Training in progress, step 15, checkpoint
Browse files
last-checkpoint/adapter_config.json
CHANGED
|
@@ -10,23 +10,23 @@
|
|
| 10 |
"layers_pattern": null,
|
| 11 |
"layers_to_transform": null,
|
| 12 |
"loftq_config": {},
|
| 13 |
-
"lora_alpha":
|
| 14 |
"lora_dropout": 0.05,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
"modules_to_save": null,
|
| 18 |
"peft_type": "LORA",
|
| 19 |
-
"r":
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"up_proj",
|
| 24 |
-
"o_proj",
|
| 25 |
"gate_proj",
|
| 26 |
-
"down_proj",
|
| 27 |
"v_proj",
|
|
|
|
|
|
|
| 28 |
"k_proj",
|
| 29 |
-
"q_proj"
|
|
|
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 10 |
"layers_pattern": null,
|
| 11 |
"layers_to_transform": null,
|
| 12 |
"loftq_config": {},
|
| 13 |
+
"lora_alpha": 32,
|
| 14 |
"lora_dropout": 0.05,
|
| 15 |
"megatron_config": null,
|
| 16 |
"megatron_core": "megatron.core",
|
| 17 |
"modules_to_save": null,
|
| 18 |
"peft_type": "LORA",
|
| 19 |
+
"r": 16,
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
| 23 |
"gate_proj",
|
|
|
|
| 24 |
"v_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
+
"up_proj",
|
| 27 |
"k_proj",
|
| 28 |
+
"q_proj",
|
| 29 |
+
"down_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5f5168c85d041eb40b539ed1ea4fa405f9e14109cf8ce608d3eac8f26f627745
|
| 3 |
+
size 45118424
|
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 sha256:87fd6f0b4471cdc416e96626857e0c5e197d9b25e78f6d528666378e766fc69f
|
| 3 |
+
size 23159290
|
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:c683a96b3a33504bcf104bd66d70d07f59ed807698ad96230e879f4b6bf5d00a
|
| 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 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b4906f488285c3d93b3de9477b5cdb50810bb8e8a714368724d443d38a5757a
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,1704 +1,171 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"eval_steps":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate":
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"eval_loss": 1.
|
| 21 |
-
"eval_runtime":
|
| 22 |
-
"eval_samples_per_second":
|
| 23 |
-
"eval_steps_per_second":
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss": 1.
|
| 31 |
"step": 2
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm":
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 1.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
"step": 3
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm":
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 1.
|
| 45 |
"step": 4
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm":
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss": 1.
|
| 52 |
"step": 5
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm":
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss": 1.
|
| 59 |
"step": 6
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"
|
| 64 |
-
"
|
| 65 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 66 |
"step": 7
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm": 0.
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss": 1.
|
| 73 |
"step": 8
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"
|
| 78 |
-
"
|
| 79 |
-
"
|
| 80 |
-
"
|
| 81 |
-
"step": 8
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"
|
| 86 |
-
"
|
| 87 |
-
"
|
|
|
|
| 88 |
"step": 9
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"epoch": 0.
|
| 92 |
-
"grad_norm": 0.
|
| 93 |
-
"learning_rate":
|
| 94 |
-
"loss": 1.
|
| 95 |
"step": 10
|
| 96 |
},
|
| 97 |
{
|
| 98 |
-
"epoch":
|
| 99 |
-
"grad_norm": 0.
|
| 100 |
-
"learning_rate":
|
| 101 |
-
"loss": 1.
|
| 102 |
"step": 11
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"grad_norm": 0.
|
| 107 |
-
"learning_rate":
|
| 108 |
-
"loss": 1.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 109 |
"step": 12
|
| 110 |
},
|
| 111 |
{
|
| 112 |
-
"epoch":
|
| 113 |
-
"grad_norm": 0.
|
| 114 |
-
"learning_rate":
|
| 115 |
-
"loss": 1.
|
| 116 |
"step": 13
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
-
"grad_norm": 0.
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 1.
|
| 123 |
"step": 14
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"grad_norm": 0.
|
| 128 |
-
"learning_rate":
|
| 129 |
-
"loss": 1.
|
| 130 |
"step": 15
|
| 131 |
},
|
| 132 |
{
|
| 133 |
-
"epoch":
|
| 134 |
-
"
|
| 135 |
-
"
|
| 136 |
-
"
|
| 137 |
-
"
|
| 138 |
-
|
| 139 |
-
{
|
| 140 |
-
"epoch": 0.5039370078740157,
|
| 141 |
-
"eval_loss": 1.2311681509017944,
|
| 142 |
-
"eval_runtime": 1.979,
|
| 143 |
-
"eval_samples_per_second": 50.531,
|
| 144 |
-
"eval_steps_per_second": 3.537,
|
| 145 |
-
"step": 16
|
| 146 |
-
},
|
| 147 |
-
{
|
| 148 |
-
"epoch": 0.5354330708661418,
|
| 149 |
-
"grad_norm": 0.9032379984855652,
|
| 150 |
-
"learning_rate": 8.500000000000002e-06,
|
| 151 |
-
"loss": 1.3434,
|
| 152 |
-
"step": 17
|
| 153 |
-
},
|
| 154 |
-
{
|
| 155 |
-
"epoch": 0.5669291338582677,
|
| 156 |
-
"grad_norm": 0.859302282333374,
|
| 157 |
-
"learning_rate": 9e-06,
|
| 158 |
-
"loss": 1.3049,
|
| 159 |
-
"step": 18
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.5984251968503937,
|
| 163 |
-
"grad_norm": 0.7455488443374634,
|
| 164 |
-
"learning_rate": 9.5e-06,
|
| 165 |
-
"loss": 1.3192,
|
| 166 |
-
"step": 19
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 0.6299212598425197,
|
| 170 |
-
"grad_norm": 0.713253915309906,
|
| 171 |
-
"learning_rate": 1e-05,
|
| 172 |
-
"loss": 1.2799,
|
| 173 |
-
"step": 20
|
| 174 |
-
},
|
| 175 |
-
{
|
| 176 |
-
"epoch": 0.6614173228346457,
|
| 177 |
-
"grad_norm": 0.5953424572944641,
|
| 178 |
-
"learning_rate": 1.05e-05,
|
| 179 |
-
"loss": 1.1514,
|
| 180 |
-
"step": 21
|
| 181 |
-
},
|
| 182 |
-
{
|
| 183 |
-
"epoch": 0.6929133858267716,
|
| 184 |
-
"grad_norm": 0.6268596053123474,
|
| 185 |
-
"learning_rate": 1.1000000000000001e-05,
|
| 186 |
-
"loss": 1.2747,
|
| 187 |
-
"step": 22
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 0.7244094488188977,
|
| 191 |
-
"grad_norm": 0.722594141960144,
|
| 192 |
-
"learning_rate": 1.1500000000000002e-05,
|
| 193 |
-
"loss": 1.2532,
|
| 194 |
-
"step": 23
|
| 195 |
-
},
|
| 196 |
-
{
|
| 197 |
-
"epoch": 0.7559055118110236,
|
| 198 |
-
"grad_norm": 0.5606786012649536,
|
| 199 |
-
"learning_rate": 1.2e-05,
|
| 200 |
-
"loss": 1.2336,
|
| 201 |
-
"step": 24
|
| 202 |
-
},
|
| 203 |
-
{
|
| 204 |
-
"epoch": 0.7559055118110236,
|
| 205 |
-
"eval_loss": 1.2047480344772339,
|
| 206 |
-
"eval_runtime": 1.9744,
|
| 207 |
-
"eval_samples_per_second": 50.649,
|
| 208 |
-
"eval_steps_per_second": 3.545,
|
| 209 |
-
"step": 24
|
| 210 |
-
},
|
| 211 |
-
{
|
| 212 |
-
"epoch": 0.7874015748031497,
|
| 213 |
-
"grad_norm": 0.5434914827346802,
|
| 214 |
-
"learning_rate": 1.25e-05,
|
| 215 |
-
"loss": 1.3273,
|
| 216 |
-
"step": 25
|
| 217 |
-
},
|
| 218 |
-
{
|
| 219 |
-
"epoch": 0.8188976377952756,
|
| 220 |
-
"grad_norm": 0.589859664440155,
|
| 221 |
-
"learning_rate": 1.3000000000000001e-05,
|
| 222 |
-
"loss": 1.2225,
|
| 223 |
-
"step": 26
|
| 224 |
-
},
|
| 225 |
-
{
|
| 226 |
-
"epoch": 0.8503937007874016,
|
| 227 |
-
"grad_norm": 0.6487225294113159,
|
| 228 |
-
"learning_rate": 1.3500000000000001e-05,
|
| 229 |
-
"loss": 1.2899,
|
| 230 |
-
"step": 27
|
| 231 |
-
},
|
| 232 |
-
{
|
| 233 |
-
"epoch": 0.8818897637795275,
|
| 234 |
-
"grad_norm": 0.5731435418128967,
|
| 235 |
-
"learning_rate": 1.4000000000000001e-05,
|
| 236 |
-
"loss": 1.237,
|
| 237 |
-
"step": 28
