Upload folder using huggingface_hub
Browse files- adapter_config.json +2 -2
- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- special_tokens_map.json +3 -19
- tokenizer.json +0 -18
- tokenizer_config.json +3 -23
- trainer_state.json +303 -303
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -17,9 +17,9 @@
|
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
"dense",
|
| 20 |
-
"dense_4h_to_h",
|
| 21 |
"query_key_value",
|
| 22 |
-
"dense_h_to_4h"
|
|
|
|
| 23 |
],
|
| 24 |
"task_type": "CAUSAL_LM"
|
| 25 |
}
|
|
|
|
| 17 |
"revision": null,
|
| 18 |
"target_modules": [
|
| 19 |
"dense",
|
|
|
|
| 20 |
"query_key_value",
|
| 21 |
+
"dense_h_to_4h",
|
| 22 |
+
"dense_4h_to_h"
|
| 23 |
],
|
| 24 |
"task_type": "CAUSAL_LM"
|
| 25 |
}
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 134235712
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:204d748906039d9c2f5b8c12a9b05b589fce0f4580b129371d45b6e3f8aab6cd
|
| 3 |
size 134235712
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 268514874
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abd76e3fa8c993edfef4560aa20cd1689a5ff44535f25d7d445ac3b3a42682f4
|
| 3 |
size 268514874
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c78496c814d1125f068ceed05684a004a2cef8c56f84ae9ea235bf1fa565449e
|
| 3 |
size 14244
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ced1bbbbd6439b37716ff72c4d2a98af0eeaed72dba8340a1ae4eb8f8b02ec40
|
| 3 |
size 1064
|
special_tokens_map.json
CHANGED
|
@@ -1,22 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false
|
| 9 |
-
},
|
| 10 |
-
{
|
| 11 |
-
"content": "<|im_start|>",
|
| 12 |
-
"lstrip": false,
|
| 13 |
-
"normalized": false,
|
| 14 |
-
"rstrip": false,
|
| 15 |
-
"single_word": false
|
| 16 |
-
}
|
| 17 |
-
],
|
| 18 |
-
"bos_token": "<|im_start|>",
|
| 19 |
-
"eos_token": "<|im_end|>",
|
| 20 |
-
"pad_token": "<|endoftext|>",
|
| 21 |
"unk_token": "<|endoftext|>"
|
| 22 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"bos_token": "<|endoftext|>",
|
| 3 |
+
"eos_token": "<|endoftext|>",
|
| 4 |
+
"pad_token": "[PAD]",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
"unk_token": "<|endoftext|>"
|
| 6 |
}
|
tokenizer.json
CHANGED
|
@@ -232,24 +232,6 @@
|
|
| 232 |
"rstrip": false,
|
| 233 |
"normalized": true,
|
| 234 |
"special": false
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"id": 50277,
|
| 238 |
-
"content": "<|im_end|>",
|
| 239 |
-
"single_word": false,
|
| 240 |
-
"lstrip": false,
|
| 241 |
-
"rstrip": false,
|
| 242 |
-
"normalized": false,
|
| 243 |
-
"special": true
|
| 244 |
-
},
|
| 245 |
-
{
|
| 246 |
-
"id": 50278,
|
| 247 |
-
"content": "<|im_start|>",
|
| 248 |
-
"single_word": false,
|
| 249 |
-
"lstrip": false,
|
| 250 |
-
"rstrip": false,
|
| 251 |
-
"normalized": false,
|
| 252 |
-
"special": true
|
| 253 |
}
|
| 254 |
],
|
| 255 |
"normalizer": {
|
|
|
|
| 232 |
"rstrip": false,
|
| 233 |
"normalized": true,
|
| 234 |
"special": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 235 |
}
|
| 236 |
],
|
| 237 |
"normalizer": {
|
tokenizer_config.json
CHANGED
|
@@ -200,33 +200,13 @@
|
|
| 200 |
"rstrip": false,
|
| 201 |
"single_word": false,
|
| 202 |
"special": false
|
| 203 |
-
},
|
| 204 |
-
"50277": {
|
| 205 |
-
"content": "<|im_end|>",
|
| 206 |
-
"lstrip": false,
|
| 207 |
-
"normalized": false,
|
| 208 |
-
"rstrip": false,
|
| 209 |
-
"single_word": false,
|
| 210 |
-
"special": true
|
| 211 |
-
},
|
| 212 |
-
"50278": {
|
| 213 |
-
"content": "<|im_start|>",
|
| 214 |
-
"lstrip": false,
|
| 215 |
-
"normalized": false,
|
| 216 |
-
"rstrip": false,
|
| 217 |
-
"single_word": false,
|
| 218 |
-
"special": true
|
| 219 |
}
|
| 220 |
},
|
| 221 |
-
"
|
| 222 |
-
"<|im_end|>",
|
| 223 |
-
"<|im_start|>"
|
| 224 |
-
],
|
| 225 |
-
"bos_token": "<|im_start|>",
|
| 226 |
"clean_up_tokenization_spaces": true,
|
| 227 |
-
"eos_token": "<|
|
| 228 |
"model_max_length": 1000000000000000019884624838656,
|
| 229 |
-
"pad_token": "
|
| 230 |
"tokenizer_class": "GPTNeoXTokenizer",
|
| 231 |
"unk_token": "<|endoftext|>"
|
| 232 |
}
|
|
|
|
| 200 |
"rstrip": false,
|
| 201 |
"single_word": false,
|
| 202 |
"special": false
