Training in progress, step 50, checkpoint
Browse files- last-checkpoint/README.md +3 -3
- last-checkpoint/adapter_config.json +9 -9
- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +209 -1641
- last-checkpoint/training_args.bin +2 -2
last-checkpoint/README.md
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
---
|
| 2 |
-
base_model:
|
| 3 |
library_name: peft
|
| 4 |
pipeline_tag: text-generation
|
| 5 |
tags:
|
| 6 |
-
- base_model:adapter:
|
| 7 |
- lora
|
| 8 |
- sft
|
| 9 |
- transformers
|
|
@@ -207,4 +207,4 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 207 |
[More Information Needed]
|
| 208 |
### Framework versions
|
| 209 |
|
| 210 |
-
- PEFT 0.18.
|
|
|
|
| 1 |
---
|
| 2 |
+
base_model: owenisas/nemotron-3-nano-reasoning
|
| 3 |
library_name: peft
|
| 4 |
pipeline_tag: text-generation
|
| 5 |
tags:
|
| 6 |
+
- base_model:adapter:owenisas/nemotron-3-nano-reasoning
|
| 7 |
- lora
|
| 8 |
- sft
|
| 9 |
- transformers
|
|
|
|
| 207 |
[More Information Needed]
|
| 208 |
### Framework versions
|
| 209 |
|
| 210 |
+
- PEFT 0.18.1
|
last-checkpoint/adapter_config.json
CHANGED
|
@@ -4,10 +4,10 @@
|
|
| 4 |
"arrow_config": null,
|
| 5 |
"auto_mapping": {
|
| 6 |
"base_model_class": "NemotronHForCausalLM",
|
| 7 |
-
"parent_library": "transformers_modules.
|
| 8 |
"unsloth_fixed": true
|
| 9 |
},
|
| 10 |
-
"base_model_name_or_path": "
|
| 11 |
"bias": "none",
|
| 12 |
"corda_config": null,
|
| 13 |
"ensure_weight_tying": false,
|
|
@@ -27,21 +27,21 @@
|
|
| 27 |
"megatron_core": "megatron.core",
|
| 28 |
"modules_to_save": null,
|
| 29 |
"peft_type": "LORA",
|
| 30 |
-
"peft_version": "0.18.
|
| 31 |
"qalora_group_size": 16,
|
| 32 |
"r": 32,
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
|
|
|
|
|
|
| 36 |
"k_proj",
|
| 37 |
-
"down_proj",
|
| 38 |
-
"gate_proj",
|
| 39 |
"out_proj",
|
| 40 |
-
"o_proj",
|
| 41 |
-
"q_proj",
|
| 42 |
"up_proj",
|
| 43 |
-
"
|
| 44 |
-
"
|
|
|
|
|
|
|
| 45 |
],
|
| 46 |
"target_parameters": null,
|
| 47 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 4 |
"arrow_config": null,
|
| 5 |
"auto_mapping": {
|
| 6 |
"base_model_class": "NemotronHForCausalLM",
|
| 7 |
+
"parent_library": "transformers_modules.owenisas.nemotron-3-nano-reasoning.c06798b01704b3d322954056e8de8bf6cae11e38.modeling_nemotron_h",
|
| 8 |
"unsloth_fixed": true
|
| 9 |
},
|
| 10 |
+
"base_model_name_or_path": "owenisas/nemotron-3-nano-reasoning",
|
| 11 |
"bias": "none",
|
| 12 |
"corda_config": null,
|
| 13 |
"ensure_weight_tying": false,
|
|
|
|
| 27 |
"megatron_core": "megatron.core",
|
| 28 |
"modules_to_save": null,
|
| 29 |
"peft_type": "LORA",
|
| 30 |
+
"peft_version": "0.18.1",
|
| 31 |
"qalora_group_size": 16,
|
| 32 |
"r": 32,
|
| 33 |
"rank_pattern": {},
|
| 34 |
"revision": null,
|
| 35 |
"target_modules": [
|
| 36 |
+
"in_proj",
|
| 37 |
+
"q_proj",
|
| 38 |
"k_proj",
|
|
|
|
|
|
|
| 39 |
"out_proj",
|
|
|
|
|
|
|
| 40 |
"up_proj",
|
| 41 |
+
"down_proj",
|
| 42 |
+
"gate_proj",
|
| 43 |
+
"v_proj",
|
| 44 |
+
"o_proj"
|
| 45 |
],
|
| 46 |
"target_parameters": null,
|
| 47 |
"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 3537299144
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c302dc1f7d4b868ed2fec7fb599c56ab89a9be3b061d10a09c33f91bc884118
|
| 3 |
size 3537299144
|
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:c82c2ea846bada76c6987cfb10fc7217cfd00b4b82d0021a138e9add209aaec9
|
| 3 |
+
size 1798933287
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ccb8eeb935749fc43744e0a5eeacdf6f0f10253be15266a497cbca0ffaa2573
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83429aff07094f43f6ae84f250d5d91c95fca2dfaf4ecddce133674cbbfe1442
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,1806 +2,374 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 50,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
-
"loss":
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
-
"epoch": 0.
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
-
"learning_rate":
|
| 23 |
-
"loss":
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
-
"learning_rate":
|
| 30 |
-
"loss":
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss":
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"grad_norm": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss":
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"grad_norm": 0.
|
| 50 |
-
"learning_rate":
|
| 51 |
-
"loss":
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
-
"epoch": 0.
|
| 56 |
-
"grad_norm": 0.
|
| 57 |
-
"learning_rate":
|
| 58 |
-
"loss":
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
-
"epoch": 0.
|
| 63 |
-
"grad_norm": 0.
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss":
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
-
"epoch": 0.
|
| 70 |
-
"grad_norm": 0.
|
| 71 |
-
"learning_rate":
|
| 72 |
-
"loss":
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
-
"epoch": 0.
|
| 77 |
-
"grad_norm": 0.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss":
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"grad_norm": 0.
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss":
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"grad_norm": 0.
|
| 92 |
-
"learning_rate":
|
| 93 |
-
"loss":
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
-
"epoch": 0.
|
| 98 |
-
"grad_norm": 0.
|
| 99 |
-
"learning_rate":
|
| 100 |
-
"loss":
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
-
"epoch": 0.
|
| 105 |
-
"grad_norm": 0.
|
| 106 |
-
"learning_rate":
|
| 107 |
-
"loss":
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch": 0.
|
| 112 |
-
"grad_norm": 0.
|
| 113 |
-
"learning_rate":
|
| 114 |
-
"loss":
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
-
"epoch": 0.
|
| 119 |
-
"grad_norm": 0.
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss":
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch": 0.
|
| 126 |
-
"grad_norm": 0.
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss":
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"grad_norm": 0.
|
| 134 |
-
"learning_rate":
|
| 135 |
-
"loss":
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"epoch": 0.
|
| 140 |
-
"grad_norm": 0.
|
| 141 |
-
"learning_rate":
|
| 142 |
-
"loss":
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
-
"epoch": 0.
|
| 147 |
-
"grad_norm": 0.
|
| 148 |
-
"learning_rate":
|
| 149 |
-
"loss":
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
-
"epoch": 0.
|
| 154 |
-
"grad_norm": 0.
|
| 155 |
-
"learning_rate":
|
| 156 |
-
"loss":
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch": 0.
|
| 161 |
-
"grad_norm": 0.
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss":
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch": 0.
|
| 168 |
-
"grad_norm": 0.
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss":
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"grad_norm": 0.
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss":
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch": 0.
|
| 182 |
-
"grad_norm": 0.
|
| 183 |
-
"learning_rate":
|
| 184 |
-
"loss":
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch": 0.
|
| 189 |
-
"grad_norm": 0.
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss":
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch": 0.
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss":
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch": 0.
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss":
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch": 0.
|
| 210 |
-
"grad_norm": 0.
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss":
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch": 0.
|
| 217 |
-
"grad_norm": 0.
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss":
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch": 0.
|
| 224 |
-
"grad_norm": 0.
|
| 225 |
-
"learning_rate":
|
| 226 |
-
"loss":
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
-
"epoch": 0.
|
| 231 |
-
"grad_norm": 0.
|
| 232 |
-
"learning_rate":
|
| 233 |
-
"loss":
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch": 0.
|
| 238 |
-
"grad_norm": 0.
|
| 239 |
-
"learning_rate":
|
| 240 |
-
"loss":
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
-
"epoch": 0.
|
| 245 |
-
"grad_norm": 0.
|
| 246 |
-
"learning_rate":
|
| 247 |
-
"loss":
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"epoch": 0.
|
| 252 |
-
"grad_norm": 0.
|
| 253 |
-
"learning_rate":
|
| 254 |
-
"loss":
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch": 0.
|
| 259 |
-
"grad_norm": 0.
|
| 260 |
-
"learning_rate":
|
| 261 |
-
"loss":
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
-
"epoch": 0.
|
| 266 |
-
"grad_norm": 0.
|
| 267 |
-
"learning_rate":
|
| 268 |
-
"loss":
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
-
"epoch": 0.
|
| 273 |
-
"grad_norm": 0.
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss":
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch": 0.
|
| 280 |
-
"grad_norm": 0.
|
| 281 |
-
"learning_rate":
|
| 282 |
-
"loss":
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
-
"epoch": 0.
|
| 287 |
-
"grad_norm": 0.
|
| 288 |
-
"learning_rate":
|
| 289 |
-
"loss":
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
-
"epoch": 0.
|
| 294 |
-
"grad_norm": 0.
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss":
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch": 0.
|
| 301 |
-
"grad_norm": 0.
|
| 302 |
-
"learning_rate":
|
| 303 |
-
"loss":
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"epoch": 0.
|
| 308 |
-
"grad_norm": 0.
|
| 309 |
-
"learning_rate":
|
| 310 |
-
"loss":
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
-
"epoch": 0.
|
| 315 |
-
"grad_norm": 0.
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss":
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch": 0.
|
| 322 |
-
"grad_norm": 0.
|
| 323 |
-
"learning_rate":
|
| 324 |
-
"loss":
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
-
"epoch": 0.
|
| 329 |
-
"grad_norm": 0.
|
| 330 |
-
"learning_rate":
|
| 331 |
-
"loss":
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
-
"epoch": 0.
|
| 336 |
-
"grad_norm": 0.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss":
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 0.
|
| 343 |
-
"grad_norm": 0.
|
| 344 |
-
"learning_rate":
|
| 345 |
-
"loss":
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
-
"epoch": 0.
|
| 350 |
-
"grad_norm": 0.
|
| 351 |
-
"learning_rate":
|
| 352 |
-
"loss":
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
-
"epoch": 0.
|
| 357 |
-
"grad_norm": 0.
|
| 358 |
-
"learning_rate":
|
| 359 |
-
"loss":
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch": 0.
|
| 364 |
-
"eval_loss": 0.
|
| 365 |
-
"eval_runtime":
|
| 366 |
-
"eval_samples_per_second": 2.
|
| 367 |
-
"eval_steps_per_second": 0.
