Upload folder using huggingface_hub
Browse files- adapter_config.json +8 -5
- adapter_model.bin +2 -2
- optimizer.pt +2 -2
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- special_tokens_map.json +22 -4
- tokenizer.json +0 -0
- tokenizer_config.json +28 -6
- trainer_state.json +153 -753
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
-
"base_model_name_or_path": "
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
|
@@ -14,10 +14,13 @@
|
|
| 14 |
"r": 64,
|
| 15 |
"revision": null,
|
| 16 |
"target_modules": [
|
| 17 |
-
"
|
| 18 |
-
"
|
| 19 |
-
"
|
| 20 |
-
"
|
|
|
|
|
|
|
|
|
|
| 21 |
],
|
| 22 |
"task_type": "CAUSAL_LM"
|
| 23 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "TinyPixel/Llama-2-7B-bf16-sharded",
|
| 4 |
"bias": "none",
|
| 5 |
"fan_in_fan_out": false,
|
| 6 |
"inference_mode": true,
|
|
|
|
| 14 |
"r": 64,
|
| 15 |
"revision": null,
|
| 16 |
"target_modules": [
|
| 17 |
+
"q_proj",
|
| 18 |
+
"up_proj",
|
| 19 |
+
"o_proj",
|
| 20 |
+
"k_proj",
|
| 21 |
+
"down_proj",
|
| 22 |
+
"gate_proj",
|
| 23 |
+
"v_proj"
|
| 24 |
],
|
| 25 |
"task_type": "CAUSAL_LM"
|
| 26 |
}
|
adapter_model.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:d9d0201f6519602fdb0a1e9bfe74c23bde2f75aa4974d9727a3ba7b8831c3c52
|
| 3 |
+
size 639792909
|
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:f828fd98d557668603427f374c6ed59c1b219cf948e90247cfcb37e20166445a
|
| 3 |
+
size 1279539973
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:275d9de1c3e5b37f047ae8355f292a2419092f9ef0035a764a03e5322591e86f
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2c1b52c97a7a8218f3e120cb3e6f62943bb3c9d99f61adb685463e5b497c02d
|
| 3 |
size 627
|
special_tokens_map.json
CHANGED
|
@@ -1,6 +1,24 @@
|
|
| 1 |
{
|
| 2 |
-
"bos_token":
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
}
|
tokenizer.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
CHANGED
|
@@ -1,9 +1,31 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
"model_max_length": 1000000000000000019884624838656,
|
| 7 |
-
"
|
| 8 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"__type": "AddedToken",
|
| 4 |
+
"content": "<s>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": true,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false
|
| 9 |
+
},
|
| 10 |
+
"clean_up_tokenization_spaces": false,
|
| 11 |
+
"eos_token": {
|
| 12 |
+
"__type": "AddedToken",
|
| 13 |
+
"content": "</s>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": true,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false
|
| 18 |
+
},
|
| 19 |
"model_max_length": 1000000000000000019884624838656,
|
| 20 |
+
"pad_token": null,
|
| 21 |
+
"sp_model_kwargs": {},
|
| 22 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"__type": "AddedToken",
|
| 25 |
+
"content": "<unk>",
|
| 26 |
+
"lstrip": false,
|
| 27 |
+
"normalized": true,
|
| 28 |
+
"rstrip": false,
|
| 29 |
+
"single_word": false
|
| 30 |
+
}
|
| 31 |
}
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -10,1507 +10,907 @@
|
|
| 10 |
{
|
| 11 |
"epoch": 0.0,
|
| 12 |
"learning_rate": 2e-05,
|
| 13 |
-
"loss": 1.
|
| 14 |
"step": 2
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.01,
|
| 18 |
"learning_rate": 2e-05,
|
| 19 |
-
"loss": 1.
|
| 20 |
"step": 4
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.01,
|
| 24 |
"learning_rate": 2e-05,
|
| 25 |
-
"loss": 1.
|
| 26 |
"step": 6
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.02,
|
| 30 |
"learning_rate": 2e-05,
|
| 31 |
-
"loss": 1.
|
| 32 |
"step": 8
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.02,
|
| 36 |
"learning_rate": 2e-05,
|
| 37 |
-
"loss": 1.
|
| 38 |
"step": 10
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.02,
|
| 42 |
"learning_rate": 2e-05,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 12
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03,
|
| 48 |
"learning_rate": 2e-05,
|
| 49 |
-
"loss": 1.
|
| 50 |
"step": 14
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.03,
|
| 54 |
"learning_rate": 2e-05,
|
| 55 |
-
"loss": 1.
|
| 56 |
"step": 16
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.03,
|
| 60 |
"learning_rate": 2e-05,
|
| 61 |
-
"loss": 1.
|
| 62 |
"step": 18
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.04,
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
-
"loss":
|
| 68 |
"step": 20
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.04,
|
| 72 |
"learning_rate": 2e-05,
|
| 73 |
-
"loss":
|
| 74 |
"step": 22
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.05,
|
| 78 |
"learning_rate": 2e-05,
|
| 79 |
-
"loss":
|
| 80 |
"step": 24
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.05,
|
| 84 |
"learning_rate": 2e-05,
|
| 85 |
-
"loss":
|
| 86 |
"step": 26
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.05,
|
| 90 |
"learning_rate": 2e-05,
|
| 91 |
-
"loss":
|
| 92 |
"step": 28
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.06,
|
| 96 |
"learning_rate": 2e-05,
|
| 97 |
-
"loss":
|
| 98 |
"step": 30
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.06,
|
| 102 |
"learning_rate": 2e-05,
|
| 103 |
-
"loss":
|
| 104 |
"step": 32
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.07,
|
| 108 |
"learning_rate": 2e-05,
|
| 109 |
-
"loss":
|
| 110 |
"step": 34
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.07,
|
| 114 |
"learning_rate": 2e-05,
|
| 115 |
-
"loss":
|
| 116 |
"step": 36
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.07,
|
| 120 |
"learning_rate": 2e-05,
|
| 121 |
-
"loss":
|
| 122 |
"step": 38
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.08,
|
| 126 |
"learning_rate": 2e-05,
|
| 127 |
-
"loss":
|
| 128 |
"step": 40
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.08,
|
| 132 |
"learning_rate": 2e-05,
|
| 133 |
-
"loss":
|
| 134 |
"step": 42
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.09,
|
| 138 |
"learning_rate": 2e-05,
|
| 139 |
-
"loss":
|
| 140 |
"step": 44
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.09,
|
| 144 |
"learning_rate": 2e-05,
|
| 145 |
-
"loss":
|
| 146 |
"step": 46
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.09,
|
| 150 |
"learning_rate": 2e-05,
|
| 151 |
-
"loss": 1.
|
| 152 |
"step": 48
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.1,
|
| 156 |
"learning_rate": 2e-05,
|
| 157 |
-
"loss":
|
| 158 |
"step": 50
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.1,
|
| 162 |
"learning_rate": 2e-05,
|
| 163 |
-
"loss":
|
| 164 |
"step": 52
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.1,
|
| 168 |
"learning_rate": 2e-05,
|
| 169 |
-
"loss": 1.
|
| 170 |
"step": 54
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.11,
|
| 174 |
"learning_rate": 2e-05,
|
| 175 |
-
"loss": 1.
|
| 176 |
"step": 56
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.11,
|
| 180 |
"learning_rate": 2e-05,
|
| 181 |
-
"loss": 1.
|
| 182 |
"step": 58
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.12,
|
| 186 |
"learning_rate": 2e-05,
|
| 187 |
-
"loss": 1.
|
| 188 |
"step": 60
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.12,
|
| 192 |
"learning_rate": 2e-05,
|
| 193 |
-
"loss": 1.
|
| 194 |
"step": 62
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.12,
|
| 198 |
"learning_rate": 2e-05,
|
| 199 |
-
"loss": 1.
|
| 200 |
"step": 64
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.13,
|
| 204 |
"learning_rate": 2e-05,
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 66
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.13,
|
| 210 |
"learning_rate": 2e-05,
|
| 211 |
-
"loss": 1.
|
| 212 |
"step": 68
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.14,
|
| 216 |
"learning_rate": 2e-05,
|
| 217 |
-
"loss": 1.
|
| 218 |
"step": 70
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.14,
|
| 222 |
"learning_rate": 2e-05,
|
| 223 |
-
"loss": 1.
|
| 224 |
"step": 72
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.14,
|
| 228 |
"learning_rate": 2e-05,
|
| 229 |
-
"loss": 1.
|
| 230 |
"step": 74
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.15,
|
| 234 |
"learning_rate": 2e-05,
|
| 235 |
-
"loss": 1.
|
| 236 |
"step": 76
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.15,
|
| 240 |
"learning_rate": 2e-05,
|
| 241 |
-
"loss":
|
| 242 |
"step": 78
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.15,
|
| 246 |
"learning_rate": 2e-05,
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 80
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.16,
|
| 252 |
"learning_rate": 2e-05,
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 82
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.16,
|
| 258 |
"learning_rate": 2e-05,
|
| 259 |
-
"loss": 1.
|
| 260 |
"step": 84
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.17,
|
| 264 |
"learning_rate": 2e-05,
|
| 265 |
-
"loss": 1.
|
| 266 |
"step": 86
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.17,
|
| 270 |
"learning_rate": 2e-05,
|
| 271 |
-
"loss": 1.
|
| 272 |
"step": 88
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 0.17,
|
| 276 |
"learning_rate": 2e-05,
|
| 277 |
-
"loss":
|
| 278 |
"step": 90
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 0.18,
|
| 282 |
"learning_rate": 2e-05,
|
| 283 |
-
"loss": 1.
|
| 284 |
"step": 92
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.18,
|
| 288 |
"learning_rate": 2e-05,
|
| 289 |
-
"loss": 1.
|
| 290 |
"step": 94
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.19,
|
| 294 |
"learning_rate": 2e-05,
|
| 295 |
-
"loss": 1.
|
| 296 |
"step": 96
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.19,
|
| 300 |
"learning_rate": 2e-05,
|
| 301 |
-
"loss": 1.
|
| 302 |
"step": 98
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.19,
|
| 306 |
"learning_rate": 2e-05,
|
| 307 |
-
"loss":
|
| 308 |
"step": 100
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.2,
|
| 312 |
"learning_rate": 2e-05,
|
| 313 |
-
"loss": 1.
