Model save
Browse files- last-checkpoint/generation_config.json → generation_config.json +0 -0
- last-checkpoint/config.json +0 -68
- last-checkpoint/merges.txt +0 -0
- last-checkpoint/optimizer.pt +0 -3
- last-checkpoint/pytorch_model.bin +0 -3
- last-checkpoint/rng_state.pth +0 -3
- last-checkpoint/scaler.pt +0 -3
- last-checkpoint/scheduler.pt +0 -3
- last-checkpoint/special_tokens_map.json +0 -753
- last-checkpoint/tokenizer.json +0 -0
- last-checkpoint/tokenizer_config.json +0 -64
- last-checkpoint/trainer_state.json +0 -476
- last-checkpoint/training_args.bin +0 -3
- last-checkpoint/vocab.json +0 -0
- pytorch_model.bin +1 -1
last-checkpoint/generation_config.json → generation_config.json
RENAMED
|
File without changes
|
last-checkpoint/config.json
DELETED
|
@@ -1,68 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"_name_or_path": "Salesforce/codet5-small",
|
| 3 |
-
"architectures": [
|
| 4 |
-
"T5ForConditionalGeneration"
|
| 5 |
-
],
|
| 6 |
-
"bos_token_id": 1,
|
| 7 |
-
"d_ff": 2048,
|
| 8 |
-
"d_kv": 64,
|
| 9 |
-
"d_model": 512,
|
| 10 |
-
"decoder_start_token_id": 0,
|
| 11 |
-
"dense_act_fn": "relu",
|
| 12 |
-
"dropout_rate": 0.1,
|
| 13 |
-
"eos_token_id": 2,
|
| 14 |
-
"feed_forward_proj": "relu",
|
| 15 |
-
"gradient_checkpointing": false,
|
| 16 |
-
"id2label": {
|
| 17 |
-
"0": "LABEL_0"
|
| 18 |
-
},
|
| 19 |
-
"initializer_factor": 1.0,
|
| 20 |
-
"is_encoder_decoder": true,
|
| 21 |
-
"is_gated_act": false,
|
| 22 |
-
"label2id": {
|
| 23 |
-
"LABEL_0": 0
|
| 24 |
-
},
|
| 25 |
-
"layer_norm_epsilon": 1e-06,
|
| 26 |
-
"model_type": "t5",
|
| 27 |
-
"n_positions": 512,
|
| 28 |
-
"num_decoder_layers": 6,
|
| 29 |
-
"num_heads": 8,
|
| 30 |
-
"num_layers": 6,
|
| 31 |
-
"output_past": true,
|
| 32 |
-
"pad_token_id": 0,
|
| 33 |
-
"relative_attention_max_distance": 128,
|
| 34 |
-
"relative_attention_num_buckets": 32,
|
| 35 |
-
"task_specific_params": {
|
| 36 |
-
"summarization": {
|
| 37 |
-
"early_stopping": true,
|
| 38 |
-
"length_penalty": 2.0,
|
| 39 |
-
"max_length": 200,
|
| 40 |
-
"min_length": 30,
|
| 41 |
-
"no_repeat_ngram_size": 3,
|
| 42 |
-
"num_beams": 4,
|
| 43 |
-
"prefix": "summarize: "
|
| 44 |
-
},
|
| 45 |
-
"translation_en_to_de": {
|
| 46 |
-
"early_stopping": true,
|
| 47 |
-
"max_length": 300,
|
| 48 |
-
"num_beams": 4,
|
| 49 |
-
"prefix": "translate English to German: "
|
| 50 |
-
},
|
| 51 |
-
"translation_en_to_fr": {
|
| 52 |
-
"early_stopping": true,
|
| 53 |
-
"max_length": 300,
|
| 54 |
-
"num_beams": 4,
|
| 55 |
-
"prefix": "translate English to French: "
|
| 56 |
-
},
|
| 57 |
-
"translation_en_to_ro": {
|
| 58 |
-
"early_stopping": true,
|
| 59 |
-
"max_length": 300,
|
| 60 |
-
"num_beams": 4,
|
| 61 |
-
"prefix": "translate English to Romanian: "
|
| 62 |
-
}
|
| 63 |
-
},
|
| 64 |
-
"torch_dtype": "float32",
|
| 65 |
-
"transformers_version": "4.27.1",
|
| 66 |
-
"use_cache": true,
|
| 67 |
-
"vocab_size": 32100
|
| 68 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/merges.txt
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|
last-checkpoint/optimizer.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:85d02e6771037a9f06ba89407268dc679aa7e28576727773f7a0ba476cc30782