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 0.9133858267716536,
|
| 241 |
-
"grad_norm": 0.6190696358680725,
|
| 242 |
-
"learning_rate": 1.45e-05,
|
| 243 |
-
"loss": 1.2299,
|
| 244 |
-
"step": 29
|
| 245 |
-
},
|
| 246 |
-
{
|
| 247 |
-
"epoch": 0.9448818897637795,
|
| 248 |
-
"grad_norm": 0.548401951789856,
|
| 249 |
-
"learning_rate": 1.5e-05,
|
| 250 |
-
"loss": 1.2117,
|
| 251 |
-
"step": 30
|
| 252 |
-
},
|
| 253 |
-
{
|
| 254 |
-
"epoch": 0.9763779527559056,
|
| 255 |
-
"grad_norm": 0.6051440834999084,
|
| 256 |
-
"learning_rate": 1.55e-05,
|
| 257 |
-
"loss": 1.2396,
|
| 258 |
-
"step": 31
|
| 259 |
-
},
|
| 260 |
-
{
|
| 261 |
-
"epoch": 1.0078740157480315,
|
| 262 |
-
"grad_norm": 0.5788630247116089,
|
| 263 |
-
"learning_rate": 1.6000000000000003e-05,
|
| 264 |
-
"loss": 1.235,
|
| 265 |
-
"step": 32
|
| 266 |
-
},
|
| 267 |
-
{
|
| 268 |
-
"epoch": 1.0078740157480315,
|
| 269 |
-
"eval_loss": 1.1931809186935425,
|
| 270 |
-
"eval_runtime": 1.9767,
|
| 271 |
-
"eval_samples_per_second": 50.589,
|
| 272 |
-
"eval_steps_per_second": 3.541,
|
| 273 |
-
"step": 32
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 1.0393700787401574,
|
| 277 |
-
"grad_norm": 0.6048874258995056,
|
| 278 |
-
"learning_rate": 1.65e-05,
|
| 279 |
-
"loss": 1.2221,
|
| 280 |
-
"step": 33
|
| 281 |
-
},
|
| 282 |
-
{
|
| 283 |
-
"epoch": 1.0708661417322836,
|
| 284 |
-
"grad_norm": 0.6334845423698425,
|
| 285 |
-
"learning_rate": 1.7000000000000003e-05,
|
| 286 |
-
"loss": 1.2613,
|
| 287 |
-
"step": 34
|
| 288 |
-
},
|
| 289 |
-
{
|
| 290 |
-
"epoch": 1.1023622047244095,
|
| 291 |
-
"grad_norm": 0.5881842970848083,
|
| 292 |
-
"learning_rate": 1.75e-05,
|
| 293 |
-
"loss": 1.264,
|
| 294 |
-
"step": 35
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 1.1338582677165354,
|
| 298 |
-
"grad_norm": 0.5609626173973083,
|
| 299 |
-
"learning_rate": 1.8e-05,
|
| 300 |
-
"loss": 1.2696,
|
| 301 |
-
"step": 36
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 1.1653543307086613,
|
| 305 |
-
"grad_norm": 0.47871723771095276,
|
| 306 |
-
"learning_rate": 1.85e-05,
|
| 307 |
-
"loss": 1.1659,
|
| 308 |
-
"step": 37
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"epoch": 1.1968503937007875,
|
| 312 |
-
"grad_norm": 0.5133270025253296,
|
| 313 |
-
"learning_rate": 1.9e-05,
|
| 314 |
-
"loss": 1.1731,
|
| 315 |
-
"step": 38
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 1.2283464566929134,
|
| 319 |
-
"grad_norm": 0.49426957964897156,
|
| 320 |
-
"learning_rate": 1.9500000000000003e-05,
|
| 321 |
-
"loss": 1.0909,
|
| 322 |
-
"step": 39
|
| 323 |
-
},
|
| 324 |
-
{
|
| 325 |
-
"epoch": 1.2598425196850394,
|
| 326 |
-
"grad_norm": 0.4927002191543579,
|
| 327 |
-
"learning_rate": 2e-05,
|
| 328 |
-
"loss": 1.2591,
|
| 329 |
-
"step": 40
|
| 330 |
-
},
|
| 331 |
-
{
|
| 332 |
-
"epoch": 1.2598425196850394,
|
| 333 |
-
"eval_loss": 1.1798231601715088,
|
| 334 |
-
"eval_runtime": 2.2274,
|
| 335 |
-
"eval_samples_per_second": 44.895,
|
| 336 |
-
"eval_steps_per_second": 3.143,
|
| 337 |
-
"step": 40
|
| 338 |
-
},
|
| 339 |
-
{
|
| 340 |
-
"epoch": 1.2913385826771653,
|
| 341 |
-
"grad_norm": 0.5016794800758362,
|
| 342 |
-
"learning_rate": 2.05e-05,
|
| 343 |
-
"loss": 1.1344,
|
| 344 |
-
"step": 41
|
| 345 |
-
},
|
| 346 |
-
{
|
| 347 |
-
"epoch": 1.3228346456692912,
|
| 348 |
-
"grad_norm": 0.5014638304710388,
|
| 349 |
-
"learning_rate": 2.1e-05,
|
| 350 |
-
"loss": 1.1438,
|
| 351 |
-
"step": 42
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"epoch": 1.3543307086614174,
|
| 355 |
-
"grad_norm": 0.4870070517063141,
|
| 356 |
-
"learning_rate": 2.15e-05,
|
| 357 |
-
"loss": 1.2946,
|
| 358 |
-
"step": 43
|
| 359 |
-
},
|
| 360 |
-
{
|
| 361 |
-
"epoch": 1.3858267716535433,
|
| 362 |
-
"grad_norm": 0.5146998763084412,
|
| 363 |
-
"learning_rate": 2.2000000000000003e-05,
|
| 364 |
-
"loss": 1.122,
|
| 365 |
-
"step": 44
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"epoch": 1.4173228346456692,
|
| 369 |
-
"grad_norm": 0.5737994313240051,
|
| 370 |
-
"learning_rate": 2.25e-05,
|
| 371 |
-
"loss": 1.2749,
|
| 372 |
-
"step": 45
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"epoch": 1.4488188976377954,
|
| 376 |
-
"grad_norm": 0.5805953145027161,
|
| 377 |
-
"learning_rate": 2.3000000000000003e-05,
|
| 378 |
-
"loss": 1.2347,
|
| 379 |
-
"step": 46
|
| 380 |
-
},
|
| 381 |
-
{
|
| 382 |
-
"epoch": 1.4803149606299213,
|
| 383 |
-
"grad_norm": 0.5060011744499207,
|
| 384 |
-
"learning_rate": 2.35e-05,
|
| 385 |
-
"loss": 1.1806,
|
| 386 |
-
"step": 47
|
| 387 |
-
},
|
| 388 |
-
{
|
| 389 |
-
"epoch": 1.5118110236220472,
|
| 390 |
-
"grad_norm": 0.5228325128555298,
|
| 391 |
-
"learning_rate": 2.4e-05,
|
| 392 |
-
"loss": 1.1836,
|
| 393 |
-
"step": 48
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"epoch": 1.5118110236220472,
|
| 397 |
-
"eval_loss": 1.1722474098205566,
|
| 398 |
-
"eval_runtime": 1.9903,
|
| 399 |
-
"eval_samples_per_second": 50.243,
|
| 400 |
-
"eval_steps_per_second": 3.517,
|
| 401 |
-
"step": 48
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 1.5433070866141732,
|
| 405 |
-
"grad_norm": 0.49606096744537354,
|
| 406 |
-
"learning_rate": 2.45e-05,
|
| 407 |
-
"loss": 1.2222,
|
| 408 |
-
"step": 49
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 1.574803149606299,
|
| 412 |
-
"grad_norm": 0.37724393606185913,
|
| 413 |
-
"learning_rate": 2.5e-05,
|
| 414 |
-
"loss": 1.0979,
|
| 415 |
-
"step": 50
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 1.6062992125984252,
|
| 419 |
-
"grad_norm": 0.42948251962661743,
|
| 420 |
-
"learning_rate": 2.5500000000000003e-05,
|
| 421 |
-
"loss": 1.0678,
|
| 422 |
-
"step": 51
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 1.6377952755905512,
|
| 426 |
-
"grad_norm": 0.4671652317047119,
|
| 427 |
-
"learning_rate": 2.6000000000000002e-05,
|
| 428 |
-
"loss": 1.18,
|
| 429 |
-
"step": 52
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 1.6692913385826773,
|
| 433 |
-
"grad_norm": 0.44847941398620605,
|
| 434 |
-
"learning_rate": 2.6500000000000004e-05,
|
| 435 |
-
"loss": 1.2253,
|
| 436 |
-
"step": 53
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 1.7007874015748032,
|
| 440 |
-
"grad_norm": 0.4691849946975708,
|
| 441 |
-
"learning_rate": 2.7000000000000002e-05,
|
| 442 |
-
"loss": 1.1719,
|
| 443 |
-
"step": 54
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 1.7322834645669292,
|
| 447 |
-
"grad_norm": 0.505724310874939,
|
| 448 |
-
"learning_rate": 2.7500000000000004e-05,
|
| 449 |
-
"loss": 1.235,
|
| 450 |
-
"step": 55
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 1.763779527559055,
|
| 454 |
-
"grad_norm": 0.48790884017944336,
|
| 455 |
-
"learning_rate": 2.8000000000000003e-05,
|
| 456 |
-
"loss": 1.2719,
|
| 457 |
-
"step": 56
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 1.763779527559055,
|
| 461 |
-
"eval_loss": 1.1640315055847168,
|
| 462 |
-
"eval_runtime": 1.9729,
|
| 463 |
-
"eval_samples_per_second": 50.688,
|
| 464 |
-
"eval_steps_per_second": 3.548,
|
| 465 |
-
"step": 56
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.795275590551181,
|
| 469 |
-
"grad_norm": 0.4780273735523224,
|
| 470 |
-
"learning_rate": 2.8499999999999998e-05,
|
| 471 |
-
"loss": 1.3099,
|
| 472 |
-
"step": 57
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 1.826771653543307,
|
| 476 |
-
"grad_norm": 0.4578011929988861,
|
| 477 |
-
"learning_rate": 2.9e-05,
|
| 478 |
-