|
| 314 |
"step": 102
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.2,
|
| 318 |
"learning_rate": 2e-05,
|
| 319 |
-
"loss": 1.
|
| 320 |
"step": 104
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.2,
|
| 324 |
"learning_rate": 2e-05,
|
| 325 |
-
"loss": 1.
|
| 326 |
"step": 106
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.21,
|
| 330 |
"learning_rate": 2e-05,
|
| 331 |
-
"loss": 1.
|
| 332 |
"step": 108
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.21,
|
| 336 |
"learning_rate": 2e-05,
|
| 337 |
-
"loss": 1.
|
| 338 |
"step": 110
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.22,
|
| 342 |
"learning_rate": 2e-05,
|
| 343 |
-
"loss": 1.
|
| 344 |
"step": 112
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.22,
|
| 348 |
"learning_rate": 2e-05,
|
| 349 |
-
"loss": 1.
|
| 350 |
"step": 114
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.22,
|
| 354 |
"learning_rate": 2e-05,
|
| 355 |
-
"loss":
|
| 356 |
"step": 116
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.23,
|
| 360 |
"learning_rate": 2e-05,
|
| 361 |
-
"loss": 1.
|
| 362 |
"step": 118
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.23,
|
| 366 |
"learning_rate": 2e-05,
|
| 367 |
-
"loss": 1.
|
| 368 |
"step": 120
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.24,
|
| 372 |
"learning_rate": 2e-05,
|
| 373 |
-
"loss": 1.
|
| 374 |
"step": 122
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.24,
|
| 378 |
"learning_rate": 2e-05,
|
| 379 |
-
"loss": 1.
|
| 380 |
"step": 124
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.24,
|
| 384 |
"learning_rate": 2e-05,
|
| 385 |
-
"loss": 1.
|
| 386 |
"step": 126
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.25,
|
| 390 |
"learning_rate": 2e-05,
|
| 391 |
-
"loss": 1.
|
| 392 |
"step": 128
|
| 393 |
},
|
| 394 |
{
|
| 395 |
"epoch": 0.25,
|
| 396 |
"learning_rate": 2e-05,
|
| 397 |
-
"loss": 1.
|
| 398 |
"step": 130
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.26,
|
| 402 |
"learning_rate": 2e-05,
|
| 403 |
-
"loss": 1.
|
| 404 |
"step": 132
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 0.26,
|
| 408 |
"learning_rate": 2e-05,
|
| 409 |
-
"loss": 1.
|
| 410 |
"step": 134
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.26,
|
| 414 |
"learning_rate": 2e-05,
|
| 415 |
-
"loss":
|
| 416 |
"step": 136
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.27,
|
| 420 |
"learning_rate": 2e-05,
|
| 421 |
-
"loss":
|
| 422 |
"step": 138
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.27,
|
| 426 |
"learning_rate": 2e-05,
|
| 427 |
-
"loss":
|
| 428 |
"step": 140
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.27,
|
| 432 |
"learning_rate": 2e-05,
|
| 433 |
-
"loss": 1.
|
| 434 |
"step": 142
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.28,
|
| 438 |
"learning_rate": 2e-05,
|
| 439 |
-
"loss": 1.
|
| 440 |
"step": 144
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.28,
|
| 444 |
"learning_rate": 2e-05,
|
| 445 |
-
"loss": 1.
|
| 446 |
"step": 146
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.29,
|
| 450 |
"learning_rate": 2e-05,
|
| 451 |
-
"loss": 1.
|
| 452 |
"step": 148
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.29,
|
| 456 |
"learning_rate": 2e-05,
|
| 457 |
-
"loss":
|
| 458 |
"step": 150
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.29,
|
| 462 |
"learning_rate": 2e-05,
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 152
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.3,
|
| 468 |
"learning_rate": 2e-05,
|
| 469 |
-
"loss": 1.
|
| 470 |
"step": 154
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.3,
|
| 474 |
"learning_rate": 2e-05,
|
| 475 |
-
"loss": 1.
|
| 476 |
"step": 156
|
| 477 |
},
|
| 478 |
{
|
| 479 |
"epoch": 0.31,
|
| 480 |
"learning_rate": 2e-05,
|
| 481 |
-
"loss": 1.
|
| 482 |
"step": 158
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.31,
|
| 486 |
"learning_rate": 2e-05,
|
| 487 |
-
"loss": 1.
|
| 488 |
"step": 160
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.31,
|
| 492 |
"learning_rate": 2e-05,
|
| 493 |
-
"loss": 1.
|
| 494 |
"step": 162
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.32,
|
| 498 |
"learning_rate": 2e-05,
|
| 499 |
-
"loss":
|
| 500 |
"step": 164
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.32,
|
| 504 |
"learning_rate": 2e-05,
|
| 505 |
-
"loss":
|
| 506 |
"step": 166
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.32,
|
| 510 |
"learning_rate": 2e-05,
|
| 511 |
-
"loss": 1.
|
| 512 |
"step": 168
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 0.33,
|
| 516 |
"learning_rate": 2e-05,
|
| 517 |
-
"loss": 1.
|
| 518 |
"step": 170
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 0.33,
|
| 522 |
"learning_rate": 2e-05,
|
| 523 |
-
"loss": 1.
|
| 524 |
"step": 172
|
| 525 |
},
|
| 526 |
{
|
| 527 |
"epoch": 0.34,
|
| 528 |
"learning_rate": 2e-05,
|
| 529 |
-
"loss": 1.
|
| 530 |
"step": 174
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.34,
|
| 534 |
"learning_rate": 2e-05,
|
| 535 |
-
"loss": 1.
|
| 536 |
"step": 176
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.34,
|
| 540 |
"learning_rate": 2e-05,
|
| 541 |
-
"loss": 1.
|
| 542 |
"step": 178
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.35,
|
| 546 |
"learning_rate": 2e-05,
|
| 547 |
-
"loss": 1.
|
| 548 |
"step": 180
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.35,
|
| 552 |
"learning_rate": 2e-05,
|
| 553 |
-
"loss": 1.
|
| 554 |
"step": 182
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 0.36,
|
| 558 |
"learning_rate": 2e-05,
|
| 559 |
-
"loss": 1.
|
| 560 |
"step": 184
|
| 561 |
},
|
| 562 |
{
|
| 563 |
"epoch": 0.36,
|
| 564 |
"learning_rate": 2e-05,
|
| 565 |
-
"loss": 1.
|
| 566 |
"step": 186
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 0.36,
|
| 570 |
"learning_rate": 2e-05,
|
| 571 |
-
"loss": 1.
|
| 572 |
"step": 188
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.37,
|
| 576 |
"learning_rate": 2e-05,
|
| 577 |
-
"loss": 1.
|
| 578 |
"step": 190
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.37,
|
| 582 |
"learning_rate": 2e-05,
|
| 583 |
-
"loss": 1.
|
| 584 |
"step": 192
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.38,
|
| 588 |
"learning_rate": 2e-05,
|
| 589 |
-
"loss": 1.
|
| 590 |
"step": 194
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.38,
|
| 594 |
"learning_rate": 2e-05,
|
| 595 |
-
"loss": 1.
|
| 596 |
"step": 196
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 0.38,
|
| 600 |
"learning_rate": 2e-05,
|
| 601 |
-
"loss": 1.
|
| 602 |
"step": 198
|
| 603 |
},
|
| 604 |
{
|
| 605 |
"epoch": 0.39,
|
| 606 |
"learning_rate": 2e-05,
|
| 607 |
-
"loss":
|
| 608 |
"step": 200
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.39,
|
| 612 |
"learning_rate": 2e-05,
|
| 613 |
-
"loss": 1.
|
| 614 |
"step": 202
|
| 615 |
},
|
| 616 |
{
|
| 617 |
"epoch": 0.39,
|
| 618 |
"learning_rate": 2e-05,
|
| 619 |
-
"loss":
|
| 620 |
"step": 204
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 0.4,
|
| 624 |
"learning_rate": 2e-05,
|
| 625 |
-
"loss": 1.
|
| 626 |
"step": 206
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.4,
|
| 630 |
"learning_rate": 2e-05,
|
| 631 |
-
"loss": 1.
|
| 632 |
"step": 208
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.41,
|
| 636 |
"learning_rate": 2e-05,
|
| 637 |
-
"loss": 1.
|
| 638 |
"step": 210
|
| 639 |
},
|
| 640 |
{
|
| 641 |
"epoch": 0.41,
|
| 642 |
"learning_rate": 2e-05,
|
| 643 |
-
"loss": 1.
|
| 644 |
"step": 212
|
| 645 |
},
|
| 646 |
{
|
| 647 |
"epoch": 0.41,
|
| 648 |
"learning_rate": 2e-05,
|
| 649 |
-
"loss":
|
| 650 |
"step": 214
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.42,
|
| 654 |
"learning_rate": 2e-05,
|
| 655 |
-
"loss": 1.
|
| 656 |
"step": 216
|
| 657 |
},
|
| 658 |
{
|
| 659 |
"epoch": 0.42,
|
| 660 |
"learning_rate": 2e-05,
|
| 661 |
-
"loss": 1.
|
| 662 |
"step": 218
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 0.43,
|
| 666 |
"learning_rate": 2e-05,
|
| 667 |
-
"loss": 1.
|
| 668 |
"step": 220
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.43,
|
| 672 |
"learning_rate": 2e-05,
|
| 673 |
-
"loss": 1.
|
| 674 |
"step": 222
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.43,
|
| 678 |
"learning_rate": 2e-05,
|
| 679 |
-
"loss": 1.
|
| 680 |
"step": 224
|
| 681 |
},
|
| 682 |
{
|
| 683 |
"epoch": 0.44,
|
| 684 |
"learning_rate": 2e-05,
|
| 685 |
-
"loss": 1.
|
| 686 |
"step": 226
|
| 687 |
},
|
| 688 |
{
|
| 689 |
"epoch": 0.44,
|
| 690 |
"learning_rate": 2e-05,
|
| 691 |
-
"loss": 1.
|
| 692 |
"step": 228
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.44,
|
| 696 |
"learning_rate": 2e-05,
|
| 697 |
-
"loss": 1.
|
| 698 |
"step": 230
|
| 699 |
},
|
| 700 |
{
|
| 701 |
"epoch": 0.45,
|
| 702 |
"learning_rate": 2e-05,
|
| 703 |
-
"loss": 1.