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 203 |
}
|
| 204 |
},
|
| 205 |
+
"bos_token": "<|endoftext|>",
|
|
|
|
|
|
|
|
|
|
|
|
|
| 206 |
"clean_up_tokenization_spaces": true,
|
| 207 |
+
"eos_token": "<|endoftext|>",
|
| 208 |
"model_max_length": 1000000000000000019884624838656,
|
| 209 |
+
"pad_token": "[PAD]",
|
| 210 |
"tokenizer_class": "GPTNeoXTokenizer",
|
| 211 |
"unk_token": "<|endoftext|>"
|
| 212 |
}
|
trainer_state.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 201,
|
| 7 |
"is_hyper_param_search": false,
|
|
@@ -9,611 +9,611 @@
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"learning_rate":
|
| 14 |
-
"loss": 2.
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate":
|
| 20 |
-
"loss": 2.
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 1.
|
| 26 |
-
"loss": 2.
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 1.
|
| 32 |
-
"loss": 2.
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 1.
|
| 38 |
-
"loss":
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate": 1.
|
| 44 |
-
"loss": 2.
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate": 1.
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate": 1.
|
| 56 |
-
"loss":
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate": 1.
|
| 62 |
-
"loss": 2.
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch": 0.
|
| 67 |
-
"learning_rate": 1.
|
| 68 |
-
"loss":
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"learning_rate": 1.
|
| 74 |
-
"loss":
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch": 0.
|
| 79 |
-
"learning_rate": 1.
|
| 80 |
-
"loss":
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch": 0.
|
| 85 |
-
"learning_rate": 1.
|
| 86 |
-
"loss":
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch": 0.
|
| 91 |
-
"learning_rate": 1.
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch": 0.
|
| 97 |
-
"learning_rate": 1.
|
| 98 |
-
"loss": 2.
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch": 0.
|
| 103 |
-
"learning_rate": 1.
|
| 104 |
-
"loss": 2.
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"learning_rate": 1.
|
| 110 |
-
"loss": 2.
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch": 0.
|
| 115 |
-
"learning_rate": 1.
|
| 116 |
-
"loss": 2.
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch": 0.
|
| 121 |
-
"learning_rate": 1.
|
| 122 |
-
"loss":
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch": 0.
|
| 127 |
-
"learning_rate": 1.
|
| 128 |
-
"loss":
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch": 0.
|
| 133 |
-
"learning_rate": 1.
|
| 134 |
-
"loss":
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch": 0.
|
| 139 |
-
"learning_rate": 1.
|
| 140 |
-
"loss": 2.
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch": 0.
|
| 145 |
-
"learning_rate": 1.
|
| 146 |
-
"loss": 2.
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 0.
|
| 151 |
-
"learning_rate": 1.
|
| 152 |
-
"loss": 2.
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch": 0.
|
| 157 |
-
"learning_rate": 1.
|
| 158 |
-
"loss": 2.
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch": 0.
|
| 163 |
-
"learning_rate": 1.
|
| 164 |
-
"loss":
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch": 0.
|
| 169 |
-
"learning_rate": 1.
|
| 170 |
-
"loss":
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch": 0.
|
| 175 |
-
"learning_rate": 1.
|
| 176 |
-
"loss":
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch": 0.
|
| 181 |
-
"learning_rate": 1.
|
| 182 |
-
"loss":
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch": 0.
|
| 187 |
-
"learning_rate": 1.
|
| 188 |
-
"loss":
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch": 0.
|
| 193 |
-
"learning_rate": 1.
|
| 194 |
-
"loss":
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch": 0.
|
| 199 |
-
"learning_rate": 1.
|
| 200 |
-
"loss": 2.
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 0.
|
| 205 |
-
"learning_rate": 1.
|
| 206 |
-
"loss":
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"learning_rate": 1.
|
| 212 |
-
"loss": 2.
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"learning_rate": 1.
|
| 218 |
-
"loss": 2.
