"grad_norm": 3.4453940391540527,
|
| 586 |
-
"learning_rate": 6.93800537755041e-05,
|
| 587 |
-
"loss": 1.8881,
|
| 588 |
-
"step": 770
|
| 589 |
-
},
|
| 590 |
-
{
|
| 591 |
-
"epoch": 0.9961685823754789,
|
| 592 |
-
"grad_norm": 2.987304925918579,
|
| 593 |
-
"learning_rate": 6.93334304372217e-05,
|
| 594 |
-
"loss": 1.8828,
|
| 595 |
-
"step": 780
|
| 596 |
-
},
|
| 597 |
-
{
|
| 598 |
-
"epoch": 1.0089399744572158,
|
| 599 |
-
"grad_norm": 2.9873013496398926,
|
| 600 |
-
"learning_rate": 6.928513373597703e-05,
|
| 601 |
-
"loss": 1.8152,
|
| 602 |
-
"step": 790
|
| 603 |
-
},
|
| 604 |
-
{
|
| 605 |
-
"epoch": 1.0217113665389528,
|
| 606 |
-
"grad_norm": 3.0838310718536377,
|
| 607 |
-
"learning_rate": 6.923516602568319e-05,
|
| 608 |
-
"loss": 1.7415,
|
| 609 |
-
"step": 800
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 1.0344827586206897,
|
| 613 |
-
"grad_norm": 3.1033427715301514,
|
| 614 |
-
"learning_rate": 6.918352974169587e-05,
|
| 615 |
-
"loss": 1.7952,
|
| 616 |
-
"step": 810
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 1.0472541507024267,
|
| 620 |
-
"grad_norm": 3.1028897762298584,
|
| 621 |
-
"learning_rate": 6.913022740069471e-05,
|
| 622 |
-
"loss": 1.7518,
|
| 623 |
-
"step": 820
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 1.0600255427841634,
|
| 627 |
-
"grad_norm": 3.008892297744751,
|
| 628 |
-
"learning_rate": 6.90752616005606e-05,
|
| 629 |
-
"loss": 1.7723,
|
| 630 |
-
"step": 830
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 1.0727969348659003,
|
| 634 |
-
"grad_norm": 3.101642370223999,
|
| 635 |
-
"learning_rate": 6.901863502024912e-05,
|
| 636 |
-
"loss": 1.7136,
|
| 637 |
-
"step": 840
|
| 638 |
-
},
|
| 639 |
-
{
|
| 640 |
-
"epoch": 1.0855683269476373,
|
| 641 |
-
"grad_norm": 3.0864851474761963,
|
| 642 |
-
"learning_rate": 6.896035041965987e-05,
|
| 643 |
-
"loss": 1.7585,
|
| 644 |
-
"step": 850
|
| 645 |
-
},
|
| 646 |
-
{
|
| 647 |
-
"epoch": 1.0983397190293742,
|
| 648 |
-
"grad_norm": 3.0635933876037598,
|
| 649 |
-
"learning_rate": 6.890041063950208e-05,
|
| 650 |
-
"loss": 1.7557,
|
| 651 |
-
"step": 860
|
| 652 |
-
},
|
| 653 |
-
{
|
| 654 |
-
"epoch": 1.1111111111111112,
|
| 655 |
-
"grad_norm": 3.0961105823516846,
|
| 656 |
-
"learning_rate": 6.883881860115608e-05,
|
| 657 |
-
"loss": 1.7491,
|
| 658 |
-
"step": 870
|
| 659 |
-
},
|
| 660 |
-
{
|
| 661 |
-
"epoch": 1.123882503192848,
|
| 662 |
-
"grad_norm": 3.00886607170105,
|
| 663 |
-
"learning_rate": 6.87755773065309e-05,
|
| 664 |
-
"loss": 1.7299,
|
| 665 |
-
"step": 880
|
| 666 |
-
},
|
| 667 |
-
{
|
| 668 |
-
"epoch": 1.136653895274585,
|
| 669 |
-
"grad_norm": 2.9253010749816895,
|
| 670 |
-
"learning_rate": 6.871068983791803e-05,
|
| 671 |
-
"loss": 1.7236,
|
| 672 |
-
"step": 890
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 1.1494252873563218,
|
| 676 |
-
"grad_norm": 3.2435286045074463,
|
| 677 |
-
"learning_rate": 6.864415935784116e-05,
|
| 678 |
-
"loss": 1.7262,
|
| 679 |
-
"step": 900
|
| 680 |
-
},
|
| 681 |
-
{
|
| 682 |
-
"epoch": 1.1494252873563218,
|
| 683 |
-
"eval_loss": 1.9637728929519653,
|
| 684 |
-
"eval_runtime": 26.2765,
|
| 685 |
-
"eval_samples_per_second": 19.028,
|
| 686 |
-
"eval_steps_per_second": 19.028,
|
| 687 |
-
"step": 900
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"epoch": 1.1621966794380587,
|
| 691 |
-
"grad_norm": 2.878432512283325,
|
| 692 |
-
"learning_rate": 6.8575989108902e-05,
|
| 693 |
-
"loss": 1.7455,
|
| 694 |
-
"step": 910
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"epoch": 1.1749680715197957,
|
| 698 |
-
"grad_norm": 2.9958291053771973,
|
| 699 |
-
"learning_rate": 6.850618241362235e-05,
|
| 700 |
-
"loss": 1.7218,
|
| 701 |
-
"step": 920
|
| 702 |
-
},
|
| 703 |
-
{
|
| 704 |
-
"epoch": 1.1877394636015326,
|
| 705 |
-
"grad_norm": 3.107431173324585,
|
| 706 |
-
"learning_rate": 6.843474267428202e-05,
|
| 707 |
-
"loss": 1.7237,
|
| 708 |
-
"step": 930
|
| 709 |
-
},
|
| 710 |
-
{
|
| 711 |
-
"epoch": 1.2005108556832695,
|
| 712 |
-
"grad_norm": 3.1416854858398438,
|
| 713 |
-
"learning_rate": 6.836167337275314e-05,
|
| 714 |
-
"loss": 1.7539,
|
| 715 |
-
"step": 940
|
| 716 |
-
},
|
| 717 |
-
{
|
| 718 |
-
"epoch": 1.2132822477650063,
|
| 719 |
-
"grad_norm": 3.1158761978149414,
|
| 720 |
-
"learning_rate": 6.828697807033038e-05,
|
| 721 |
-
"loss": 1.7637,
|
| 722 |
-
"step": 950
|
| 723 |
-
},
|
| 724 |
-
{
|
| 725 |
-
"epoch": 1.2260536398467432,
|
| 726 |
-
"grad_norm": 2.7915549278259277,
|
| 727 |
-
"learning_rate": 6.821066040755737e-05,
|
| 728 |
-
"loss": 1.688,
|
| 729 |
-
"step": 960
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 1.2388250319284801,
|
| 733 |
-
"grad_norm": 3.083923816680908,
|
| 734 |
-
"learning_rate": 6.813272410404936e-05,
|
| 735 |
-
"loss": 1.7765,
|
| 736 |
-
"step": 970
|
| 737 |
-
},
|
| 738 |
-
{
|
| 739 |
-
"epoch": 1.251596424010217,
|
| 740 |
-
"grad_norm": 3.111654281616211,
|
| 741 |
-
"learning_rate": 6.805317295831182e-05,
|
| 742 |
-
"loss": 1.7157,
|
| 743 |
-
"step": 980
|
| 744 |
-
},
|
| 745 |
-
{
|
| 746 |
-
"epoch": 1.264367816091954,
|
| 747 |
-
"grad_norm": 2.9396989345550537,
|
| 748 |
-
"learning_rate": 6.797201084755538e-05,
|
| 749 |
-
"loss": 1.7147,
|
| 750 |
-
"step": 990
|
| 751 |
-
},
|
| 752 |
-
{
|
| 753 |
-
"epoch": 1.277139208173691,
|
| 754 |
-
"grad_norm": 3.1540298461914062,
|
| 755 |
-
"learning_rate": 6.788924172750679e-05,
|
| 756 |
-
"loss": 1.755,
|
| 757 |
-
"step": 1000
|
| 758 |
-
},
|
| 759 |
-
{
|
| 760 |
-
"epoch": 1.289910600255428,
|
| 761 |
-
"grad_norm": 3.1720669269561768,
|
| 762 |
-
"learning_rate": 6.78048696322162e-05,
|
| 763 |
-
"loss": 1.8097,
|
| 764 |
-
"step": 1010
|
| 765 |
-
},
|
| 766 |
-
{
|
| 767 |
-
"epoch": 1.3026819923371646,
|
| 768 |
-
"grad_norm": 3.1514675617218018,
|
| 769 |
-
"learning_rate": 6.77188986738605e-05,
|
| 770 |
-
"loss": 1.7045,