|
| 704 |
"step": 232
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 0.45,
|
| 708 |
"learning_rate": 2e-05,
|
| 709 |
-
"loss": 1.
|
| 710 |
"step": 234
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.46,
|
| 714 |
"learning_rate": 2e-05,
|
| 715 |
-
"loss": 1.
|
| 716 |
"step": 236
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.46,
|
| 720 |
"learning_rate": 2e-05,
|
| 721 |
-
"loss": 1.
|
| 722 |
"step": 238
|
| 723 |
},
|
| 724 |
{
|
| 725 |
"epoch": 0.46,
|
| 726 |
"learning_rate": 2e-05,
|
| 727 |
-
"loss": 1.
|
| 728 |
"step": 240
|
| 729 |
},
|
| 730 |
{
|
| 731 |
"epoch": 0.47,
|
| 732 |
"learning_rate": 2e-05,
|
| 733 |
-
"loss": 1.
|
| 734 |
"step": 242
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.47,
|
| 738 |
"learning_rate": 2e-05,
|
| 739 |
-
"loss": 1.
|
| 740 |
"step": 244
|
| 741 |
},
|
| 742 |
{
|
| 743 |
"epoch": 0.48,
|
| 744 |
"learning_rate": 2e-05,
|
| 745 |
-
"loss": 1.
|
| 746 |
"step": 246
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 0.48,
|
| 750 |
"learning_rate": 2e-05,
|
| 751 |
-
"loss": 1.
|
| 752 |
"step": 248
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.48,
|
| 756 |
"learning_rate": 2e-05,
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 250
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.49,
|
| 762 |
"learning_rate": 2e-05,
|
| 763 |
-
"loss":
|
| 764 |
"step": 252
|
| 765 |
},
|
| 766 |
{
|
| 767 |
"epoch": 0.49,
|
| 768 |
"learning_rate": 2e-05,
|
| 769 |
-
"loss": 1.
|
| 770 |
"step": 254
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.5,
|
| 774 |
"learning_rate": 2e-05,
|
| 775 |
-
"loss": 1.
|
| 776 |
"step": 256
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 0.5,
|
| 780 |
"learning_rate": 2e-05,
|
| 781 |
-
"loss":
|
| 782 |
"step": 258
|
| 783 |
},
|
| 784 |
{
|
| 785 |
"epoch": 0.5,
|
| 786 |
"learning_rate": 2e-05,
|
| 787 |
-
"loss": 1.
|
| 788 |
"step": 260
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 0.51,
|
| 792 |
"learning_rate": 2e-05,
|
| 793 |
-
"loss": 1.
|
| 794 |
"step": 262
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.51,
|
| 798 |
"learning_rate": 2e-05,
|
| 799 |
-
"loss": 1.
|
| 800 |
"step": 264
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.51,
|
| 804 |
"learning_rate": 2e-05,
|
| 805 |
-
"loss": 1.
|
| 806 |
"step": 266
|
| 807 |
},
|
| 808 |
{
|
| 809 |
"epoch": 0.52,
|
| 810 |
"learning_rate": 2e-05,
|
| 811 |
-
"loss": 1.
|
| 812 |
"step": 268
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.52,
|
| 816 |
"learning_rate": 2e-05,
|
| 817 |
-
"loss": 1.
|
| 818 |
"step": 270
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 0.53,
|
| 822 |
"learning_rate": 2e-05,
|
| 823 |
-
"loss": 1.
|
| 824 |
"step": 272
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.53,
|
| 828 |
"learning_rate": 2e-05,
|
| 829 |
-
"loss": 1.
|
| 830 |
"step": 274
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 0.53,
|
| 834 |
"learning_rate": 2e-05,
|
| 835 |
-
"loss": 1.
|
| 836 |
"step": 276
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.54,
|
| 840 |
"learning_rate": 2e-05,
|
| 841 |
-
"loss": 1.
|
| 842 |
"step": 278
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.54,
|
| 846 |
"learning_rate": 2e-05,
|
| 847 |
-
"loss": 1.
|
| 848 |
"step": 280
|
| 849 |
},
|
| 850 |
{
|
| 851 |
"epoch": 0.55,
|
| 852 |
"learning_rate": 2e-05,
|
| 853 |
-
"loss": 1.
|
| 854 |
"step": 282
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.55,
|
| 858 |
"learning_rate": 2e-05,
|
| 859 |
-
"loss": 1.
|
| 860 |
"step": 284
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 0.55,
|
| 864 |
"learning_rate": 2e-05,
|
| 865 |
-
"loss":
|
| 866 |
"step": 286
|
| 867 |
},
|
| 868 |
{
|
| 869 |
"epoch": 0.56,
|
| 870 |
"learning_rate": 2e-05,
|
| 871 |
-
"loss": 1.
|
| 872 |
"step": 288
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 0.56,
|
| 876 |
"learning_rate": 2e-05,
|
| 877 |
-
"loss": 1.
|
| 878 |
"step": 290
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.56,
|
| 882 |
"learning_rate": 2e-05,
|
| 883 |
-
"loss": 1.
|
| 884 |
"step": 292
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.57,
|
| 888 |
"learning_rate": 2e-05,
|
| 889 |
-
"loss": 1.
|
| 890 |
"step": 294
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 0.57,
|
| 894 |
"learning_rate": 2e-05,
|
| 895 |
-
"loss": 1.
|
| 896 |
"step": 296
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.58,
|
| 900 |
"learning_rate": 2e-05,
|
| 901 |
-
"loss": 1.
|
| 902 |
"step": 298
|
| 903 |
},
|
| 904 |
{
|
| 905 |
"epoch": 0.58,
|
| 906 |
"learning_rate": 2e-05,
|
| 907 |
-
"loss":
|
| 908 |
"step": 300
|
| 909 |
-
},
|
| 910 |
-
{
|
| 911 |
-
"epoch": 0.58,
|
| 912 |
-
"learning_rate": 2e-05,
|
| 913 |
-
"loss": 1.6262,
|
| 914 |
-
"step": 302
|
| 915 |
-
},
|
| 916 |
-
{
|
| 917 |
-
"epoch": 0.59,
|
| 918 |
-
"learning_rate": 2e-05,
|
| 919 |
-
"loss": 1.7146,
|
| 920 |
-
"step": 304
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.59,
|
| 924 |
-
"learning_rate": 2e-05,
|
| 925 |
-
"loss": 1.5797,
|
| 926 |
-
"step": 306
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 0.6,
|
| 930 |
-
"learning_rate": 2e-05,
|
| 931 |
-
"loss": 1.6793,
|
| 932 |
-
"step": 308
|
| 933 |
-
},
|
| 934 |
-
{
|
| 935 |
-
"epoch": 0.6,
|
| 936 |
-
"learning_rate": 2e-05,
|
| 937 |
-
"loss": 1.5755,
|
| 938 |
-
"step": 310
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 0.6,
|
| 942 |
-
"learning_rate": 2e-05,
|
| 943 |
-
"loss": 1.5813,
|
| 944 |
-
"step": 312
|
| 945 |
-
},
|
| 946 |
-
{
|
| 947 |
-
"epoch": 0.61,
|
| 948 |
-
"learning_rate": 2e-05,
|
| 949 |
-
"loss": 1.6024,
|
| 950 |
-
"step": 314
|
| 951 |
-
},
|
| 952 |
-
{
|
| 953 |
-
"epoch": 0.61,
|
| 954 |
-
"learning_rate": 2e-05,
|
| 955 |
-
"loss": 1.7725,
|
| 956 |
-
"step": 316
|
| 957 |
-
},
|
| 958 |
-
{
|
| 959 |
-
"epoch": 0.61,
|
| 960 |
-
"learning_rate": 2e-05,
|
| 961 |
-
"loss": 1.6265,
|
| 962 |
-
"step": 318
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 0.62,
|
| 966 |
-
"learning_rate": 2e-05,
|
| 967 |
-
"loss": 1.5773,
|
| 968 |
-
"step": 320
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 0.62,
|
| 972 |
-
"learning_rate": 2e-05,
|
| 973 |
-
"loss": 1.6883,
|
| 974 |
-
"step": 322
|
| 975 |
-
},
|
| 976 |
-
{
|
| 977 |
-
"epoch": 0.63,
|
| 978 |
-
"learning_rate": 2e-05,
|
| 979 |
-
"loss": 1.7982,
|
| 980 |
-
"step": 324
|
| 981 |
-
},
|
| 982 |
-
{
|
| 983 |
-
"epoch": 0.63,
|
| 984 |
-
"learning_rate": 2e-05,
|
| 985 |
-
"loss": 1.7559,
|
| 986 |
-
"step": 326
|
| 987 |
-
},
|
| 988 |
-
{
|
| 989 |
-
"epoch": 0.63,
|
| 990 |
-
"learning_rate": 2e-05,
|
| 991 |
-
"loss": 1.8074,
|
| 992 |
-
"step": 328
|
| 993 |
-
},
|
| 994 |
-
{
|
| 995 |
-
"epoch": 0.64,
|
| 996 |
-
"learning_rate": 2e-05,
|
| 997 |
-
"loss": 2.0078,
|
| 998 |
-
"step": 330