"loss": 1.3569,
|
| 479 |
-
"step": 58
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 1.858267716535433,
|
| 483 |
-
"grad_norm": 0.5303736925125122,
|
| 484 |
-
"learning_rate": 2.95e-05,
|
| 485 |
-
"loss": 1.145,
|
| 486 |
-
"step": 59
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 1.889763779527559,
|
| 490 |
-
"grad_norm": 0.5604854226112366,
|
| 491 |
-
"learning_rate": 3e-05,
|
| 492 |
-
"loss": 1.231,
|
| 493 |
-
"step": 60
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.9212598425196852,
|
| 497 |
-
"grad_norm": 0.4924694895744324,
|
| 498 |
-
"learning_rate": 3.05e-05,
|
| 499 |
-
"loss": 1.1343,
|
| 500 |
-
"step": 61
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.952755905511811,
|
| 504 |
-
"grad_norm": 0.4921957850456238,
|
| 505 |
-
"learning_rate": 3.1e-05,
|
| 506 |
-
"loss": 1.3127,
|
| 507 |
-
"step": 62
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.984251968503937,
|
| 511 |
-
"grad_norm": 0.5097357034683228,
|
| 512 |
-
"learning_rate": 3.15e-05,
|
| 513 |
-
"loss": 1.2336,
|
| 514 |
-
"step": 63
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 2.015748031496063,
|
| 518 |
-
"grad_norm": 0.46658650040626526,
|
| 519 |
-
"learning_rate": 3.2000000000000005e-05,
|
| 520 |
-
"loss": 1.2083,
|
| 521 |
-
"step": 64
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 2.015748031496063,
|
| 525 |
-
"eval_loss": 1.1596944332122803,
|
| 526 |
-
"eval_runtime": 1.9712,
|
| 527 |
-
"eval_samples_per_second": 50.731,
|
| 528 |
-
"eval_steps_per_second": 3.551,
|
| 529 |
-
"step": 64
|
| 530 |
-
},
|
| 531 |
-
{
|
| 532 |
-
"epoch": 2.047244094488189,
|
| 533 |
-
"grad_norm": 0.46166104078292847,
|
| 534 |
-
"learning_rate": 3.2500000000000004e-05,
|
| 535 |
-
"loss": 1.1131,
|
| 536 |
-
"step": 65
|
| 537 |
-
},
|
| 538 |
-
{
|
| 539 |
-
"epoch": 2.078740157480315,
|
| 540 |
-
"grad_norm": 0.5125811696052551,
|
| 541 |
-
"learning_rate": 3.3e-05,
|
| 542 |
-
"loss": 1.0556,
|
| 543 |
-
"step": 66
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 2.1102362204724407,
|
| 547 |
-
"grad_norm": 0.5640822649002075,
|
| 548 |
-
"learning_rate": 3.35e-05,
|
| 549 |
-
"loss": 1.2563,
|
| 550 |
-
"step": 67
|
| 551 |
-
},
|
| 552 |
-
{
|
| 553 |
-
"epoch": 2.141732283464567,
|
| 554 |
-
"grad_norm": 0.5031111836433411,
|
| 555 |
-
"learning_rate": 3.4000000000000007e-05,
|
| 556 |
-
"loss": 1.0555,
|
| 557 |
-
"step": 68
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 2.173228346456693,
|
| 561 |
-
"grad_norm": 0.5319817066192627,
|
| 562 |
-
"learning_rate": 3.45e-05,
|
| 563 |
-
"loss": 1.149,
|
| 564 |
-
"step": 69
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 2.204724409448819,
|
| 568 |
-
"grad_norm": 0.49313291907310486,
|
| 569 |
-
"learning_rate": 3.5e-05,
|
| 570 |
-
"loss": 1.1457,
|
| 571 |
-
"step": 70
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 2.236220472440945,
|
| 575 |
-
"grad_norm": 0.48414379358291626,
|
| 576 |
-
"learning_rate": 3.55e-05,
|
| 577 |
-
"loss": 1.2644,
|
| 578 |
-
"step": 71
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 2.267716535433071,
|
| 582 |
-
"grad_norm": 0.5062035918235779,
|
| 583 |
-
"learning_rate": 3.6e-05,
|
| 584 |
-
"loss": 1.2119,
|
| 585 |
-
"step": 72
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 2.267716535433071,
|
| 589 |
-
"eval_loss": 1.1549252271652222,
|
| 590 |
-
"eval_runtime": 1.9635,
|
| 591 |
-
"eval_samples_per_second": 50.93,
|
| 592 |
-
"eval_steps_per_second": 3.565,
|
| 593 |
-
"step": 72
|
| 594 |
-
},
|
| 595 |
-
{
|
| 596 |
-
"epoch": 2.2992125984251968,
|
| 597 |
-
"grad_norm": 0.47118309140205383,
|
| 598 |
-
"learning_rate": 3.65e-05,
|
| 599 |
-
"loss": 1.1578,
|
| 600 |
-
"step": 73
|
| 601 |
-
},
|
| 602 |
-
{
|
| 603 |
-
"epoch": 2.3307086614173227,
|
| 604 |
-
"grad_norm": 0.5640192627906799,
|
| 605 |
-
"learning_rate": 3.7e-05,
|
| 606 |
-
"loss": 1.1214,
|
| 607 |
-
"step": 74
|
| 608 |
-
},
|
| 609 |
-
{
|
| 610 |
-
"epoch": 2.362204724409449,
|
| 611 |
-
"grad_norm": 0.4832814633846283,
|
| 612 |
-
"learning_rate": 3.7500000000000003e-05,
|
| 613 |
-
"loss": 1.2154,
|
| 614 |
-
"step": 75
|
| 615 |
-
},
|
| 616 |
-
{
|
| 617 |
-
"epoch": 2.393700787401575,
|
| 618 |
-
"grad_norm": 0.48993921279907227,
|
| 619 |
-
"learning_rate": 3.8e-05,
|
| 620 |
-
"loss": 1.1292,
|
| 621 |
-
"step": 76
|
| 622 |
-
},
|
| 623 |
-
{
|
| 624 |
-
"epoch": 2.425196850393701,
|
| 625 |
-
"grad_norm": 0.4585767388343811,
|
| 626 |
-
"learning_rate": 3.85e-05,
|
| 627 |
-
"loss": 1.2205,
|
| 628 |
-
"step": 77
|
| 629 |
-
},
|
| 630 |
-
{
|
| 631 |
-
"epoch": 2.456692913385827,
|
| 632 |
-
"grad_norm": 0.47851210832595825,
|
| 633 |
-
"learning_rate": 3.9000000000000006e-05,
|
| 634 |
-
"loss": 1.1434,
|
| 635 |
-
"step": 78
|
| 636 |
-
},
|
| 637 |
-
{
|
| 638 |
-
"epoch": 2.4881889763779528,
|
| 639 |
-
"grad_norm": 0.48592087626457214,
|
| 640 |
-
"learning_rate": 3.9500000000000005e-05,
|
| 641 |
-
"loss": 1.1855,
|
| 642 |
-
"step": 79
|
| 643 |
-
},
|
| 644 |
-
{
|
| 645 |
-
"epoch": 2.5196850393700787,
|
| 646 |
-
"grad_norm": 0.5699076652526855,
|
| 647 |
-
"learning_rate": 4e-05,
|
| 648 |
-
"loss": 1.2574,
|
| 649 |
-
"step": 80
|
| 650 |
-
},
|
| 651 |
-
{
|
| 652 |
-
"epoch": 2.5196850393700787,
|
| 653 |
-
"eval_loss": 1.15181565284729,
|
| 654 |
-
"eval_runtime": 2.2403,
|
| 655 |
-
"eval_samples_per_second": 44.638,
|
| 656 |
-
"eval_steps_per_second": 3.125,
|
| 657 |
-
"step": 80
|
| 658 |
-
},
|
| 659 |
-
{
|
| 660 |
-
"epoch": 2.5511811023622046,
|
| 661 |
-
"grad_norm": 0.5299522876739502,
|
| 662 |
-
"learning_rate": 4.05e-05,
|
| 663 |
-
"loss": 1.0995,
|
| 664 |
-
"step": 81
|
| 665 |
-
},
|
| 666 |
-
{
|
| 667 |
-
"epoch": 2.5826771653543306,
|
| 668 |
-
"grad_norm": 0.5705897808074951,
|
| 669 |
-
"learning_rate": 4.1e-05,
|
| 670 |
-
"loss": 1.2274,
|
| 671 |
-
"step": 82
|
| 672 |
-
},
|
| 673 |
-
{
|
| 674 |
-
"epoch": 2.6141732283464565,
|
| 675 |
-
"grad_norm": 0.48036718368530273,
|
| 676 |
-
"learning_rate": 4.15e-05,
|
| 677 |
-
"loss": 1.0861,
|
| 678 |
-
"step": 83
|
| 679 |
-
},
|
| 680 |
-
{
|
| 681 |
-
"epoch": 2.6456692913385824,
|
| 682 |
-
"grad_norm": 0.49471601843833923,
|
| 683 |
-
"learning_rate": 4.2e-05,
|
| 684 |
-
"loss": 1.1628,
|
| 685 |
-
"step": 84
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 2.677165354330709,
|
| 689 |
-
"grad_norm": 0.49998700618743896,
|
| 690 |
-
"learning_rate": 4.25e-05,
|
| 691 |
-
"loss": 1.0709,
|
| 692 |
-
"step": 85
|
| 693 |
-
},
|
| 694 |
-
{
|
| 695 |
-
"epoch": 2.7086614173228347,
|
| 696 |
-
"grad_norm": 0.5527383685112,
|
| 697 |
-
"learning_rate": 4.3e-05,
|
| 698 |
-
"loss": 1.201,
|
| 699 |
-
"step": 86
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 2.7401574803149606,
|
| 703 |
-
"grad_norm": 0.6283419728279114,
|
| 704 |
-
"learning_rate": 4.35e-05,
|
| 705 |
-
"loss": 1.237,
|
| 706 |
-
"step": 87
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 2.7716535433070866,
|
| 710 |
-
"grad_norm": 0.6050185561180115,
|
| 711 |
-
"learning_rate": 4.4000000000000006e-05,
|
| 712 |
-
"loss": 1.025,
|
| 713 |
-
"step": 88
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"epoch": 2.7716535433070866,
|
| 717 |
-
"eval_loss": 1.1485306024551392,
|
| 718 |
-
"eval_runtime": 2.0753,
|
| 719 |
-
"eval_samples_per_second": 48.185,
|
| 720 |
-
"eval_steps_per_second": 3.373,
|
| 721 |
-
"step": 88
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 2.8031496062992125,