|
| 771 |
-
"step": 1020
|
| 772 |
-
},
|
| 773 |
-
{
|
| 774 |
-
"epoch": 1.3154533844189016,
|
| 775 |
-
"grad_norm": 2.926434278488159,
|
| 776 |
-
"learning_rate": 6.763133304254292e-05,
|
| 777 |
-
"loss": 1.7047,
|
| 778 |
-
"step": 1030
|
| 779 |
-
},
|
| 780 |
-
{
|
| 781 |
-
"epoch": 1.3282247765006385,
|
| 782 |
-
"grad_norm": 3.011573314666748,
|
| 783 |
-
"learning_rate": 6.75421770060888e-05,
|
| 784 |
-
"loss": 1.7246,
|
| 785 |
-
"step": 1040
|
| 786 |
-
},
|
| 787 |
-
{
|
| 788 |
-
"epoch": 1.3409961685823755,
|
| 789 |
-
"grad_norm": 3.360208749771118,
|
| 790 |
-
"learning_rate": 6.745143490983756e-05,
|
| 791 |
-
"loss": 1.823,
|
| 792 |
-
"step": 1050
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 1.3409961685823755,
|
| 796 |
-
"eval_loss": 1.9604660272598267,
|
| 797 |
-
"eval_runtime": 28.9496,
|
| 798 |
-
"eval_samples_per_second": 17.271,
|
| 799 |
-
"eval_steps_per_second": 17.271,
|
| 800 |
-
"step": 1050
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 1.3537675606641124,
|
| 804 |
-
"grad_norm": 3.0648999214172363,
|
| 805 |
-
"learning_rate": 6.735911117643095e-05,
|
| 806 |
-
"loss": 1.7675,
|
| 807 |
-
"step": 1060
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 1.3665389527458494,
|
| 811 |
-
"grad_norm": 2.986990451812744,
|
| 812 |
-
"learning_rate": 6.726521030559751e-05,
|
| 813 |
-
"loss": 1.7806,
|
| 814 |
-
"step": 1070
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 1.3793103448275863,
|
| 818 |
-
"grad_norm": 2.9186227321624756,
|
| 819 |
-
"learning_rate": 6.71697368739332e-05,
|
| 820 |
-
"loss": 1.7838,
|
| 821 |
-
"step": 1080
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 1.392081736909323,
|
| 825 |
-
"grad_norm": 2.8783702850341797,
|
| 826 |
-
"learning_rate": 6.707269553467838e-05,
|
| 827 |
-
"loss": 1.7621,
|
| 828 |
-
"step": 1090
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 1.40485312899106,
|
| 832 |
-
"grad_norm": 3.190941333770752,
|
| 833 |
-
"learning_rate": 6.697409101749102e-05,
|
| 834 |
-
"loss": 1.7883,
|
| 835 |
-
"step": 1100
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 1.417624521072797,
|
| 839 |
-
"grad_norm": 3.060194969177246,
|
| 840 |
-
"learning_rate": 6.687392812821619e-05,
|
| 841 |
-
"loss": 1.7676,
|
| 842 |
-
"step": 1110
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 1.4303959131545338,
|
| 846 |
-
"grad_norm": 3.0321028232574463,
|
| 847 |
-
"learning_rate": 6.677221174865179e-05,
|
| 848 |
-
"loss": 1.7914,
|
| 849 |
-
"step": 1120
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 1.4431673052362708,
|
| 853 |
-
"grad_norm": 2.939692497253418,
|
| 854 |
-
"learning_rate": 6.666894683631068e-05,
|
| 855 |
-
"loss": 1.7739,
|
| 856 |
-
"step": 1130
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 1.4559386973180077,
|
| 860 |
-
"grad_norm": 2.909308433532715,
|
| 861 |
-
"learning_rate": 6.656413842417897e-05,
|
| 862 |
-
"loss": 1.7782,
|
| 863 |
-
"step": 1140
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 1.4687100893997447,
|
| 867 |
-
"grad_norm": 3.142610788345337,
|
| 868 |
-
"learning_rate": 6.645779162047084e-05,
|
| 869 |
-
"loss": 1.7468,
|
| 870 |
-
"step": 1150
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 1.4814814814814814,
|
| 874 |
-
"grad_norm": 3.127857208251953,
|
| 875 |
-
"learning_rate": 6.634991160837945e-05,
|
| 876 |
-
"loss": 1.8058,
|
| 877 |
-
"step": 1160
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 1.4942528735632183,
|
| 881 |
-
"grad_norm": 2.9502968788146973,
|
| 882 |
-
"learning_rate": 6.624050364582439e-05,
|
| 883 |
-
"loss": 1.7747,
|
| 884 |
-
"step": 1170
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 1.5070242656449553,
|
| 888 |
-
"grad_norm": 3.0171525478363037,
|
| 889 |
-
"learning_rate": 6.612957306519541e-05,
|
| 890 |
-
"loss": 1.8112,
|
| 891 |
-
"step": 1180
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 1.5197956577266922,
|
| 895 |
-
"grad_norm": 3.0701234340667725,
|
| 896 |
-
"learning_rate": 6.60171252730925e-05,
|
| 897 |
-
"loss": 1.7723,
|
| 898 |
-
"step": 1190
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 1.5325670498084292,
|
| 902 |
-
"grad_norm": 3.009382486343384,
|
| 903 |
-
"learning_rate": 6.590316575006243e-05,
|
| 904 |
-
"loss": 1.815,
|
| 905 |
-
"step": 1200
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 1.5325670498084292,
|
| 909 |
-
"eval_loss": 1.9615885019302368,
|
| 910 |
-
"eval_runtime": 27.2206,
|
| 911 |
-
"eval_samples_per_second": 18.368,
|
| 912 |
-
"eval_steps_per_second": 18.368,
|
| 913 |
-
"step": 1200
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 1.545338441890166,
|
| 917 |
-
"grad_norm": 2.813608407974243,
|
| 918 |
-
"learning_rate": 6.578770005033157e-05,
|
| 919 |
-
"loss": 1.7336,
|
| 920 |
-
"step": 1210
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 1.558109833971903,
|
| 924 |
-
"grad_norm": 3.030839443206787,
|
| 925 |
-
"learning_rate": 6.567073380153521e-05,
|
| 926 |
-
"loss": 1.7409,
|
| 927 |
-
"step": 1220
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 1.5708812260536398,
|
| 931 |
-
"grad_norm": 2.8138840198516846,
|
| 932 |
-
"learning_rate": 6.555227270444334e-05,
|
| 933 |
-
"loss": 1.7933,
|
| 934 |
-
"step": 1230
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 1.5836526181353767,
|
| 938 |
-
"grad_norm": 2.8430979251861572,
|
| 939 |
-
"learning_rate": 6.543232253268266e-05,
|
| 940 |
-
"loss": 1.7454,
|
| 941 |
-
"step": 1240
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 1.5964240102171137,
|
| 945 |
-
"grad_norm": 3.018444538116455,
|
| 946 |
-
"learning_rate": 6.531088913245535e-05,
|
| 947 |
-
"loss": 1.7844,
|
| 948 |
-
"step": 1250
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 1.6091954022988506,
|
| 952 |
-
"grad_norm": 2.859196662902832,
|
| 953 |
-
"learning_rate": 6.518797842225401e-05,
|
| 954 |
-
"loss": 1.7814,