|
| 999 |
-
},
|
| 1000 |
-
{
|
| 1001 |
-
"epoch": 0.64,
|
| 1002 |
-
"learning_rate": 2e-05,
|
| 1003 |
-
"loss": 1.9758,
|
| 1004 |
-
"step": 332
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 0.65,
|
| 1008 |
-
"learning_rate": 2e-05,
|
| 1009 |
-
"loss": 1.8103,
|
| 1010 |
-
"step": 334
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 0.65,
|
| 1014 |
-
"learning_rate": 2e-05,
|
| 1015 |
-
"loss": 2.0364,
|
| 1016 |
-
"step": 336
|
| 1017 |
-
},
|
| 1018 |
-
{
|
| 1019 |
-
"epoch": 0.65,
|
| 1020 |
-
"learning_rate": 2e-05,
|
| 1021 |
-
"loss": 1.8717,
|
| 1022 |
-
"step": 338
|
| 1023 |
-
},
|
| 1024 |
-
{
|
| 1025 |
-
"epoch": 0.66,
|
| 1026 |
-
"learning_rate": 2e-05,
|
| 1027 |
-
"loss": 1.8783,
|
| 1028 |
-
"step": 340
|
| 1029 |
-
},
|
| 1030 |
-
{
|
| 1031 |
-
"epoch": 0.66,
|
| 1032 |
-
"learning_rate": 2e-05,
|
| 1033 |
-
"loss": 1.954,
|
| 1034 |
-
"step": 342
|
| 1035 |
-
},
|
| 1036 |
-
{
|
| 1037 |
-
"epoch": 0.67,
|
| 1038 |
-
"learning_rate": 2e-05,
|
| 1039 |
-
"loss": 1.8943,
|
| 1040 |
-
"step": 344
|
| 1041 |
-
},
|
| 1042 |
-
{
|
| 1043 |
-
"epoch": 0.67,
|
| 1044 |
-
"learning_rate": 2e-05,
|
| 1045 |
-
"loss": 1.7534,
|
| 1046 |
-
"step": 346
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 0.67,
|
| 1050 |
-
"learning_rate": 2e-05,
|
| 1051 |
-
"loss": 1.606,
|
| 1052 |
-
"step": 348
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 0.68,
|
| 1056 |
-
"learning_rate": 2e-05,
|
| 1057 |
-
"loss": 1.0254,
|
| 1058 |
-
"step": 350
|
| 1059 |
-
},
|
| 1060 |
-
{
|
| 1061 |
-
"epoch": 0.68,
|
| 1062 |
-
"learning_rate": 2e-05,
|
| 1063 |
-
"loss": 1.5422,
|
| 1064 |
-
"step": 352
|
| 1065 |
-
},
|
| 1066 |
-
{
|
| 1067 |
-
"epoch": 0.68,
|
| 1068 |
-
"learning_rate": 2e-05,
|
| 1069 |
-
"loss": 1.6437,
|
| 1070 |
-
"step": 354
|
| 1071 |
-
},
|
| 1072 |
-
{
|
| 1073 |
-
"epoch": 0.69,
|
| 1074 |
-
"learning_rate": 2e-05,
|
| 1075 |
-
"loss": 1.5021,
|
| 1076 |
-
"step": 356
|
| 1077 |
-
},
|
| 1078 |
-
{
|
| 1079 |
-
"epoch": 0.69,
|
| 1080 |
-
"learning_rate": 2e-05,
|
| 1081 |
-
"loss": 1.7269,
|
| 1082 |
-
"step": 358
|
| 1083 |
-
},
|
| 1084 |
-
{
|
| 1085 |
-
"epoch": 0.7,
|
| 1086 |
-
"learning_rate": 2e-05,
|
| 1087 |
-
"loss": 1.6228,
|
| 1088 |
-
"step": 360
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.7,
|
| 1092 |
-
"learning_rate": 2e-05,
|
| 1093 |
-
"loss": 1.631,
|
| 1094 |
-
"step": 362
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 0.7,
|
| 1098 |
-
"learning_rate": 2e-05,
|
| 1099 |
-
"loss": 1.7287,
|
| 1100 |
-
"step": 364
|
| 1101 |
-
},
|
| 1102 |
-
{
|
| 1103 |
-
"epoch": 0.71,
|
| 1104 |
-
"learning_rate": 2e-05,
|
| 1105 |
-
"loss": 1.6234,
|
| 1106 |
-
"step": 366
|
| 1107 |
-
},
|
| 1108 |
-
{
|
| 1109 |
-
"epoch": 0.71,
|
| 1110 |
-
"learning_rate": 2e-05,
|
| 1111 |
-
"loss": 1.6457,
|
| 1112 |
-
"step": 368
|
| 1113 |
-
},
|
| 1114 |
-
{
|
| 1115 |
-
"epoch": 0.72,
|
| 1116 |
-
"learning_rate": 2e-05,
|
| 1117 |
-
"loss": 1.6345,
|
| 1118 |
-
"step": 370
|
| 1119 |
-
},
|
| 1120 |
-
{
|
| 1121 |
-
"epoch": 0.72,
|
| 1122 |
-
"learning_rate": 2e-05,
|
| 1123 |
-
"loss": 1.6274,
|
| 1124 |
-
"step": 372
|
| 1125 |
-
},
|
| 1126 |
-
{
|
| 1127 |
-
"epoch": 0.72,
|
| 1128 |
-
"learning_rate": 2e-05,
|
| 1129 |
-
"loss": 1.65,
|
| 1130 |
-
"step": 374
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.73,
|
| 1134 |
-
"learning_rate": 2e-05,
|
| 1135 |
-
"loss": 1.7637,
|
| 1136 |
-
"step": 376
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.73,
|
| 1140 |
-
"learning_rate": 2e-05,
|
| 1141 |
-
"loss": 1.8289,
|
| 1142 |
-
"step": 378
|
| 1143 |
-
},
|
| 1144 |
-
{
|
| 1145 |
-
"epoch": 0.73,
|
| 1146 |
-
"learning_rate": 2e-05,
|
| 1147 |
-
"loss": 1.9179,
|
| 1148 |
-
"step": 380
|
| 1149 |
-
},
|
| 1150 |
-
{
|
| 1151 |
-
"epoch": 0.74,
|
| 1152 |
-
"learning_rate": 2e-05,
|
| 1153 |
-
"loss": 2.0096,
|
| 1154 |
-
"step": 382
|
| 1155 |
-
},
|
| 1156 |
-
{
|
| 1157 |
-
"epoch": 0.74,
|
| 1158 |
-
"learning_rate": 2e-05,
|
| 1159 |
-
"loss": 1.9448,
|
| 1160 |
-
"step": 384
|
| 1161 |
-
},
|
| 1162 |
-
{
|
| 1163 |
-
"epoch": 0.75,
|
| 1164 |
-
"learning_rate": 2e-05,
|
| 1165 |
-
"loss": 1.9791,
|
| 1166 |
-
"step": 386
|
| 1167 |
-
},
|
| 1168 |
-
{
|
| 1169 |
-
"epoch": 0.75,
|
| 1170 |
-
"learning_rate": 2e-05,
|
| 1171 |
-
"loss": 1.9646,
|
| 1172 |
-
"step": 388
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.75,
|
| 1176 |
-
"learning_rate": 2e-05,
|
| 1177 |
-
"loss": 1.9525,
|
| 1178 |
-
"step": 390
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 0.76,
|
| 1182 |
-
"learning_rate": 2e-05,
|
| 1183 |
-
"loss": 1.9584,
|
| 1184 |
-
"step": 392
|
| 1185 |
-
},
|
| 1186 |
-
{
|
| 1187 |
-
"epoch": 0.76,
|
| 1188 |
-
"learning_rate": 2e-05,
|
| 1189 |
-
"loss": 1.8401,
|
| 1190 |
-
"step": 394
|
| 1191 |
-
},
|
| 1192 |
-
{
|
| 1193 |
-
"epoch": 0.77,
|
| 1194 |
-
"learning_rate": 2e-05,
|
| 1195 |
-
"loss": 1.7071,
|
| 1196 |
-
"step": 396
|
| 1197 |
-
},
|
| 1198 |
-
{
|
| 1199 |
-
"epoch": 0.77,
|
| 1200 |
-
"learning_rate": 2e-05,
|
| 1201 |
-
"loss": 1.5105,
|
| 1202 |
-
"step": 398
|
| 1203 |
-
},
|
| 1204 |
-
{
|
| 1205 |
-
"epoch": 0.77,
|
| 1206 |
-
"learning_rate": 2e-05,
|
| 1207 |
-
"loss": 0.9903,
|
| 1208 |
-
"step": 400
|
| 1209 |
-
},
|
| 1210 |
-
{
|
| 1211 |
-
"epoch": 0.78,
|
| 1212 |
-
"learning_rate": 2e-05,
|
| 1213 |
-
"loss": 1.6691,
|
| 1214 |
-
"step": 402
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.78,
|
| 1218 |
-
"learning_rate": 2e-05,
|
| 1219 |
-
"loss": 1.5754,
|
| 1220 |
-
"step": 404
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 0.79,
|
| 1224 |
-
"learning_rate": 2e-05,
|
| 1225 |
-
"loss": 1.7317,
|
| 1226 |
-
"step": 406
|
| 1227 |
-
},
|
| 1228 |
-
{
|
| 1229 |
-
"epoch": 0.79,
|
| 1230 |
-
"learning_rate": 2e-05,
|
| 1231 |
-
"loss": 1.6006,
|
| 1232 |
-
"step": 408
|
| 1233 |
-
},
|
| 1234 |
-
{
|
| 1235 |
-
"epoch": 0.79,
|
| 1236 |
-
"learning_rate": 2e-05,
|
| 1237 |
-
"loss": 1.6372,
|
| 1238 |
-
"step": 410
|
| 1239 |
-
},
|
| 1240 |
-
{
|
| 1241 |
-
"epoch": 0.8,
|
| 1242 |
-
"learning_rate": 2e-05,
|
| 1243 |
-
"loss": 1.4847,
|
| 1244 |
-
"step": 412
|
| 1245 |
-
},
|
| 1246 |
-
{
|
| 1247 |
-
"epoch": 0.8,
|
| 1248 |
-
"learning_rate": 2e-05,
|
| 1249 |
-
"loss": 1.5862,
|
| 1250 |
-
"step": 414
|
| 1251 |
-
},
|
| 1252 |
-
{
|
| 1253 |
-
"epoch": 0.8,
|
| 1254 |
-
"learning_rate": 2e-05,
|
| 1255 |
-
"loss": 1.6069,
|
| 1256 |
-
"step": 416
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 0.81,
|
| 1260 |
-
"learning_rate": 2e-05,
|
| 1261 |
-
"loss": 1.6557,
|
| 1262 |
-
"step": 418
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 0.81,
|
| 1266 |
-
"learning_rate": 2e-05,
|
| 1267 |
-
"loss": 1.7094,
|
| 1268 |
-
"step": 420
|
| 1269 |
-
},
|
| 1270 |
-
{
|
| 1271 |
-