|
| 725 |
-
"grad_norm": 0.5350843667984009,
|
| 726 |
-
"learning_rate": 4.4500000000000004e-05,
|
| 727 |
-
"loss": 1.1582,
|
| 728 |
-
"step": 89
|
| 729 |
-
},
|
| 730 |
-
{
|
| 731 |
-
"epoch": 2.8346456692913384,
|
| 732 |
-
"grad_norm": 0.5085338950157166,
|
| 733 |
-
"learning_rate": 4.5e-05,
|
| 734 |
-
"loss": 1.254,
|
| 735 |
-
"step": 90
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 2.866141732283465,
|
| 739 |
-
"grad_norm": 0.5343408584594727,
|
| 740 |
-
"learning_rate": 4.55e-05,
|
| 741 |
-
"loss": 1.0941,
|
| 742 |
-
"step": 91
|
| 743 |
-
},
|
| 744 |
-
{
|
| 745 |
-
"epoch": 2.8976377952755907,
|
| 746 |
-
"grad_norm": 0.5207767486572266,
|
| 747 |
-
"learning_rate": 4.600000000000001e-05,
|
| 748 |
-
"loss": 1.176,
|
| 749 |
-
"step": 92
|
| 750 |
-
},
|
| 751 |
-
{
|
| 752 |
-
"epoch": 2.9291338582677167,
|
| 753 |
-
"grad_norm": 0.6931973695755005,
|
| 754 |
-
"learning_rate": 4.6500000000000005e-05,
|
| 755 |
-
"loss": 1.2044,
|
| 756 |
-
"step": 93
|
| 757 |
-
},
|
| 758 |
-
{
|
| 759 |
-
"epoch": 2.9606299212598426,
|
| 760 |
-
"grad_norm": 0.5525833964347839,
|
| 761 |
-
"learning_rate": 4.7e-05,
|
| 762 |
-
"loss": 1.1156,
|
| 763 |
-
"step": 94
|
| 764 |
-
},
|
| 765 |
-
{
|
| 766 |
-
"epoch": 2.9921259842519685,
|
| 767 |
-
"grad_norm": 0.5206693410873413,
|
| 768 |
-
"learning_rate": 4.75e-05,
|
| 769 |
-
"loss": 1.1831,
|
| 770 |
-
"step": 95
|
| 771 |
-
},
|
| 772 |
-
{
|
| 773 |
-
"epoch": 3.0236220472440944,
|
| 774 |
-
"grad_norm": 0.6194021105766296,
|
| 775 |
-
"learning_rate": 4.8e-05,
|
| 776 |
-
"loss": 1.2232,
|
| 777 |
-
"step": 96
|
| 778 |
-
},
|
| 779 |
-
{
|
| 780 |
-
"epoch": 3.0236220472440944,
|
| 781 |
-
"eval_loss": 1.1487455368041992,
|
| 782 |
-
"eval_runtime": 1.9591,
|
| 783 |
-
"eval_samples_per_second": 51.044,
|
| 784 |
-
"eval_steps_per_second": 3.573,
|
| 785 |
-
"step": 96
|
| 786 |
-
},
|
| 787 |
-
{
|
| 788 |
-
"epoch": 3.0551181102362204,
|
| 789 |
-
"grad_norm": 0.47633522748947144,
|
| 790 |
-
"learning_rate": 4.85e-05,
|
| 791 |
-
"loss": 1.0667,
|
| 792 |
-
"step": 97
|
| 793 |
-
},
|
| 794 |
-
{
|
| 795 |
-
"epoch": 3.0866141732283463,
|
| 796 |
-
"grad_norm": 0.49245792627334595,
|
| 797 |
-
"learning_rate": 4.9e-05,
|
| 798 |
-
"loss": 1.0812,
|
| 799 |
-
"step": 98
|
| 800 |
-
},
|
| 801 |
-
{
|
| 802 |
-
"epoch": 3.1181102362204722,
|
| 803 |
-
"grad_norm": 0.49651336669921875,
|
| 804 |
-
"learning_rate": 4.9500000000000004e-05,
|
| 805 |
-
"loss": 1.1153,
|
| 806 |
-
"step": 99
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 3.1496062992125986,
|
| 810 |
-
"grad_norm": 0.5725173354148865,
|
| 811 |
-
"learning_rate": 5e-05,
|
| 812 |
-
"loss": 1.1583,
|
| 813 |
-
"step": 100
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 3.1811023622047245,
|
| 817 |
-
"grad_norm": 0.5892531871795654,
|
| 818 |
-
"learning_rate": 4.9999567360675626e-05,
|
| 819 |
-
"loss": 1.1323,
|
| 820 |
-
"step": 101
|
| 821 |
-
},
|
| 822 |
-
{
|
| 823 |
-
"epoch": 3.2125984251968505,
|
| 824 |
-
"grad_norm": 0.5684159994125366,
|
| 825 |
-
"learning_rate": 4.999826945767665e-05,
|
| 826 |
-
"loss": 1.2006,
|
| 827 |
-
"step": 102
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 3.2440944881889764,
|
| 831 |
-
"grad_norm": 0.5696210265159607,
|
| 832 |
-
"learning_rate": 4.999610633592496e-05,
|
| 833 |
-
"loss": 1.1272,
|
| 834 |
-
"step": 103
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 3.2755905511811023,
|
| 838 |
-
"grad_norm": 0.6086538434028625,
|
| 839 |
-
"learning_rate": 4.999307807028871e-05,
|
| 840 |
-
"loss": 1.1621,
|
| 841 |
-
"step": 104
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 3.2755905511811023,
|
| 845 |
-
"eval_loss": 1.1464533805847168,
|
| 846 |
-
"eval_runtime": 1.9521,
|
| 847 |
-
"eval_samples_per_second": 51.226,
|
| 848 |
-
"eval_steps_per_second": 3.586,
|
| 849 |
-
"step": 104
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 3.3070866141732282,
|
| 853 |
-
"grad_norm": 0.543544590473175,
|
| 854 |
-
"learning_rate": 4.998918476557963e-05,
|
| 855 |
-
"loss": 1.1049,
|
| 856 |
-
"step": 105
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 3.338582677165354,
|
| 860 |
-
"grad_norm": 0.6197894811630249,
|
| 861 |
-
"learning_rate": 4.9984426556549456e-05,
|
| 862 |
-
"loss": 1.0589,
|
| 863 |
-
"step": 106
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 3.3700787401574805,
|
| 867 |
-
"grad_norm": 0.553490400314331,
|
| 868 |
-
"learning_rate": 4.997880360788526e-05,
|
| 869 |
-
"loss": 1.0755,
|
| 870 |
-
"step": 107
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 3.4015748031496065,
|
| 874 |
-
"grad_norm": 0.5474947094917297,
|
| 875 |
-
"learning_rate": 4.997231611420373e-05,
|
| 876 |
-
"loss": 0.9718,
|
| 877 |
-
"step": 108
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 3.4330708661417324,
|
| 881 |
-
"grad_norm": 0.6409115791320801,
|
| 882 |
-
"learning_rate": 4.996496430004446e-05,
|
| 883 |
-
"loss": 1.0829,
|
| 884 |
-
"step": 109
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 3.4645669291338583,
|
| 888 |
-
"grad_norm": 0.5846779346466064,
|
| 889 |
-
"learning_rate": 4.995674841986217e-05,
|
| 890 |
-
"loss": 1.0633,
|
| 891 |
-
"step": 110
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 3.4960629921259843,
|
| 895 |
-
"grad_norm": 0.6325021982192993,
|
| 896 |
-
"learning_rate": 4.9947668758017884e-05,
|
| 897 |
-
"loss": 1.0721,
|
| 898 |
-
"step": 111
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 3.52755905511811,
|
| 902 |
-
"grad_norm": 0.6479124426841736,
|
| 903 |
-
"learning_rate": 4.9937725628769094e-05,
|
| 904 |
-
"loss": 1.031,
|
| 905 |
-
"step": 112
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 3.52755905511811,
|
| 909 |
-
"eval_loss": 1.1493302583694458,
|
| 910 |
-
"eval_runtime": 2.0087,
|
| 911 |
-
"eval_samples_per_second": 49.783,
|
| 912 |
-
"eval_steps_per_second": 3.485,
|
| 913 |
-
"step": 112
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 3.559055118110236,
|
| 917 |
-
"grad_norm": 0.663337767124176,
|
| 918 |
-
"learning_rate": 4.9926919376258916e-05,
|
| 919 |
-
"loss": 1.0875,
|
| 920 |
-
"step": 113
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 3.590551181102362,
|
| 924 |
-
"grad_norm": 0.609626293182373,
|
| 925 |
-
"learning_rate": 4.991525037450412e-05,
|
| 926 |
-
"loss": 1.0576,
|
| 927 |
-
"step": 114
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 3.622047244094488,
|
| 931 |
-
"grad_norm": 0.7602643966674805,
|
| 932 |
-
"learning_rate": 4.990271902738223e-05,
|
| 933 |
-
"loss": 1.1482,
|
| 934 |
-
"step": 115
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 3.653543307086614,
|
| 938 |
-
"grad_norm": 0.7129984498023987,
|
| 939 |
-
"learning_rate": 4.9889325768617536e-05,
|
| 940 |
-
"loss": 1.0728,
|
| 941 |
-
"step": 116
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 3.6850393700787403,
|
| 945 |
-
"grad_norm": 0.6811193823814392,
|
| 946 |
-
"learning_rate": 4.987507106176606e-05,
|
| 947 |
-
"loss": 1.126,
|
| 948 |
-
"step": 117
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 3.716535433070866,
|
| 952 |
-
"grad_norm": 0.6760783195495605,
|
| 953 |
-
"learning_rate": 4.985995540019955e-05,
|
| 954 |
-
"loss": 1.3472,
|
| 955 |
-
"step": 118
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 3.748031496062992,
|
| 959 |
-
"grad_norm": 0.7298296689987183,
|
| 960 |
-
"learning_rate": 4.984397930708838e-05,