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
-
"epoch":
|
| 223 |
-
"learning_rate": 1.
|
| 224 |
-
"loss":
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
-
"epoch":
|
| 229 |
-
"learning_rate": 1.
|
| 230 |
-
"loss":
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch":
|
| 235 |
-
"learning_rate": 1.
|
| 236 |
-
"loss":
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
-
"epoch":
|
| 241 |
-
"learning_rate": 1.
|
| 242 |
-
"loss":
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
-
"learning_rate": 1.
|
| 248 |
-
"loss":
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
-
"epoch":
|
| 253 |
-
"learning_rate": 1.
|
| 254 |
-
"loss": 2.
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
-
"epoch":
|
| 259 |
-
"learning_rate": 1.
|
| 260 |
-
"loss": 2.
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"learning_rate": 1.
|
| 266 |
-
"loss": 2.
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
-
"epoch":
|
| 271 |
-
"learning_rate": 1.
|
| 272 |
-
"loss":
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
-
"learning_rate": 1.
|
| 278 |
-
"loss":
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
-
"epoch":
|
| 283 |
-
"learning_rate": 1.
|
| 284 |
-
"loss":
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch":
|
| 289 |
-
"learning_rate": 1.
|
| 290 |
-
"loss":
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"learning_rate": 1.
|
| 296 |
-
"loss":
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
-
"learning_rate": 1.
|
| 302 |
-
"loss":
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"learning_rate": 1.
|
| 308 |
-
"loss":
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch": 1.
|
| 313 |
-
"learning_rate": 1.
|
| 314 |
-
"loss": 2.
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch": 1.
|
| 319 |
-
"learning_rate":
|
| 320 |
-
"loss":
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch": 1.
|
| 325 |
-
"learning_rate":
|
| 326 |
-
"loss":
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch": 1.
|
| 331 |
-
"learning_rate":
|
| 332 |
-
"loss":
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch": 1.
|
| 337 |
-
"learning_rate":
|
| 338 |
-
"loss":
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch": 1.
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss":
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch": 1.
|
| 349 |
-
"learning_rate":
|
| 350 |
-
"loss": 2.
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
-
"epoch": 1.
|
| 355 |
-
"learning_rate":
|
| 356 |
-
"loss": 2.
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
-
"epoch": 1.
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss": 2.
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch": 1.
|
| 367 |
-
"learning_rate":
|
| 368 |
-
"loss":
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
-
"epoch": 1.
|
| 373 |
-
"learning_rate":
|
| 374 |
-
"loss":
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
-
"epoch": 1.
|
| 379 |
-
"learning_rate":
|
| 380 |
-
"loss":
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
-
"epoch": 1.
|
| 385 |
-
"learning_rate":
|
| 386 |
-
"loss":
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
-
"epoch": 1.
|
| 391 |
-
"learning_rate":
|
| 392 |
-
"loss":
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
-
"epoch": 1.
|
| 397 |
-
"learning_rate":
|
| 398 |
-
"loss": 2.
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
-
"epoch": 1.
|
| 403 |
-
"learning_rate":
|
| 404 |
-
"loss":
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
-
"epoch":
|
| 409 |
-
"learning_rate":
|
| 410 |
-
"loss":
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
-
"epoch":
|
| 415 |
-
"learning_rate":
|
| 416 |
-
"loss": 2.
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
-
"epoch":
|
| 421 |
-
"learning_rate":
|
| 422 |
-
"loss":
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss":
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss":
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch":
|
| 439 |
-
"learning_rate":
|
| 440 |
-
"loss":
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch":
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss":
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss":
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss":
|
| 459 |
"step": 150
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 2.
|
| 465 |
"step": 152
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 2.
|
| 471 |
"step": 154
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"learning_rate":
|
| 476 |
-
"loss":
|
| 477 |
"step": 156
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"learning_rate":
|
| 482 |
-
"loss":
|
| 483 |
"step": 158
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
-
"learning_rate":
|
| 488 |
-
"loss":
|
| 489 |
"step": 160
|
| 490 |
},
|
| 491 |
{
|
| 492 |
-
"epoch":
|
| 493 |
-
"learning_rate":
|
| 494 |
-
"loss":
|
| 495 |
"step": 162
|
| 496 |
},
|
| 497 |
{
|
| 498 |
-
"epoch":
|
| 499 |
-
"learning_rate":
|
| 500 |
-
"loss":
|
| 501 |
"step": 164
|
| 502 |
},
|
| 503 |
{
|
| 504 |
-
"epoch":
|
| 505 |
-
"learning_rate":
|
| 506 |
-
"loss":
|
| 507 |
"step": 166
|
| 508 |
},
|
| 509 |
{
|
| 510 |
-
"epoch":
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 2.