"epoch": 0.82,
|
| 1272 |
-
"learning_rate": 2e-05,
|
| 1273 |
-
"loss": 1.8511,
|
| 1274 |
-
"step": 422
|
| 1275 |
-
},
|
| 1276 |
-
{
|
| 1277 |
-
"epoch": 0.82,
|
| 1278 |
-
"learning_rate": 2e-05,
|
| 1279 |
-
"loss": 1.8803,
|
| 1280 |
-
"step": 424
|
| 1281 |
-
},
|
| 1282 |
-
{
|
| 1283 |
-
"epoch": 0.82,
|
| 1284 |
-
"learning_rate": 2e-05,
|
| 1285 |
-
"loss": 1.857,
|
| 1286 |
-
"step": 426
|
| 1287 |
-
},
|
| 1288 |
-
{
|
| 1289 |
-
"epoch": 0.83,
|
| 1290 |
-
"learning_rate": 2e-05,
|
| 1291 |
-
"loss": 1.7882,
|
| 1292 |
-
"step": 428
|
| 1293 |
-
},
|
| 1294 |
-
{
|
| 1295 |
-
"epoch": 0.83,
|
| 1296 |
-
"learning_rate": 2e-05,
|
| 1297 |
-
"loss": 1.7939,
|
| 1298 |
-
"step": 430
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 0.84,
|
| 1302 |
-
"learning_rate": 2e-05,
|
| 1303 |
-
"loss": 1.8605,
|
| 1304 |
-
"step": 432
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.84,
|
| 1308 |
-
"learning_rate": 2e-05,
|
| 1309 |
-
"loss": 1.9402,
|
| 1310 |
-
"step": 434
|
| 1311 |
-
},
|
| 1312 |
-
{
|
| 1313 |
-
"epoch": 0.84,
|
| 1314 |
-
"learning_rate": 2e-05,
|
| 1315 |
-
"loss": 1.7993,
|
| 1316 |
-
"step": 436
|
| 1317 |
-
},
|
| 1318 |
-
{
|
| 1319 |
-
"epoch": 0.85,
|
| 1320 |
-
"learning_rate": 2e-05,
|
| 1321 |
-
"loss": 2.0032,
|
| 1322 |
-
"step": 438
|
| 1323 |
-
},
|
| 1324 |
-
{
|
| 1325 |
-
"epoch": 0.85,
|
| 1326 |
-
"learning_rate": 2e-05,
|
| 1327 |
-
"loss": 1.9246,
|
| 1328 |
-
"step": 440
|
| 1329 |
-
},
|
| 1330 |
-
{
|
| 1331 |
-
"epoch": 0.85,
|
| 1332 |
-
"learning_rate": 2e-05,
|
| 1333 |
-
"loss": 1.8394,
|
| 1334 |
-
"step": 442
|
| 1335 |
-
},
|
| 1336 |
-
{
|
| 1337 |
-
"epoch": 0.86,
|
| 1338 |
-
"learning_rate": 2e-05,
|
| 1339 |
-
"loss": 1.7374,
|
| 1340 |
-
"step": 444
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 0.86,
|
| 1344 |
-
"learning_rate": 2e-05,
|
| 1345 |
-
"loss": 1.6768,
|
| 1346 |
-
"step": 446
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 0.87,
|
| 1350 |
-
"learning_rate": 2e-05,
|
| 1351 |
-
"loss": 1.4662,
|
| 1352 |
-
"step": 448
|
| 1353 |
-
},
|
| 1354 |
-
{
|
| 1355 |
-
"epoch": 0.87,
|
| 1356 |
-
"learning_rate": 2e-05,
|
| 1357 |
-
"loss": 0.9718,
|
| 1358 |
-
"step": 450
|
| 1359 |
-
},
|
| 1360 |
-
{
|
| 1361 |
-
"epoch": 0.87,
|
| 1362 |
-
"learning_rate": 2e-05,
|
| 1363 |
-
"loss": 1.4755,
|
| 1364 |
-
"step": 452
|
| 1365 |
-
},
|
| 1366 |
-
{
|
| 1367 |
-
"epoch": 0.88,
|
| 1368 |
-
"learning_rate": 2e-05,
|
| 1369 |
-
"loss": 1.6846,
|
| 1370 |
-
"step": 454
|
| 1371 |
-
},
|
| 1372 |
-
{
|
| 1373 |
-
"epoch": 0.88,
|
| 1374 |
-
"learning_rate": 2e-05,
|
| 1375 |
-
"loss": 1.6875,
|
| 1376 |
-
"step": 456
|
| 1377 |
-
},
|
| 1378 |
-
{
|
| 1379 |
-
"epoch": 0.89,
|
| 1380 |
-
"learning_rate": 2e-05,
|
| 1381 |
-
"loss": 1.4627,
|
| 1382 |
-
"step": 458
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 0.89,
|
| 1386 |
-
"learning_rate": 2e-05,
|
| 1387 |
-
"loss": 1.6094,
|
| 1388 |
-
"step": 460
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.89,
|
| 1392 |
-
"learning_rate": 2e-05,
|
| 1393 |
-
"loss": 1.4555,
|
| 1394 |
-
"step": 462
|
| 1395 |
-
},
|
| 1396 |
-
{
|
| 1397 |
-
"epoch": 0.9,
|
| 1398 |
-
"learning_rate": 2e-05,
|
| 1399 |
-
"loss": 1.6874,
|
| 1400 |
-
"step": 464
|
| 1401 |
-
},
|
| 1402 |
-
{
|
| 1403 |
-
"epoch": 0.9,
|
| 1404 |
-
"learning_rate": 2e-05,
|
| 1405 |
-
"loss": 1.5308,
|
| 1406 |
-
"step": 466
|
| 1407 |
-
},
|
| 1408 |
-
{
|
| 1409 |
-
"epoch": 0.91,
|
| 1410 |
-
"learning_rate": 2e-05,
|
| 1411 |
-
"loss": 1.7519,
|
| 1412 |
-
"step": 468
|
| 1413 |
-
},
|
| 1414 |
-
{
|
| 1415 |
-
"epoch": 0.91,
|
| 1416 |
-
"learning_rate": 2e-05,
|
| 1417 |
-
"loss": 1.6585,
|
| 1418 |
-
"step": 470
|
| 1419 |
-
},
|
| 1420 |
-
{
|
| 1421 |
-
"epoch": 0.91,
|
| 1422 |
-
"learning_rate": 2e-05,
|
| 1423 |
-
"loss": 1.726,
|
| 1424 |
-
"step": 472
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 0.92,
|
| 1428 |
-
"learning_rate": 2e-05,
|
| 1429 |
-
"loss": 1.6701,
|
| 1430 |
-
"step": 474
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 0.92,
|
| 1434 |
-
"learning_rate": 2e-05,
|
| 1435 |
-
"loss": 1.8123,
|
| 1436 |
-
"step": 476
|
| 1437 |
-
},
|
| 1438 |
-
{
|
| 1439 |
-
"epoch": 0.92,
|
| 1440 |
-
"learning_rate": 2e-05,
|
| 1441 |
-
"loss": 1.7762,
|
| 1442 |
-
"step": 478
|
| 1443 |
-
},
|
| 1444 |
-
{
|
| 1445 |
-
"epoch": 0.93,
|
| 1446 |
-
"learning_rate": 2e-05,
|
| 1447 |
-
"loss": 1.796,
|
| 1448 |
-
"step": 480
|
| 1449 |
-
},
|
| 1450 |
-
{
|
| 1451 |
-
"epoch": 0.93,
|
| 1452 |
-
"learning_rate": 2e-05,
|
| 1453 |
-
"loss": 1.8989,
|
| 1454 |
-
"step": 482
|
| 1455 |
-
},
|
| 1456 |
-
{
|
| 1457 |
-
"epoch": 0.94,
|
| 1458 |
-
"learning_rate": 2e-05,
|
| 1459 |
-
"loss": 1.7805,
|
| 1460 |
-
"step": 484
|
| 1461 |
-
},
|
| 1462 |
-
{
|
| 1463 |
-
"epoch": 0.94,
|
| 1464 |
-
"learning_rate": 2e-05,
|
| 1465 |
-
"loss": 1.911,
|
| 1466 |
-
"step": 486
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 0.94,
|
| 1470 |
-
"learning_rate": 2e-05,
|
| 1471 |
-
"loss": 1.9393,
|
| 1472 |
-
"step": 488
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.95,
|
| 1476 |
-
"learning_rate": 2e-05,
|
| 1477 |
-
"loss": 1.9401,
|
| 1478 |
-
"step": 490
|
| 1479 |
-
},
|
| 1480 |
-
{
|
| 1481 |
-
"epoch": 0.95,
|
| 1482 |
-
"learning_rate": 2e-05,
|
| 1483 |
-
"loss": 1.8289,
|
| 1484 |
-
"step": 492
|
| 1485 |
-
},
|
| 1486 |
-
{
|
| 1487 |
-
"epoch": 0.96,
|
| 1488 |
-
"learning_rate": 2e-05,
|
| 1489 |
-
"loss": 1.8784,
|
| 1490 |
-
"step": 494
|
| 1491 |
-
},
|
| 1492 |
-
{
|
| 1493 |
-
"epoch": 0.96,
|
| 1494 |
-
"learning_rate": 2e-05,
|
| 1495 |
-
"loss": 1.7339,
|
| 1496 |
-
"step": 496
|
| 1497 |
-
},
|
| 1498 |
-
{
|
| 1499 |
-
"epoch": 0.96,
|
| 1500 |
-
"learning_rate": 2e-05,
|
| 1501 |
-
"loss": 1.4148,
|
| 1502 |
-
"step": 498
|
| 1503 |
-
},
|
| 1504 |
-
{
|
| 1505 |
-
"epoch": 0.97,
|
| 1506 |
-
"learning_rate": 2e-05,
|
| 1507 |
-
"loss": 0.9561,
|
| 1508 |
-
"step": 500
|
| 1509 |
}
|
| 1510 |
],
|
| 1511 |
"max_steps": 1000,
|
| 1512 |
"num_train_epochs": 2,
|
| 1513 |
-
"total_flos":
|
| 1514 |
"trial_name": null,
|
| 1515 |
"trial_params": null
|
| 1516 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.580130529369108,
|
| 5 |
+
"global_step": 300,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 10 |
{
|
| 11 |
"epoch": 0.0,
|
| 12 |
"learning_rate": 2e-05,
|
| 13 |
+
"loss": 1.3732,
|
| 14 |
"step": 2
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.01,
|
| 18 |
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 1.1638,
|
| 20 |
"step": 4
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"epoch": 0.01,
|
| 24 |
"learning_rate": 2e-05,
|
| 25 |
+
"loss": 1.1545,
|
| 26 |
"step": 6
|
| 27 |
},
|
| 28 |
{
|
| 29 |
"epoch": 0.02,
|
| 30 |
"learning_rate": 2e-05,
|
| 31 |
+
"loss": 1.2039,