|
| 961 |
-
"loss": 1.1568,
|
| 962 |
-
"step": 119
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 3.779527559055118,
|
| 966 |
-
"grad_norm": 0.6873138546943665,
|
| 967 |
-
"learning_rate": 4.982714333538343e-05,
|
| 968 |
-
"loss": 1.0529,
|
| 969 |
-
"step": 120
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 3.779527559055118,
|
| 973 |
-
"eval_loss": 1.1472485065460205,
|
| 974 |
-
"eval_runtime": 2.0184,
|
| 975 |
-
"eval_samples_per_second": 49.543,
|
| 976 |
-
"eval_steps_per_second": 3.468,
|
| 977 |
-
"step": 120
|
| 978 |
-
},
|
| 979 |
-
{
|
| 980 |
-
"epoch": 3.811023622047244,
|
| 981 |
-
"grad_norm": 0.5837106108665466,
|
| 982 |
-
"learning_rate": 4.9809448067796974e-05,
|
| 983 |
-
"loss": 1.1478,
|
| 984 |
-
"step": 121
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 3.84251968503937,
|
| 988 |
-
"grad_norm": 0.721906304359436,
|
| 989 |
-
"learning_rate": 4.9790894116782514e-05,
|
| 990 |
-
"loss": 1.2416,
|
| 991 |
-
"step": 122
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 3.8740157480314963,
|
| 995 |
-
"grad_norm": 0.5964561700820923,
|
| 996 |
-
"learning_rate": 4.977148212451354e-05,
|
| 997 |
-
"loss": 1.1178,
|
| 998 |
-
"step": 123
|
| 999 |
-
},
|
| 1000 |
-
{
|
| 1001 |
-
"epoch": 3.905511811023622,
|
| 1002 |
-
"grad_norm": 0.6578372120857239,
|
| 1003 |
-
"learning_rate": 4.975121276286136e-05,
|
| 1004 |
-
"loss": 1.1086,
|
| 1005 |
-
"step": 124
|
| 1006 |
-
},
|
| 1007 |
-
{
|
| 1008 |
-
"epoch": 3.937007874015748,
|
| 1009 |
-
"grad_norm": 0.7385239601135254,
|
| 1010 |
-
"learning_rate": 4.973008673337181e-05,
|
| 1011 |
-
"loss": 1.1771,
|
| 1012 |
-
"step": 125
|
| 1013 |
-
},
|
| 1014 |
-
{
|
| 1015 |
-
"epoch": 3.968503937007874,
|
| 1016 |
-
"grad_norm": 0.648508608341217,
|
| 1017 |
-
"learning_rate": 4.970810476724097e-05,
|
| 1018 |
-
"loss": 1.1603,
|
| 1019 |
-
"step": 126
|
| 1020 |
-
},
|
| 1021 |
-
{
|
| 1022 |
-
"epoch": 4.0,
|
| 1023 |
-
"grad_norm": 0.7525383830070496,
|
| 1024 |
-
"learning_rate": 4.9685267625289886e-05,
|
| 1025 |
-
"loss": 1.1405,
|
| 1026 |
-
"step": 127
|
| 1027 |
-
},
|
| 1028 |
-
{
|
| 1029 |
-
"epoch": 4.031496062992126,
|
| 1030 |
-
"grad_norm": 0.6271669864654541,
|
| 1031 |
-
"learning_rate": 4.96615760979382e-05,
|
| 1032 |
-
"loss": 1.1122,
|
| 1033 |
-
"step": 128
|
| 1034 |
-
},
|
| 1035 |
-
{
|
| 1036 |
-
"epoch": 4.031496062992126,
|
| 1037 |
-
"eval_loss": 1.1474381685256958,
|
| 1038 |
-
"eval_runtime": 1.9734,
|
| 1039 |
-
"eval_samples_per_second": 50.675,
|
| 1040 |
-
"eval_steps_per_second": 3.547,
|
| 1041 |
-
"step": 128
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 4.062992125984252,
|
| 1045 |
-
"grad_norm": 0.6840550899505615,
|
| 1046 |
-
"learning_rate": 4.963703100517684e-05,
|
| 1047 |
-
"loss": 0.9849,
|
| 1048 |
-
"step": 129
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 4.094488188976378,
|
| 1052 |
-
"grad_norm": 0.6858397722244263,
|
| 1053 |
-
"learning_rate": 4.9611633196539584e-05,
|
| 1054 |
-
"loss": 1.095,
|
| 1055 |
-
"step": 130
|
| 1056 |
-
},
|
| 1057 |
-
{
|
| 1058 |
-
"epoch": 4.125984251968504,
|
| 1059 |
-
"grad_norm": 0.8086863160133362,
|
| 1060 |
-
"learning_rate": 4.9585383551073694e-05,
|
| 1061 |
-
"loss": 1.0421,
|
| 1062 |
-
"step": 131
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 4.15748031496063,
|
| 1066 |
-
"grad_norm": 0.6306203007698059,
|
| 1067 |
-
"learning_rate": 4.955828297730949e-05,
|
| 1068 |
-
"loss": 1.0594,
|
| 1069 |
-
"step": 132
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 4.188976377952756,
|
| 1073 |
-
"grad_norm": 0.6755027174949646,
|
| 1074 |
-
"learning_rate": 4.953033241322886e-05,
|
| 1075 |
-
"loss": 1.1747,
|
| 1076 |
-
"step": 133
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 4.2204724409448815,
|
| 1080 |
-
"grad_norm": 0.7362006902694702,
|
| 1081 |
-
"learning_rate": 4.950153282623289e-05,
|
| 1082 |
-
"loss": 1.1125,
|
| 1083 |
-
"step": 134
|
| 1084 |
-
},
|
| 1085 |
-
{
|
| 1086 |
-
"epoch": 4.251968503937007,
|
| 1087 |
-
"grad_norm": 0.6847573518753052,
|
| 1088 |
-
"learning_rate": 4.9471885213108274e-05,
|
| 1089 |
-
"loss": 1.0188,
|
| 1090 |
-
"step": 135
|
| 1091 |
-
},
|
| 1092 |
-
{
|
| 1093 |
-
"epoch": 4.283464566929134,
|
| 1094 |
-
"grad_norm": 0.6339418292045593,
|
| 1095 |
-
"learning_rate": 4.9441390599992864e-05,
|
| 1096 |
-
"loss": 1.0234,
|
| 1097 |
-
"step": 136
|
| 1098 |
-
},
|
| 1099 |
-
{
|
| 1100 |
-
"epoch": 4.283464566929134,
|
| 1101 |
-
"eval_loss": 1.1559439897537231,
|
| 1102 |
-
"eval_runtime": 1.9613,
|
| 1103 |
-
"eval_samples_per_second": 50.987,
|
| 1104 |
-
"eval_steps_per_second": 3.569,
|
| 1105 |
-
"step": 136
|
| 1106 |
-
},
|
| 1107 |
-
{
|
| 1108 |
-
"epoch": 4.31496062992126,
|
| 1109 |
-
"grad_norm": 0.7027537822723389,
|
| 1110 |
-
"learning_rate": 4.941005004234018e-05,
|
| 1111 |
-
"loss": 1.1565,
|
| 1112 |
-
"step": 137
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 4.346456692913386,
|
| 1116 |
-
"grad_norm": 0.7245253324508667,
|
| 1117 |
-
"learning_rate": 4.937786462488284e-05,
|
| 1118 |
-
"loss": 1.2024,
|
| 1119 |
-
"step": 138
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 4.377952755905512,
|
| 1123 |
-
"grad_norm": 0.7252236008644104,
|
| 1124 |
-
"learning_rate": 4.9344835461595014e-05,
|
| 1125 |
-
"loss": 1.0075,
|
| 1126 |
-
"step": 139
|
| 1127 |
-
},
|
| 1128 |
-
{
|
| 1129 |
-
"epoch": 4.409448818897638,
|
| 1130 |
-
"grad_norm": 0.6999779939651489,
|
| 1131 |
-
"learning_rate": 4.93109636956539e-05,
|
| 1132 |
-
"loss": 0.9823,
|
| 1133 |
-
"step": 140
|
| 1134 |
-
},
|
| 1135 |
-
{
|
| 1136 |
-
"epoch": 4.440944881889764,
|
| 1137 |
-
"grad_norm": 0.7879251837730408,
|
| 1138 |
-
"learning_rate": 4.927625049940013e-05,
|
| 1139 |
-
"loss": 1.0108,
|
| 1140 |
-
"step": 141
|
| 1141 |
-
},
|
| 1142 |
-
{
|
| 1143 |
-
"epoch": 4.47244094488189,
|
| 1144 |
-
"grad_norm": 0.7063820958137512,
|
| 1145 |
-
"learning_rate": 4.9240697074297206e-05,
|
| 1146 |
-
"loss": 1.043,
|
| 1147 |
-
"step": 142
|
| 1148 |
-
},
|
| 1149 |
-
{
|
| 1150 |
-
"epoch": 4.503937007874016,
|
| 1151 |
-
"grad_norm": 0.710659921169281,
|
| 1152 |
-
"learning_rate": 4.9204304650889915e-05,
|
| 1153 |
-
"loss": 1.0339,
|
| 1154 |
-
"step": 143
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 4.535433070866142,
|
| 1158 |
-
"grad_norm": 0.772346019744873,
|
| 1159 |
-
"learning_rate": 4.9167074488761735e-05,
|
| 1160 |
-
"loss": 0.9867,
|
| 1161 |
-
"step": 144
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 4.535433070866142,
|
| 1165 |
-
"eval_loss": 1.1680593490600586,
|
| 1166 |
-
"eval_runtime": 1.9681,
|
| 1167 |
-
"eval_samples_per_second": 50.81,
|
| 1168 |
-
"eval_steps_per_second": 3.557,
|
| 1169 |
-
"step": 144
|
| 1170 |
-
},
|
| 1171 |
-
{
|
| 1172 |
-
"epoch": 4.566929133858268,
|
| 1173 |
-
"grad_norm": 0.8781888484954834,
|
| 1174 |
-
"learning_rate": 4.912900787649124e-05,
|
| 1175 |
-
"loss": 1.0622,
|
| 1176 |
-
"step": 145
|
| 1177 |
-
},
|
| 1178 |
-
{
|
| 1179 |
-
"epoch": 4.5984251968503935,
|
| 1180 |
-
"grad_norm": 0.9489847421646118,
|
| 1181 |
-
"learning_rate": 4.90901061316075e-05,
|
| 1182 |
-
"loss": 1.0378,
|
| 1183 |
-
"step": 146
|
| 1184 |
-
},
|
| 1185 |
-
{
|
| 1186 |
-
"epoch": 4.6299212598425195,
|
| 1187 |
-
"grad_norm": 0.8057307600975037,
|
| 1188 |
-
"learning_rate": 4.90503706005445e-05,
|
| 1189 |
-
"loss": 1.0674,
|
| 1190 |
-
"step": 147
|