|
| 955 |
-
"step": 1260
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 1.6219667943805876,
|
| 959 |
-
"grad_norm": 2.9246723651885986,
|
| 960 |
-
"learning_rate": 6.506359639257325e-05,
|
| 961 |
-
"loss": 1.7289,
|
| 962 |
-
"step": 1270
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 1.6347381864623243,
|
| 966 |
-
"grad_norm": 3.0037896633148193,
|
| 967 |
-
"learning_rate": 6.493774910561772e-05,
|
| 968 |
-
"loss": 1.7259,
|
| 969 |
-
"step": 1280
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 1.6475095785440614,
|
| 973 |
-
"grad_norm": 3.0603103637695312,
|
| 974 |
-
"learning_rate": 6.481044269500665e-05,
|
| 975 |
-
"loss": 1.6821,
|
| 976 |
-
"step": 1290
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 1.6602809706257982,
|
| 980 |
-
"grad_norm": 3.065397262573242,
|
| 981 |
-
"learning_rate": 6.46816833654749e-05,
|
| 982 |
-
"loss": 1.8109,
|
| 983 |
-
"step": 1300
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 1.673052362707535,
|
| 987 |
-
"grad_norm": 3.0101401805877686,
|
| 988 |
-
"learning_rate": 6.455147739257053e-05,
|
| 989 |
-
"loss": 1.7583,
|
| 990 |
-
"step": 1310
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 1.685823754789272,
|
| 994 |
-
"grad_norm": 3.0239782333374023,
|
| 995 |
-
"learning_rate": 6.441983112234894e-05,
|
| 996 |
-
"loss": 1.759,
|
| 997 |
-
"step": 1320
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 1.698595146871009,
|
| 1001 |
-
"grad_norm": 3.068173408508301,
|
| 1002 |
-
"learning_rate": 6.428675097106366e-05,
|
| 1003 |
-
"loss": 1.7957,
|
| 1004 |
-
"step": 1330
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 1.711366538952746,
|
| 1008 |
-
"grad_norm": 3.2432665824890137,
|
| 1009 |
-
"learning_rate": 6.415224342485348e-05,
|
| 1010 |
-
"loss": 1.7558,
|
| 1011 |
-
"step": 1340
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 1.7241379310344827,
|
| 1015 |
-
"grad_norm": 3.144869327545166,
|
| 1016 |
-
"learning_rate": 6.401631503942645e-05,
|
| 1017 |
-
"loss": 1.7613,
|
| 1018 |
-
"step": 1350
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 1.7241379310344827,
|
| 1022 |
-
"eval_loss": 1.9588252305984497,
|
| 1023 |
-
"eval_runtime": 27.7722,
|
| 1024 |
-
"eval_samples_per_second": 18.004,
|
| 1025 |
-
"eval_steps_per_second": 18.004,
|
| 1026 |
-
"step": 1350
|
| 1027 |
-
},
|
| 1028 |
-
{
|
| 1029 |
-
"epoch": 1.7369093231162198,
|
| 1030 |
-
"grad_norm": 2.8430140018463135,
|
| 1031 |
-
"learning_rate": 6.387897243974032e-05,
|
| 1032 |
-
"loss": 1.767,
|
| 1033 |
-
"step": 1360
|
| 1034 |
-
},
|
| 1035 |
-
{
|
| 1036 |
-
"epoch": 1.7496807151979565,
|
| 1037 |
-
"grad_norm": 3.0153555870056152,
|
| 1038 |
-
"learning_rate": 6.374022231967963e-05,
|
| 1039 |
-
"loss": 1.7671,
|
| 1040 |
-
"step": 1370
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 1.7624521072796935,
|
| 1044 |
-
"grad_norm": 2.876494884490967,
|
| 1045 |
-
"learning_rate": 6.360007144172949e-05,
|
| 1046 |
-
"loss": 1.7545,
|
| 1047 |
-
"step": 1380
|
| 1048 |
-
},
|
| 1049 |
-
{
|
| 1050 |
-
"epoch": 1.7752234993614304,
|
| 1051 |
-
"grad_norm": 3.244321823120117,
|
| 1052 |
-
"learning_rate": 6.345852663664596e-05,
|
| 1053 |
-
"loss": 1.8216,
|
| 1054 |
-
"step": 1390
|
| 1055 |
-
},
|
| 1056 |
-
{
|
| 1057 |
-
"epoch": 1.7879948914431671,
|
| 1058 |
-
"grad_norm": 2.9006967544555664,
|
| 1059 |
-
"learning_rate": 6.331559480312315e-05,
|
| 1060 |
-
"loss": 1.7792,
|
| 1061 |
-
"step": 1400
|
| 1062 |
-
},
|
| 1063 |
-
{
|
| 1064 |
-
"epoch": 1.8007662835249043,
|
| 1065 |
-
"grad_norm": 2.7696609497070312,
|
| 1066 |
-
"learning_rate": 6.317128290745699e-05,
|
| 1067 |
-
"loss": 1.7941,
|
| 1068 |
-
"step": 1410
|
| 1069 |
-
},
|
| 1070 |
-
{
|
| 1071 |
-
"epoch": 1.813537675606641,
|
| 1072 |
-
"grad_norm": 2.7425568103790283,
|
| 1073 |
-
"learning_rate": 6.302559798320566e-05,
|
| 1074 |
-
"loss": 1.7169,
|
| 1075 |
-
"step": 1420
|
| 1076 |
-
},
|
| 1077 |
-
{
|
| 1078 |
-
"epoch": 1.8263090676883782,
|
| 1079 |
-
"grad_norm": 3.130302906036377,
|
| 1080 |
-
"learning_rate": 6.287854713084686e-05,
|
| 1081 |
-
"loss": 1.7979,
|
| 1082 |
-
"step": 1430
|
| 1083 |
-
},
|
| 1084 |
-
{
|
| 1085 |
-
"epoch": 1.839080459770115,
|
| 1086 |
-
"grad_norm": 2.8706188201904297,
|
| 1087 |
-
"learning_rate": 6.273013751743166e-05,
|
| 1088 |
-
"loss": 1.7407,
|
| 1089 |
-
"step": 1440
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 1.8518518518518519,
|
| 1093 |
-
"grad_norm": 3.3041722774505615,
|
| 1094 |
-
"learning_rate": 6.258037637623526e-05,
|
| 1095 |
-
"loss": 1.7932,
|
| 1096 |
-
"step": 1450
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 1.8646232439335888,
|
| 1100 |
-
"grad_norm": 3.403945207595825,
|
| 1101 |
-
"learning_rate": 6.242927100640439e-05,
|
| 1102 |
-
"loss": 1.7524,
|
| 1103 |
-
"step": 1460
|
| 1104 |
-
},
|
| 1105 |
-
{
|
| 1106 |
-
"epoch": 1.8773946360153255,
|
| 1107 |
-
"grad_norm": 3.074903964996338,
|
| 1108 |
-
"learning_rate": 6.22768287726016e-05,
|
| 1109 |
-
"loss": 1.8342,
|
| 1110 |
-
"step": 1470
|
| 1111 |
-
},
|
| 1112 |
-
{
|
| 1113 |
-
"epoch": 1.8901660280970627,
|
| 1114 |
-
"grad_norm": 2.789820671081543,
|
| 1115 |
-
"learning_rate": 6.212305710464628e-05,
|
| 1116 |
-
"loss": 1.7446,
|
| 1117 |
-
"step": 1480
|
| 1118 |
-
},
|
| 1119 |
-
{
|
| 1120 |
-
"epoch": 1.9029374201787994,
|
| 1121 |
-
"grad_norm": 3.0101871490478516,
|
| 1122 |
-
"learning_rate": 6.196796349715262e-05,
|
| 1123 |
-
"loss": 1.7563,
|
| 1124 |
-
"step": 1490
|
| 1125 |
-
},
|
| 1126 |
-
{
|
| 1127 |
-
"epoch": 1.9157088122605364,
|
| 1128 |
-
"grad_norm": 2.9494054317474365,
|
| 1129 |
-
"learning_rate": 6.181155550916422e-05,
|
| 1130 |
-
"loss": 1.7892,
|
| 1131 |
-
"step": 1500
|
| 1132 |
-
},
|
| 1133 |
-
{
|
| 1134 |
-
"epoch": 1.9157088122605364,
|
| 1135 |
-
"eval_loss": 1.9544912576675415,
|
| 1136 |
-
"eval_runtime": 26.8049,