|
| 32 |
"step": 8
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.02,
|
| 36 |
"learning_rate": 2e-05,
|
| 37 |
+
"loss": 1.2867,
|
| 38 |
"step": 10
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.02,
|
| 42 |
"learning_rate": 2e-05,
|
| 43 |
+
"loss": 1.2554,
|
| 44 |
"step": 12
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03,
|
| 48 |
"learning_rate": 2e-05,
|
| 49 |
+
"loss": 1.2596,
|
| 50 |
"step": 14
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"epoch": 0.03,
|
| 54 |
"learning_rate": 2e-05,
|
| 55 |
+
"loss": 1.2628,
|
| 56 |
"step": 16
|
| 57 |
},
|
| 58 |
{
|
| 59 |
"epoch": 0.03,
|
| 60 |
"learning_rate": 2e-05,
|
| 61 |
+
"loss": 1.3257,
|
| 62 |
"step": 18
|
| 63 |
},
|
| 64 |
{
|
| 65 |
"epoch": 0.04,
|
| 66 |
"learning_rate": 2e-05,
|
| 67 |
+
"loss": 1.3706,
|
| 68 |
"step": 20
|
| 69 |
},
|
| 70 |
{
|
| 71 |
"epoch": 0.04,
|
| 72 |
"learning_rate": 2e-05,
|
| 73 |
+
"loss": 1.6044,
|
| 74 |
"step": 22
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.05,
|
| 78 |
"learning_rate": 2e-05,
|
| 79 |
+
"loss": 1.2826,
|
| 80 |
"step": 24
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 0.05,
|
| 84 |
"learning_rate": 2e-05,
|
| 85 |
+
"loss": 1.3722,
|
| 86 |
"step": 26
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.05,
|
| 90 |
"learning_rate": 2e-05,
|
| 91 |
+
"loss": 1.4654,
|
| 92 |
"step": 28
|
| 93 |
},
|
| 94 |
{
|
| 95 |
"epoch": 0.06,
|
| 96 |
"learning_rate": 2e-05,
|
| 97 |
+
"loss": 1.6323,
|
| 98 |
"step": 30
|
| 99 |
},
|
| 100 |
{
|
| 101 |
"epoch": 0.06,
|
| 102 |
"learning_rate": 2e-05,
|
| 103 |
+
"loss": 1.683,
|
| 104 |
"step": 32
|
| 105 |
},
|
| 106 |
{
|
| 107 |
"epoch": 0.07,
|
| 108 |
"learning_rate": 2e-05,
|
| 109 |
+
"loss": 1.5204,
|
| 110 |
"step": 34
|
| 111 |
},
|
| 112 |
{
|
| 113 |
"epoch": 0.07,
|
| 114 |
"learning_rate": 2e-05,
|
| 115 |
+
"loss": 1.679,
|
| 116 |
"step": 36
|
| 117 |
},
|
| 118 |
{
|
| 119 |
"epoch": 0.07,
|
| 120 |
"learning_rate": 2e-05,
|
| 121 |
+
"loss": 1.4497,
|
| 122 |
"step": 38
|
| 123 |
},
|
| 124 |
{
|
| 125 |
"epoch": 0.08,
|
| 126 |
"learning_rate": 2e-05,
|
| 127 |
+
"loss": 1.5898,
|
| 128 |
"step": 40
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.08,
|
| 132 |
"learning_rate": 2e-05,
|
| 133 |
+
"loss": 1.5762,
|
| 134 |
"step": 42
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 0.09,
|
| 138 |
"learning_rate": 2e-05,
|
| 139 |
+
"loss": 1.6473,
|
| 140 |
"step": 44
|
| 141 |
},
|
| 142 |
{
|
| 143 |
"epoch": 0.09,
|
| 144 |
"learning_rate": 2e-05,
|
| 145 |
+
"loss": 1.4771,
|
| 146 |
"step": 46
|
| 147 |
},
|
| 148 |
{
|
| 149 |
"epoch": 0.09,
|
| 150 |
"learning_rate": 2e-05,
|
| 151 |
+
"loss": 1.3425,
|
| 152 |
"step": 48
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"epoch": 0.1,
|
| 156 |
"learning_rate": 2e-05,
|
| 157 |
+
"loss": 0.8668,
|
| 158 |
"step": 50
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.1,
|
| 162 |
"learning_rate": 2e-05,
|
| 163 |
+
"loss": 0.9341,
|
| 164 |
"step": 52
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.1,
|
| 168 |
"learning_rate": 2e-05,
|
| 169 |
+
"loss": 1.0335,
|
| 170 |
"step": 54
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.11,
|
| 174 |
"learning_rate": 2e-05,
|
| 175 |
+
"loss": 1.1254,
|
| 176 |
"step": 56
|
| 177 |
},
|
| 178 |
{
|
| 179 |
"epoch": 0.11,
|
| 180 |
"learning_rate": 2e-05,
|
| 181 |
+
"loss": 1.0025,
|
| 182 |
"step": 58
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"epoch": 0.12,
|
| 186 |
"learning_rate": 2e-05,
|
| 187 |
+
"loss": 1.081,
|
| 188 |
"step": 60
|
| 189 |
},
|
| 190 |
{
|
| 191 |
"epoch": 0.12,
|
| 192 |
"learning_rate": 2e-05,
|
| 193 |
+
"loss": 1.029,
|
| 194 |
"step": 62
|
| 195 |
},
|
| 196 |
{
|
| 197 |
"epoch": 0.12,
|
| 198 |
"learning_rate": 2e-05,
|
| 199 |
+
"loss": 1.0352,
|
| 200 |
"step": 64
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"epoch": 0.13,
|
| 204 |
"learning_rate": 2e-05,
|
| 205 |
+
"loss": 1.156,
|
| 206 |
"step": 66
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.13,
|
| 210 |
"learning_rate": 2e-05,
|
| 211 |
+
"loss": 1.0304,
|
| 212 |
"step": 68
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.14,
|
| 216 |
"learning_rate": 2e-05,
|
| 217 |
+
"loss": 1.0913,
|
| 218 |
"step": 70
|
| 219 |
},
|
| 220 |
{
|
| 221 |
"epoch": 0.14,
|
| 222 |
"learning_rate": 2e-05,
|
| 223 |
+
"loss": 1.239,
|
| 224 |
"step": 72
|
| 225 |
},
|
| 226 |
{
|
| 227 |
"epoch": 0.14,
|
| 228 |
"learning_rate": 2e-05,
|
| 229 |
+
"loss": 1.0982,
|
| 230 |
"step": 74
|
| 231 |
},
|
| 232 |
{
|
| 233 |
"epoch": 0.15,
|
| 234 |
"learning_rate": 2e-05,
|
| 235 |
+
"loss": 1.2554,
|
| 236 |
"step": 76
|
| 237 |
},
|
| 238 |
{
|
| 239 |
"epoch": 0.15,
|
| 240 |
"learning_rate": 2e-05,
|
| 241 |
+
"loss": 1.3172,
|
| 242 |
"step": 78
|
| 243 |
},
|
| 244 |
{
|
| 245 |
"epoch": 0.15,
|
| 246 |
"learning_rate": 2e-05,
|
| 247 |
+
"loss": 1.3916,
|
| 248 |
"step": 80
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.16,
|
| 252 |
"learning_rate": 2e-05,
|
| 253 |
+
"loss": 1.1953,
|
| 254 |
"step": 82
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.16,
|
| 258 |
"learning_rate": 2e-05,
|
| 259 |
+
"loss": 1.3974,
|
| 260 |
"step": 84
|
| 261 |
},
|
| 262 |
{
|
| 263 |
"epoch": 0.17,
|
| 264 |
"learning_rate": 2e-05,
|
| 265 |
+
"loss": 1.3862,
|
| 266 |
"step": 86
|
| 267 |
},
|
| 268 |
{
|
| 269 |
"epoch": 0.17,
|
| 270 |
"learning_rate": 2e-05,
|
| 271 |
+
"loss": 1.2707,
|
| 272 |
"step": 88
|
| 273 |
},
|
| 274 |
{
|
| 275 |
"epoch": 0.17,
|
| 276 |
"learning_rate": 2e-05,
|
| 277 |
+
"loss": 1.4924,
|
| 278 |
"step": 90
|
| 279 |
},
|
| 280 |
{
|
| 281 |
"epoch": 0.18,
|
| 282 |
"learning_rate": 2e-05,
|
| 283 |
+
"loss": 1.5437,
|
| 284 |
"step": 92
|
| 285 |
},
|
| 286 |
{
|
| 287 |
"epoch": 0.18,
|
| 288 |
"learning_rate": 2e-05,
|
| 289 |
+
"loss": 1.35,
|
| 290 |
"step": 94
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.19,
|
| 294 |
"learning_rate": 2e-05,
|
| 295 |
+
"loss": 1.3366,
|
| 296 |
"step": 96
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.19,
|
| 300 |
"learning_rate": 2e-05,
|
| 301 |
+
"loss": 1.2631,
|
| 302 |
"step": 98
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 0.19,
|
| 306 |
"learning_rate": 2e-05,
|
| 307 |
+
"loss": 0.814,
|
| 308 |
"step": 100
|
| 309 |
},
|
| 310 |
{
|
| 311 |
"epoch": 0.2,
|
| 312 |
"learning_rate": 2e-05,
|
| 313 |
+
"loss": 1.1012,
|
| 314 |
"step": 102
|
| 315 |
},
|
| 316 |
{
|
| 317 |
"epoch": 0.2,
|
| 318 |
"learning_rate": 2e-05,
|
| 319 |
+
"loss": 1.0838,
|
| 320 |
"step": 104
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"epoch": 0.2,
|
| 324 |
"learning_rate": 2e-05,
|
| 325 |
+
"loss": 1.026,
|
| 326 |
"step": 106
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.21,
|
| 330 |