| 1191 |
-
},
|
| 1192 |
-
{
|
| 1193 |
-
"epoch": 4.661417322834645,
|
| 1194 |
-
"grad_norm": 0.9859374761581421,
|
| 1195 |
-
"learning_rate": 4.900980265859448e-05,
|
| 1196 |
-
"loss": 1.0841,
|
| 1197 |
-
"step": 148
|
| 1198 |
-
},
|
| 1199 |
-
{
|
| 1200 |
-
"epoch": 4.692913385826771,
|
| 1201 |
-
"grad_norm": 0.8140759468078613,
|
| 1202 |
-
"learning_rate": 4.896840370986042e-05,
|
| 1203 |
-
"loss": 1.0617,
|
| 1204 |
-
"step": 149
|
| 1205 |
-
},
|
| 1206 |
-
{
|
| 1207 |
-
"epoch": 4.724409448818898,
|
| 1208 |
-
"grad_norm": 0.8298155665397644,
|
| 1209 |
-
"learning_rate": 4.892617518720737e-05,
|
| 1210 |
-
"loss": 1.0421,
|
| 1211 |
-
"step": 150
|
| 1212 |
-
},
|
| 1213 |
-
{
|
| 1214 |
-
"epoch": 4.755905511811024,
|
| 1215 |
-
"grad_norm": 0.8481395840644836,
|
| 1216 |
-
"learning_rate": 4.888311855221289e-05,
|
| 1217 |
-
"loss": 1.0279,
|
| 1218 |
-
"step": 151
|
| 1219 |
-
},
|
| 1220 |
-
{
|
| 1221 |
-
"epoch": 4.78740157480315,
|
| 1222 |
-
"grad_norm": 0.7747954726219177,
|
| 1223 |
-
"learning_rate": 4.883923529511646e-05,
|
| 1224 |
-
"loss": 0.9092,
|
| 1225 |
-
"step": 152
|
| 1226 |
-
},
|
| 1227 |
-
{
|
| 1228 |
-
"epoch": 4.78740157480315,
|
| 1229 |
-
"eval_loss": 1.1637229919433594,
|
| 1230 |
-
"eval_runtime": 1.9694,
|
| 1231 |
-
"eval_samples_per_second": 50.777,
|
| 1232 |
-
"eval_steps_per_second": 3.554,
|
| 1233 |
-
"step": 152
|
| 1234 |
-
},
|
| 1235 |
-
{
|
| 1236 |
-
"epoch": 4.818897637795276,
|
| 1237 |
-
"grad_norm": 0.8691744208335876,
|
| 1238 |
-
"learning_rate": 4.8794526934767894e-05,
|
| 1239 |
-
"loss": 1.0147,
|
| 1240 |
-
"step": 153
|
| 1241 |
-
},
|
| 1242 |
-
{
|
| 1243 |
-
"epoch": 4.850393700787402,
|
| 1244 |
-
"grad_norm": 0.8221555352210999,
|
| 1245 |
-
"learning_rate": 4.874899501857477e-05,
|
| 1246 |
-
"loss": 1.0088,
|
| 1247 |
-
"step": 154
|
| 1248 |
-
},
|
| 1249 |
-
{
|
| 1250 |
-
"epoch": 4.881889763779528,
|
| 1251 |
-
"grad_norm": 0.8129372000694275,
|
| 1252 |
-
"learning_rate": 4.87026411224489e-05,
|
| 1253 |
-
"loss": 1.0464,
|
| 1254 |
-
"step": 155
|
| 1255 |
-
},
|
| 1256 |
-
{
|
| 1257 |
-
"epoch": 4.913385826771654,
|
| 1258 |
-
"grad_norm": 0.8731322884559631,
|
| 1259 |
-
"learning_rate": 4.865546685075174e-05,
|
| 1260 |
-
"loss": 1.0963,
|
| 1261 |
-
"step": 156
|
| 1262 |
-
},
|
| 1263 |
-
{
|
| 1264 |
-
"epoch": 4.94488188976378,
|
| 1265 |
-
"grad_norm": 0.9633178114891052,
|
| 1266 |
-
"learning_rate": 4.860747383623889e-05,
|
| 1267 |
-
"loss": 1.0273,
|
| 1268 |
-
"step": 157
|
| 1269 |
-
},
|
| 1270 |
-
{
|
| 1271 |
-
"epoch": 4.9763779527559056,
|
| 1272 |
-
"grad_norm": 0.8986226320266724,
|
| 1273 |
-
"learning_rate": 4.85586637400036e-05,
|
| 1274 |
-
"loss": 1.0807,
|
| 1275 |
-
"step": 158
|
| 1276 |
-
},
|
| 1277 |
-
{
|
| 1278 |
-
"epoch": 5.0078740157480315,
|
| 1279 |
-
"grad_norm": 0.9396881461143494,
|
| 1280 |
-
"learning_rate": 4.8509038251419196e-05,
|
| 1281 |
-
"loss": 0.9067,
|
| 1282 |
-
"step": 159
|
| 1283 |
-
},
|
| 1284 |
-
{
|
| 1285 |
-
"epoch": 5.039370078740157,
|
| 1286 |
-
"grad_norm": 0.8420762419700623,
|
| 1287 |
-
"learning_rate": 4.8458599088080735e-05,
|
| 1288 |
-
"loss": 0.947,
|
| 1289 |
-
"step": 160
|
| 1290 |
-
},
|
| 1291 |
-
{
|
| 1292 |
-
"epoch": 5.039370078740157,
|
| 1293 |
-
"eval_loss": 1.1647462844848633,
|
| 1294 |
-
"eval_runtime": 1.983,
|
| 1295 |
-
"eval_samples_per_second": 50.43,
|
| 1296 |
-
"eval_steps_per_second": 3.53,
|
| 1297 |
-
"step": 160
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 5.070866141732283,
|
| 1301 |
-
"grad_norm": 0.8484461307525635,
|
| 1302 |
-
"learning_rate": 4.840734799574546e-05,
|
| 1303 |
-
"loss": 0.9632,
|
| 1304 |
-
"step": 161
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 5.102362204724409,
|
| 1308 |
-
"grad_norm": 0.8203635215759277,
|
| 1309 |
-
"learning_rate": 4.83552867482724e-05,
|
| 1310 |
-
"loss": 0.8487,
|
| 1311 |
-
"step": 162
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 5.133858267716535,
|
| 1315 |
-
"grad_norm": 0.9048157930374146,
|
| 1316 |
-
"learning_rate": 4.830241714756099e-05,
|
| 1317 |
-
"loss": 1.011,
|
| 1318 |
-
"step": 163
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 5.165354330708661,
|
| 1322 |
-
"grad_norm": 0.832145094871521,
|
| 1323 |
-
"learning_rate": 4.82487410234887e-05,
|
| 1324 |
-
"loss": 0.9208,
|
| 1325 |
-
"step": 164
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 5.196850393700787,
|
| 1329 |
-
"grad_norm": 0.8314303755760193,
|
| 1330 |
-
"learning_rate": 4.8194260233847695e-05,
|
| 1331 |
-
"loss": 1.048,
|
| 1332 |
-
"step": 165
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 5.228346456692913,
|
| 1336 |
-
"grad_norm": 0.8757374882698059,
|
| 1337 |
-
"learning_rate": 4.8138976664280536e-05,
|
| 1338 |
-
"loss": 0.9722,
|
| 1339 |
-
"step": 166
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 5.259842519685039,
|
| 1343 |
-
"grad_norm": 0.818915605545044,
|
| 1344 |
-
"learning_rate": 4.8082892228214906e-05,
|
| 1345 |
-
"loss": 0.9581,
|
| 1346 |
-
"step": 167
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 5.291338582677166,
|
| 1350 |
-
"grad_norm": 0.8921651244163513,
|
| 1351 |
-
"learning_rate": 4.8026008866797423e-05,
|
| 1352 |
-
"loss": 0.9651,
|
| 1353 |
-
"step": 168
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 5.291338582677166,
|
| 1357 |
-
"eval_loss": 1.1861763000488281,
|
| 1358 |
-
"eval_runtime": 3.1874,
|
| 1359 |
-
"eval_samples_per_second": 31.374,
|
| 1360 |
-
"eval_steps_per_second": 2.196,
|
| 1361 |
-
"step": 168
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 5.322834645669292,
|
| 1365 |
-
"grad_norm": 0.8719169497489929,
|
| 1366 |
-
"learning_rate": 4.79683285488264e-05,
|
| 1367 |
-
"loss": 0.9496,
|
| 1368 |
-
"step": 169
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 5.354330708661418,
|
| 1372 |
-
"grad_norm": 0.8905276656150818,
|
| 1373 |
-
"learning_rate": 4.7909853270683756e-05,
|
| 1374 |
-
"loss": 1.0808,
|
| 1375 |
-
"step": 170
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 5.3858267716535435,
|
| 1379 |
-
"grad_norm": 0.9556466341018677,
|
| 1380 |
-
"learning_rate": 4.785058505626587e-05,
|
| 1381 |
-
"loss": 0.8845,
|
| 1382 |
-
"step": 171
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 5.417322834645669,
|
| 1386 |
-
"grad_norm": 0.9415910840034485,
|
| 1387 |
-
"learning_rate": 4.779052595691355e-05,
|
| 1388 |
-
"loss": 0.9678,
|
| 1389 |
-
"step": 172
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 5.448818897637795,
|
| 1393 |
-
"grad_norm": 0.8661073446273804,
|
| 1394 |
-
"learning_rate": 4.772967805134106e-05,
|
| 1395 |
-
"loss": 1.0017,
|
| 1396 |
-
"step": 173
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 5.480314960629921,
|
| 1400 |
-
"grad_norm": 0.9943151473999023,
|
| 1401 |
-
"learning_rate": 4.7668043445564134e-05,
|
| 1402 |
-
"loss": 0.946,
|
| 1403 |
-
"step": 174
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 5.511811023622047,
|
| 1407 |
-
"grad_norm": 0.9678016304969788,
|
| 1408 |
-
"learning_rate": 4.7605624272827126e-05,
|
| 1409 |
-
"loss": 0.9621,
|
| 1410 |
-
"step": 175
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 5.543307086614173,
|
| 1414 |
-
"grad_norm": 0.9622290134429932,
|
| 1415 |
-
"learning_rate": 4.754242269352912e-05,
|
| 1416 |
-
"loss": 0.8979,
|
| 1417 |
-
"step": 176
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 5.543307086614173,
|