|
| 513 |
"step": 168
|
| 514 |
},
|
| 515 |
{
|
| 516 |
-
"epoch":
|
| 517 |
-
"learning_rate":
|
| 518 |
-
"loss":
|
| 519 |
"step": 170
|
| 520 |
},
|
| 521 |
{
|
| 522 |
-
"epoch":
|
| 523 |
-
"learning_rate":
|
| 524 |
-
"loss":
|
| 525 |
"step": 172
|
| 526 |
},
|
| 527 |
{
|
| 528 |
-
"epoch":
|
| 529 |
-
"learning_rate":
|
| 530 |
-
"loss":
|
| 531 |
"step": 174
|
| 532 |
},
|
| 533 |
{
|
| 534 |
-
"epoch":
|
| 535 |
-
"learning_rate":
|
| 536 |
-
"loss":
|
| 537 |
"step": 176
|
| 538 |
},
|
| 539 |
{
|
| 540 |
-
"epoch":
|
| 541 |
-
"learning_rate":
|
| 542 |
-
"loss":
|
| 543 |
"step": 178
|
| 544 |
},
|
| 545 |
{
|
| 546 |
-
"epoch":
|
| 547 |
-
"learning_rate":
|
| 548 |
-
"loss":
|
| 549 |
"step": 180
|
| 550 |
},
|
| 551 |
{
|
| 552 |
-
"epoch":
|
| 553 |
-
"learning_rate":
|
| 554 |
-
"loss": 2.
|
| 555 |
"step": 182
|
| 556 |
},
|
| 557 |
{
|
| 558 |
-
"epoch":
|
| 559 |
-
"learning_rate":
|
| 560 |
-
"loss": 2.
|
| 561 |
"step": 184
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
-
"learning_rate":
|
| 566 |
-
"loss":
|
| 567 |
"step": 186
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch":
|
| 571 |
-
"learning_rate":
|
| 572 |
-
"loss":
|
| 573 |
"step": 188
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch":
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss":
|
| 579 |
"step": 190
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch":
|
| 583 |
-
"learning_rate":
|
| 584 |
-
"loss":
|
| 585 |
"step": 192
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch":
|
| 589 |
-
"learning_rate":
|
| 590 |
-
"loss":
|
| 591 |
"step": 194
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 2.
|
| 597 |
"step": 196
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"learning_rate":
|
| 602 |
-
"loss":
|
| 603 |
"step": 198
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch":
|
| 607 |
-
"learning_rate":
|
| 608 |
-
"loss":
|
| 609 |
"step": 200
|
| 610 |
}
|
| 611 |
],
|
| 612 |
"logging_steps": 2,
|
| 613 |
-
"max_steps":
|
| 614 |
"num_train_epochs": 3,
|
| 615 |
"save_steps": 500,
|
| 616 |
-
"total_flos":
|
| 617 |
"trial_name": null,
|
| 618 |
"trial_params": null
|
| 619 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9972041006523766,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 201,
|
| 7 |
"is_hyper_param_search": false,
|
|
|
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 14 |
+
"loss": 2.7224,
|
| 15 |
"step": 2
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.06,
|
| 19 |
+
"learning_rate": 1.1428571428571429e-05,
|
| 20 |
+
"loss": 2.0227,
|
| 21 |
"step": 4
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.09,
|
| 25 |
+
"learning_rate": 1.7142857142857142e-05,
|
| 26 |
+
"loss": 2.0095,
|
| 27 |
"step": 6
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.12,
|
| 31 |
+
"learning_rate": 1.9998688836656322e-05,
|
| 32 |
+
"loss": 2.0443,
|
| 33 |
"step": 8
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.15,
|
| 37 |
+
"learning_rate": 1.998820159279591e-05,
|
| 38 |
+
"loss": 1.7957,
|
| 39 |
"step": 10
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.18,
|
| 43 |
+
"learning_rate": 1.9967238104745695e-05,
|
| 44 |
+
"loss": 2.2818,
|
| 45 |
"step": 12
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.21,
|
| 49 |
+
"learning_rate": 1.993582036030978e-05,
|
| 50 |
+
"loss": 2.5503,
|
| 51 |
"step": 14
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.24,
|
| 55 |
+
"learning_rate": 1.9893981312363563e-05,
|
| 56 |
+
"loss": 3.1289,
|
| 57 |
"step": 16
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.27,
|
| 61 |
+
"learning_rate": 1.9841764844290744e-05,
|
| 62 |
+
"loss": 2.1579,
|
| 63 |
"step": 18
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.3,