"learning_rate": 2e-05,
|
| 331 |
+
"loss": 1.0317,
|
| 332 |
"step": 108
|
| 333 |
},
|
| 334 |
{
|
| 335 |
"epoch": 0.21,
|
| 336 |
"learning_rate": 2e-05,
|
| 337 |
+
"loss": 1.0449,
|
| 338 |
"step": 110
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.22,
|
| 342 |
"learning_rate": 2e-05,
|
| 343 |
+
"loss": 1.1489,
|
| 344 |
"step": 112
|
| 345 |
},
|
| 346 |
{
|
| 347 |
"epoch": 0.22,
|
| 348 |
"learning_rate": 2e-05,
|
| 349 |
+
"loss": 1.1748,
|
| 350 |
"step": 114
|
| 351 |
},
|
| 352 |
{
|
| 353 |
"epoch": 0.22,
|
| 354 |
"learning_rate": 2e-05,
|
| 355 |
+
"loss": 0.9617,
|
| 356 |
"step": 116
|
| 357 |
},
|
| 358 |
{
|
| 359 |
"epoch": 0.23,
|
| 360 |
"learning_rate": 2e-05,
|
| 361 |
+
"loss": 1.2164,
|
| 362 |
"step": 118
|
| 363 |
},
|
| 364 |
{
|
| 365 |
"epoch": 0.23,
|
| 366 |
"learning_rate": 2e-05,
|
| 367 |
+
"loss": 1.1345,
|
| 368 |
"step": 120
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.24,
|
| 372 |
"learning_rate": 2e-05,
|
| 373 |
+
"loss": 1.1359,
|
| 374 |
"step": 122
|
| 375 |
},
|
| 376 |
{
|
| 377 |
"epoch": 0.24,
|
| 378 |
"learning_rate": 2e-05,
|
| 379 |
+
"loss": 1.2059,
|
| 380 |
"step": 124
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.24,
|
| 384 |
"learning_rate": 2e-05,
|
| 385 |
+
"loss": 1.1603,
|
| 386 |
"step": 126
|
| 387 |
},
|
| 388 |
{
|
| 389 |
"epoch": 0.25,
|
| 390 |
"learning_rate": 2e-05,
|
| 391 |
+
"loss": 1.298,
|
| 392 |
"step": 128
|
| 393 |
},
|
| 394 |
{
|
| 395 |
"epoch": 0.25,
|
| 396 |
"learning_rate": 2e-05,
|
| 397 |
+
"loss": 1.1893,
|
| 398 |
"step": 130
|
| 399 |
},
|
| 400 |
{
|
| 401 |
"epoch": 0.26,
|
| 402 |
"learning_rate": 2e-05,
|
| 403 |
+
"loss": 1.3392,
|
| 404 |
"step": 132
|
| 405 |
},
|
| 406 |
{
|
| 407 |
"epoch": 0.26,
|
| 408 |
"learning_rate": 2e-05,
|
| 409 |
+
"loss": 1.4176,
|
| 410 |
"step": 134
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.26,
|
| 414 |
"learning_rate": 2e-05,
|
| 415 |
+
"loss": 1.4475,
|
| 416 |
"step": 136
|
| 417 |
},
|
| 418 |
{
|
| 419 |
"epoch": 0.27,
|
| 420 |
"learning_rate": 2e-05,
|
| 421 |
+
"loss": 1.4052,
|
| 422 |
"step": 138
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.27,
|
| 426 |
"learning_rate": 2e-05,
|
| 427 |
+
"loss": 1.464,
|
| 428 |
"step": 140
|
| 429 |
},
|
| 430 |
{
|
| 431 |
"epoch": 0.27,
|
| 432 |
"learning_rate": 2e-05,
|
| 433 |
+
"loss": 1.4211,
|
| 434 |
"step": 142
|
| 435 |
},
|
| 436 |
{
|
| 437 |
"epoch": 0.28,
|
| 438 |
"learning_rate": 2e-05,
|
| 439 |
+
"loss": 1.2525,
|
| 440 |
"step": 144
|
| 441 |
},
|
| 442 |
{
|
| 443 |
"epoch": 0.28,
|
| 444 |
"learning_rate": 2e-05,
|
| 445 |
+
"loss": 1.3329,
|
| 446 |
"step": 146
|
| 447 |
},
|
| 448 |
{
|
| 449 |
"epoch": 0.29,
|
| 450 |
"learning_rate": 2e-05,
|
| 451 |
+
"loss": 1.135,
|
| 452 |
"step": 148
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.29,
|
| 456 |
"learning_rate": 2e-05,
|
| 457 |
+
"loss": 0.8317,
|
| 458 |
"step": 150
|
| 459 |
},
|
| 460 |
{
|
| 461 |
"epoch": 0.29,
|
| 462 |
"learning_rate": 2e-05,
|
| 463 |
+
"loss": 1.03,
|
| 464 |
"step": 152
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.3,
|
| 468 |
"learning_rate": 2e-05,
|
| 469 |
+
"loss": 1.0428,
|
| 470 |
"step": 154
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 0.3,
|
| 474 |
"learning_rate": 2e-05,
|
| 475 |
+
"loss": 1.1478,
|
| 476 |
"step": 156
|
| 477 |
},
|
| 478 |
{
|
| 479 |
"epoch": 0.31,
|
| 480 |
"learning_rate": 2e-05,
|
| 481 |
+
"loss": 1.0514,
|
| 482 |
"step": 158
|
| 483 |
},
|
| 484 |
{
|
| 485 |
"epoch": 0.31,
|
| 486 |
"learning_rate": 2e-05,
|
| 487 |
+
"loss": 1.0284,
|
| 488 |
"step": 160
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.31,
|
| 492 |
"learning_rate": 2e-05,
|
| 493 |
+
"loss": 1.0399,
|
| 494 |
"step": 162
|
| 495 |
},
|
| 496 |
{
|
| 497 |
"epoch": 0.32,
|
| 498 |
"learning_rate": 2e-05,
|
| 499 |
+
"loss": 0.9585,
|
| 500 |
"step": 164
|
| 501 |
},
|
| 502 |
{
|
| 503 |
"epoch": 0.32,
|
| 504 |
"learning_rate": 2e-05,
|
| 505 |
+
"loss": 0.984,
|
| 506 |
"step": 166
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.32,
|
| 510 |
"learning_rate": 2e-05,
|
| 511 |
+
"loss": 1.0095,
|
| 512 |
"step": 168
|
| 513 |
},
|
| 514 |
{
|
| 515 |
"epoch": 0.33,
|
| 516 |
"learning_rate": 2e-05,
|
| 517 |
+
"loss": 1.0371,
|
| 518 |
"step": 170
|
| 519 |
},
|
| 520 |
{
|
| 521 |
"epoch": 0.33,
|
| 522 |
"learning_rate": 2e-05,
|
| 523 |
+
"loss": 1.1832,
|
| 524 |
"step": 172
|
| 525 |
},
|
| 526 |
{
|
| 527 |
"epoch": 0.34,
|
| 528 |
"learning_rate": 2e-05,
|
| 529 |
+
"loss": 1.1151,
|
| 530 |
"step": 174
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.34,
|
| 534 |
"learning_rate": 2e-05,
|
| 535 |
+
"loss": 1.2616,
|
| 536 |
"step": 176
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 0.34,
|
| 540 |
"learning_rate": 2e-05,
|
| 541 |
+
"loss": 1.2923,
|
| 542 |
"step": 178
|
| 543 |
},
|
| 544 |
{
|
| 545 |
"epoch": 0.35,
|
| 546 |
"learning_rate": 2e-05,
|
| 547 |
+
"loss": 1.1316,
|
| 548 |
"step": 180
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.35,
|
| 552 |
"learning_rate": 2e-05,
|
| 553 |
+
"loss": 1.2669,
|
| 554 |
"step": 182
|
| 555 |
},
|
| 556 |
{
|
| 557 |
"epoch": 0.36,
|
| 558 |
"learning_rate": 2e-05,
|
| 559 |
+
"loss": 1.3682,
|
| 560 |
"step": 184
|
| 561 |
},
|
| 562 |
{
|
| 563 |
"epoch": 0.36,
|
| 564 |
"learning_rate": 2e-05,
|
| 565 |
+
"loss": 1.297,
|
| 566 |
"step": 186
|
| 567 |
},
|
| 568 |
{
|
| 569 |
"epoch": 0.36,
|
| 570 |
"learning_rate": 2e-05,
|
| 571 |
+
"loss": 1.2828,
|
| 572 |
"step": 188
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.37,
|
| 576 |
"learning_rate": 2e-05,
|
| 577 |
+
"loss": 1.2625,
|
| 578 |
"step": 190
|
| 579 |
},
|
| 580 |
{
|
| 581 |
"epoch": 0.37,
|
| 582 |
"learning_rate": 2e-05,
|
| 583 |
+
"loss": 1.4606,
|
| 584 |
"step": 192
|
| 585 |
},
|
| 586 |
{
|
| 587 |
"epoch": 0.38,
|
| 588 |
"learning_rate": 2e-05,
|
| 589 |
+
"loss": 1.2795,
|
| 590 |
"step": 194
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.38,
|
| 594 |
"learning_rate": 2e-05,
|
| 595 |
+
"loss": 1.4136,
|
| 596 |
"step": 196
|
| 597 |
},
|
| 598 |
{
|
| 599 |
"epoch": 0.38,
|
| 600 |
"learning_rate": 2e-05,
|
| 601 |
+
"loss": 1.0876,
|
| 602 |
"step": 198
|
| 603 |
},
|
| 604 |
{
|
| 605 |
"epoch": 0.39,
|
| 606 |
"learning_rate": 2e-05,
|
| 607 |
+
"loss": 0.7762,
|
| 608 |
"step": 200
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.39,
|
| 612 |
"learning_rate": 2e-05,
|
| 613 |
+
"loss": 1.0827,
|
| 614 |
"step": 202
|
| 615 |
},
|
| 616 |
{
|
| 617 |
"epoch": 0.39,
|
| 618 |
"learning_rate": 2e-05,
|
| 619 |
+
"loss": 0.9613,
|
| 620 |
"step": 204
|
| 621 |
},
|
| 622 |
{
|
| 623 |
"epoch": 0.4,
|
| 624 |
"learning_rate": 2e-05,
|
| 625 |
+
"loss": 1.0078,
|
| 626 |
"step": 206
|
| 627 |
},
|
| 628 |
{
|
| 629 |