|
| 1137 |
-
"eval_samples_per_second": 18.653,
|
| 1138 |
-
"eval_steps_per_second": 18.653,
|
| 1139 |
-
"step": 1500
|
| 1140 |
-
},
|
| 1141 |
-
{
|
| 1142 |
-
"epoch": 1.9284802043422733,
|
| 1143 |
-
"grad_norm": 2.912105083465576,
|
| 1144 |
-
"learning_rate": 6.165384076378578e-05,
|
| 1145 |
-
"loss": 1.7858,
|
| 1146 |
-
"step": 1510
|
| 1147 |
-
},
|
| 1148 |
-
{
|
| 1149 |
-
"epoch": 1.9412515964240102,
|
| 1150 |
-
"grad_norm": 3.1038968563079834,
|
| 1151 |
-
"learning_rate": 6.149482694781147e-05,
|
| 1152 |
-
"loss": 1.8045,
|
| 1153 |
-
"step": 1520
|
| 1154 |
-
},
|
| 1155 |
-
{
|
| 1156 |
-
"epoch": 1.9540229885057472,
|
| 1157 |
-
"grad_norm": 3.214099407196045,
|
| 1158 |
-
"learning_rate": 6.133452181135035e-05,
|
| 1159 |
-
"loss": 1.771,
|
| 1160 |
-
"step": 1530
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 1.966794380587484,
|
| 1164 |
-
"grad_norm": 2.9663312435150146,
|
| 1165 |
-
"learning_rate": 6.117293316744862e-05,
|
| 1166 |
-
"loss": 1.7892,
|
| 1167 |
-
"step": 1540
|
| 1168 |
-
},
|
| 1169 |
-
{
|
| 1170 |
-
"epoch": 1.979565772669221,
|
| 1171 |
-
"grad_norm": 2.9719676971435547,
|
| 1172 |
-
"learning_rate": 6.101006889170879e-05,
|
| 1173 |
-
"loss": 1.773,
|
| 1174 |
-
"step": 1550
|
| 1175 |
-
},
|
| 1176 |
-
{
|
| 1177 |
-
"epoch": 1.9923371647509578,
|
| 1178 |
-
"grad_norm": 3.2106828689575195,
|
| 1179 |
-
"learning_rate": 6.0845936921905935e-05,
|
| 1180 |
-
"loss": 1.7839,
|
| 1181 |
-
"step": 1560
|
| 1182 |
-
},
|
| 1183 |
-
{
|
| 1184 |
-
"epoch": 2.005108556832695,
|
| 1185 |
-
"grad_norm": 2.85646390914917,
|
| 1186 |
-
"learning_rate": 6.068054525760066e-05,
|
| 1187 |
-
"loss": 1.6438,
|
| 1188 |
-
"step": 1570
|
| 1189 |
-
},
|
| 1190 |
-
{
|
| 1191 |
-
"epoch": 2.0178799489144317,
|
| 1192 |
-
"grad_norm": 2.613004684448242,
|
| 1193 |
-
"learning_rate": 6.0513901959749396e-05,
|
| 1194 |
-
"loss": 1.594,
|
| 1195 |
-
"step": 1580
|
| 1196 |
-
},
|
| 1197 |
-
{
|
| 1198 |
-
"epoch": 2.0306513409961684,
|
| 1199 |
-
"grad_norm": 2.919440984725952,
|
| 1200 |
-
"learning_rate": 6.0346015150311366e-05,
|
| 1201 |
-
"loss": 1.6577,
|
| 1202 |
-
"step": 1590
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 2.0434227330779056,
|
| 1206 |
-
"grad_norm": 2.744115114212036,
|
| 1207 |
-
"learning_rate": 6.017689301185278e-05,
|
| 1208 |
-
"loss": 1.6052,
|
| 1209 |
-
"step": 1600
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 2.0561941251596423,
|
| 1213 |
-
"grad_norm": 2.954751491546631,
|
| 1214 |
-
"learning_rate": 6.000654378714811e-05,
|
| 1215 |
-
"loss": 1.6132,
|
| 1216 |
-
"step": 1610
|
| 1217 |
-
},
|
| 1218 |
-
{
|
| 1219 |
-
"epoch": 2.0689655172413794,
|
| 1220 |
-
"grad_norm": 2.783085823059082,
|
| 1221 |
-
"learning_rate": 5.983497577877823e-05,
|
| 1222 |
-
"loss": 1.6281,
|
| 1223 |
-
"step": 1620
|
| 1224 |
-
},
|
| 1225 |
-
{
|
| 1226 |
-
"epoch": 2.081736909323116,
|
| 1227 |
-
"grad_norm": 2.958265781402588,
|
| 1228 |
-
"learning_rate": 5.966219734872581e-05,
|
| 1229 |
-
"loss": 1.5537,
|
| 1230 |
-
"step": 1630
|
| 1231 |
-
},
|
| 1232 |
-
{
|
| 1233 |
-
"epoch": 2.0945083014048533,
|
| 1234 |
-
"grad_norm": 2.761991262435913,
|
| 1235 |
-
"learning_rate": 5.9488216917967784e-05,
|
| 1236 |
-
"loss": 1.6174,
|
| 1237 |
-
"step": 1640
|
| 1238 |
-
},
|
| 1239 |
-
{
|
| 1240 |
-
"epoch": 2.10727969348659,
|
| 1241 |
-
"grad_norm": 2.8865253925323486,
|
| 1242 |
-
"learning_rate": 5.9313042966064896e-05,
|
| 1243 |
-
"loss": 1.6407,
|
| 1244 |
-
"step": 1650
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 2.10727969348659,
|
| 1248 |
-
"eval_loss": 1.9608972072601318,
|
| 1249 |
-
"eval_runtime": 26.9926,
|
| 1250 |
-
"eval_samples_per_second": 18.524,
|
| 1251 |
-
"eval_steps_per_second": 18.524,
|
| 1252 |
-
"step": 1650
|
| 1253 |
-
},
|
| 1254 |
-
{
|
| 1255 |
-
"epoch": 2.1200510855683268,
|
| 1256 |
-
"grad_norm": 2.831836700439453,
|
| 1257 |
-
"learning_rate": 5.9136684030748436e-05,
|
| 1258 |
-
"loss": 1.6225,
|
| 1259 |
-
"step": 1660
|
| 1260 |
-
},
|
| 1261 |
-
{
|
| 1262 |
-
"epoch": 2.132822477650064,
|
| 1263 |
-
"grad_norm": 2.761368751525879,
|
| 1264 |
-
"learning_rate": 5.89591487075041e-05,
|
| 1265 |
-
"loss": 1.6014,
|
| 1266 |
-
"step": 1670
|
| 1267 |
-
},
|
| 1268 |
-
{
|
| 1269 |
-
"epoch": 2.1455938697318007,
|
| 1270 |
-
"grad_norm": 2.7214226722717285,
|
| 1271 |
-
"learning_rate": 5.8780445649153075e-05,
|
| 1272 |
-
"loss": 1.6009,
|
| 1273 |
-
"step": 1680
|
| 1274 |
-
},
|
| 1275 |
-
{
|
| 1276 |
-
"epoch": 2.158365261813538,
|
| 1277 |
-
"grad_norm": 3.010241746902466,
|
| 1278 |
-
"learning_rate": 5.860058356543031e-05,
|
| 1279 |
-
"loss": 1.5901,
|
| 1280 |
-
"step": 1690
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 2.1711366538952745,
|
| 1284 |
-
"grad_norm": 3.019549608230591,
|
| 1285 |
-
"learning_rate": 5.8419571222560034e-05,
|
| 1286 |
-
"loss": 1.5947,
|
| 1287 |
-
"step": 1700
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 2.1839080459770113,
|
| 1291 |
-
"grad_norm": 2.638648509979248,
|
| 1292 |
-
"learning_rate": 5.823741744282845e-05,
|
| 1293 |
-
"loss": 1.566,
|
| 1294 |
-
"step": 1710
|
| 1295 |
-
},
|
| 1296 |
-
{
|
| 1297 |
-
"epoch": 2.1966794380587484,
|
| 1298 |
-
"grad_norm": 2.9191718101501465,
|
| 1299 |
-
"learning_rate": 5.805413110415381e-05,
|
| 1300 |
-
"loss": 1.5797,
|
| 1301 |
-
"step": 1720
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 2.209450830140485,
|
| 1305 |
-
"grad_norm": 3.0825226306915283,
|
| 1306 |
-
"learning_rate": 5.786972113965369e-05,
|
| 1307 |
-
"loss": 1.6247,
|
| 1308 |
-
"step": 1730
|
| 1309 |
-
},
|
| 1310 |
-
{
|
| 1311 |
-
"epoch": 2.2222222222222223,
|
| 1312 |
-
"grad_norm": 3.186307191848755,
|
| 1313 |
-
"learning_rate": 5.7684196537209574e-05,
|
| 1314 |
-
"loss": 1.6044,
|
| 1315 |
-
"step": 1740
|
| 1316 |
-
},
|
| 1317 |
-
{
|
| 1318 |
-