|
| 67 |
+
"learning_rate": 1.977922572395571e-05,
|
| 68 |
+
"loss": 1.9154,
|
| 69 |
"step": 20
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.33,
|
| 73 |
+
"learning_rate": 1.9706429546259592e-05,
|
| 74 |
+
"loss": 1.9413,
|
| 75 |
"step": 22
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.36,
|
| 79 |
+
"learning_rate": 1.9623452664340305e-05,
|
| 80 |
+
"loss": 1.9034,
|
| 81 |
"step": 24
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.39,
|
| 85 |
+
"learning_rate": 1.953038210948861e-05,
|
| 86 |
+
"loss": 1.9566,
|
| 87 |
"step": 26
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.42,
|
| 91 |
+
"learning_rate": 1.9427315499864345e-05,
|
| 92 |
+
"loss": 2.2827,
|
| 93 |
"step": 28
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.45,
|
| 97 |
+
"learning_rate": 1.9314360938108427e-05,
|
| 98 |
+
"loss": 2.5939,
|
| 99 |
"step": 30
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.48,
|
| 103 |
+
"learning_rate": 1.9191636897958123e-05,
|
| 104 |
+
"loss": 2.8079,
|
| 105 |
"step": 32
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.51,
|
| 109 |
+
"learning_rate": 1.905927209998447e-05,
|
| 110 |
+
"loss": 2.5432,
|
| 111 |
"step": 34
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.54,
|
| 115 |
+
"learning_rate": 1.8917405376582144e-05,
|
| 116 |
+
"loss": 2.1705,
|
| 117 |
"step": 36
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.57,
|
| 121 |
+
"learning_rate": 1.876618552635348e-05,
|
| 122 |
+
"loss": 1.857,
|
| 123 |
"step": 38
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.6,
|
| 127 |
+
"learning_rate": 1.8605771158039253e-05,
|
| 128 |
+
"loss": 1.8961,
|
| 129 |
"step": 40
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.63,
|
| 133 |
+
"learning_rate": 1.8436330524160048e-05,
|
| 134 |
+
"loss": 1.7572,
|
| 135 |
"step": 42
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.66,
|
| 139 |
+
"learning_rate": 1.8258041344542567e-05,
|
| 140 |
+
"loss": 2.0378,
|
| 141 |
"step": 44
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.69,
|
| 145 |
+
"learning_rate": 1.8071090619916095e-05,
|
| 146 |
+
"loss": 2.3258,
|
| 147 |
"step": 46
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.72,
|
| 151 |
+
"learning_rate": 1.7875674435774546e-05,
|
| 152 |
+
"loss": 2.4449,
|
| 153 |
"step": 48
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.75,
|
| 157 |
+
"learning_rate": 1.767199775670986e-05,
|
| 158 |
+
"loss": 2.6689,
|
| 159 |
"step": 50
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.78,
|
| 163 |
+
"learning_rate": 1.7460274211432463e-05,
|
| 164 |
+
"loss": 1.9396,
|
| 165 |
"step": 52
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.81,
|
| 169 |
+
"learning_rate": 1.7240725868704218e-05,
|
| 170 |
+
"loss": 1.8322,
|
| 171 |
"step": 54
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 0.84,
|
| 175 |
+
"learning_rate": 1.7013583004418994e-05,
|
| 176 |
+
"loss": 1.8619,
|
| 177 |
"step": 56
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.86,
|
| 181 |
+
"learning_rate": 1.6779083860075032e-05,
|
| 182 |
+
"loss": 1.8185,
|
| 183 |
"step": 58
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 0.89,
|
| 187 |
+
"learning_rate": 1.6537474392892527e-05,
|
| 188 |
+
"loss": 1.7511,
|
| 189 |
"step": 60
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.92,
|
| 193 |
+
"learning_rate": 1.6289008017838447e-05,
|
| 194 |
+
"loss": 1.9468,
|
| 195 |
"step": 62
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 0.95,
|
| 199 |
+
"learning_rate": 1.603394534182925e-05,
|
| 200 |
+
"loss": 2.1683,
|
| 201 |
"step": 64
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 0.98,
|
| 205 |
+