|
| 368 |
"step": 50
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 0.19154929577464788,
|
| 372 |
-
"grad_norm": 0.5213260650634766,
|
| 373 |
-
"learning_rate": 0.00018579211906908215,
|
| 374 |
-
"loss": 3.6026,
|
| 375 |
-
"step": 51
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 0.19530516431924883,
|
| 379 |
-
"grad_norm": 0.4835609495639801,
|
| 380 |
-
"learning_rate": 0.00018516990084166442,
|
| 381 |
-
"loss": 4.178,
|
| 382 |
-
"step": 52
|
| 383 |
-
},
|
| 384 |
-
{
|
| 385 |
-
"epoch": 0.19906103286384977,
|
| 386 |
-
"grad_norm": 0.6534969210624695,
|
| 387 |
-
"learning_rate": 0.0001845354370697482,
|
| 388 |
-
"loss": 4.6021,
|
| 389 |
-
"step": 53
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 0.2028169014084507,
|
| 393 |
-
"grad_norm": 0.5575639605522156,
|
| 394 |
-
"learning_rate": 0.000183888818975165,
|
| 395 |
-
"loss": 4.2967,
|
| 396 |
-
"step": 54
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 0.20657276995305165,
|
| 400 |
-
"grad_norm": 0.5025990009307861,
|
| 401 |
-
"learning_rate": 0.00018323013952726875,
|
| 402 |
-
"loss": 3.9018,
|
| 403 |
-
"step": 55
|
| 404 |
-
},
|
| 405 |
-
{
|
| 406 |
-
"epoch": 0.21032863849765257,
|
| 407 |
-
"grad_norm": 0.5214390754699707,
|
| 408 |
-
"learning_rate": 0.00018255949342956863,
|
| 409 |
-
"loss": 4.2738,
|
| 410 |
-
"step": 56
|
| 411 |
-
},
|
| 412 |
-
{
|
| 413 |
-
"epoch": 0.2140845070422535,
|
| 414 |
-
"grad_norm": 0.46630731225013733,
|
| 415 |
-
"learning_rate": 0.00018187697710611298,
|
| 416 |
-
"loss": 3.6578,
|
| 417 |
-
"step": 57
|
| 418 |
-
},
|
| 419 |
-
{
|
| 420 |
-
"epoch": 0.21784037558685446,
|
| 421 |
-
"grad_norm": 0.47853055596351624,
|
| 422 |
-
"learning_rate": 0.00018118268868762546,
|
| 423 |
-
"loss": 3.8583,
|
| 424 |
-
"step": 58
|
| 425 |
-
},
|
| 426 |
-
{
|
| 427 |
-
"epoch": 0.2215962441314554,
|
| 428 |
-
"grad_norm": 0.6015180349349976,
|
| 429 |
-
"learning_rate": 0.00018047672799739628,
|
| 430 |
-
"loss": 4.4667,
|
| 431 |
-
"step": 59
|
| 432 |
-
},
|
| 433 |
-
{
|
| 434 |
-
"epoch": 0.22535211267605634,
|
| 435 |
-
"grad_norm": 0.6465438604354858,
|
| 436 |
-
"learning_rate": 0.0001797591965369296,
|
| 437 |
-
"loss": 4.6973,
|
| 438 |
-
"step": 60
|
| 439 |
-
},
|
| 440 |
-
{
|
| 441 |
-
"epoch": 0.2291079812206573,
|
| 442 |
-
"grad_norm": 0.6278983950614929,
|
| 443 |
-
"learning_rate": 0.00017903019747134998,
|
| 444 |
-
"loss": 4.3331,
|
| 445 |
-
"step": 61
|
| 446 |
-
},
|
| 447 |
-
{
|
| 448 |
-
"epoch": 0.23286384976525823,
|
| 449 |
-
"grad_norm": 0.5308983325958252,
|
| 450 |
-
"learning_rate": 0.00017828983561456941,
|
| 451 |
-
"loss": 4.1813,
|
| 452 |
-
"step": 62
|
| 453 |
-
},
|
| 454 |
-
{
|
| 455 |
-
"epoch": 0.23661971830985915,
|
| 456 |
-
"grad_norm": 0.5170332789421082,
|
| 457 |
-
"learning_rate": 0.00017753821741421769,
|
| 458 |
-
"loss": 3.9088,
|
| 459 |
-
"step": 63
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 0.2403755868544601,
|
| 463 |
-
"grad_norm": 0.560627818107605,
|
| 464 |
-
"learning_rate": 0.00017677545093633713,
|
| 465 |
-
"loss": 4.1723,
|
| 466 |
-
"step": 64
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 0.24413145539906103,
|
| 470 |
-
"grad_norm": 0.6482558846473694,
|
| 471 |
-
"learning_rate": 0.00017600164584984546,
|
| 472 |
-
"loss": 4.4782,
|
| 473 |
-
"step": 65
|
| 474 |
-
},
|
| 475 |
-
{
|
| 476 |
-
"epoch": 0.24788732394366197,
|
| 477 |
-
"grad_norm": 0.55652916431427,
|
| 478 |
-
"learning_rate": 0.00017521691341076774,
|
| 479 |
-
"loss": 4.0827,
|
| 480 |
-
"step": 66
|
| 481 |
-
},
|
| 482 |
-
{
|
| 483 |
-
"epoch": 0.2516431924882629,
|
| 484 |
-
"grad_norm": 0.5317822098731995,
|
| 485 |
-
"learning_rate": 0.00017442136644624015,
|
| 486 |
-
"loss": 3.8659,
|
| 487 |
-
"step": 67
|
| 488 |
-
},
|
| 489 |
-
{
|
| 490 |
-
"epoch": 0.25539906103286386,
|
| 491 |
-
"grad_norm": 0.5009008049964905,
|
| 492 |
-
"learning_rate": 0.00017361511933828801,
|
| 493 |
-
"loss": 3.9116,
|
| 494 |
-
"step": 68
|
| 495 |
-
},
|
| 496 |
-
{
|
| 497 |
-
"epoch": 0.2591549295774648,
|
| 498 |
-
"grad_norm": 0.4998956620693207,
|
| 499 |
-
"learning_rate": 0.00017279828800738017,
|
| 500 |
-
"loss": 3.6624,
|
| 501 |
-
"step": 69
|
| 502 |
-
},
|
| 503 |
-
{
|
| 504 |
-
"epoch": 0.26291079812206575,
|
| 505 |
-
"grad_norm": 0.6255429983139038,
|
| 506 |
-
"learning_rate": 0.00017197098989576222,
|
| 507 |
-
"loss": 4.8328,
|
| 508 |
-
"step": 70
|
| 509 |
-
},
|
| 510 |
-
{
|
| 511 |
-
"epoch": 0.26666666666666666,
|
| 512 |
-
"grad_norm": 0.4984213411808014,
|
| 513 |
-
"learning_rate": 0.00017113334395057087,
|
| 514 |
-
"loss": 3.8159,
|
| 515 |
-
"step": 71
|
| 516 |
-
},
|
| 517 |
-
{
|
| 518 |
-
"epoch": 0.2704225352112676,
|
| 519 |
-
"grad_norm": 0.6359574794769287,
|
| 520 |
-
"learning_rate": 0.000170285470606732,
|
| 521 |
-
"loss": 4.6995,
|
| 522 |
-
"step": 72
|
| 523 |
-
},
|
| 524 |
-
{
|
| 525 |
-
"epoch": 0.27417840375586855,
|
| 526 |
-
"grad_norm": 0.4945514500141144,
|
| 527 |
-
"learning_rate": 0.0001694274917696448,
|
| 528 |
-
"loss": 3.6875,
|
| 529 |
-
"step": 73
|
| 530 |
-
},
|
| 531 |
-
{
|
| 532 |
-
"epoch": 0.27793427230046946,
|
| 533 |
-
"grad_norm": 0.5279414057731628,
|
| 534 |
-
"learning_rate": 0.00016855953079765448,
|
| 535 |
-
"loss": 3.8174,
|
| 536 |
-
"step": 74
|
| 537 |
-
},
|
| 538 |
-
{
|
| 539 |
-
"epoch": 0.28169014084507044,
|
| 540 |
-
"grad_norm": 0.6026564240455627,
|
| 541 |
-
"learning_rate": 0.00016768171248431602,
|
| 542 |
-
"loss": 4.3851,
|
| 543 |
-
"step": 75
|
| 544 |
-
},
|
| 545 |
-
{
|
| 546 |
-
"epoch": 0.28544600938967135,
|
| 547 |
-
"grad_norm": 0.5192970633506775,
|
| 548 |
-
"learning_rate": 0.0001667941630404517,
|
| 549 |
-
"loss": 3.7024,
|
| 550 |
-
"step": 76
|
| 551 |
-
},
|
| 552 |
-
{
|
| 553 |
-
"epoch": 0.2892018779342723,
|
| 554 |
-
"grad_norm": 0.48926350474357605,
|
| 555 |
-
"learning_rate": 0.00016589701007600476,
|
| 556 |
-
"loss": 4.0638,
|
| 557 |
-
"step": 77
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 0.29295774647887324,
|
| 561 |
-
"grad_norm": 0.6420406699180603,
|
| 562 |
-
"learning_rate": 0.0001649903825816918,
|
| 563 |
-
"loss": 4.751,
|
| 564 |
-
"step": 78
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 0.29671361502347415,
|
| 568 |
-
"grad_norm": 0.5534292459487915,
|
| 569 |
-
"learning_rate": 0.00016407441091045706,
|
| 570 |
-
"loss": 4.0464,
|
| 571 |
-
"step": 79
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 0.3004694835680751,
|
| 575 |
-
"grad_norm": 0.5345770120620728,
|
| 576 |
-
"learning_rate": 0.0001631492267587301,
|
| 577 |
-
"loss": 4.0831,
|
| 578 |
-
"step": 80
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 0.30422535211267604,
|
| 582 |
-
"grad_norm": 0.4912847876548767,
|
| 583 |
-
"learning_rate": 0.0001622149631474913,
|
| 584 |
-
"loss": 3.537,
|
| 585 |
-
"step": 81
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 0.307981220657277,
|
| 589 |
-
"grad_norm": 0.5144412517547607,
|
| 590 |
-
"learning_rate": 0.00016127175440314596,
|
| 591 |
-
"loss": 4.0086,
|
| 592 |
-
"step": 82
|
| 593 |
-
},
|
| 594 |
-
{
|
| 595 |
-
"epoch": 0.3117370892018779,
|
| 596 |
-
"grad_norm": 0.4543435573577881,
|
| 597 |
-
"learning_rate": 0.0001603197361382114,
|
| 598 |
-
"loss": 3.5774,
|
| 599 |
-
"step": 83
|
| 600 |
-
},
|
| 601 |
-
{
|
| 602 |
-
"epoch": 0.3154929577464789,
|
| 603 |
-
"grad_norm": 0.5636236667633057,
|
| 604 |
-
"learning_rate": 0.0001593590452318187,
|
| 605 |
-
"loss": 4.0208,
|
| 606 |
-
"step": 84
|
| 607 |
-
},
|
| 608 |
-
{
|
| 609 |
-
"epoch": 0.3192488262910798,
|
| 610 |
-
"grad_norm": 0.47736960649490356,
|
| 611 |
-
"learning_rate": 0.00015838981981003273,
|
| 612 |
-
"loss": 3.7864,
|
| 613 |
-
"step": 85
|
| 614 |
-
},
|
| 615 |
-
{
|
| 616 |
-
"epoch": 0.32300469483568073,
|
| 617 |
-
"grad_norm": 0.5701449513435364,
|
| 618 |
-
"learning_rate": 0.00015741219922599253,
|
| 619 |
-
"loss": 4.2106,
|
| 620 |
-
"step": 86
|
| 621 |
-
},
|
| 622 |
-
{
|
| 623 |
-
"epoch": 0.3267605633802817,
|
| 624 |
-
"grad_norm": 0.4872874617576599,
|
| 625 |
-
"learning_rate": 0.00015642632403987535,
|
| 626 |
-
"loss": 3.5756,
|
| 627 |
-
"step": 87
|
| 628 |
-
},
|
| 629 |
-
{
|
| 630 |
-
"epoch": 0.3305164319248826,
|
| 631 |
-
"grad_norm": 0.49332892894744873,