"epoch": 2.234993614303959,
|
| 1319 |
-
"grad_norm": 2.94689679145813,
|
| 1320 |
-
"learning_rate": 5.749756633902887e-05,
|
| 1321 |
-
"loss": 1.5743,
|
| 1322 |
-
"step": 1750
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 2.247765006385696,
|
| 1326 |
-
"grad_norm": 3.0586421489715576,
|
| 1327 |
-
"learning_rate": 5.7309839641204136e-05,
|
| 1328 |
-
"loss": 1.6237,
|
| 1329 |
-
"step": 1760
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 2.260536398467433,
|
| 1333 |
-
"grad_norm": 3.0845377445220947,
|
| 1334 |
-
"learning_rate": 5.7121025593269777e-05,
|
| 1335 |
-
"loss": 1.6194,
|
| 1336 |
-
"step": 1770
|
| 1337 |
-
},
|
| 1338 |
-
{
|
| 1339 |
-
"epoch": 2.27330779054917,
|
| 1340 |
-
"grad_norm": 3.0147876739501953,
|
| 1341 |
-
"learning_rate": 5.693113339775611e-05,
|
| 1342 |
-
"loss": 1.5865,
|
| 1343 |
-
"step": 1780
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 2.286079182630907,
|
| 1347 |
-
"grad_norm": 3.1859536170959473,
|
| 1348 |
-
"learning_rate": 5.674017230974085e-05,
|
| 1349 |
-
"loss": 1.6086,
|
| 1350 |
-
"step": 1790
|
| 1351 |
-
},
|
| 1352 |
-
{
|
| 1353 |
-
"epoch": 2.2988505747126435,
|
| 1354 |
-
"grad_norm": 3.106766939163208,
|
| 1355 |
-
"learning_rate": 5.654815163639803e-05,
|
| 1356 |
-
"loss": 1.5977,
|
| 1357 |
-
"step": 1800
|
| 1358 |
-
},
|
| 1359 |
-
{
|
| 1360 |
-
"epoch": 2.2988505747126435,
|
| 1361 |
-
"eval_loss": 1.9625232219696045,
|
| 1362 |
-
"eval_runtime": 28.5711,
|
| 1363 |
-
"eval_samples_per_second": 17.5,
|
| 1364 |
-
"eval_steps_per_second": 17.5,
|
| 1365 |
-
"step": 1800
|
| 1366 |
-
},
|
| 1367 |
-
{
|
| 1368 |
-
"epoch": 2.3116219667943807,
|
| 1369 |
-
"grad_norm": 2.9333529472351074,
|
| 1370 |
-
"learning_rate": 5.654815163639803e-07,
|
| 1371 |
-
"loss": 1.5856,
|
| 1372 |
-
"step": 1810
|
| 1373 |
-
},
|
| 1374 |
-
{
|
| 1375 |
-
"epoch": 2.3243933588761174,
|
| 1376 |
-
"grad_norm": 2.9890449047088623,
|
| 1377 |
-
"learning_rate": 1.1309630327279607e-06,
|
| 1378 |
-
"loss": 1.584,
|
| 1379 |
-
"step": 1820
|
| 1380 |
-
},
|
| 1381 |
-
{
|
| 1382 |
-
"epoch": 2.3371647509578546,
|
| 1383 |
-
"grad_norm": 3.300159215927124,
|
| 1384 |
-
"learning_rate": 1.6964445490919409e-06,
|
| 1385 |
-
"loss": 1.5921,
|
| 1386 |
-
"step": 1830
|
| 1387 |
-
},
|
| 1388 |
-
{
|
| 1389 |
-
"epoch": 2.3499361430395913,
|
| 1390 |
-
"grad_norm": 3.0725886821746826,
|
| 1391 |
-
"learning_rate": 2.2619260654559213e-06,
|
| 1392 |
-
"loss": 1.5305,
|
| 1393 |
-
"step": 1840
|
| 1394 |
-
},
|
| 1395 |
-
{
|
| 1396 |
-
"epoch": 2.362707535121328,
|
| 1397 |
-
"grad_norm": 2.9211957454681396,
|
| 1398 |
-
"learning_rate": 2.8274075818199017e-06,
|
| 1399 |
-
"loss": 1.6247,
|
| 1400 |
-
"step": 1850
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 2.375478927203065,
|
| 1404 |
-
"grad_norm": 2.942033052444458,
|
| 1405 |
-
"learning_rate": 3.3928890981838817e-06,
|
| 1406 |
-
"loss": 1.5717,
|
| 1407 |
-
"step": 1860
|
| 1408 |
-
},
|
| 1409 |
-
{
|
| 1410 |
-
"epoch": 2.388250319284802,
|
| 1411 |
-
"grad_norm": 2.8808839321136475,
|
| 1412 |
-
"learning_rate": 3.958370614547863e-06,
|
| 1413 |
-
"loss": 1.606,
|
| 1414 |
-
"step": 1870
|
| 1415 |
-
},
|
| 1416 |
-
{
|
| 1417 |
-
"epoch": 2.401021711366539,
|
| 1418 |
-
"grad_norm": 2.9812419414520264,
|
| 1419 |
-
"learning_rate": 4.523852130911843e-06,
|
| 1420 |
-
"loss": 1.5825,
|
| 1421 |
-
"step": 1880
|
| 1422 |
-
},
|
| 1423 |
-
{
|
| 1424 |
-
"epoch": 2.413793103448276,
|
| 1425 |
-
"grad_norm": 3.1213929653167725,
|
| 1426 |
-
"learning_rate": 5.089333647275823e-06,
|
| 1427 |
-
"loss": 1.6172,
|
| 1428 |
-
"step": 1890
|
| 1429 |
-
},
|
| 1430 |
-
{
|
| 1431 |
-
"epoch": 2.4265644955300125,
|
| 1432 |
-
"grad_norm": 2.865854501724243,
|
| 1433 |
-
"learning_rate": 5.6548151636398035e-06,
|
| 1434 |
-
"loss": 1.6028,
|
| 1435 |
-
"step": 1900
|
| 1436 |
-
},
|
| 1437 |
-
{
|
| 1438 |
-
"epoch": 2.4393358876117497,
|
| 1439 |
-
"grad_norm": 3.07641863822937,
|
| 1440 |
-
"learning_rate": 6.220296680003784e-06,
|
| 1441 |
-
"loss": 1.6674,
|
| 1442 |
-
"step": 1910
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 2.4521072796934864,
|
| 1446 |
-
"grad_norm": 2.7488229274749756,
|
| 1447 |
-
"learning_rate": 6.7857781963677635e-06,
|
| 1448 |
-
"loss": 1.5994,
|
| 1449 |
-
"step": 1920
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 2.4648786717752236,
|
| 1453 |
-
"grad_norm": 2.993471384048462,
|
| 1454 |
-
"learning_rate": 7.351259712731745e-06,
|
| 1455 |
-
"loss": 1.6039,
|
| 1456 |
-
"step": 1930
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 2.4776500638569603,
|
| 1460 |
-
"grad_norm": 2.7299869060516357,
|
| 1461 |
-
"learning_rate": 7.916741229095726e-06,
|
| 1462 |
-
"loss": 1.5046,
|
| 1463 |
-
"step": 1940
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 2.4904214559386975,
|
| 1467 |
-
"grad_norm": 3.036618947982788,
|
| 1468 |
-
"learning_rate": 8.482222745459704e-06,
|
| 1469 |
-
"loss": 1.6052,
|
| 1470 |
-
"step": 1950
|
| 1471 |
-
},
|
| 1472 |
-
{
|
| 1473 |
-
"epoch": 2.4904214559386975,
|
| 1474 |
-
"eval_loss": 1.9600526094436646,
|
| 1475 |
-
"eval_runtime": 26.6264,
|
| 1476 |
-
"eval_samples_per_second": 18.778,
|
| 1477 |
-
"eval_steps_per_second": 18.778,
|
| 1478 |
-
"step": 1950
|
| 1479 |
-
},
|
| 1480 |
-
{
|
| 1481 |
-
"epoch": 2.503192848020434,
|
| 1482 |
-
"grad_norm": 3.1929354667663574,
|
| 1483 |
-
"learning_rate": 9.047704261823685e-06,
|
| 1484 |
-
"loss": 1.632,
|
| 1485 |
-
"step": 1960
|
| 1486 |
-
},
|
| 1487 |
-
{
|
| 1488 |
-
"epoch": 2.5159642401021713,
|
| 1489 |
-
"grad_norm": 3.241421699523926,
|
| 1490 |
-
"learning_rate": 9.613185778187667e-06,
|
| 1491 |
-
"loss": 1.5757,
|
| 1492 |
-
"step": 1970
|
| 1493 |
-
},
|
| 1494 |
-
{
|
| 1495 |
-
"epoch": 2.528735632183908,
|
| 1496 |
-
"grad_norm": 2.8949759006500244,
|
| 1497 |
-
"learning_rate": 1.0178667294551646e-05,