"learning_rate": 1.5772553890390196e-05,
|
| 206 |
+
"loss": 1.7529,
|
| 207 |
"step": 66
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 1.01,
|
| 211 |
+
"learning_rate": 1.5505107827058038e-05,
|
| 212 |
+
"loss": 2.3314,
|
| 213 |
"step": 68
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 1.04,
|
| 217 |
+
"learning_rate": 1.52318876658213e-05,
|
| 218 |
+
"loss": 2.0392,
|
| 219 |
"step": 70
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 1.07,
|
| 223 |
+
"learning_rate": 1.4953179976899878e-05,
|
| 224 |
+
"loss": 1.7551,
|
| 225 |
"step": 72
|
| 226 |
},
|
| 227 |
{
|
| 228 |
+
"epoch": 1.1,
|
| 229 |
+
"learning_rate": 1.4669277086172406e-05,
|
| 230 |
+
"loss": 1.9161,
|
| 231 |
"step": 74
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 1.13,
|
| 235 |
+
"learning_rate": 1.4380476768566825e-05,
|
| 236 |
+
"loss": 1.6602,
|
| 237 |
"step": 76
|
| 238 |
},
|
| 239 |
{
|
| 240 |
+
"epoch": 1.16,
|
| 241 |
+
"learning_rate": 1.4087081935735565e-05,
|
| 242 |
+
"loss": 1.652,
|
| 243 |
"step": 78
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 1.19,
|
| 247 |
+
"learning_rate": 1.378940031834307e-05,
|
| 248 |
+
"loss": 1.9199,
|
| 249 |
"step": 80
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 1.22,
|
| 253 |
+
"learning_rate": 1.3487744143298822e-05,
|
| 254 |
+
"loss": 2.0564,
|
| 255 |
"step": 82
|
| 256 |
},
|
| 257 |
{
|
| 258 |
+
"epoch": 1.25,
|
| 259 |
+
"learning_rate": 1.3182429806274442e-05,
|
| 260 |
+
"loss": 2.1871,
|
| 261 |
"step": 84
|
| 262 |
},
|
| 263 |
{
|
| 264 |
+
"epoch": 1.28,
|
| 265 |
+
"learning_rate": 1.2873777539848284e-05,
|
| 266 |
+
"loss": 2.0577,
|
| 267 |
"step": 86
|
| 268 |
},
|
| 269 |
{
|
| 270 |
+
"epoch": 1.31,
|
| 271 |
+
"learning_rate": 1.2562111077625723e-05,
|
| 272 |
+
"loss": 1.8106,
|
| 273 |
"step": 88
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 1.34,
|
| 277 |
+
"learning_rate": 1.2247757314687296e-05,
|
| 278 |
+
"loss": 1.7766,
|
| 279 |
"step": 90
|
| 280 |
},
|
| 281 |
{
|
| 282 |
+
"epoch": 1.37,
|
| 283 |
+
"learning_rate": 1.1931045964720882e-05,
|
| 284 |
+
"loss": 1.7395,
|
| 285 |
"step": 92
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 1.4,
|
| 289 |
+
"learning_rate": 1.1612309214197599e-05,
|
| 290 |
+
"loss": 1.6721,
|
| 291 |
"step": 94
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 1.43,
|
| 295 |
+
"learning_rate": 1.1291881373954066e-05,
|
| 296 |
+
"loss": 1.7987,
|
| 297 |
"step": 96
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 1.46,
|
| 301 |
+
"learning_rate": 1.0970098528546482e-05,
|
| 302 |
+
"loss": 1.9993,
|
| 303 |
"step": 98
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 1.49,
|
| 307 |
+
"learning_rate": 1.0647298183744359e-05,
|
| 308 |
+
"loss": 1.9175,
|
| 309 |
"step": 100
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 1.52,
|
| 313 |
+
"learning_rate": 1.0323818912533561e-05,
|
| 314 |
+
"loss": 2.5671,
|
| 315 |
"step": 102
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 1.55,
|
| 319 |
+
"learning_rate": 1e-05,
|
| 320 |
+
"loss": 1.7011,
|
| 321 |
"step": 104
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 1.58,
|
| 325 |
+
"learning_rate": 9.676181087466444e-06,
|
| 326 |
+
"loss": 1.7352,
|
| 327 |
"step": 106
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 1.61,
|
| 331 |
+
"learning_rate": 9.352701816255643e-06,
|
| 332 |
+
"loss": 1.669,
|
| 333 |
"step": 108
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 1.64,
|
| 337 |
+
"learning_rate": 9.02990147145352e-06,
|
| 338 |
+
"loss": 1.5447,
|
| 339 |
"step": 110
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 1.67,
|
| 343 |
+
"learning_rate": 8.708118626045939e-06,
|
| 344 |
+
"loss": 1.7499,
|
| 345 |