|
| 632 |
-
"learning_rate": 0.00015543233599868742,
|
| 633 |
-
"loss": 4.0097,
|
| 634 |
-
"step": 88
|
| 635 |
-
},
|
| 636 |
-
{
|
| 637 |
-
"epoch": 0.3342723004694836,
|
| 638 |
-
"grad_norm": 0.5447995662689209,
|
| 639 |
-
"learning_rate": 0.0001544303780158837,
|
| 640 |
-
"loss": 4.0947,
|
| 641 |
-
"step": 89
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 0.3380281690140845,
|
| 645 |
-
"grad_norm": 0.46900174021720886,
|
| 646 |
-
"learning_rate": 0.0001534205941508202,
|
| 647 |
-
"loss": 3.8602,
|
| 648 |
-
"step": 90
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 0.34178403755868547,
|
| 652 |
-
"grad_norm": 0.5882900953292847,
|
| 653 |
-
"learning_rate": 0.00015240312958804132,
|
| 654 |
-
"loss": 4.2984,
|
| 655 |
-
"step": 91
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 0.3455399061032864,
|
| 659 |
-
"grad_norm": 0.5628681182861328,
|
| 660 |
-
"learning_rate": 0.00015137813061640563,
|
| 661 |
-
"loss": 4.0235,
|
| 662 |
-
"step": 92
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 0.3492957746478873,
|
| 666 |
-
"grad_norm": 0.5416899919509888,
|
| 667 |
-
"learning_rate": 0.00015034574460805279,
|
| 668 |
-
"loss": 4.003,
|
| 669 |
-
"step": 93
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.3530516431924883,
|
| 673 |
-
"grad_norm": 0.5500250458717346,
|
| 674 |
-
"learning_rate": 0.00014930611999721457,
|
| 675 |
-
"loss": 3.8399,
|
| 676 |
-
"step": 94
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 0.3568075117370892,
|
| 680 |
-
"grad_norm": 0.5311074256896973,
|
| 681 |
-
"learning_rate": 0.00014825940625887342,
|
| 682 |
-
"loss": 3.7204,
|
| 683 |
-
"step": 95
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 0.36056338028169016,
|
| 687 |
-
"grad_norm": 0.5872887969017029,
|
| 688 |
-
"learning_rate": 0.00014720575388727132,
|
| 689 |
-
"loss": 4.2717,
|
| 690 |
-
"step": 96
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 0.3643192488262911,
|
| 694 |
-
"grad_norm": 0.5307738184928894,
|
| 695 |
-
"learning_rate": 0.0001461453143742718,
|
| 696 |
-
"loss": 4.0727,
|
| 697 |
-
"step": 97
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 0.36807511737089205,
|
| 701 |
-
"grad_norm": 0.5269085168838501,
|
| 702 |
-
"learning_rate": 0.00014507824018757906,
|
| 703 |
-
"loss": 3.703,
|
| 704 |
-
"step": 98
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 0.37183098591549296,
|
| 708 |
-
"grad_norm": 0.5249464511871338,
|
| 709 |
-
"learning_rate": 0.0001440046847488163,
|
| 710 |
-
"loss": 4.1674,
|
| 711 |
-
"step": 99
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 0.3755868544600939,
|
| 715 |
-
"grad_norm": 0.5709621906280518,
|
| 716 |
-
"learning_rate": 0.00014292480241146716,
|
| 717 |
-
"loss": 4.2529,
|
| 718 |
-
"step": 100
|
| 719 |
-
},
|
| 720 |
-
{
|
| 721 |
-
"epoch": 0.3755868544600939,
|
| 722 |
-
"eval_loss": 0.5099202990531921,
|
| 723 |
-
"eval_runtime": 361.6043,
|
| 724 |
-
"eval_samples_per_second": 2.619,
|
| 725 |
-
"eval_steps_per_second": 0.655,
|
| 726 |
-
"step": 100
|
| 727 |
-
},
|
| 728 |
-
{
|
| 729 |
-
"epoch": 0.37934272300469485,
|
| 730 |
-
"grad_norm": 0.5334945321083069,
|
| 731 |
-
"learning_rate": 0.00014183874843868313,
|
| 732 |
-
"loss": 4.042,
|
| 733 |
-
"step": 101
|
| 734 |
-
},
|
| 735 |
-
{
|
| 736 |
-
"epoch": 0.38309859154929576,
|
| 737 |
-
"grad_norm": 0.5117276310920715,
|
| 738 |
-
"learning_rate": 0.0001407466789809601,
|
| 739 |
-
"loss": 3.7999,
|
| 740 |
-
"step": 102
|
| 741 |
-
},
|
| 742 |
-
{
|
| 743 |
-
"epoch": 0.38685446009389673,
|
| 744 |
-
"grad_norm": 0.527893602848053,
|
| 745 |
-
"learning_rate": 0.0001396487510536874,
|
| 746 |
-
"loss": 3.8513,
|
| 747 |
-
"step": 103
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 0.39061032863849765,
|
| 751 |
-
"grad_norm": 0.49835067987442017,
|
| 752 |
-
"learning_rate": 0.00013854512251457247,
|
| 753 |
-
"loss": 3.8276,
|
| 754 |
-
"step": 104
|
| 755 |
-
},
|
| 756 |
-
{
|
| 757 |
-
"epoch": 0.39436619718309857,
|
| 758 |
-
"grad_norm": 0.6509612202644348,
|
| 759 |
-
"learning_rate": 0.0001374359520409444,
|
| 760 |
-
"loss": 4.2759,
|
| 761 |
-
"step": 105
|
| 762 |
-
},
|
| 763 |
-
{
|
| 764 |
-
"epoch": 0.39812206572769954,
|
| 765 |
-
"grad_norm": 0.4912853538990021,
|
| 766 |
-
"learning_rate": 0.0001363213991069397,
|
| 767 |
-
"loss": 3.5757,
|
| 768 |
-
"step": 106
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 0.40187793427230045,
|
| 772 |
-
"grad_norm": 0.5202915668487549,
|
| 773 |
-
"learning_rate": 0.00013520162396057342,
|
| 774 |
-
"loss": 4.0784,
|
| 775 |
-
"step": 107
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 0.4056338028169014,
|
| 779 |
-
"grad_norm": 0.5149546265602112,
|
| 780 |
-
"learning_rate": 0.00013407678760069891,
|
| 781 |
-
"loss": 3.7496,
|
| 782 |
-
"step": 108
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 0.40938967136150234,
|
| 786 |
-
"grad_norm": 0.5628048777580261,
|
| 787 |
-
"learning_rate": 0.00013294705175386003,
|
| 788 |
-
"loss": 4.3535,
|
| 789 |
-
"step": 109
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 0.4131455399061033,
|
| 793 |
-
"grad_norm": 0.5484233498573303,
|
| 794 |
-
"learning_rate": 0.00013181257885103818,
|
| 795 |
-
"loss": 3.9337,
|
| 796 |
-
"step": 110
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 0.4169014084507042,
|
| 800 |
-
"grad_norm": 0.5825195908546448,
|
| 801 |
-
"learning_rate": 0.00013067353200429857,
|
| 802 |
-
"loss": 4.0801,
|
| 803 |
-
"step": 111
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 0.42065727699530514,
|
| 807 |
-
"grad_norm": 0.5359321236610413,
|
| 808 |
-
"learning_rate": 0.00012953007498333808,
|
| 809 |
-
"loss": 4.1705,
|
| 810 |
-
"step": 112
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 0.4244131455399061,
|
| 814 |
-
"grad_norm": 0.6541722416877747,
|
| 815 |
-
"learning_rate": 0.00012838237219193896,
|
| 816 |
-
"loss": 4.6486,
|
| 817 |
-
"step": 113
|
| 818 |
-
},
|
| 819 |
-
{
|
| 820 |
-
"epoch": 0.428169014084507,
|
| 821 |
-
"grad_norm": 0.6243426203727722,
|
| 822 |
-
"learning_rate": 0.00012723058864433118,
|
| 823 |
-
"loss": 4.1711,
|
| 824 |
-
"step": 114
|
| 825 |
-
},
|
| 826 |
-
{
|
| 827 |
-
"epoch": 0.431924882629108,
|
| 828 |
-
"grad_norm": 0.585813581943512,
|
| 829 |
-
"learning_rate": 0.00012607488994146704,
|
| 830 |
-
"loss": 4.0612,
|
| 831 |
-
"step": 115
|
| 832 |
-
},
|
| 833 |
-
{
|
| 834 |
-
"epoch": 0.4356807511737089,
|
| 835 |
-
"grad_norm": 0.5607698559761047,
|
| 836 |
-
"learning_rate": 0.00012491544224721136,
|
| 837 |
-
"loss": 4.0229,
|
| 838 |
-
"step": 116
|
| 839 |
-
},
|
| 840 |
-
{
|
| 841 |
-
"epoch": 0.4394366197183099,
|
| 842 |
-
"grad_norm": 0.519707441329956,
|
| 843 |
-
"learning_rate": 0.00012375241226445088,
|
| 844 |
-
"loss": 4.0728,
|
| 845 |
-
"step": 117
|
| 846 |
-
},
|
| 847 |
-
{
|
| 848 |
-
"epoch": 0.4431924882629108,
|
| 849 |
-
"grad_norm": 0.44573110342025757,
|
| 850 |
-
"learning_rate": 0.00012258596721112608,
|
| 851 |
-
"loss": 3.5927,
|
| 852 |
-
"step": 118
|
| 853 |
-
},
|
| 854 |
-
{
|
| 855 |
-
"epoch": 0.4469483568075117,
|
| 856 |
-
"grad_norm": 0.527217447757721,
|
| 857 |
-
"learning_rate": 0.00012141627479618885,
|
| 858 |
-
"loss": 4.0032,
|
| 859 |
-
"step": 119
|
| 860 |
-
},
|
| 861 |
-
{
|
| 862 |
-
"epoch": 0.4507042253521127,
|
| 863 |
-
"grad_norm": 0.5384316444396973,
|
| 864 |
-
"learning_rate": 0.00012024350319548976,
|
| 865 |
-
"loss": 3.8763,
|
| 866 |
-
"step": 120
|
| 867 |
-
},
|
| 868 |
-
{
|
| 869 |
-
"epoch": 0.4544600938967136,
|
| 870 |
-
"grad_norm": 0.5317234992980957,
|
| 871 |
-
"learning_rate": 0.00011906782102759808,
|
| 872 |
-
"loss": 4.1505,
|
| 873 |
-
"step": 121
|
| 874 |
-
},
|
| 875 |
-
{
|
| 876 |
-
"epoch": 0.4582159624413146,
|
| 877 |
-
"grad_norm": 0.6159288287162781,
|
| 878 |
-
"learning_rate": 0.0001178893973295581,
|
| 879 |
-
"loss": 4.286,
|
| 880 |
-
"step": 122
|
| 881 |
-
},
|
| 882 |
-
{
|
| 883 |
-
"epoch": 0.4619718309859155,
|
| 884 |
-
"grad_norm": 0.5571346282958984,
|
| 885 |
-
"learning_rate": 0.00011670840153258547,
|
| 886 |
-
"loss": 3.8812,
|
| 887 |
-
"step": 123
|
| 888 |
-
},
|
| 889 |
-
{
|
| 890 |
-
"epoch": 0.46572769953051646,
|
| 891 |
-
"grad_norm": 0.5357509851455688,
|
| 892 |
-
"learning_rate": 0.00011552500343770658,