|
| 1498 |
-
"loss": 1.6781,
|
| 1499 |
-
"step": 1980
|
| 1500 |
-
},
|
| 1501 |
-
{
|
| 1502 |
-
"epoch": 2.541507024265645,
|
| 1503 |
-
"grad_norm": 2.983260154724121,
|
| 1504 |
-
"learning_rate": 1.0744148810915626e-05,
|
| 1505 |
-
"loss": 1.6403,
|
| 1506 |
-
"step": 1990
|
| 1507 |
-
},
|
| 1508 |
-
{
|
| 1509 |
-
"epoch": 2.554278416347382,
|
| 1510 |
-
"grad_norm": 3.2770559787750244,
|
| 1511 |
-
"learning_rate": 1.1309630327279607e-05,
|
| 1512 |
-
"loss": 1.6642,
|
| 1513 |
-
"step": 2000
|
| 1514 |
-
},
|
| 1515 |
-
{
|
| 1516 |
-
"epoch": 2.5670498084291187,
|
| 1517 |
-
"grad_norm": 2.995584487915039,
|
| 1518 |
-
"learning_rate": 1.1875111843643587e-05,
|
| 1519 |
-
"loss": 1.6282,
|
| 1520 |
-
"step": 2010
|
| 1521 |
-
},
|
| 1522 |
-
{
|
| 1523 |
-
"epoch": 2.579821200510856,
|
| 1524 |
-
"grad_norm": 3.113140106201172,
|
| 1525 |
-
"learning_rate": 1.2440593360007568e-05,
|
| 1526 |
-
"loss": 1.5997,
|
| 1527 |
-
"step": 2020
|
| 1528 |
-
},
|
| 1529 |
-
{
|
| 1530 |
-
"epoch": 2.5925925925925926,
|
| 1531 |
-
"grad_norm": 2.9088339805603027,
|
| 1532 |
-
"learning_rate": 1.3006074876371547e-05,
|
| 1533 |
-
"loss": 1.6134,
|
| 1534 |
-
"step": 2030
|
| 1535 |
-
},
|
| 1536 |
-
{
|
| 1537 |
-
"epoch": 2.6053639846743293,
|
| 1538 |
-
"grad_norm": 2.9703259468078613,
|
| 1539 |
-
"learning_rate": 1.3571556392735527e-05,
|
| 1540 |
-
"loss": 1.5381,
|
| 1541 |
-
"step": 2040
|
| 1542 |
-
},
|
| 1543 |
-
{
|
| 1544 |
-
"epoch": 2.6181353767560664,
|
| 1545 |
-
"grad_norm": 3.0287814140319824,
|
| 1546 |
-
"learning_rate": 1.4137037909099508e-05,
|
| 1547 |
-
"loss": 1.6708,
|
| 1548 |
-
"step": 2050
|
| 1549 |
-
},
|
| 1550 |
-
{
|
| 1551 |
-
"epoch": 2.630906768837803,
|
| 1552 |
-
"grad_norm": 3.0678138732910156,
|
| 1553 |
-
"learning_rate": 1.470251942546349e-05,
|
| 1554 |
-
"loss": 1.6232,
|
| 1555 |
-
"step": 2060
|
| 1556 |
-
},
|
| 1557 |
-
{
|
| 1558 |
-
"epoch": 2.6436781609195403,
|
| 1559 |
-
"grad_norm": 2.8624658584594727,
|
| 1560 |
-
"learning_rate": 1.526800094182747e-05,
|
| 1561 |
-
"loss": 1.5495,
|
| 1562 |
-
"step": 2070
|
| 1563 |
-
},
|
| 1564 |
-
{
|
| 1565 |
-
"epoch": 2.656449553001277,
|
| 1566 |
-
"grad_norm": 3.0500433444976807,
|
| 1567 |
-
"learning_rate": 1.5833482458191452e-05,
|
| 1568 |
-
"loss": 1.5992,
|
| 1569 |
-
"step": 2080
|
| 1570 |
-
},
|
| 1571 |
-
{
|
| 1572 |
-
"epoch": 2.6692209450830138,
|
| 1573 |
-
"grad_norm": 2.9770045280456543,
|
| 1574 |
-
"learning_rate": 1.639896397455543e-05,
|
| 1575 |
-
"loss": 1.6213,
|
| 1576 |
-
"step": 2090
|
| 1577 |
-
},
|
| 1578 |
-
{
|
| 1579 |
-
"epoch": 2.681992337164751,
|
| 1580 |
-
"grad_norm": 3.110860586166382,
|
| 1581 |
-
"learning_rate": 1.6964445490919408e-05,
|
| 1582 |
-
"loss": 1.5852,
|
| 1583 |
-
"step": 2100
|
| 1584 |
-
},
|
| 1585 |
-
{
|
| 1586 |
-
"epoch": 2.681992337164751,
|
| 1587 |
-
"eval_loss": 1.957663893699646,
|
| 1588 |
-
"eval_runtime": 26.8577,
|
| 1589 |
-
"eval_samples_per_second": 18.617,
|
| 1590 |
-
"eval_steps_per_second": 18.617,
|
| 1591 |
-
"step": 2100
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 2.694763729246488,
|
| 1595 |
-
"grad_norm": 2.987915515899658,
|
| 1596 |
-
"learning_rate": 1.752992700728339e-05,
|
| 1597 |
-
"loss": 1.613,
|
| 1598 |
-
"step": 2110
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 2.707535121328225,
|
| 1602 |
-
"grad_norm": 3.1154093742370605,
|
| 1603 |
-
"learning_rate": 1.809540852364737e-05,
|
| 1604 |
-
"loss": 1.6002,
|
| 1605 |
-
"step": 2120
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 2.7203065134099615,
|
| 1609 |
-
"grad_norm": 3.175532817840576,
|
| 1610 |
-
"learning_rate": 1.8660890040011352e-05,
|
| 1611 |
-
"loss": 1.5809,
|
| 1612 |
-
"step": 2130
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 2.7330779054916987,
|
| 1616 |
-
"grad_norm": 2.9489593505859375,
|
| 1617 |
-
"learning_rate": 1.9226371556375333e-05,
|
| 1618 |
-
"loss": 1.5821,
|
| 1619 |
-
"step": 2140
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 2.7458492975734354,
|
| 1623 |
-
"grad_norm": 2.855994939804077,
|
| 1624 |
-
"learning_rate": 1.979185307273931e-05,
|
| 1625 |
-
"loss": 1.5858,
|
| 1626 |
-
"step": 2150
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 2.7586206896551726,
|
| 1630 |
-
"grad_norm": 2.9239962100982666,
|
| 1631 |
-
"learning_rate": 2.0357334589103292e-05,
|
| 1632 |
-
"loss": 1.5816,
|
| 1633 |
-
"step": 2160
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 2.7713920817369093,
|
| 1637 |
-
"grad_norm": 2.9092483520507812,
|
| 1638 |
-
"learning_rate": 2.0922816105467273e-05,
|
| 1639 |
-
"loss": 1.6278,
|
| 1640 |
-
"step": 2170
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 2.784163473818646,
|
| 1644 |
-
"grad_norm": 3.0730059146881104,
|
| 1645 |
-
"learning_rate": 2.148829762183125e-05,
|
| 1646 |
-
"loss": 1.587,
|
| 1647 |
-
"step": 2180
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 2.796934865900383,
|
| 1651 |
-
"grad_norm": 2.9733469486236572,
|
| 1652 |
-
"learning_rate": 2.2053779138195233e-05,
|
| 1653 |
-
"loss": 1.5931,
|
| 1654 |
-
"step": 2190
|
| 1655 |
-
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 2.80970625798212,
|
| 1658 |
-
"grad_norm": 2.882889986038208,
|
| 1659 |
-
"learning_rate": 2.2619260654559214e-05,
|
| 1660 |
-
"loss": 1.6206,
|
| 1661 |
-
"step": 2200
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 2.822477650063857,
|
| 1665 |
-
"grad_norm": 3.2219135761260986,
|
| 1666 |
-
"learning_rate": 2.3184742170923192e-05,
|
| 1667 |
-
"loss": 1.5782,
|
| 1668 |
-
"step": 2210
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 2.835249042145594,
|
| 1672 |
-
"grad_norm": 3.1879076957702637,
|
| 1673 |
-
"learning_rate": 2.3750223687287173e-05,
|
| 1674 |
-
"loss": 1.6398,
|
| 1675 |
-
"step": 2220
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 2.8480204342273305,
|