| 1421 |
-
"eval_loss": 1.1995248794555664,
|
| 1422 |
-
"eval_runtime": 3.0442,
|
| 1423 |
-
"eval_samples_per_second": 32.849,
|
| 1424 |
-
"eval_steps_per_second": 2.299,
|
| 1425 |
-
"step": 176
|
| 1426 |
-
},
|
| 1427 |
-
{
|
| 1428 |
-
"epoch": 5.574803149606299,
|
| 1429 |
-
"grad_norm": 0.9539031386375427,
|
| 1430 |
-
"learning_rate": 4.747844089514919e-05,
|
| 1431 |
-
"loss": 1.0158,
|
| 1432 |
-
"step": 177
|
| 1433 |
-
},
|
| 1434 |
-
{
|
| 1435 |
-
"epoch": 5.606299212598425,
|
| 1436 |
-
"grad_norm": 1.084650993347168,
|
| 1437 |
-
"learning_rate": 4.7413681092170715e-05,
|
| 1438 |
-
"loss": 0.9219,
|
| 1439 |
-
"step": 178
|
| 1440 |
-
},
|
| 1441 |
-
{
|
| 1442 |
-
"epoch": 5.637795275590551,
|
| 1443 |
-
"grad_norm": 0.9811678528785706,
|
| 1444 |
-
"learning_rate": 4.734814552600469e-05,
|
| 1445 |
-
"loss": 1.0474,
|
| 1446 |
-
"step": 179
|
| 1447 |
-
},
|
| 1448 |
-
{
|
| 1449 |
-
"epoch": 5.669291338582677,
|
| 1450 |
-
"grad_norm": 0.8820479512214661,
|
| 1451 |
-
"learning_rate": 4.728183646491214e-05,
|
| 1452 |
-
"loss": 0.9529,
|
| 1453 |
-
"step": 180
|
| 1454 |
-
},
|
| 1455 |
-
{
|
| 1456 |
-
"epoch": 5.700787401574803,
|
| 1457 |
-
"grad_norm": 1.0495096445083618,
|
| 1458 |
-
"learning_rate": 4.7214756203925676e-05,
|
| 1459 |
-
"loss": 0.9337,
|
| 1460 |
-
"step": 181
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 5.73228346456693,
|
| 1464 |
-
"grad_norm": 1.0566222667694092,
|
| 1465 |
-
"learning_rate": 4.7146907064769994e-05,
|
| 1466 |
-
"loss": 0.9897,
|
| 1467 |
-
"step": 182
|
| 1468 |
-
},
|
| 1469 |
-
{
|
| 1470 |
-
"epoch": 5.7637795275590555,
|
| 1471 |
-
"grad_norm": 1.2577701807022095,
|
| 1472 |
-
"learning_rate": 4.7078291395781554e-05,
|
| 1473 |
-
"loss": 1.1532,
|
| 1474 |
-
"step": 183
|
| 1475 |
-
},
|
| 1476 |
-
{
|
| 1477 |
-
"epoch": 5.7952755905511815,
|
| 1478 |
-
"grad_norm": 0.9294533729553223,
|
| 1479 |
-
"learning_rate": 4.700891157182729e-05,
|
| 1480 |
-
"loss": 1.0119,
|
| 1481 |
-
"step": 184
|
| 1482 |
-
},
|
| 1483 |
-
{
|
| 1484 |
-
"epoch": 5.7952755905511815,
|
| 1485 |
-
"eval_loss": 1.192854881286621,
|
| 1486 |
-
"eval_runtime": 3.055,
|
| 1487 |
-
"eval_samples_per_second": 32.733,
|
| 1488 |
-
"eval_steps_per_second": 2.291,
|
| 1489 |
-
"step": 184
|
| 1490 |
-
},
|
| 1491 |
-
{
|
| 1492 |
-
"epoch": 5.826771653543307,
|
| 1493 |
-
"grad_norm": 0.9633037447929382,
|
| 1494 |
-
"learning_rate": 4.693876999422241e-05,
|
| 1495 |
-
"loss": 0.9447,
|
| 1496 |
-
"step": 185
|
| 1497 |
-
},
|
| 1498 |
-
{
|
| 1499 |
-
"epoch": 5.858267716535433,
|
| 1500 |
-
"grad_norm": 0.9411280751228333,
|
| 1501 |
-
"learning_rate": 4.686786909064729e-05,
|
| 1502 |
-
"loss": 0.9115,
|
| 1503 |
-
"step": 186
|
| 1504 |
-
},
|
| 1505 |
-
{
|
| 1506 |
-
"epoch": 5.889763779527559,
|
| 1507 |
-
"grad_norm": 0.9875217080116272,
|
| 1508 |
-
"learning_rate": 4.679621131506347e-05,
|
| 1509 |
-
"loss": 1.0157,
|
| 1510 |
-
"step": 187
|
| 1511 |
-
},
|
| 1512 |
-
{
|
| 1513 |
-
"epoch": 5.921259842519685,
|
| 1514 |
-
"grad_norm": 0.9757392406463623,
|
| 1515 |
-
"learning_rate": 4.6723799147628666e-05,
|
| 1516 |
-
"loss": 0.9835,
|
| 1517 |
-
"step": 188
|
| 1518 |
-
},
|
| 1519 |
-
{
|
| 1520 |
-
"epoch": 5.952755905511811,
|
| 1521 |
-
"grad_norm": 0.987448513507843,
|
| 1522 |
-
"learning_rate": 4.665063509461097e-05,
|
| 1523 |
-
"loss": 1.0326,
|
| 1524 |
-
"step": 189
|
| 1525 |
-
},
|
| 1526 |
-
{
|
| 1527 |
-
"epoch": 5.984251968503937,
|
| 1528 |
-
"grad_norm": 0.985375702381134,
|
| 1529 |
-
"learning_rate": 4.6576721688302105e-05,
|
| 1530 |
-
"loss": 0.8626,
|
| 1531 |
-
"step": 190
|
| 1532 |
-
},
|
| 1533 |
-
{
|
| 1534 |
-
"epoch": 6.015748031496063,
|
| 1535 |
-
"grad_norm": 1.0300512313842773,
|
| 1536 |
-
"learning_rate": 4.650206148692977e-05,
|
| 1537 |
-
"loss": 0.9618,
|
| 1538 |
-
"step": 191
|
| 1539 |
-
},
|
| 1540 |
-
{
|
| 1541 |
-
"epoch": 6.047244094488189,
|
| 1542 |
-
"grad_norm": 0.9991244077682495,
|
| 1543 |
-
"learning_rate": 4.642665707456908e-05,
|
| 1544 |
-
"loss": 0.9045,
|
| 1545 |
-
"step": 192
|
| 1546 |
-
},
|
| 1547 |
-
{
|
| 1548 |
-
"epoch": 6.047244094488189,
|
| 1549 |
-
"eval_loss": 1.2067475318908691,
|
| 1550 |
-
"eval_runtime": 3.6768,
|
| 1551 |
-
"eval_samples_per_second": 27.197,
|
| 1552 |
-
"eval_steps_per_second": 1.904,
|
| 1553 |
-
"step": 192
|
| 1554 |
-
},
|
| 1555 |
-
{
|
| 1556 |
-
"epoch": 6.078740157480315,
|
| 1557 |
-
"grad_norm": 0.8374089598655701,
|
| 1558 |
-
"learning_rate": 4.635051106105316e-05,
|
| 1559 |
-
"loss": 0.867,
|
| 1560 |
-
"step": 193
|
| 1561 |
-
},
|
| 1562 |
-
{
|
| 1563 |
-
"epoch": 6.110236220472441,
|
| 1564 |
-
"grad_norm": 1.0568664073944092,
|
| 1565 |
-
"learning_rate": 4.6273626081882805e-05,
|
| 1566 |
-
"loss": 0.8759,
|
| 1567 |
-
"step": 194
|
| 1568 |
-
},
|
| 1569 |
-
{
|
| 1570 |
-
"epoch": 6.141732283464567,
|
| 1571 |
-
"grad_norm": 1.0698866844177246,
|
| 1572 |
-
"learning_rate": 4.619600479813524e-05,
|
| 1573 |
-
"loss": 0.8402,
|
| 1574 |
-
"step": 195
|
| 1575 |
-
},
|
| 1576 |
-
{
|
| 1577 |
-
"epoch": 6.173228346456693,
|
| 1578 |
-
"grad_norm": 1.1773091554641724,
|
| 1579 |
-
"learning_rate": 4.611764989637205e-05,
|
| 1580 |
-
"loss": 0.9268,
|
| 1581 |
-
"step": 196
|
| 1582 |
-
},
|
| 1583 |
-
{
|
| 1584 |
-
"epoch": 6.2047244094488185,
|
| 1585 |
-
"grad_norm": 1.2211183309555054,
|
| 1586 |
-
"learning_rate": 4.603856408854618e-05,
|
| 1587 |
-
"loss": 0.8635,
|
| 1588 |
-
"step": 197
|
| 1589 |
-
},
|
| 1590 |
-
{
|
| 1591 |
-
"epoch": 6.2362204724409445,
|
| 1592 |
-
"grad_norm": 1.0713794231414795,
|
| 1593 |
-
"learning_rate": 4.595875011190807e-05,
|
| 1594 |
-
"loss": 0.8548,
|
| 1595 |
-
"step": 198
|
| 1596 |
-
},
|
| 1597 |
-
{
|
| 1598 |
-
"epoch": 6.267716535433071,
|
| 1599 |
-
"grad_norm": 1.2533146142959595,
|
| 1600 |
-
"learning_rate": 4.5878210728910894e-05,
|
| 1601 |
-
"loss": 0.8742,
|
| 1602 |
-
"step": 199
|
| 1603 |
-
},
|
| 1604 |
-
{
|
| 1605 |
-
"epoch": 6.299212598425197,
|
| 1606 |
-
"grad_norm": 1.1741169691085815,
|
| 1607 |
-
"learning_rate": 4.579694872711501e-05,
|
| 1608 |
-
"loss": 0.9117,
|
| 1609 |
-
"step": 200
|
| 1610 |
-
},
|
| 1611 |
-
{
|
| 1612 |
-
"epoch": 6.299212598425197,
|
| 1613 |
-
"eval_loss": 1.2343026399612427,
|
| 1614 |
-
"eval_runtime": 3.7714,
|
| 1615 |
-
"eval_samples_per_second": 26.515,
|
| 1616 |
-
"eval_steps_per_second": 1.856,
|
| 1617 |
-
"step": 200
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 6.330708661417323,
|
| 1621 |
-
"grad_norm": 1.1068271398544312,
|
| 1622 |
-
"learning_rate": 4.5714966919091415e-05,
|
| 1623 |
-
"loss": 0.8904,
|
| 1624 |
-
"step": 201
|
| 1625 |
-
},
|
| 1626 |
-
{
|
| 1627 |
-
"epoch": 6.362204724409449,
|
| 1628 |
-
"grad_norm": 1.2471762895584106,
|
| 1629 |
-
"learning_rate": 4.563226814232444e-05,
|
| 1630 |
-
"loss": 0.7685,
|
| 1631 |
-
"step": 202
|
| 1632 |
-
},
|
| 1633 |
-
{
|
| 1634 |
-
"epoch": 6.393700787401575,
|
| 1635 |
-
"grad_norm": 1.0919934511184692,
|
| 1636 |
-
"learning_rate": 4.554885525911351e-05,
|
| 1637 |
-
"loss": 0.7867,
|
| 1638 |
-
"step": 203
|
| 1639 |
-
},
|
| 1640 |
-
{
|
| 1641 |
-
"epoch": 6.425196850393701,
|
| 1642 |
-
"grad_norm": 1.1424957513809204,
|
| 1643 |
-
"learning_rate": 4.5464731156474094e-05,
|
| 1644 |
-
"loss": 0.9329,
|
| 1645 |
-
"step": 204
|
| 1646 |
-
},
|
| 1647 |
-
{
|
| 1648 |
-
"epoch": 6.456692913385827,
|
| 1649 |
-
"grad_norm": 1.190438985824585,