|
| 893 |
-
"loss": 3.9433,
|
| 894 |
-
"step": 124
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 0.4694835680751174,
|
| 898 |
-
"grad_norm": 0.5376076698303223,
|
| 899 |
-
"learning_rate": 0.00011433937319134511,
|
| 900 |
-
"loss": 3.6673,
|
| 901 |
-
"step": 125
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 0.4732394366197183,
|
| 905 |
-
"grad_norm": 0.5321595072746277,
|
| 906 |
-
"learning_rate": 0.00011315168126085857,
|
| 907 |
-
"loss": 3.6064,
|
| 908 |
-
"step": 126
|
| 909 |
-
},
|
| 910 |
-
{
|
| 911 |
-
"epoch": 0.47699530516431926,
|
| 912 |
-
"grad_norm": 0.5370662808418274,
|
| 913 |
-
"learning_rate": 0.00011196209841002909,
|
| 914 |
-
"loss": 3.7798,
|
| 915 |
-
"step": 127
|
| 916 |
-
},
|
| 917 |
-
{
|
| 918 |
-
"epoch": 0.4807511737089202,
|
| 919 |
-
"grad_norm": 0.47796905040740967,
|
| 920 |
-
"learning_rate": 0.00011077079567451111,
|
| 921 |
-
"loss": 3.4657,
|
| 922 |
-
"step": 128
|
| 923 |
-
},
|
| 924 |
-
{
|
| 925 |
-
"epoch": 0.48450704225352115,
|
| 926 |
-
"grad_norm": 0.5204142332077026,
|
| 927 |
-
"learning_rate": 0.00010957794433724051,
|
| 928 |
-
"loss": 3.6982,
|
| 929 |
-
"step": 129
|
| 930 |
-
},
|
| 931 |
-
{
|
| 932 |
-
"epoch": 0.48826291079812206,
|
| 933 |
-
"grad_norm": 0.5429206490516663,
|
| 934 |
-
"learning_rate": 0.00010838371590380765,
|
| 935 |
-
"loss": 3.9558,
|
| 936 |
-
"step": 130
|
| 937 |
-
},
|
| 938 |
-
{
|
| 939 |
-
"epoch": 0.492018779342723,
|
| 940 |
-
"grad_norm": 0.5927493572235107,
|
| 941 |
-
"learning_rate": 0.00010718828207779894,
|
| 942 |
-
"loss": 4.3142,
|
| 943 |
-
"step": 131
|
| 944 |
-
},
|
| 945 |
-
{
|
| 946 |
-
"epoch": 0.49577464788732395,
|
| 947 |
-
"grad_norm": 0.5089852213859558,
|
| 948 |
-
"learning_rate": 0.0001059918147361094,
|
| 949 |
-
"loss": 4.186,
|
| 950 |
-
"step": 132
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 0.49953051643192486,
|
| 954 |
-
"grad_norm": 0.5470064282417297,
|
| 955 |
-
"learning_rate": 0.00010479448590423082,
|
| 956 |
-
"loss": 3.5952,
|
| 957 |
-
"step": 133
|
| 958 |
-
},
|
| 959 |
-
{
|
| 960 |
-
"epoch": 0.5032863849765258,
|
| 961 |
-
"grad_norm": 0.5791990160942078,
|
| 962 |
-
"learning_rate": 0.00010359646773151814,
|
| 963 |
-
"loss": 3.9606,
|
| 964 |
-
"step": 134
|
| 965 |
-
},
|
| 966 |
-
{
|
| 967 |
-
"epoch": 0.5070422535211268,
|
| 968 |
-
"grad_norm": 0.6007618308067322,
|
| 969 |
-
"learning_rate": 0.00010239793246643819,
|
| 970 |
-
"loss": 4.0543,
|
| 971 |
-
"step": 135
|
| 972 |
-
},
|
| 973 |
-
{
|
| 974 |
-
"epoch": 0.5107981220657277,
|
| 975 |
-
"grad_norm": 0.5693303942680359,
|
| 976 |
-
"learning_rate": 0.00010119905243180432,
|
| 977 |
-
"loss": 4.4878,
|
| 978 |
-
"step": 136
|
| 979 |
-
},
|
| 980 |
-
{
|
| 981 |
-
"epoch": 0.5145539906103287,
|
| 982 |
-
"grad_norm": 0.5487916469573975,
|
| 983 |
-
"learning_rate": 0.0001,
|
| 984 |
-
"loss": 3.8789,
|
| 985 |
-
"step": 137
|
| 986 |
-
},
|
| 987 |
-
{
|
| 988 |
-
"epoch": 0.5183098591549296,
|
| 989 |
-
"grad_norm": 0.5099664330482483,
|
| 990 |
-
"learning_rate": 9.880094756819572e-05,
|
| 991 |
-
"loss": 3.9367,
|
| 992 |
-
"step": 138
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 0.5220657276995305,
|
| 996 |
-
"grad_norm": 0.5488762259483337,
|
| 997 |
-
"learning_rate": 9.760206753356184e-05,
|
| 998 |
-
"loss": 3.8747,
|
| 999 |
-
"step": 139
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.5258215962441315,
|
| 1003 |
-
"grad_norm": 0.6004945039749146,
|
| 1004 |
-
"learning_rate": 9.64035322684819e-05,
|
| 1005 |
-
"loss": 4.6112,
|
| 1006 |
-
"step": 140
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 0.5295774647887324,
|
| 1010 |
-
"grad_norm": 0.4959471821784973,
|
| 1011 |
-
"learning_rate": 9.520551409576919e-05,
|
| 1012 |
-
"loss": 3.5334,
|
| 1013 |
-
"step": 141
|
| 1014 |
-
},
|
| 1015 |
-
{
|
| 1016 |
-
"epoch": 0.5333333333333333,
|
| 1017 |
-
"grad_norm": 0.5801951885223389,
|
| 1018 |
-
"learning_rate": 9.400818526389063e-05,
|
| 1019 |
-
"loss": 4.0366,
|
| 1020 |
-
"step": 142
|
| 1021 |
-
},
|
| 1022 |
-
{
|
| 1023 |
-
"epoch": 0.5370892018779343,
|
| 1024 |
-
"grad_norm": 0.6023839116096497,
|
| 1025 |
-
"learning_rate": 9.281171792220107e-05,
|
| 1026 |
-
"loss": 4.0858,
|
| 1027 |
-
"step": 143
|
| 1028 |
-
},
|
| 1029 |
-
{
|
| 1030 |
-
"epoch": 0.5408450704225352,
|
| 1031 |
-
"grad_norm": 0.5538918972015381,
|
| 1032 |
-
"learning_rate": 9.161628409619236e-05,
|
| 1033 |
-
"loss": 3.8907,
|
| 1034 |
-
"step": 144
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 0.5446009389671361,
|
| 1038 |
-
"grad_norm": 0.5293076038360596,
|
| 1039 |
-
"learning_rate": 9.042205566275951e-05,
|
| 1040 |
-
"loss": 3.6444,
|
| 1041 |
-
"step": 145
|
| 1042 |
-
},
|
| 1043 |
-
{
|
| 1044 |
-
"epoch": 0.5483568075117371,
|
| 1045 |
-
"grad_norm": 0.637414276599884,
|
| 1046 |
-
"learning_rate": 8.92292043254889e-05,
|
| 1047 |
-
"loss": 4.5763,
|
| 1048 |
-
"step": 146
|
| 1049 |
-
},
|
| 1050 |
-
{
|
| 1051 |
-
"epoch": 0.5521126760563381,
|
| 1052 |
-
"grad_norm": 0.5648563504219055,
|
| 1053 |
-
"learning_rate": 8.803790158997095e-05,
|
| 1054 |
-
"loss": 4.3616,
|
| 1055 |
-
"step": 147
|
| 1056 |
-
},
|
| 1057 |
-
{
|
| 1058 |
-
"epoch": 0.5558685446009389,
|
| 1059 |
-
"grad_norm": 0.5116698741912842,
|
| 1060 |
-
"learning_rate": 8.684831873914145e-05,
|
| 1061 |
-
"loss": 3.7008,
|
| 1062 |
-
"step": 148
|
| 1063 |
-
},
|
| 1064 |
-
{
|
| 1065 |
-
"epoch": 0.5596244131455399,
|
| 1066 |
-
"grad_norm": 0.6268083453178406,
|
| 1067 |
-
"learning_rate": 8.566062680865494e-05,
|
| 1068 |
-
"loss": 4.224,
|
| 1069 |
-
"step": 149
|
| 1070 |
-
},
|
| 1071 |
-
{
|
| 1072 |
-
"epoch": 0.5633802816901409,
|
| 1073 |
-
"grad_norm": 0.47161865234375,
|
| 1074 |
-
"learning_rate": 8.447499656229344e-05,
|
| 1075 |
-
"loss": 3.5188,
|
| 1076 |
-
"step": 150
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 0.5633802816901409,
|
| 1080 |
-
"eval_loss": 0.5007660388946533,
|
| 1081 |
-
"eval_runtime": 373.0551,
|
| 1082 |
-
"eval_samples_per_second": 2.538,
|
| 1083 |
-
"eval_steps_per_second": 0.635,
|
| 1084 |
-
"step": 150
|
| 1085 |
-
},
|
| 1086 |
-
{
|
| 1087 |
-
"epoch": 0.5671361502347417,
|
| 1088 |
-
"grad_norm": 0.5017052292823792,
|
| 1089 |
-
"learning_rate": 8.329159846741457e-05,
|
| 1090 |
-
"loss": 3.7126,
|
| 1091 |
-
"step": 151
|
| 1092 |
-
},
|
| 1093 |
-
{
|
| 1094 |
-
"epoch": 0.5708920187793427,
|
| 1095 |
-
"grad_norm": 0.6328552961349487,
|
| 1096 |
-
"learning_rate": 8.211060267044191e-05,
|
| 1097 |
-
"loss": 4.3953,
|
| 1098 |
-
"step": 152
|
| 1099 |
-
},
|
| 1100 |
-
{
|
| 1101 |
-
"epoch": 0.5746478873239437,
|
| 1102 |
-
"grad_norm": 0.6981328725814819,
|
| 1103 |
-
"learning_rate": 8.093217897240195e-05,
|
| 1104 |
-
"loss": 4.7259,
|
| 1105 |
-
"step": 153
|
| 1106 |
-
},
|
| 1107 |
-
{
|
| 1108 |
-
"epoch": 0.5784037558685446,
|
| 1109 |
-
"grad_norm": 0.5479520559310913,
|
| 1110 |
-
"learning_rate": 7.975649680451024e-05,
|
| 1111 |
-
"loss": 3.9748,
|
| 1112 |
-
"step": 154
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 0.5821596244131455,
|
| 1116 |
-
"grad_norm": 0.5974417328834534,
|
| 1117 |
-
"learning_rate": 7.858372520381119e-05,
|
| 1118 |
-
"loss": 4.0999,
|
| 1119 |
-
"step": 155
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 0.5859154929577465,
|
| 1123 |
-
"grad_norm": 0.5800752639770508,
|
| 1124 |
-
"learning_rate": 7.741403278887397e-05,
|
| 1125 |
-
"loss": 4.3825,
|
| 1126 |
-
"step": 156
|
| 1127 |
-
},
|
| 1128 |
-
{
|
| 1129 |
-
"epoch": 0.5896713615023474,
|
| 1130 |
-
"grad_norm": 0.5555791258811951,
|
| 1131 |
-
"learning_rate": 7.624758773554914e-05,
|
| 1132 |
-
"loss": 4.0412,
|
| 1133 |
-
"step": 157
|
| 1134 |
-
},
|
| 1135 |
-
{
|
| 1136 |
-
"epoch": 0.5934272300469483,
|
| 1137 |
-
"grad_norm": 0.5661817789077759,
|
| 1138 |
-
"learning_rate": 7.508455775278867e-05,
|
| 1139 |
-
"loss": 3.8501,
|
| 1140 |
-
"step": 158
|
| 1141 |
-
},
|
| 1142 |
-
{
|
| 1143 |
-
"epoch": 0.5971830985915493,
|
| 1144 |
-
"grad_norm": 0.5640938878059387,
|
| 1145 |
-
"learning_rate": 7.392511005853297e-05,
|
| 1146 |
-
"loss": 4.0421,
|
| 1147 |
-
"step": 159
|
| 1148 |
-
},
|
| 1149 |
-
{
|
| 1150 |
-
"epoch": 0.6009389671361502,
|
| 1151 |
-
"grad_norm": 0.5214723944664001,
|
| 1152 |