| 1679 |
-
"grad_norm": 3.0046067237854004,
|
| 1680 |
-
"learning_rate": 2.4315705203651154e-05,
|
| 1681 |
-
"loss": 1.6471,
|
| 1682 |
-
"step": 2230
|
| 1683 |
-
},
|
| 1684 |
-
{
|
| 1685 |
-
"epoch": 2.8607918263090677,
|
| 1686 |
-
"grad_norm": 2.837573766708374,
|
| 1687 |
-
"learning_rate": 2.4881186720015136e-05,
|
| 1688 |
-
"loss": 1.5785,
|
| 1689 |
-
"step": 2240
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 2.873563218390805,
|
| 1693 |
-
"grad_norm": 2.998067855834961,
|
| 1694 |
-
"learning_rate": 2.5446668236379117e-05,
|
| 1695 |
-
"loss": 1.6123,
|
| 1696 |
-
"step": 2250
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 2.873563218390805,
|
| 1700 |
-
"eval_loss": 1.9578022956848145,
|
| 1701 |
-
"eval_runtime": 27.614,
|
| 1702 |
-
"eval_samples_per_second": 18.107,
|
| 1703 |
-
"eval_steps_per_second": 18.107,
|
| 1704 |
-
"step": 2250
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 2.8863346104725416,
|
| 1708 |
-
"grad_norm": 3.0444719791412354,
|
| 1709 |
-
"learning_rate": 2.6012149752743095e-05,
|
| 1710 |
-
"loss": 1.5877,
|
| 1711 |
-
"step": 2260
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 2.8991060025542783,
|
| 1715 |
-
"grad_norm": 3.109495162963867,
|
| 1716 |
-
"learning_rate": 2.6577631269107073e-05,
|
| 1717 |
-
"loss": 1.6581,
|
| 1718 |
-
"step": 2270
|
| 1719 |
-
},
|
| 1720 |
-
{
|
| 1721 |
-
"epoch": 2.9118773946360155,
|
| 1722 |
-
"grad_norm": 2.8879830837249756,
|
| 1723 |
-
"learning_rate": 2.7143112785471054e-05,
|
| 1724 |
-
"loss": 1.6034,
|
| 1725 |
-
"step": 2280
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 2.924648786717752,
|
| 1729 |
-
"grad_norm": 3.3881685733795166,
|
| 1730 |
-
"learning_rate": 2.7708594301835035e-05,
|
| 1731 |
-
"loss": 1.6445,
|
| 1732 |
-
"step": 2290
|
| 1733 |
-
},
|
| 1734 |
-
{
|
| 1735 |
-
"epoch": 2.9374201787994894,
|
| 1736 |
-
"grad_norm": 2.869450330734253,
|
| 1737 |
-
"learning_rate": 2.8274075818199017e-05,
|
| 1738 |
-
"loss": 1.5976,
|
| 1739 |
-
"step": 2300
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 2.950191570881226,
|
| 1743 |
-
"grad_norm": 2.9405651092529297,
|
| 1744 |
-
"learning_rate": 2.8273731308557363e-05,
|
| 1745 |
-
"loss": 1.5806,
|
| 1746 |
-
"step": 2310
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 2.962962962962963,
|
| 1750 |
-
"grad_norm": 3.227057695388794,
|
| 1751 |
-
"learning_rate": 2.8272697796423325e-05,
|
| 1752 |
-
"loss": 1.6023,
|
| 1753 |
-
"step": 2320
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 2.9757343550447,
|
| 1757 |
-
"grad_norm": 3.0668394565582275,
|
| 1758 |
-
"learning_rate": 2.8270975332168814e-05,
|
| 1759 |
-
"loss": 1.5949,
|
| 1760 |
-
"step": 2330
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 2.9885057471264367,
|
| 1764 |
-
"grad_norm": 3.4752612113952637,
|
| 1765 |
-
"learning_rate": 2.8268563999744315e-05,
|
| 1766 |
-
"loss": 1.6142,
|
| 1767 |
-
"step": 2340
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 3.001277139208174,
|
| 1771 |
-
"grad_norm": 2.851846694946289,
|
| 1772 |
-
"learning_rate": 2.8265463916674774e-05,
|
| 1773 |
-
"loss": 1.6006,
|
| 1774 |
-
"step": 2350
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 3.0140485312899106,
|
| 1778 |
-
"grad_norm": 2.914916753768921,
|
| 1779 |
-
"learning_rate": 2.8261675234053857e-05,
|
| 1780 |
-
"loss": 1.5599,
|
| 1781 |
-
"step": 2360
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 3.0268199233716473,
|
| 1785 |
-
"grad_norm": 3.022381544113159,
|
| 1786 |
-
"learning_rate": 2.825719813653661e-05,
|
| 1787 |
-
"loss": 1.561,
|
| 1788 |
-
"step": 2370
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 3.0395913154533845,
|
| 1792 |
-
"grad_norm": 2.812488079071045,
|
| 1793 |
-
"learning_rate": 2.8252032842330455e-05,
|
| 1794 |
-
"loss": 1.5051,
|
| 1795 |
-
"step": 2380
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 3.052362707535121,
|
| 1799 |
-
"grad_norm": 2.879337787628174,
|
| 1800 |
-
"learning_rate": 2.8246179603184542e-05,
|
| 1801 |
-
"loss": 1.5499,
|
| 1802 |
-
"step": 2390
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 3.0651340996168583,
|
| 1806 |
-
"grad_norm": 3.1652944087982178,
|
| 1807 |
-
"learning_rate": 2.823963870437749e-05,
|
| 1808 |
-
"loss": 1.5814,
|
| 1809 |
-
"step": 2400
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 3.0651340996168583,
|
| 1813 |
-
"eval_loss": 1.9594498872756958,
|
| 1814 |
-
"eval_runtime": 27.4191,
|
| 1815 |
-
"eval_samples_per_second": 18.235,
|
| 1816 |
-
"eval_steps_per_second": 18.235,
|
| 1817 |
-
"step": 2400
|
| 1818 |
-
},
|
| 1819 |
-
{
|
| 1820 |
-
"epoch": 3.077905491698595,
|
| 1821 |
-
"grad_norm": 3.0777502059936523,
|
| 1822 |
-
"learning_rate": 2.8239638704377494e-07,
|
| 1823 |
-
"loss": 1.5266,
|
| 1824 |
-
"step": 2410
|
| 1825 |
-
},
|
| 1826 |
-
{
|
| 1827 |
-
"epoch": 3.0906768837803322,
|
| 1828 |
-
"grad_norm": 2.999093770980835,
|
| 1829 |
-
"learning_rate": 5.647927740875499e-07,
|
| 1830 |
-
"loss": 1.5265,
|
| 1831 |
-
"step": 2420
|
| 1832 |
-
},
|
| 1833 |
-
{
|
| 1834 |
-
"epoch": 3.103448275862069,
|
| 1835 |
-
"grad_norm": 3.03286075592041,
|
| 1836 |
-
"learning_rate": 8.471891611313247e-07,
|
| 1837 |
-
"loss": 1.5492,
|
| 1838 |
-
"step": 2430
|
| 1839 |
-
},
|
| 1840 |
-
{
|
| 1841 |
-
"epoch": 3.1162196679438057,
|
| 1842 |
-
"grad_norm": 3.1120502948760986,
|
| 1843 |
-
"learning_rate": 1.1295855481750998e-06,
|
| 1844 |
-
"loss": 1.5494,
|
| 1845 |
-
"step": 2440
|
| 1846 |
-
},
|
| 1847 |
-
{
|
| 1848 |
-
"epoch": 3.128991060025543,
|
| 1849 |
-
"grad_norm": 3.289019823074341,
|
| 1850 |
-
"learning_rate": 1.4119819352188746e-06,
|
| 1851 |
-
"loss": 1.5846,
|
| 1852 |
-
"step": 2450
|
| 1853 |
-
},
|
| 1854 |
-
{
|
| 1855 |
-
"epoch": 3.1417624521072796,
|
| 1856 |
-
"grad_norm": 2.9034361839294434,
|
| 1857 |
-
"learning_rate": 1.6943783222626494e-06,
|