"step": 112
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 1.7,
|
| 349 |
+
"learning_rate": 8.387690785802403e-06,
|
| 350 |
+
"loss": 2.0995,
|
| 351 |
"step": 114
|
| 352 |
},
|
| 353 |
{
|
| 354 |
+
"epoch": 1.73,
|
| 355 |
+
"learning_rate": 8.068954035279121e-06,
|
| 356 |
+
"loss": 2.3553,
|
| 357 |
"step": 116
|
| 358 |
},
|
| 359 |
{
|
| 360 |
+
"epoch": 1.76,
|
| 361 |
+
"learning_rate": 7.752242685312709e-06,
|
| 362 |
+
"loss": 2.3198,
|
| 363 |
"step": 118
|
| 364 |
},
|
| 365 |
{
|
| 366 |
+
"epoch": 1.79,
|
| 367 |
+
"learning_rate": 7.4378889223742766e-06,
|
| 368 |
+
"loss": 1.6883,
|
| 369 |
"step": 120
|
| 370 |
},
|
| 371 |
{
|
| 372 |
+
"epoch": 1.82,
|
| 373 |
+
"learning_rate": 7.126222460151719e-06,
|
| 374 |
+
"loss": 1.7977,
|
| 375 |
"step": 122
|
| 376 |
},
|
| 377 |
{
|
| 378 |
+
"epoch": 1.85,
|
| 379 |
+
"learning_rate": 6.8175701937255645e-06,
|
| 380 |
+
"loss": 1.6274,
|
| 381 |
"step": 124
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 1.88,
|
| 385 |
+
"learning_rate": 6.5122558567011775e-06,
|
| 386 |
+
"loss": 1.5804,
|
| 387 |
"step": 126
|
| 388 |
},
|
| 389 |
{
|
| 390 |
+
"epoch": 1.91,
|
| 391 |
+
"learning_rate": 6.210599681656933e-06,
|
| 392 |
+
"loss": 1.8931,
|
| 393 |
"step": 128
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 1.94,
|
| 397 |
+
"learning_rate": 5.912918064264441e-06,
|
| 398 |
+
"loss": 2.0786,
|
| 399 |
"step": 130
|
| 400 |
},
|
| 401 |
{
|
| 402 |
+
"epoch": 1.97,
|
| 403 |
+
"learning_rate": 5.619523231433177e-06,
|
| 404 |
+
"loss": 1.628,
|
| 405 |
"step": 132
|
| 406 |
},
|
| 407 |
{
|
| 408 |
+
"epoch": 2.0,
|
| 409 |
+
"learning_rate": 5.330722913827594e-06,
|
| 410 |
+
"loss": 1.7865,
|
| 411 |
"step": 134
|
| 412 |
},
|
| 413 |
{
|
| 414 |
+
"epoch": 2.03,
|
| 415 |
+
"learning_rate": 5.046820023100129e-06,
|
| 416 |
+
"loss": 2.2694,
|
| 417 |
"step": 136
|
| 418 |
},
|
| 419 |
{
|
| 420 |
+
"epoch": 2.06,
|
| 421 |
+
"learning_rate": 4.7681123341787e-06,
|
| 422 |
+
"loss": 1.8822,
|
| 423 |
"step": 138
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 2.09,
|
| 427 |
+
"learning_rate": 4.494892172941965e-06,
|
| 428 |
+
"loss": 1.7278,
|
| 429 |
"step": 140
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 2.12,
|
| 433 |
+
"learning_rate": 4.2274461096098085e-06,
|
| 434 |
+
"loss": 1.7028,
|
| 435 |
"step": 142
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 2.15,
|
| 439 |
+
"learning_rate": 3.966054658170754e-06,
|
| 440 |
+
"loss": 1.7108,
|
| 441 |
"step": 144
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 2.18,
|
| 445 |
+
"learning_rate": 3.7109919821615546e-06,
|
| 446 |
+
"loss": 1.7017,
|
| 447 |
"step": 146
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 2.21,
|
| 451 |
+
"learning_rate": 3.4625256071074776e-06,
|
| 452 |
+
"loss": 1.8465,
|
| 453 |
"step": 148
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 2.24,
|
| 457 |
+
"learning_rate": 3.2209161399249677e-06,
|
| 458 |
+
"loss": 1.7957,
|
| 459 |
"step": 150
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 2.27,
|
| 463 |
+
"learning_rate": 2.9864169955810085e-06,
|
| 464 |
+
"loss": 2.444,
|
| 465 |
"step": 152
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 2.3,
|
| 469 |
+
"learning_rate": 2.759274131295787e-06,
|
| 470 |
+
"loss": 2.2676,
|
| 471 |
"step": 154
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 2.33,
|
| 475 |
+
"learning_rate": 2.5397257885675396e-06,
|
| 476 |
+
"loss": 1.6855,
|
| 477 |
"step": 156
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 2.36,
|
| 481 |
+
"learning_rate": 2.328002243290138e-06,
|
| 482 |
+
"loss": 1.7492,
|
| 483 |
"step": 158
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 2.39,