|
| 1650 |
-
"learning_rate": 4.5379898746037804e-05,
|
| 1651 |
-
"loss": 0.8624,
|
| 1652 |
-
"step": 205
|
| 1653 |
-
},
|
| 1654 |
-
{
|
| 1655 |
-
"epoch": 6.488188976377953,
|
| 1656 |
-
"grad_norm": 1.210954189300537,
|
| 1657 |
-
"learning_rate": 4.529436096395156e-05,
|
| 1658 |
-
"loss": 0.9547,
|
| 1659 |
-
"step": 206
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 6.519685039370079,
|
| 1663 |
-
"grad_norm": 1.2978620529174805,
|
| 1664 |
-
"learning_rate": 4.520812077077604e-05,
|
| 1665 |
-
"loss": 0.8868,
|
| 1666 |
-
"step": 207
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 6.551181102362205,
|
| 1670 |
-
"grad_norm": 1.081527590751648,
|
| 1671 |
-
"learning_rate": 4.5121181151383143e-05,
|
| 1672 |
-
"loss": 0.7841,
|
| 1673 |
-
"step": 208
|
| 1674 |
-
},
|
| 1675 |
-
{
|
| 1676 |
-
"epoch": 6.551181102362205,
|
| 1677 |
-
"eval_loss": 1.2477138042449951,
|
| 1678 |
-
"eval_runtime": 3.8991,
|
| 1679 |
-
"eval_samples_per_second": 25.647,
|
| 1680 |
-
"eval_steps_per_second": 1.795,
|
| 1681 |
-
"step": 208
|
| 1682 |
-
},
|
| 1683 |
-
{
|
| 1684 |
-
"epoch": 6.582677165354331,
|
| 1685 |
-
"grad_norm": 1.1496070623397827,
|
| 1686 |
-
"learning_rate": 4.503354511485273e-05,
|
| 1687 |
-
"loss": 0.9487,
|
| 1688 |
-
"step": 209
|
| 1689 |
-
},
|
| 1690 |
-
{
|
| 1691 |
-
"epoch": 6.6141732283464565,
|
| 1692 |
-
"grad_norm": 1.188530683517456,
|
| 1693 |
-
"learning_rate": 4.494521569436845e-05,
|
| 1694 |
-
"loss": 0.9439,
|
| 1695 |
-
"step": 210
|
| 1696 |
}
|
| 1697 |
],
|
| 1698 |
"logging_steps": 1,
|
| 1699 |
-
"max_steps":
|
| 1700 |
"num_input_tokens_seen": 0,
|
| 1701 |
-
"num_train_epochs":
|
| 1702 |
"save_steps": 5,
|
| 1703 |
"stateful_callbacks": {
|
| 1704 |
"TrainerControl": {
|
|
@@ -1712,8 +179,8 @@
|
|
| 1712 |
"attributes": {}
|
| 1713 |
}
|
| 1714 |
},
|
| 1715 |
-
"total_flos":
|
| 1716 |
-
"train_batch_size":
|
| 1717 |
"trial_name": null,
|
| 1718 |
"trial_params": null
|
| 1719 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.4047619047619047,
|
| 5 |
+
"eval_steps": 3,
|
| 6 |
+
"global_step": 15,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.09523809523809523,
|
| 13 |
+
"grad_norm": 0.5343478322029114,
|
| 14 |
+
"learning_rate": 2e-05,
|
| 15 |
+
"loss": 1.3356,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.09523809523809523,
|
| 20 |
+
"eval_loss": 1.2671657800674438,
|
| 21 |
+
"eval_runtime": 20.0994,
|
| 22 |
+
"eval_samples_per_second": 4.975,
|
| 23 |
+
"eval_steps_per_second": 2.488,
|
| 24 |
"step": 1
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.19047619047619047,
|
| 28 |
+
"grad_norm": 0.5495722889900208,
|
| 29 |
+
"learning_rate": 4e-05,
|
| 30 |
+
"loss": 1.3576,
|
| 31 |
"step": 2
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.2857142857142857,
|
| 35 |
+
"grad_norm": 0.48776495456695557,
|
| 36 |
+
"learning_rate": 6e-05,
|
| 37 |
+
"loss": 1.2159,
|
| 38 |
+
"step": 3
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.2857142857142857,
|
| 42 |
+
"eval_loss": 1.2533905506134033,
|
| 43 |
+
"eval_runtime": 24.7126,
|
| 44 |
+
"eval_samples_per_second": 4.047,
|
| 45 |
+
"eval_steps_per_second": 2.023,
|
| 46 |
"step": 3
|
| 47 |
},
|
| 48 |
{
|
| 49 |
+
"epoch": 0.38095238095238093,
|
| 50 |
+
"grad_norm": 0.4593156576156616,
|
| 51 |
+
"learning_rate": 8e-05,
|
| 52 |
+
"loss": 1.2203,
|
| 53 |
"step": 4
|
| 54 |
},
|
| 55 |
{
|
| 56 |
+
"epoch": 0.47619047619047616,
|
| 57 |
+
"grad_norm": 0.3861481249332428,
|
| 58 |
+
"learning_rate": 0.0001,
|
| 59 |
+
"loss": 1.278,
|
| 60 |
"step": 5
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"epoch": 0.5714285714285714,
|
| 64 |
+
"grad_norm": 0.26556482911109924,
|
| 65 |
+
"learning_rate": 0.00012,
|
| 66 |
+
"loss": 1.2716,
|
| 67 |
"step": 6
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"epoch": 0.5714285714285714,
|
| 71 |
+
"eval_loss": 1.209787130355835,
|
| 72 |
+
"eval_runtime": 20.5346,
|
| 73 |
+
"eval_samples_per_second": 4.87,
|
| 74 |
+
"eval_steps_per_second": 2.435,
|
| 75 |
+
"step": 6
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.6666666666666666,
|
| 79 |
+
"grad_norm": 0.3640616536140442,
|
| 80 |
+
"learning_rate": 0.00014,
|
| 81 |
+
"loss": 1.2188,
|
| 82 |
"step": 7
|
| 83 |
},
|
| 84 |
{
|
| 85 |
+
"epoch": 0.7619047619047619,
|
| 86 |
+
"grad_norm": 0.49822962284088135,
|
| 87 |
+
"learning_rate": 0.00016,
|
| 88 |
+
"loss": 1.2624,
|
| 89 |
"step": 8
|
| 90 |
},
|
| 91 |
{
|
| 92 |
+
"epoch": 0.8571428571428571,
|
| 93 |
+
"grad_norm": 0.42707493901252747,
|
| 94 |
+
"learning_rate": 0.00018,
|
| 95 |
+
"loss": 1.2697,
|
| 96 |
+
"step": 9
|
|
|
|
| 97 |
},
|
| 98 |
{
|
| 99 |
+
"epoch": 0.8571428571428571,
|
| 100 |
+
"eval_loss": 1.1909722089767456,
|
| 101 |
+
"eval_runtime": 20.1688,
|
| 102 |
+
"eval_samples_per_second": 4.958,
|
| 103 |
+
"eval_steps_per_second": 2.479,
|
| 104 |
"step": 9
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 0.9523809523809523,
|
| 108 |
+
"grad_norm": 0.27150899171829224,
|
| 109 |
+
"learning_rate": 0.0002,
|
| 110 |
+
"loss": 1.215,
|
| 111 |
"step": 10
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 1.0238095238095237,
|
| 115 |
+
"grad_norm": 0.23839783668518066,
|
| 116 |
+
"learning_rate": 0.00019876883405951377,
|
| 117 |
+
"loss": 1.2365,
|
| 118 |
"step": 11
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 1.119047619047619,
|
| 122 |
+
"grad_norm": 0.20644636452198029,
|
| 123 |
+
"learning_rate": 0.00019510565162951537,
|
| 124 |
+
"loss": 1.2243,
|
| 125 |
+
"step": 12
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 1.119047619047619,
|
| 129 |
+
"eval_loss": 1.1754947900772095,
|
| 130 |
+
"eval_runtime": 21.1588,
|
| 131 |
+
"eval_samples_per_second": 4.726,
|
| 132 |
+
"eval_steps_per_second": 2.363,
|
| 133 |
"step": 12
|
| 134 |
},
|
| 135 |
{
|
| 136 |
+
"epoch": 1.2142857142857142,
|
| 137 |
+
"grad_norm": 0.23169651627540588,
|
| 138 |
+
"learning_rate": 0.0001891006524188368,
|
| 139 |
+
"loss": 1.1474,
|
| 140 |
"step": 13
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 1.3095238095238095,
|
| 144 |
+
"grad_norm": 0.21209821105003357,
|
| 145 |
+
"learning_rate": 0.00018090169943749476,
|
| 146 |
+
"loss": 1.1395,
|
| 147 |
"step": 14
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 1.4047619047619047,
|
| 151 |
+
"grad_norm": 0.1954081654548645,
|
| 152 |
+
"learning_rate": 0.00017071067811865476,
|
| 153 |
+
"loss": 1.1981,
|
| 154 |
"step": 15
|
| 155 |
},
|
| 156 |
{
|
| 157 |
+
"epoch": 1.4047619047619047,
|
| 158 |
+
"eval_loss": 1.1603792905807495,
|
| 159 |
+
"eval_runtime": 16.71,
|
| 160 |
+
"eval_samples_per_second": 5.984,
|
| 161 |
+
"eval_steps_per_second": 2.992,
|
| 162 |
+
"step": 15
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 163 |
}
|
| 164 |
],
|
| 165 |
"logging_steps": 1,
|
| 166 |
+
"max_steps": 30,
|
| 167 |
"num_input_tokens_seen": 0,
|
| 168 |
+
"num_train_epochs": 3,
|
| 169 |
"save_steps": 5,
|
| 170 |
"stateful_callbacks": {
|
| 171 |
"TrainerControl": {
|
|
|
|
| 179 |
"attributes": {}
|
| 180 |
}
|
| 181 |
},
|
| 182 |
+
"total_flos": 2835124715520000.0,
|
| 183 |
+
"train_batch_size": 2,
|
| 184 |
"trial_name": null,
|
| 185 |
"trial_params": null
|
| 186 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6648
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c62cd4a5f5b5d98624b05b91ea66842c68cee52403c2495c0459cf13d17cfcc
|
| 3 |
size 6648
|