-
"learning_rate": 7.276941135566884e-05,
|
| 1153 |
-
"loss": 3.6713,
|
| 1154 |
-
"step": 160
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 0.6046948356807512,
|
| 1158 |
-
"grad_norm": 0.6109952926635742,
|
| 1159 |
-
"learning_rate": 7.161762780806103e-05,
|
| 1160 |
-
"loss": 4.1644,
|
| 1161 |
-
"step": 161
|
| 1162 |
-
},
|
| 1163 |
-
{
|
| 1164 |
-
"epoch": 0.6084507042253521,
|
| 1165 |
-
"grad_norm": 0.4427598714828491,
|
| 1166 |
-
"learning_rate": 7.046992501666195e-05,
|
| 1167 |
-
"loss": 3.3083,
|
| 1168 |
-
"step": 162
|
| 1169 |
-
},
|
| 1170 |
-
{
|
| 1171 |
-
"epoch": 0.612206572769953,
|
| 1172 |
-
"grad_norm": 0.593859076499939,
|
| 1173 |
-
"learning_rate": 6.932646799570144e-05,
|
| 1174 |
-
"loss": 4.0755,
|
| 1175 |
-
"step": 163
|
| 1176 |
-
},
|
| 1177 |
-
{
|
| 1178 |
-
"epoch": 0.615962441314554,
|
| 1179 |
-
"grad_norm": 0.5258002281188965,
|
| 1180 |
-
"learning_rate": 6.818742114896184e-05,
|
| 1181 |
-
"loss": 3.9657,
|
| 1182 |
-
"step": 164
|
| 1183 |
-
},
|
| 1184 |
-
{
|
| 1185 |
-
"epoch": 0.6197183098591549,
|
| 1186 |
-
"grad_norm": 0.6475521922111511,
|
| 1187 |
-
"learning_rate": 6.705294824614004e-05,
|
| 1188 |
-
"loss": 4.8808,
|
| 1189 |
-
"step": 165
|
| 1190 |
-
},
|
| 1191 |
-
{
|
| 1192 |
-
"epoch": 0.6234741784037559,
|
| 1193 |
-
"grad_norm": 0.45996585488319397,
|
| 1194 |
-
"learning_rate": 6.592321239930112e-05,
|
| 1195 |
-
"loss": 3.2131,
|
| 1196 |
-
"step": 166
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 0.6272300469483568,
|
| 1200 |
-
"grad_norm": 0.505383312702179,
|
| 1201 |
-
"learning_rate": 6.479837603942665e-05,
|
| 1202 |
-
"loss": 3.8211,
|
| 1203 |
-
"step": 167
|
| 1204 |
-
},
|
| 1205 |
-
{
|
| 1206 |
-
"epoch": 0.6309859154929578,
|
| 1207 |
-
"grad_norm": 0.5987924933433533,
|
| 1208 |
-
"learning_rate": 6.367860089306028e-05,
|
| 1209 |
-
"loss": 4.073,
|
| 1210 |
-
"step": 168
|
| 1211 |
-
},
|
| 1212 |
-
{
|
| 1213 |
-
"epoch": 0.6347417840375587,
|
| 1214 |
-
"grad_norm": 0.5297787189483643,
|
| 1215 |
-
"learning_rate": 6.256404795905561e-05,
|
| 1216 |
-
"loss": 3.8255,
|
| 1217 |
-
"step": 169
|
| 1218 |
-
},
|
| 1219 |
-
{
|
| 1220 |
-
"epoch": 0.6384976525821596,
|
| 1221 |
-
"grad_norm": 0.5636645555496216,
|
| 1222 |
-
"learning_rate": 6.145487748542753e-05,
|
| 1223 |
-
"loss": 4.4,
|
| 1224 |
-
"step": 170
|
| 1225 |
-
},
|
| 1226 |
-
{
|
| 1227 |
-
"epoch": 0.6422535211267606,
|
| 1228 |
-
"grad_norm": 0.584668755531311,
|
| 1229 |
-
"learning_rate": 6.035124894631263e-05,
|
| 1230 |
-
"loss": 3.9181,
|
| 1231 |
-
"step": 171
|
| 1232 |
-
},
|
| 1233 |
-
{
|
| 1234 |
-
"epoch": 0.6460093896713615,
|
| 1235 |
-
"grad_norm": 0.6228652000427246,
|
| 1236 |
-
"learning_rate": 5.925332101903994e-05,
|
| 1237 |
-
"loss": 4.1693,
|
| 1238 |
-
"step": 172
|
| 1239 |
-
},
|
| 1240 |
-
{
|
| 1241 |
-
"epoch": 0.6497652582159624,
|
| 1242 |
-
"grad_norm": 0.6707691550254822,
|
| 1243 |
-
"learning_rate": 5.816125156131691e-05,
|
| 1244 |
-
"loss": 4.8515,
|
| 1245 |
-
"step": 173
|
| 1246 |
-
},
|
| 1247 |
-
{
|
| 1248 |
-
"epoch": 0.6535211267605634,
|
| 1249 |
-
"grad_norm": 0.5788469314575195,
|
| 1250 |
-
"learning_rate": 5.707519758853288e-05,
|
| 1251 |
-
"loss": 3.9823,
|
| 1252 |
-
"step": 174
|
| 1253 |
-
},
|
| 1254 |
-
{
|
| 1255 |
-
"epoch": 0.6572769953051644,
|
| 1256 |
-
"grad_norm": 0.5524411201477051,
|
| 1257 |
-
"learning_rate": 5.5995315251183734e-05,
|
| 1258 |
-
"loss": 3.8754,
|
| 1259 |
-
"step": 175
|
| 1260 |
-
},
|
| 1261 |
-
{
|
| 1262 |
-
"epoch": 0.6610328638497652,
|
| 1263 |
-
"grad_norm": 0.5607343912124634,
|
| 1264 |
-
"learning_rate": 5.492175981242097e-05,
|
| 1265 |
-
"loss": 4.0384,
|
| 1266 |
-
"step": 176
|
| 1267 |
-
},
|
| 1268 |
-
{
|
| 1269 |
-
"epoch": 0.6647887323943662,
|
| 1270 |
-
"grad_norm": 0.5071132183074951,
|
| 1271 |
-
"learning_rate": 5.385468562572823e-05,
|
| 1272 |
-
"loss": 3.4805,
|
| 1273 |
-
"step": 177
|
| 1274 |
-
},
|
| 1275 |
-
{
|
| 1276 |
-
"epoch": 0.6685446009389672,
|
| 1277 |
-
"grad_norm": 0.5974758267402649,
|
| 1278 |
-
"learning_rate": 5.279424611272873e-05,
|
| 1279 |
-
"loss": 4.2209,
|
| 1280 |
-
"step": 178
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 0.672300469483568,
|
| 1284 |
-
"grad_norm": 0.48757869005203247,
|
| 1285 |
-
"learning_rate": 5.174059374112657e-05,
|
| 1286 |
-
"loss": 3.6575,
|
| 1287 |
-
"step": 179
|
| 1288 |
-
},
|
| 1289 |
-
{
|
| 1290 |
-
"epoch": 0.676056338028169,
|
| 1291 |
-
"grad_norm": 0.6194841861724854,
|
| 1292 |
-
"learning_rate": 5.0693880002785456e-05,
|
| 1293 |
-
"loss": 4.5966,
|
| 1294 |
-
"step": 180
|
| 1295 |
-
},
|
| 1296 |
-
{
|
| 1297 |
-
"epoch": 0.67981220657277,
|
| 1298 |
-
"grad_norm": 0.5687793493270874,
|
| 1299 |
-
"learning_rate": 4.965425539194726e-05,
|
| 1300 |
-
"loss": 3.8884,
|
| 1301 |
-
"step": 181
|
| 1302 |
-
},
|
| 1303 |
-
{
|
| 1304 |
-
"epoch": 0.6835680751173709,
|
| 1305 |
-
"grad_norm": 0.5817456245422363,
|
| 1306 |
-
"learning_rate": 4.8621869383594406e-05,
|
| 1307 |
-
"loss": 4.2781,
|
| 1308 |
-
"step": 182
|
| 1309 |
-
},
|
| 1310 |
-
{
|
| 1311 |
-
"epoch": 0.6873239436619718,
|
| 1312 |
-
"grad_norm": 0.4698617458343506,
|
| 1313 |
-
"learning_rate": 4.759687041195874e-05,
|
| 1314 |
-
"loss": 3.5443,
|
| 1315 |
-
"step": 183
|
| 1316 |
-
},
|
| 1317 |
-
{
|
| 1318 |
-
"epoch": 0.6910798122065728,
|
| 1319 |
-
"grad_norm": 0.5563104748725891,
|
| 1320 |
-
"learning_rate": 4.657940584917983e-05,
|
| 1321 |
-
"loss": 3.8302,
|
| 1322 |
-
"step": 184
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 0.6948356807511737,
|
| 1326 |
-
"grad_norm": 0.5313715934753418,
|
| 1327 |
-
"learning_rate": 4.556962198411631e-05,
|
| 1328 |
-
"loss": 3.7517,
|
| 1329 |
-
"step": 185
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 0.6985915492957746,
|
| 1333 |
-
"grad_norm": 0.47465792298316956,
|
| 1334 |
-
"learning_rate": 4.45676640013126e-05,
|
| 1335 |
-
"loss": 3.4732,
|
| 1336 |
-
"step": 186
|
| 1337 |
-
},
|
| 1338 |
-
{
|
| 1339 |
-
"epoch": 0.7023474178403756,
|
| 1340 |
-
"grad_norm": 0.49628129601478577,
|
| 1341 |
-
"learning_rate": 4.3573675960124684e-05,
|
| 1342 |
-
"loss": 3.7668,
|
| 1343 |
-
"step": 187
|
| 1344 |
-
},
|
| 1345 |
-
{
|
| 1346 |
-
"epoch": 0.7061032863849765,
|
| 1347 |
-
"grad_norm": 0.5268252491950989,
|
| 1348 |
-
"learning_rate": 4.258780077400748e-05,
|
| 1349 |
-
"loss": 3.5747,
|
| 1350 |
-
"step": 188
|
| 1351 |
-
},
|
| 1352 |
-
{
|
| 1353 |
-
"epoch": 0.7098591549295775,
|
| 1354 |
-
"grad_norm": 0.6187554001808167,
|
| 1355 |
-
"learning_rate": 4.161018018996727e-05,
|
| 1356 |
-
"loss": 4.3695,
|
| 1357 |
-
"step": 189
|
| 1358 |
-
},
|
| 1359 |
-
{
|
| 1360 |
-
"epoch": 0.7136150234741784,
|
| 1361 |
-
"grad_norm": 0.5253807902336121,
|
| 1362 |
-
"learning_rate": 4.064095476818133e-05,
|
| 1363 |
-
"loss": 3.8376,
|
| 1364 |
-
"step": 190
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 0.7173708920187793,
|
| 1368 |
-
"grad_norm": 0.5611537098884583,
|
| 1369 |
-
"learning_rate": 3.968026386178867e-05,
|
| 1370 |
-
"loss": 3.8718,
|
| 1371 |
-
"step": 191
|
| 1372 |
-
},
|
| 1373 |
-
{
|
| 1374 |
-
"epoch": 0.7211267605633803,
|
| 1375 |
-
"grad_norm": 0.6236064434051514,
|
| 1376 |
-
"learning_rate": 3.87282455968541e-05,
|
| 1377 |
-
"loss": 4.4724,
|
| 1378 |
-
"step": 192
|
| 1379 |
-
},
|
| 1380 |
-
{
|
| 1381 |
-
"epoch": 0.7248826291079812,
|
| 1382 |
-
"grad_norm": 0.4799625277519226,
|
| 1383 |
-
"learning_rate": 3.778503685250873e-05,
|
| 1384 |
-
"loss": 3.6452,
|
| 1385 |
-
"step": 193
|
| 1386 |
-
},
|
| 1387 |
-
{
|
| 1388 |
-
"epoch": 0.7286384976525822,
|
| 1389 |
-
"grad_norm": 0.5699834227561951,
|
| 1390 |
-
"learning_rate": 3.685077324126992e-05,
|
| 1391 |
-
"loss": 3.9373,
|
| 1392 |
-
"step": 194
|
| 1393 |
-
},
|
| 1394 |
-
{
|
| 1395 |
-
"epoch": 0.7323943661971831,
|
| 1396 |
-
"grad_norm": 0.49022650718688965,
|
| 1397 |
-
"learning_rate": 3.592558908954295e-05,
|
| 1398 |
-
"loss": 3.3991,
|
| 1399 |
-
"step": 195
|
| 1400 |
-
},
|
| 1401 |
-
{
|
| 1402 |
-
"epoch": 0.7361502347417841,
|
| 1403 |
-
"grad_norm": 0.5775969624519348,
|
| 1404 |
-
"learning_rate": 3.500961741830821e-05,
|
| 1405 |
-
"loss": 4.2728,
|
| 1406 |
-
"step": 196
|
| 1407 |
-
},
|
| 1408 |
-
{
|
| 1409 |
-