"epoch": 0.4,
|
| 630 |
"learning_rate": 2e-05,
|
| 631 |
+
"loss": 1.0128,
|
| 632 |
"step": 208
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.41,
|
| 636 |
"learning_rate": 2e-05,
|
| 637 |
+
"loss": 1.063,
|
| 638 |
"step": 210
|
| 639 |
},
|
| 640 |
{
|
| 641 |
"epoch": 0.41,
|
| 642 |
"learning_rate": 2e-05,
|
| 643 |
+
"loss": 1.0805,
|
| 644 |
"step": 212
|
| 645 |
},
|
| 646 |
{
|
| 647 |
"epoch": 0.41,
|
| 648 |
"learning_rate": 2e-05,
|
| 649 |
+
"loss": 0.9611,
|
| 650 |
"step": 214
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.42,
|
| 654 |
"learning_rate": 2e-05,
|
| 655 |
+
"loss": 1.0106,
|
| 656 |
"step": 216
|
| 657 |
},
|
| 658 |
{
|
| 659 |
"epoch": 0.42,
|
| 660 |
"learning_rate": 2e-05,
|
| 661 |
+
"loss": 1.0785,
|
| 662 |
"step": 218
|
| 663 |
},
|
| 664 |
{
|
| 665 |
"epoch": 0.43,
|
| 666 |
"learning_rate": 2e-05,
|
| 667 |
+
"loss": 1.1339,
|
| 668 |
"step": 220
|
| 669 |
},
|
| 670 |
{
|
| 671 |
"epoch": 0.43,
|
| 672 |
"learning_rate": 2e-05,
|
| 673 |
+
"loss": 1.1942,
|
| 674 |
"step": 222
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.43,
|
| 678 |
"learning_rate": 2e-05,
|
| 679 |
+
"loss": 1.1281,
|
| 680 |
"step": 224
|
| 681 |
},
|
| 682 |
{
|
| 683 |
"epoch": 0.44,
|
| 684 |
"learning_rate": 2e-05,
|
| 685 |
+
"loss": 1.1421,
|
| 686 |
"step": 226
|
| 687 |
},
|
| 688 |
{
|
| 689 |
"epoch": 0.44,
|
| 690 |
"learning_rate": 2e-05,
|
| 691 |
+
"loss": 1.1084,
|
| 692 |
"step": 228
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.44,
|
| 696 |
"learning_rate": 2e-05,
|
| 697 |
+
"loss": 1.3067,
|
| 698 |
"step": 230
|
| 699 |
},
|
| 700 |
{
|
| 701 |
"epoch": 0.45,
|
| 702 |
"learning_rate": 2e-05,
|
| 703 |
+
"loss": 1.3573,
|
| 704 |
"step": 232
|
| 705 |
},
|
| 706 |
{
|
| 707 |
"epoch": 0.45,
|
| 708 |
"learning_rate": 2e-05,
|
| 709 |
+
"loss": 1.2424,
|
| 710 |
"step": 234
|
| 711 |
},
|
| 712 |
{
|
| 713 |
"epoch": 0.46,
|
| 714 |
"learning_rate": 2e-05,
|
| 715 |
+
"loss": 1.2454,
|
| 716 |
"step": 236
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.46,
|
| 720 |
"learning_rate": 2e-05,
|
| 721 |
+
"loss": 1.3976,
|
| 722 |
"step": 238
|
| 723 |
},
|
| 724 |
{
|
| 725 |
"epoch": 0.46,
|
| 726 |
"learning_rate": 2e-05,
|
| 727 |
+
"loss": 1.2691,
|
| 728 |
"step": 240
|
| 729 |
},
|
| 730 |
{
|
| 731 |
"epoch": 0.47,
|
| 732 |
"learning_rate": 2e-05,
|
| 733 |
+
"loss": 1.4862,
|
| 734 |
"step": 242
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.47,
|
| 738 |
"learning_rate": 2e-05,
|
| 739 |
+
"loss": 1.3623,
|
| 740 |
"step": 244
|
| 741 |
},
|
| 742 |
{
|
| 743 |
"epoch": 0.48,
|
| 744 |
"learning_rate": 2e-05,
|
| 745 |
+
"loss": 1.2709,
|
| 746 |
"step": 246
|
| 747 |
},
|
| 748 |
{
|
| 749 |
"epoch": 0.48,
|
| 750 |
"learning_rate": 2e-05,
|
| 751 |
+
"loss": 1.0707,
|
| 752 |
"step": 248
|
| 753 |
},
|
| 754 |
{
|
| 755 |
"epoch": 0.48,
|
| 756 |
"learning_rate": 2e-05,
|
| 757 |
+
"loss": 0.6945,
|
| 758 |
"step": 250
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.49,
|
| 762 |
"learning_rate": 2e-05,
|
| 763 |
+
"loss": 0.9516,
|
| 764 |
"step": 252
|
| 765 |
},
|
| 766 |
{
|
| 767 |
"epoch": 0.49,
|
| 768 |
"learning_rate": 2e-05,
|
| 769 |
+
"loss": 1.0244,
|
| 770 |
"step": 254
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.5,
|
| 774 |
"learning_rate": 2e-05,
|
| 775 |
+
"loss": 1.1423,
|
| 776 |
"step": 256
|
| 777 |
},
|
| 778 |
{
|
| 779 |
"epoch": 0.5,
|
| 780 |
"learning_rate": 2e-05,
|
| 781 |
+
"loss": 0.9696,
|
| 782 |
"step": 258
|
| 783 |
},
|
| 784 |
{
|
| 785 |
"epoch": 0.5,
|
| 786 |
"learning_rate": 2e-05,
|
| 787 |
+
"loss": 1.0191,
|
| 788 |
"step": 260
|
| 789 |
},
|
| 790 |
{
|
| 791 |
"epoch": 0.51,
|
| 792 |
"learning_rate": 2e-05,
|
| 793 |
+
"loss": 1.0291,
|
| 794 |
"step": 262
|
| 795 |
},
|
| 796 |
{
|
| 797 |
"epoch": 0.51,
|
| 798 |
"learning_rate": 2e-05,
|
| 799 |
+
"loss": 1.0665,
|
| 800 |
"step": 264
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.51,
|
| 804 |
"learning_rate": 2e-05,
|
| 805 |
+
"loss": 1.0518,
|
| 806 |
"step": 266
|
| 807 |
},
|
| 808 |
{
|
| 809 |
"epoch": 0.52,
|
| 810 |
"learning_rate": 2e-05,
|
| 811 |
+
"loss": 1.0239,
|
| 812 |
"step": 268
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.52,
|
| 816 |
"learning_rate": 2e-05,
|
| 817 |
+
"loss": 1.12,
|
| 818 |
"step": 270
|
| 819 |
},
|
| 820 |
{
|
| 821 |
"epoch": 0.53,
|
| 822 |
"learning_rate": 2e-05,
|
| 823 |
+
"loss": 1.126,
|
| 824 |
"step": 272
|
| 825 |
},
|
| 826 |
{
|
| 827 |
"epoch": 0.53,
|
| 828 |
"learning_rate": 2e-05,
|
| 829 |
+
"loss": 1.0559,
|
| 830 |
"step": 274
|
| 831 |
},
|
| 832 |
{
|
| 833 |
"epoch": 0.53,
|
| 834 |
"learning_rate": 2e-05,
|
| 835 |
+
"loss": 1.257,
|
| 836 |
"step": 276
|
| 837 |
},
|
| 838 |
{
|
| 839 |
"epoch": 0.54,
|
| 840 |
"learning_rate": 2e-05,
|
| 841 |
+
"loss": 1.3178,
|
| 842 |
"step": 278
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.54,
|
| 846 |
"learning_rate": 2e-05,
|
| 847 |
+
"loss": 1.32,
|
| 848 |
"step": 280
|
| 849 |
},
|
| 850 |
{
|
| 851 |
"epoch": 0.55,
|
| 852 |
"learning_rate": 2e-05,
|
| 853 |
+
"loss": 1.153,
|
| 854 |
"step": 282
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.55,
|
| 858 |
"learning_rate": 2e-05,
|
| 859 |
+
"loss": 1.4046,
|
| 860 |
"step": 284
|
| 861 |
},
|
| 862 |
{
|
| 863 |
"epoch": 0.55,
|
| 864 |
"learning_rate": 2e-05,
|
| 865 |
+
"loss": 1.4531,
|
| 866 |
"step": 286
|
| 867 |
},
|
| 868 |
{
|
| 869 |
"epoch": 0.56,
|
| 870 |
"learning_rate": 2e-05,
|
| 871 |
+
"loss": 1.3236,
|
| 872 |
"step": 288
|
| 873 |
},
|
| 874 |
{
|
| 875 |
"epoch": 0.56,
|
| 876 |
"learning_rate": 2e-05,
|
| 877 |
+
"loss": 1.3427,
|
| 878 |
"step": 290
|
| 879 |
},
|
| 880 |
{
|
| 881 |
"epoch": 0.56,
|
| 882 |
"learning_rate": 2e-05,
|
| 883 |
+
"loss": 1.4536,
|
| 884 |
"step": 292
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.57,
|
| 888 |
"learning_rate": 2e-05,
|
| 889 |
+
"loss": 1.2831,
|
| 890 |
"step": 294
|
| 891 |
},
|
| 892 |
{
|
| 893 |
"epoch": 0.57,
|
| 894 |
"learning_rate": 2e-05,
|
| 895 |
+
"loss": 1.2384,
|
| 896 |
"step": 296
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.58,
|
| 900 |
"learning_rate": 2e-05,
|
| 901 |
+
"loss": 1.0424,
|
| 902 |
"step": 298
|
| 903 |
},
|
| 904 |
{
|
| 905 |
"epoch": 0.58,
|
| 906 |
"learning_rate": 2e-05,
|
| 907 |
+
"loss": 0.7491,
|
| 908 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 909 |
}
|
| 910 |
],
|
| 911 |
"max_steps": 1000,
|
| 912 |
"num_train_epochs": 2,
|
| 913 |
+
"total_flos": 3.0851676306210816e+16,
|
| 914 |
"trial_name": null,
|
| 915 |
"trial_params": null
|
| 916 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3963
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b982e90943b5865c4d2d304aa9bfce32dad4e18f6ed906af992890773d4badb
|
| 3 |
size 3963
|