|
| 487 |
+
"learning_rate": 2.124325564225458e-06,
|
| 488 |
+
"loss": 1.5845,
|
| 489 |
"step": 160
|
| 490 |
},
|
| 491 |
{
|
| 492 |
+
"epoch": 2.42,
|
| 493 |
+
"learning_rate": 1.9289093800839067e-06,
|
| 494 |
+
"loss": 1.7231,
|
| 495 |
"step": 162
|
| 496 |
},
|
| 497 |
{
|
| 498 |
+
"epoch": 2.45,
|
| 499 |
+
"learning_rate": 1.7419586554574364e-06,
|
| 500 |
+
"loss": 1.9025,
|
| 501 |
"step": 164
|
| 502 |
},
|
| 503 |
{
|
| 504 |
+
"epoch": 2.48,
|
| 505 |
+
"learning_rate": 1.5636694758399563e-06,
|
| 506 |
+
"loss": 1.964,
|
| 507 |
"step": 166
|
| 508 |
},
|
| 509 |
{
|
| 510 |
+
"epoch": 2.51,
|
| 511 |
+
"learning_rate": 1.3942288419607476e-06,
|
| 512 |
+
"loss": 2.2618,
|
| 513 |
"step": 168
|
| 514 |
},
|
| 515 |
{
|
| 516 |
+
"epoch": 2.53,
|
| 517 |
+
"learning_rate": 1.233814473646524e-06,
|
| 518 |
+
"loss": 1.7035,
|
| 519 |
"step": 170
|
| 520 |
},
|
| 521 |
{
|
| 522 |
+
"epoch": 2.56,
|
| 523 |
+
"learning_rate": 1.0825946234178575e-06,
|
| 524 |
+
"loss": 1.8286,
|
| 525 |
"step": 172
|
| 526 |
},
|
| 527 |
{
|
| 528 |
+
"epoch": 2.59,
|
| 529 |
+
"learning_rate": 9.407279000155311e-07,
|
| 530 |
+
"loss": 1.8184,
|
| 531 |
"step": 174
|
| 532 |
},
|
| 533 |
{
|
| 534 |
+
"epoch": 2.62,
|
| 535 |
+
"learning_rate": 8.083631020418792e-07,
|
| 536 |
+
"loss": 1.5351,
|
| 537 |
"step": 176
|
| 538 |
},
|
| 539 |
{
|
| 540 |
+
"epoch": 2.65,
|
| 541 |
+
"learning_rate": 6.856390618915775e-07,
|
| 542 |
+
"loss": 1.6511,
|
| 543 |
"step": 178
|
| 544 |
},
|
| 545 |
{
|
| 546 |
+
"epoch": 2.68,
|
| 547 |
+
"learning_rate": 5.726845001356573e-07,
|
| 548 |
+
"loss": 1.8842,
|
| 549 |
"step": 180
|
| 550 |
},
|
| 551 |
{
|
| 552 |
+
"epoch": 2.71,
|
| 553 |
+
"learning_rate": 4.696178905113913e-07,
|
| 554 |
+
"loss": 2.0207,
|
| 555 |
"step": 182
|
| 556 |
},
|
| 557 |
{
|
| 558 |
+
"epoch": 2.74,
|
| 559 |
+
"learning_rate": 3.7654733565969826e-07,
|
| 560 |
+
"loss": 2.6261,
|
| 561 |
"step": 184
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 2.77,
|
| 565 |
+
"learning_rate": 2.935704537404083e-07,
|
| 566 |
+
"loss": 1.7627,
|
| 567 |
"step": 186
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 2.8,
|
| 571 |
+
"learning_rate": 2.2077427604429435e-07,
|
| 572 |
+
"loss": 1.7416,
|
| 573 |
"step": 188
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 2.83,
|
| 577 |
+
"learning_rate": 1.5823515570925763e-07,
|
| 578 |
+
"loss": 1.7302,
|
| 579 |
"step": 190
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 2.86,
|
| 583 |
+
"learning_rate": 1.0601868763643997e-07,
|
| 584 |
+
"loss": 1.3737,
|
| 585 |
"step": 192
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 2.89,
|
| 589 |
+
"learning_rate": 6.417963969022389e-08,
|
| 590 |
+
"loss": 1.7739,
|
| 591 |
"step": 194
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 2.92,
|
| 595 |
+
"learning_rate": 3.2761895254306285e-08,
|
| 596 |
+
"loss": 2.0054,
|
| 597 |
"step": 196
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 2.95,
|
| 601 |
+
"learning_rate": 1.179840720409331e-08,
|
| 602 |
+
"loss": 1.9487,
|
| 603 |
"step": 198
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 2.98,
|
| 607 |
+
"learning_rate": 1.3111633436779792e-09,
|
| 608 |
+
"loss": 1.5353,
|
| 609 |
"step": 200
|
| 610 |
}
|
| 611 |
],
|
| 612 |
"logging_steps": 2,
|
| 613 |
+
"max_steps": 201,
|
| 614 |
"num_train_epochs": 3,
|
| 615 |
"save_steps": 500,
|
| 616 |
+
"total_flos": 8891281688641536.0,
|
| 617 |
"trial_name": null,
|
| 618 |
"trial_params": null
|
| 619 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4600
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:84e77ea3d2d70f689d692853b926d355519b2a679c28946fbaa23532b4bebaeb
|
| 3 |
size 4600
|