"epoch": 0.739906103286385,
|
| 1410 |
-
"grad_norm": 0.5632807612419128,
|
| 1411 |
-
"learning_rate": 3.410298992399524e-05,
|
| 1412 |
-
"loss": 4.1647,
|
| 1413 |
-
"step": 197
|
| 1414 |
-
},
|
| 1415 |
-
{
|
| 1416 |
-
"epoch": 0.7436619718309859,
|
| 1417 |
-
"grad_norm": 0.4752277135848999,
|
| 1418 |
-
"learning_rate": 3.3205836959548296e-05,
|
| 1419 |
-
"loss": 3.3707,
|
| 1420 |
-
"step": 198
|
| 1421 |
-
},
|
| 1422 |
-
{
|
| 1423 |
-
"epoch": 0.7474178403755869,
|
| 1424 |
-
"grad_norm": 0.5167598724365234,
|
| 1425 |
-
"learning_rate": 3.231828751568401e-05,
|
| 1426 |
-
"loss": 3.6365,
|
| 1427 |
-
"step": 199
|
| 1428 |
-
},
|
| 1429 |
-
{
|
| 1430 |
-
"epoch": 0.7511737089201878,
|
| 1431 |
-
"grad_norm": 0.5540789365768433,
|
| 1432 |
-
"learning_rate": 3.144046920234553e-05,
|
| 1433 |
-
"loss": 3.8104,
|
| 1434 |
-
"step": 200
|
| 1435 |
-
},
|
| 1436 |
-
{
|
| 1437 |
-
"epoch": 0.7511737089201878,
|
| 1438 |
-
"eval_loss": 0.4954932928085327,
|
| 1439 |
-
"eval_runtime": 367.7455,
|
| 1440 |
-
"eval_samples_per_second": 2.575,
|
| 1441 |
-
"eval_steps_per_second": 0.644,
|
| 1442 |
-
"step": 200
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 0.7549295774647887,
|
| 1446 |
-
"grad_norm": 0.4975653886795044,
|
| 1447 |
-
"learning_rate": 3.0572508230355246e-05,
|
| 1448 |
-
"loss": 3.763,
|
| 1449 |
-
"step": 201
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 0.7586854460093897,
|
| 1453 |
-
"grad_norm": 0.5943359136581421,
|
| 1454 |
-
"learning_rate": 2.971452939326802e-05,
|
| 1455 |
-
"loss": 4.1011,
|
| 1456 |
-
"step": 202
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 0.7624413145539906,
|
| 1460 |
-
"grad_norm": 0.5947958827018738,
|
| 1461 |
-
"learning_rate": 2.8866656049429162e-05,
|
| 1462 |
-
"loss": 3.837,
|
| 1463 |
-
"step": 203
|
| 1464 |
-
},
|
| 1465 |
-
{
|
| 1466 |
-
"epoch": 0.7661971830985915,
|
| 1467 |
-
"grad_norm": 0.55486661195755,
|
| 1468 |
-
"learning_rate": 2.8029010104237785e-05,
|
| 1469 |
-
"loss": 3.773,
|
| 1470 |
-
"step": 204
|
| 1471 |
-
},
|
| 1472 |
-
{
|
| 1473 |
-
"epoch": 0.7699530516431925,
|
| 1474 |
-
"grad_norm": 0.6001894474029541,
|
| 1475 |
-
"learning_rate": 2.720171199261987e-05,
|
| 1476 |
-
"loss": 4.1092,
|
| 1477 |
-
"step": 205
|
| 1478 |
-
},
|
| 1479 |
-
{
|
| 1480 |
-
"epoch": 0.7737089201877935,
|
| 1481 |
-
"grad_norm": 0.611171305179596,
|
| 1482 |
-
"learning_rate": 2.638488066171201e-05,
|
| 1483 |
-
"loss": 4.2872,
|
| 1484 |
-
"step": 206
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 0.7774647887323943,
|
| 1488 |
-
"grad_norm": 0.5929466485977173,
|
| 1489 |
-
"learning_rate": 2.5578633553759878e-05,
|
| 1490 |
-
"loss": 4.0139,
|
| 1491 |
-
"step": 207
|
| 1492 |
-
},
|
| 1493 |
-
{
|
| 1494 |
-
"epoch": 0.7812206572769953,
|
| 1495 |
-
"grad_norm": 0.5859886407852173,
|
| 1496 |
-
"learning_rate": 2.4783086589232295e-05,
|
| 1497 |
-
"loss": 3.9495,
|
| 1498 |
-
"step": 208
|
| 1499 |
-
},
|
| 1500 |
-
{
|
| 1501 |
-
"epoch": 0.7849765258215963,
|
| 1502 |
-
"grad_norm": 0.5463722348213196,
|
| 1503 |
-
"learning_rate": 2.3998354150154555e-05,
|
| 1504 |
-
"loss": 3.7008,
|
| 1505 |
-
"step": 209
|
| 1506 |
-
},
|
| 1507 |
-
{
|
| 1508 |
-
"epoch": 0.7887323943661971,
|
| 1509 |
-
"grad_norm": 0.5370416045188904,
|
| 1510 |
-
"learning_rate": 2.3224549063662927e-05,
|
| 1511 |
-
"loss": 3.9123,
|
| 1512 |
-
"step": 210
|
| 1513 |
-
},
|
| 1514 |
-
{
|
| 1515 |
-
"epoch": 0.7924882629107981,
|
| 1516 |
-
"grad_norm": 0.5654124021530151,
|
| 1517 |
-
"learning_rate": 2.246178258578234e-05,
|
| 1518 |
-
"loss": 3.816,
|
| 1519 |
-
"step": 211
|
| 1520 |
-
},
|
| 1521 |
-
{
|
| 1522 |
-
"epoch": 0.7962441314553991,
|
| 1523 |
-
"grad_norm": 0.5404929518699646,
|
| 1524 |
-
"learning_rate": 2.171016438543059e-05,
|
| 1525 |
-
"loss": 3.943,
|
| 1526 |
-
"step": 212
|
| 1527 |
-
},
|
| 1528 |
-
{
|
| 1529 |
-
"epoch": 0.8,
|
| 1530 |
-
"grad_norm": 0.5264220237731934,
|
| 1531 |
-
"learning_rate": 2.096980252865005e-05,
|
| 1532 |
-
"loss": 3.8148,
|
| 1533 |
-
"step": 213
|
| 1534 |
-
},
|
| 1535 |
-
{
|
| 1536 |
-
"epoch": 0.8037558685446009,
|
| 1537 |
-
"grad_norm": 0.5364089012145996,
|
| 1538 |
-
"learning_rate": 2.0240803463070425e-05,
|
| 1539 |
-
"loss": 4.0956,
|
| 1540 |
-
"step": 214
|
| 1541 |
-
},
|
| 1542 |
-
{
|
| 1543 |
-
"epoch": 0.8075117370892019,
|
| 1544 |
-
"grad_norm": 0.49832502007484436,
|
| 1545 |
-
"learning_rate": 1.9523272002603742e-05,
|
| 1546 |
-
"loss": 3.5919,
|
| 1547 |
-
"step": 215
|
| 1548 |
-
},
|
| 1549 |
-
{
|
| 1550 |
-
"epoch": 0.8112676056338028,
|
| 1551 |
-
"grad_norm": 0.5661212205886841,
|
| 1552 |
-
"learning_rate": 1.8817311312374564e-05,
|
| 1553 |
-
"loss": 3.9309,
|
| 1554 |
-
"step": 216
|
| 1555 |
-
},
|
| 1556 |
-
{
|
| 1557 |
-
"epoch": 0.8150234741784037,
|
| 1558 |
-
"grad_norm": 0.6174516677856445,
|
| 1559 |
-
"learning_rate": 1.8123022893887065e-05,
|
| 1560 |
-
"loss": 4.4702,
|
| 1561 |
-
"step": 217
|
| 1562 |
-
},
|
| 1563 |
-
{
|
| 1564 |
-
"epoch": 0.8187793427230047,
|
| 1565 |
-
"grad_norm": 0.5399917364120483,
|
| 1566 |
-
"learning_rate": 1.744050657043137e-05,
|
| 1567 |
-
"loss": 3.8469,
|
| 1568 |
-
"step": 218
|
| 1569 |
-
},
|
| 1570 |
-
{
|
| 1571 |
-
"epoch": 0.8225352112676056,
|
| 1572 |
-
"grad_norm": 0.48354753851890564,
|
| 1573 |
-
"learning_rate": 1.6769860472731257e-05,
|
| 1574 |
-
"loss": 3.5587,
|
| 1575 |
-
"step": 219
|
| 1576 |
-
},
|
| 1577 |
-
{
|
| 1578 |
-
"epoch": 0.8262910798122066,
|
| 1579 |
-
"grad_norm": 0.5603431463241577,
|
| 1580 |
-
"learning_rate": 1.6111181024835e-05,
|
| 1581 |
-
"loss": 4.3805,
|
| 1582 |
-
"step": 220
|
| 1583 |
-
},
|
| 1584 |
-
{
|
| 1585 |
-
"epoch": 0.8300469483568075,
|
| 1586 |
-
"grad_norm": 0.5792990326881409,
|
| 1587 |
-
"learning_rate": 1.5464562930251814e-05,
|
| 1588 |
-
"loss": 4.2204,
|
| 1589 |
-
"step": 221
|
| 1590 |
-
},
|
| 1591 |
-
{
|
| 1592 |
-
"epoch": 0.8338028169014085,
|
| 1593 |
-
"grad_norm": 0.5376021862030029,
|
| 1594 |
-
"learning_rate": 1.4830099158335563e-05,
|
| 1595 |
-
"loss": 3.8365,
|
| 1596 |
-
"step": 222
|
| 1597 |
-
},
|
| 1598 |
-
{
|
| 1599 |
-
"epoch": 0.8375586854460094,
|
| 1600 |
-
"grad_norm": 0.5793043971061707,
|
| 1601 |
-
"learning_rate": 1.4207880930917871e-05,
|
| 1602 |
-
"loss": 4.064,
|
| 1603 |
-
"step": 223
|
| 1604 |
-
},
|
| 1605 |
-
{
|
| 1606 |
-
"epoch": 0.8413145539906103,
|
| 1607 |
-
"grad_norm": 0.5597378611564636,
|
| 1608 |
-
"learning_rate": 1.3597997709192378e-05,
|
| 1609 |
-
"loss": 3.8224,
|
| 1610 |
-
"step": 224
|
| 1611 |
-
},
|
| 1612 |
-
{
|
| 1613 |
-
"epoch": 0.8450704225352113,
|
| 1614 |
-
"grad_norm": 0.5336353182792664,
|
| 1615 |
-
"learning_rate": 1.3000537180852212e-05,
|
| 1616 |
-
"loss": 3.7203,
|
| 1617 |
-
"step": 225
|
| 1618 |
-
},
|
| 1619 |
-
{
|
| 1620 |
-
"epoch": 0.8488262910798122,
|
| 1621 |
-
"grad_norm": 0.640953004360199,
|
| 1622 |
-
"learning_rate": 1.2415585247482498e-05,
|
| 1623 |
-
"loss": 4.3212,
|
| 1624 |
-
"step": 226
|
| 1625 |
-
},
|
| 1626 |
-
{
|
| 1627 |
-
"epoch": 0.8525821596244132,
|
| 1628 |
-
"grad_norm": 0.45982062816619873,
|
| 1629 |
-
"learning_rate": 1.1843226012209529e-05,
|
| 1630 |
-
"loss": 3.6229,
|
| 1631 |
-
"step": 227
|
| 1632 |
-
},
|
| 1633 |
-
{
|
| 1634 |
-
"epoch": 0.856338028169014,
|
| 1635 |
-
"grad_norm": 0.5055301189422607,
|
| 1636 |
-
"learning_rate": 1.128354176760873e-05,
|
| 1637 |
-
"loss": 3.6906,
|
| 1638 |
-
"step": 228
|
| 1639 |
-
},
|
| 1640 |
-
{
|
| 1641 |
-
"epoch": 0.860093896713615,
|
| 1642 |
-
"grad_norm": 0.4451459050178528,
|
| 1643 |
-
"learning_rate": 1.073661298387265e-05,
|
| 1644 |
-
"loss": 3.3596,
|
| 1645 |
-
"step": 229
|
| 1646 |
-
},
|
| 1647 |
-
{
|
| 1648 |
-
"epoch": 0.863849765258216,
|
| 1649 |
-
"grad_norm": 0.6167091727256775,
|
| 1650 |
-
"learning_rate": 1.0202518297241237e-05,
|
| 1651 |
-
"loss": 4.6817,
|
| 1652 |
-
"step": 230
|
| 1653 |
-
},
|
| 1654 |
-
{
|
| 1655 |
-
"epoch": 0.8676056338028169,
|
| 1656 |
-
"grad_norm": 0.5457577705383301,
|
| 1657 |
-
"learning_rate": 9.681334498695648e-06,
|
| 1658 |
-
"loss": 4.2546,
|
| 1659 |
-
"step": 231
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 0.8713615023474178,
|
| 1663 |
-
"grad_norm": 0.49405384063720703,
|
| 1664 |
-
"learning_rate": 9.173136522917457e-06,