| 1858 |
-
"loss": 1.553,
|
| 1859 |
-
"step": 2460
|
| 1860 |
-
},
|
| 1861 |
-
{
|
| 1862 |
-
"epoch": 3.1545338441890167,
|
| 1863 |
-
"grad_norm": 3.0615646839141846,
|
| 1864 |
-
"learning_rate": 1.9767747093064247e-06,
|
| 1865 |
-
"loss": 1.5219,
|
| 1866 |
-
"step": 2470
|
| 1867 |
-
},
|
| 1868 |
-
{
|
| 1869 |
-
"epoch": 3.1673052362707534,
|
| 1870 |
-
"grad_norm": 2.944612741470337,
|
| 1871 |
-
"learning_rate": 2.2591710963501995e-06,
|
| 1872 |
-
"loss": 1.5853,
|
| 1873 |
-
"step": 2480
|
| 1874 |
-
},
|
| 1875 |
-
{
|
| 1876 |
-
"epoch": 3.1800766283524906,
|
| 1877 |
-
"grad_norm": 2.9914486408233643,
|
| 1878 |
-
"learning_rate": 2.5415674833939743e-06,
|
| 1879 |
-
"loss": 1.5752,
|
| 1880 |
-
"step": 2490
|
| 1881 |
-
},
|
| 1882 |
-
{
|
| 1883 |
-
"epoch": 3.1928480204342273,
|
| 1884 |
-
"grad_norm": 2.6826584339141846,
|
| 1885 |
-
"learning_rate": 2.823963870437749e-06,
|
| 1886 |
-
"loss": 1.5066,
|
| 1887 |
-
"step": 2500
|
| 1888 |
-
},
|
| 1889 |
-
{
|
| 1890 |
-
"epoch": 3.205619412515964,
|
| 1891 |
-
"grad_norm": 3.1901533603668213,
|
| 1892 |
-
"learning_rate": 3.106360257481524e-06,
|
| 1893 |
-
"loss": 1.5461,
|
| 1894 |
-
"step": 2510
|
| 1895 |
-
},
|
| 1896 |
-
{
|
| 1897 |
-
"epoch": 3.218390804597701,
|
| 1898 |
-
"grad_norm": 3.0013110637664795,
|
| 1899 |
-
"learning_rate": 3.388756644525299e-06,
|
| 1900 |
-
"loss": 1.5591,
|
| 1901 |
-
"step": 2520
|
| 1902 |
-
},
|
| 1903 |
-
{
|
| 1904 |
-
"epoch": 3.231162196679438,
|
| 1905 |
-
"grad_norm": 3.266486406326294,
|
| 1906 |
-
"learning_rate": 3.671153031569074e-06,
|
| 1907 |
-
"loss": 1.5211,
|
| 1908 |
-
"step": 2530
|
| 1909 |
-
},
|
| 1910 |
-
{
|
| 1911 |
-
"epoch": 3.243933588761175,
|
| 1912 |
-
"grad_norm": 2.939565896987915,
|
| 1913 |
-
"learning_rate": 3.953549418612849e-06,
|
| 1914 |
-
"loss": 1.5392,
|
| 1915 |
-
"step": 2540
|
| 1916 |
-
},
|
| 1917 |
-
{
|
| 1918 |
-
"epoch": 3.256704980842912,
|
| 1919 |
-
"grad_norm": 2.8810532093048096,
|
| 1920 |
-
"learning_rate": 4.235945805656624e-06,
|
| 1921 |
-
"loss": 1.5439,
|
| 1922 |
-
"step": 2550
|
| 1923 |
-
},
|
| 1924 |
-
{
|
| 1925 |
-
"epoch": 3.256704980842912,
|
| 1926 |
-
"eval_loss": 1.9594852924346924,
|
| 1927 |
-
"eval_runtime": 27.7206,
|
| 1928 |
-
"eval_samples_per_second": 18.037,
|
| 1929 |
-
"eval_steps_per_second": 18.037,
|
| 1930 |
-
"step": 2550
|
| 1931 |
}
|
| 1932 |
],
|
| 1933 |
"logging_steps": 10,
|
|
@@ -1947,7 +139,7 @@
|
|
| 1947 |
"attributes": {}
|
| 1948 |
}
|
| 1949 |
},
|
| 1950 |
-
"total_flos":
|
| 1951 |
"train_batch_size": 16,
|
| 1952 |
"trial_name": null,
|
| 1953 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.9780572652816772,
|
| 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": 2.1011273860931396,
|
| 14 |
+
"learning_rate": 4.666666666666666e-06,
|
| 15 |
+
"loss": 1.9177,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.02554278416347382,
|
| 20 |
+
"grad_norm": 2.0189812183380127,
|
| 21 |
+
"learning_rate": 9.333333333333333e-06,
|
| 22 |
+
"loss": 1.9419,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.038314176245210725,
|
| 27 |
+
"grad_norm": 2.073760509490967,
|
| 28 |
+
"learning_rate": 1.4e-05,
|
| 29 |
+
"loss": 1.9122,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.05108556832694764,
|
| 34 |
+
"grad_norm": 1.955664873123169,
|
| 35 |
+
"learning_rate": 1.8666666666666665e-05,
|
| 36 |
+
"loss": 1.8872,
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.06385696040868455,
|
| 41 |
+
"grad_norm": 2.6010475158691406,
|
| 42 |
+
"learning_rate": 2.333333333333333e-05,
|
| 43 |
+
"loss": 1.9779,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.07662835249042145,
|
| 48 |
+
"grad_norm": 2.0808260440826416,
|
| 49 |
+
"learning_rate": 2.8e-05,
|
| 50 |
+
"loss": 1.933,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.08939974457215837,
|
| 55 |
+
"grad_norm": 1.969761848449707,
|
| 56 |
+
"learning_rate": 3.266666666666666e-05,
|
| 57 |
+
"loss": 1.946,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.10217113665389528,
|
| 62 |
+
"grad_norm": 2.136836290359497,
|
| 63 |
+
"learning_rate": 3.733333333333333e-05,
|
| 64 |
+
"loss": 1.9441,
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.11494252873563218,
|
| 69 |
+
"grad_norm": 2.056912899017334,
|
| 70 |
+
"learning_rate": 4.2e-05,
|
| 71 |
+
"loss": 1.9262,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.1277139208173691,
|
| 76 |
+
"grad_norm": 2.1491384506225586,
|
| 77 |
+
"learning_rate": 4.666666666666666e-05,
|
| 78 |
+
"loss": 1.9775,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.140485312899106,
|
| 83 |
+
"grad_norm": 1.8882553577423096,
|
| 84 |
+
"learning_rate": 5.1333333333333325e-05,
|
| 85 |
+
"loss": 1.9233,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.1532567049808429,
|
| 90 |
+
"grad_norm": 2.0507898330688477,
|
| 91 |
+
"learning_rate": 5.6e-05,
|
| 92 |
+
"loss": 1.9408,
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.16602809706257982,
|
| 97 |
+
"grad_norm": 2.2763912677764893,
|
| 98 |
+
"learning_rate": 6.0666666666666666e-05,
|
| 99 |
+
"loss": 1.9429,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.17879948914431673,
|
| 104 |
+
"grad_norm": 2.1040444374084473,
|
| 105 |
+
"learning_rate": 6.533333333333333e-05,
|
| 106 |
+
"loss": 1.9193,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.19157088122605365,
|
| 111 |
+
"grad_norm": 2.0764999389648438,
|
| 112 |
+
"learning_rate": 7e-05,
|
| 113 |
+
"loss": 1.9405,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.19157088122605365,
|
| 118 |
+
"eval_loss": 1.9780572652816772,
|
| 119 |
+
"eval_runtime": 24.3012,
|
| 120 |
+
"eval_samples_per_second": 20.575,
|
| 121 |
+
"eval_steps_per_second": 20.575,
|
| 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 sha256:73d63faa96a8013f73d4d225b5f62be5f6f1a8819c12a7d65e93c26570162b6b
|
| 3 |
size 5496
|