|
| 1665 |
-
"loss": 3.7713,
|
| 1666 |
-
"step": 232
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 0.8751173708920188,
|
| 1670 |
-
"grad_norm": 0.5279140472412109,
|
| 1671 |
-
"learning_rate": 8.677997437514629e-06,
|
| 1672 |
-
"loss": 3.7468,
|
| 1673 |
-
"step": 233
|
| 1674 |
-
},
|
| 1675 |
-
{
|
| 1676 |
-
"epoch": 0.8788732394366198,
|
| 1677 |
-
"grad_norm": 0.5161781311035156,
|
| 1678 |
-
"learning_rate": 8.195988432516078e-06,
|
| 1679 |
-
"loss": 4.2746,
|
| 1680 |
-
"step": 234
|
| 1681 |
-
},
|
| 1682 |
-
{
|
| 1683 |
-
"epoch": 0.8826291079812206,
|
| 1684 |
-
"grad_norm": 0.5855900049209595,
|
| 1685 |
-
"learning_rate": 7.727178810136093e-06,
|
| 1686 |
-
"loss": 4.1113,
|
| 1687 |
-
"step": 235
|
| 1688 |
-
},
|
| 1689 |
-
{
|
| 1690 |
-
"epoch": 0.8863849765258216,
|
| 1691 |
-
"grad_norm": 0.4686482548713684,
|
| 1692 |
-
"learning_rate": 7.27163597481022e-06,
|
| 1693 |
-
"loss": 3.3821,
|
| 1694 |
-
"step": 236
|
| 1695 |
-
},
|
| 1696 |
-
{
|
| 1697 |
-
"epoch": 0.8901408450704226,
|
| 1698 |
-
"grad_norm": 0.5629131197929382,
|
| 1699 |
-
"learning_rate": 6.829425423504021e-06,
|
| 1700 |
-
"loss": 4.1901,
|
| 1701 |
-
"step": 237
|
| 1702 |
-
},
|
| 1703 |
-
{
|
| 1704 |
-
"epoch": 0.8938967136150234,
|
| 1705 |
-
"grad_norm": 0.5782991647720337,
|
| 1706 |
-
"learning_rate": 6.4006107362960195e-06,
|
| 1707 |
-
"loss": 4.3302,
|
| 1708 |
-
"step": 238
|
| 1709 |
-
},
|
| 1710 |
-
{
|
| 1711 |
-
"epoch": 0.8976525821596244,
|
| 1712 |
-
"grad_norm": 0.5707590579986572,
|
| 1713 |
-
"learning_rate": 5.985253567236304e-06,
|
| 1714 |
-
"loss": 3.9955,
|
| 1715 |
-
"step": 239
|
| 1716 |
-
},
|
| 1717 |
-
{
|
| 1718 |
-
"epoch": 0.9014084507042254,
|
| 1719 |
-
"grad_norm": 0.4625610411167145,
|
| 1720 |
-
"learning_rate": 5.583413635482082e-06,
|
| 1721 |
-
"loss": 3.5662,
|
| 1722 |
-
"step": 240
|
| 1723 |
-
},
|
| 1724 |
-
{
|
| 1725 |
-
"epoch": 0.9051643192488263,
|
| 1726 |
-
"grad_norm": 0.6621753573417664,
|
| 1727 |
-
"learning_rate": 5.19514871671134e-06,
|
| 1728 |
-
"loss": 4.5634,
|
| 1729 |
-
"step": 241
|
| 1730 |
-
},
|
| 1731 |
-
{
|
| 1732 |
-
"epoch": 0.9089201877934272,
|
| 1733 |
-
"grad_norm": 0.4976242482662201,
|
| 1734 |
-
"learning_rate": 4.82051463481602e-06,
|
| 1735 |
-
"loss": 3.5897,
|
| 1736 |
-
"step": 242
|
| 1737 |
-
},
|
| 1738 |
-
{
|
| 1739 |
-
"epoch": 0.9126760563380282,
|
| 1740 |
-
"grad_norm": 0.51161789894104,
|
| 1741 |
-
"learning_rate": 4.45956525387573e-06,
|
| 1742 |
-
"loss": 3.6594,
|
| 1743 |
-
"step": 243
|
| 1744 |
-
},
|
| 1745 |
-
{
|
| 1746 |
-
"epoch": 0.9164319248826291,
|
| 1747 |
-
"grad_norm": 0.5785262584686279,
|
| 1748 |
-
"learning_rate": 4.112352470413328e-06,
|
| 1749 |
-
"loss": 4.031,
|
| 1750 |
-
"step": 244
|
| 1751 |
-
},
|
| 1752 |
-
{
|
| 1753 |
-
"epoch": 0.92018779342723,
|
| 1754 |
-
"grad_norm": 0.5122177004814148,
|
| 1755 |
-
"learning_rate": 3.778926205933342e-06,
|
| 1756 |
-
"loss": 3.6733,
|
| 1757 |
-
"step": 245
|
| 1758 |
-
},
|
| 1759 |
-
{
|
| 1760 |
-
"epoch": 0.923943661971831,
|
| 1761 |
-
"grad_norm": 0.5668466687202454,
|
| 1762 |
-
"learning_rate": 3.459334399744374e-06,
|
| 1763 |
-
"loss": 3.8761,
|
| 1764 |
-
"step": 246
|
| 1765 |
-
},
|
| 1766 |
-
{
|
| 1767 |
-
"epoch": 0.927699530516432,
|
| 1768 |
-
"grad_norm": 0.5304160714149475,
|
| 1769 |
-
"learning_rate": 3.1536230020664417e-06,
|
| 1770 |
-
"loss": 3.3638,
|
| 1771 |
-
"step": 247
|
| 1772 |
-
},
|
| 1773 |
-
{
|
| 1774 |
-
"epoch": 0.9314553990610329,
|
| 1775 |
-
"grad_norm": 0.5929594039916992,
|
| 1776 |
-
"learning_rate": 2.861835967424409e-06,
|
| 1777 |
-
"loss": 4.1158,
|
| 1778 |
-
"step": 248
|
| 1779 |
-
},
|
| 1780 |
-
{
|
| 1781 |
-
"epoch": 0.9352112676056338,
|
| 1782 |
-
"grad_norm": 0.5661305785179138,
|
| 1783 |
-
"learning_rate": 2.5840152483282752e-06,
|
| 1784 |
-
"loss": 3.8846,
|
| 1785 |
-
"step": 249
|
| 1786 |
-
},
|
| 1787 |
-
{
|
| 1788 |
-
"epoch": 0.9389671361502347,
|
| 1789 |
-
"grad_norm": 0.5555335879325867,
|
| 1790 |
-
"learning_rate": 2.3202007892413447e-06,
|
| 1791 |
-
"loss": 3.9409,
|
| 1792 |
-
"step": 250
|
| 1793 |
-
},
|
| 1794 |
-
{
|
| 1795 |
-
"epoch": 0.9389671361502347,
|
| 1796 |
-
"eval_loss": 0.4938514232635498,
|
| 1797 |
-
"eval_runtime": 365.2814,
|
| 1798 |
-
"eval_samples_per_second": 2.593,
|
| 1799 |
-
"eval_steps_per_second": 0.649,
|
| 1800 |
-
"step": 250
|
| 1801 |
}
|
| 1802 |
],
|
| 1803 |
"logging_steps": 1,
|
| 1804 |
-
"max_steps":
|
| 1805 |
"num_input_tokens_seen": 0,
|
| 1806 |
"num_train_epochs": 1,
|
| 1807 |
"save_steps": 50,
|
|
@@ -1817,8 +385,8 @@
|
|
| 1817 |
"attributes": {}
|
| 1818 |
}
|
| 1819 |
},
|
| 1820 |
-
"total_flos":
|
| 1821 |
-
"train_batch_size":
|
| 1822 |
"trial_name": null,
|
| 1823 |
"trial_params": null
|
| 1824 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.3454231433506045,
|
| 6 |
"eval_steps": 50,
|
| 7 |
+
"global_step": 50,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.0069084628670120895,
|
| 14 |
+
"grad_norm": 0.25255295634269714,
|
| 15 |
"learning_rate": 0.0,
|
| 16 |
+
"loss": 1.9819,
|
| 17 |
"step": 1
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.013816925734024179,
|
| 21 |
+
"grad_norm": 0.26424452662467957,
|
| 22 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 23 |
+
"loss": 2.0092,
|
| 24 |
"step": 2
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.02072538860103627,
|
| 28 |
+
"grad_norm": 0.2828506827354431,
|
| 29 |
+
"learning_rate": 4.000000000000001e-06,
|
| 30 |
+
"loss": 2.0253,
|
| 31 |
"step": 3
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.027633851468048358,
|
| 35 |
+
"grad_norm": 0.2540068030357361,
|
| 36 |
+
"learning_rate": 6e-06,
|
| 37 |
+
"loss": 1.9012,
|
| 38 |
"step": 4
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.03454231433506045,
|
| 42 |
+
"grad_norm": 0.22753603756427765,
|
| 43 |
+
"learning_rate": 8.000000000000001e-06,
|
| 44 |
+
"loss": 1.8407,
|
| 45 |
"step": 5
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.04145077720207254,
|
| 49 |
+
"grad_norm": 0.279053270816803,
|
| 50 |
+
"learning_rate": 1e-05,
|
| 51 |
+
"loss": 2.012,
|
| 52 |
"step": 6
|
| 53 |
},
|
| 54 |
{
|
| 55 |
+
"epoch": 0.04835924006908463,
|
| 56 |
+
"grad_norm": 0.25448864698410034,
|
| 57 |
+
"learning_rate": 9.998741174712534e-06,
|
| 58 |
+
"loss": 1.9668,
|
| 59 |
"step": 7
|
| 60 |
},
|
| 61 |
{
|
| 62 |
+
"epoch": 0.055267702936096716,
|
| 63 |
+
"grad_norm": 0.25880831480026245,
|
| 64 |
+
"learning_rate": 9.994965332706574e-06,
|
| 65 |
+
"loss": 2.1026,
|
| 66 |
"step": 8
|
| 67 |
},
|
| 68 |
{
|
| 69 |
+
"epoch": 0.06217616580310881,
|
| 70 |
+
"grad_norm": 0.2563261389732361,
|
| 71 |
+
"learning_rate": 9.98867437523228e-06,
|
| 72 |
+
"loss": 1.8686,
|
| 73 |
"step": 9
|
| 74 |
},
|
| 75 |
{
|
| 76 |
+
"epoch": 0.0690846286701209,
|
| 77 |
+
"grad_norm": 0.22925390303134918,
|
| 78 |
+
"learning_rate": 9.979871469976197e-06,
|
| 79 |
+
"loss": 1.8824,
|
| 80 |
"step": 10
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.07599309153713299,
|
| 84 |
+
"grad_norm": 0.22950085997581482,
|
| 85 |
+
"learning_rate": 9.968561049466214e-06,
|
| 86 |
+
"loss": 1.7878,
|
| 87 |
"step": 11
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.08290155440414508,
|
| 91 |
+
"grad_norm": 0.2901078462600708,
|
| 92 |
+
"learning_rate": 9.954748808839675e-06,
|
| 93 |
+
"loss": 2.2594,
|
| 94 |
"step": 12
|
| 95 |
},
|
| 96 |
{
|
| 97 |
+
"epoch": 0.08981001727115717,
|
| 98 |
+
"grad_norm": 0.24290603399276733,
|
| 99 |
+
"learning_rate": 9.938441702975689e-06,
|
| 100 |
+
"loss": 1.8975,
|
| 101 |
"step": 13
|
| 102 |
},
|
| 103 |
{
|
| 104 |
+
"epoch": 0.09671848013816926,
|
| 105 |
+
"grad_norm": 0.27432599663734436,
|
| 106 |
+
"learning_rate": 9.91964794299315e-06,
|
| 107 |
+
"loss": 2.3578,
|
| 108 |
"step": 14
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 0.10362694300518134,
|
| 112 |
+
"grad_norm": 0.23735301196575165,
|
| 113 |
+
"learning_rate": 9.898376992116179e-06,
|
| 114 |
+
"loss": 1.8863,
|
| 115 |
"step": 15
|
| 116 |
},
|
| 117 |
{
|
| 118 |
+
"epoch": 0.11053540587219343,
|
| 119 |
+
"grad_norm": 0.22492671012878418,
|
| 120 |
+
"learning_rate": 9.874639560909118e-06,
|
| 121 |
+
"loss": 1.8989,
|
| 122 |
"step": 16
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.11744386873920552,
|
| 126 |
+
"grad_norm": 0.21187926828861237,
|
| 127 |
+
"learning_rate": 9.848447601883436e-06,
|
| 128 |
+
"loss": 1.9164,
|
| 129 |
"step": 17
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.12435233160621761,
|
| 133 |
+
"grad_norm": 0.23231491446495056,
|
| 134 |
+
"learning_rate": 9.819814303479268e-06,
|
| 135 |
+
"loss": 1.8666,
|
| 136 |
"step": 18
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"epoch": 0.13126079447322972,
|
| 140 |
+
"grad_norm": 0.2294367104768753,
|
| 141 |
+
"learning_rate": 9.788754083424654e-06,
|
| 142 |
+
"loss": 1.885,
|
| 143 |
"step": 19
|
| 144 |
},
|
| 145 |
{
|
| 146 |
+
"epoch": 0.1381692573402418,
|
| 147 |
+
"grad_norm": 0.23653824627399445,
|
| 148 |
+
"learning_rate": 9.755282581475769e-06,
|
| 149 |
+
"loss": 1.9521,
|
| 150 |
"step": 20
|
| 151 |
},
|
| 152 |
{
|
| 153 |
+
"epoch": 0.14507772020725387,
|
| 154 |
+
"grad_norm": 0.2559220790863037,
|
| 155 |
+
"learning_rate": 9.719416651541839e-06,
|
| 156 |
+
"loss": 2.0136,
|
| 157 |
"step": 21
|
| 158 |
},
|
| 159 |
{
|
| 160 |
+
"epoch": 0.15198618307426598,
|
| 161 |
+
"grad_norm": 0.24212689697742462,
|
| 162 |
+
"learning_rate": 9.681174353198687e-06,
|
| 163 |
+
"loss": 2.0389,
|
| 164 |
"step": 22
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.15889464594127806,
|
| 168 |
+
"grad_norm": 0.21532879769802094,
|
| 169 |
+
"learning_rate": 9.640574942595195e-06,
|
| 170 |
+
"loss": 1.7763,
|
| 171 |
"step": 23
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.16580310880829016,
|
| 175 |
+
"grad_norm": 0.25662222504615784,
|
| 176 |
+
"learning_rate": 9.597638862757255e-06,
|
| 177 |
+
"loss": 2.1307,
|
| 178 |
"step": 24
|
| 179 |
},
|
| 180 |
{
|
| 181 |
+
"epoch": 0.17271157167530224,
|
| 182 |
+
"grad_norm": 0.2255883365869522,
|
| 183 |
+
"learning_rate": 9.552387733294081e-06,
|
| 184 |
+
"loss": 1.9851,
|
| 185 |
"step": 25
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 0.17962003454231434,
|
| 189 |
+
"grad_norm": 0.2256392389535904,
|
| 190 |
+
"learning_rate": 9.504844339512096e-06,
|
| 191 |
+
"loss": 1.8682,
|
| 192 |
"step": 26
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 0.18652849740932642,
|
| 196 |
+
"grad_norm": 0.2532212436199188,
|
| 197 |
+
"learning_rate": 9.45503262094184e-06,
|
| 198 |
+
"loss": 2.0718,
|
| 199 |
"step": 27
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 0.19343696027633853,
|
| 203 |
+
"grad_norm": 0.2326337695121765,
|
| 204 |
+
"learning_rate": 9.40297765928369e-06,
|
| 205 |
+
"loss": 1.7779,
|
| 206 |
"step": 28
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 0.2003454231433506,
|
| 210 |
+
"grad_norm": 0.2295856773853302,
|
| 211 |
+
"learning_rate": 9.348705665778479e-06,
|
| 212 |
+
"loss": 2.1006,
|
| 213 |
"step": 29
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 0.20725388601036268,
|
| 217 |
+
"grad_norm": 0.2527850270271301,
|
| 218 |
+
"learning_rate": 9.292243968009332e-06,
|
| 219 |
+
"loss": 2.2097,
|
| 220 |
"step": 30
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 0.2141623488773748,
|
| 224 |
+
"grad_norm": 0.22618888318538666,
|
| 225 |
+
"learning_rate": 9.233620996141421e-06,
|
| 226 |
+
"loss": 1.7951,
|
| 227 |
"step": 31
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 0.22107081174438686,
|
| 231 |
+
"grad_norm": 0.2514853775501251,
|
| 232 |
+
"learning_rate": 9.172866268606514e-06,
|
| 233 |
+
"loss": 2.2897,
|
| 234 |
"step": 32
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 0.22797927461139897,
|
| 238 |
+
"grad_norm": 0.2353752851486206,
|
| 239 |
+
"learning_rate": 9.110010377239552e-06,
|
| 240 |
+
"loss": 1.8954,
|
| 241 |
"step": 33
|
| 242 |
},
|
| 243 |
{
|
| 244 |
+
"epoch": 0.23488773747841105,
|
| 245 |
+
"grad_norm": 0.2222089171409607,
|
| 246 |
+
"learning_rate": 9.045084971874738e-06,
|
| 247 |
+
"loss": 1.8893,
|
| 248 |
"step": 34
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"epoch": 0.24179620034542315,
|
| 252 |
+
"grad_norm": 0.2845269739627838,
|
| 253 |
+
"learning_rate": 8.978122744408905e-06,
|
| 254 |
+
"loss": 2.1425,
|
| 255 |
"step": 35
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 0.24870466321243523,
|
| 259 |
+
"grad_norm": 0.2125595360994339,
|
| 260 |
+
"learning_rate": 8.90915741234015e-06,
|
| 261 |
+
"loss": 1.8447,
|
| 262 |
"step": 36
|
| 263 |
},
|
| 264 |
{
|
| 265 |
+
"epoch": 0.2556131260794473,
|
| 266 |
+
"grad_norm": 0.23252736032009125,
|
| 267 |
+
"learning_rate": 8.838223701790057e-06,
|
| 268 |
+
"loss": 1.8057,
|
| 269 |
"step": 37
|
| 270 |
},
|
| 271 |
{
|
| 272 |
+
"epoch": 0.26252158894645944,
|
| 273 |
+
"grad_norm": 0.22627419233322144,
|
| 274 |
+
"learning_rate": 8.765357330018056e-06,
|
| 275 |
+
"loss": 1.9897,
|
| 276 |
"step": 38
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 0.2694300518134715,
|
| 280 |
+
"grad_norm": 0.22737424075603485,
|
| 281 |
+
"learning_rate": 8.690594987436705e-06,
|
| 282 |
+
"loss": 1.8672,
|
| 283 |
"step": 39
|
| 284 |
},
|
| 285 |
{
|
| 286 |
+
"epoch": 0.2763385146804836,
|
| 287 |
+
"grad_norm": 0.25408855080604553,
|
| 288 |
+
"learning_rate": 8.613974319136959e-06,
|
| 289 |
+
"loss": 2.0794,
|
| 290 |
"step": 40
|
| 291 |
},
|
| 292 |
{
|
| 293 |
+
"epoch": 0.28324697754749567,
|
| 294 |
+
"grad_norm": 0.2922523319721222,
|
| 295 |
+
"learning_rate": 8.535533905932739e-06,
|
| 296 |
+
"loss": 2.4457,
|
| 297 |
"step": 41
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 0.29015544041450775,
|
| 301 |
+
"grad_norm": 0.23074638843536377,
|
| 302 |
+
"learning_rate": 8.455313244934324e-06,
|
| 303 |
+
"loss": 1.8467,
|
| 304 |
"step": 42
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"epoch": 0.2970639032815199,
|
| 308 |
+
"grad_norm": 0.21250127255916595,
|
| 309 |
+
"learning_rate": 8.373352729660373e-06,
|
| 310 |
+
"loss": 1.7373,
|
| 311 |
"step": 43
|
| 312 |
},
|
| 313 |
{
|
| 314 |
+
"epoch": 0.30397236614853196,
|
| 315 |
+
"grad_norm": 0.2267821580171585,
|
| 316 |
+
"learning_rate": 8.289693629698564e-06,
|
| 317 |
+
"loss": 1.8409,
|
| 318 |
"step": 44
|
| 319 |
},
|
| 320 |
{
|
| 321 |
+
"epoch": 0.31088082901554404,
|
| 322 |
+
"grad_norm": 0.23144274950027466,
|
| 323 |
+
"learning_rate": 8.204378069925121e-06,
|
| 324 |
+
"loss": 1.8812,
|
| 325 |
"step": 45
|
| 326 |
},
|
| 327 |
{
|
| 328 |
+
"epoch": 0.3177892918825561,
|
| 329 |
+
"grad_norm": 0.245137557387352,
|
| 330 |
+
"learning_rate": 8.117449009293668e-06,
|
| 331 |
+
"loss": 1.9027,
|
| 332 |
"step": 46
|
| 333 |
},
|
| 334 |
{
|
| 335 |
+
"epoch": 0.32469775474956825,
|
| 336 |
+
"grad_norm": 0.27354151010513306,
|
| 337 |
+
"learning_rate": 8.0289502192041e-06,
|
| 338 |
+
"loss": 2.2673,
|
| 339 |
"step": 47
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 0.3316062176165803,
|
| 343 |
+
"grad_norm": 0.23882536590099335,
|
| 344 |
+
"learning_rate": 7.938926261462366e-06,
|
| 345 |
+
"loss": 1.9599,
|
| 346 |
"step": 48
|
| 347 |
},
|
| 348 |
{
|
| 349 |
+
"epoch": 0.3385146804835924,
|
| 350 |
+
"grad_norm": 0.25785377621650696,
|
| 351 |
+
"learning_rate": 7.84742246584226e-06,
|
| 352 |
+
"loss": 2.1052,
|
| 353 |
"step": 49
|
| 354 |
},
|
| 355 |
{
|
| 356 |
+
"epoch": 0.3454231433506045,
|
| 357 |
+
"grad_norm": 0.2514020502567291,
|
| 358 |
+
"learning_rate": 7.754484907260513e-06,
|
| 359 |
+
"loss": 2.0549,
|
| 360 |
"step": 50
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 0.3454231433506045,
|
| 364 |
+
"eval_loss": 0.5051040649414062,
|
| 365 |
+
"eval_runtime": 212.9606,
|
| 366 |
+
"eval_samples_per_second": 2.414,
|
| 367 |
+
"eval_steps_per_second": 0.606,
|
| 368 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 369 |
}
|
| 370 |
],
|
| 371 |
"logging_steps": 1,
|
| 372 |
+
"max_steps": 145,
|
| 373 |
"num_input_tokens_seen": 0,
|
| 374 |
"num_train_epochs": 1,
|
| 375 |
"save_steps": 50,
|
|
|
|
| 385 |
"attributes": {}
|
| 386 |
}
|
| 387 |
},
|
| 388 |
+
"total_flos": 1.0220481364790016e+18,
|
| 389 |
+
"train_batch_size": 8,
|
| 390 |
"trial_name": null,
|
| 391 |
"trial_params": null
|
| 392 |
}
|
last-checkpoint/training_args.bin
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 sha256:702294f2ba6032bb021dc32bbc9dbd5ee8f2ef55f4eb6c78b41cc0994567f4e2
|
| 3 |
+
size 6289
|