|
| 3 |
-
size 484047087
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/pytorch_model.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:e00849037f634126c4c8ea164eca1af190ebbc00eb6999c45ba97618c53348bd
|
| 3 |
-
size 242012923
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/rng_state.pth
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:e9a42b6f4c9624cf70bf813b62833d1ee0fae061fe52362dc0a07eb7621e6632
|
| 3 |
-
size 14503
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/scaler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:939de8c42f529424a66409039d99578e88624f64cffcc79c6bdcefc64f9ac0e1
|
| 3 |
-
size 559
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/scheduler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:e8c966cd8905004d1a1a7a30c56fa8132477a97d86ce93bbd4f7f057044e33f8
|
| 3 |
-
size 623
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/special_tokens_map.json
DELETED
|
@@ -1,753 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"additional_special_tokens": [
|
| 3 |
-
{
|
| 4 |
-
"content": "<extra_id_99>",
|
| 5 |
-
"lstrip": true,
|
| 6 |
-
"normalized": true,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false
|
| 9 |
-
},
|
| 10 |
-
{
|
| 11 |
-
"content": "<extra_id_98>",
|
| 12 |
-
"lstrip": true,
|
| 13 |
-
"normalized": true,
|
| 14 |
-
"rstrip": false,
|
| 15 |
-
"single_word": false
|
| 16 |
-
},
|
| 17 |
-
{
|
| 18 |
-
"content": "<extra_id_97>",
|
| 19 |
-
"lstrip": true,
|
| 20 |
-
"normalized": true,
|
| 21 |
-
"rstrip": false,
|
| 22 |
-
"single_word": false
|
| 23 |
-
},
|
| 24 |
-
{
|
| 25 |
-
"content": "<extra_id_96>",
|
| 26 |
-
"lstrip": true,
|
| 27 |
-
"normalized": true,
|
| 28 |
-
"rstrip": false,
|
| 29 |
-
"single_word": false
|
| 30 |
-
},
|
| 31 |
-
{
|
| 32 |
-
"content": "<extra_id_95>",
|
| 33 |
-
"lstrip": true,
|
| 34 |
-
"normalized": true,
|
| 35 |
-
"rstrip": false,
|
| 36 |
-
"single_word": false
|
| 37 |
-
},
|
| 38 |
-
{
|
| 39 |
-
"content": "<extra_id_94>",
|
| 40 |
-
"lstrip": true,
|
| 41 |
-
"normalized": true,
|
| 42 |
-
"rstrip": false,
|
| 43 |
-
"single_word": false
|
| 44 |
-
},
|
| 45 |
-
{
|
| 46 |
-
"content": "<extra_id_93>",
|
| 47 |
-
"lstrip": true,
|
| 48 |
-
"normalized": true,
|
| 49 |
-
"rstrip": false,
|
| 50 |
-
"single_word": false
|
| 51 |
-
},
|
| 52 |
-
{
|
| 53 |
-
"content": "<extra_id_92>",
|
| 54 |
-
"lstrip": true,
|
| 55 |
-
"normalized": true,
|
| 56 |
-
"rstrip": false,
|
| 57 |
-
"single_word": false
|
| 58 |
-
},
|
| 59 |
-
{
|
| 60 |
-
"content": "<extra_id_91>",
|
| 61 |
-
"lstrip": true,
|
| 62 |
-
"normalized": true,
|
| 63 |
-
"rstrip": false,
|
| 64 |
-
"single_word": false
|
| 65 |
-
},
|
| 66 |
-
{
|
| 67 |
-
"content": "<extra_id_90>",
|
| 68 |
-
"lstrip": true,
|
| 69 |
-
"normalized": true,
|
| 70 |
-
"rstrip": false,
|
| 71 |
-
"single_word": false
|
| 72 |
-
},
|
| 73 |
-
{
|
| 74 |
-
"content": "<extra_id_89>",
|
| 75 |
-
"lstrip": true,
|
| 76 |
-
"normalized": true,
|
| 77 |
-
"rstrip": false,
|
| 78 |
-
"single_word": false
|
| 79 |
-
},
|
| 80 |
-
{
|
| 81 |
-
"content": "<extra_id_88>",
|
| 82 |
-
"lstrip": true,
|
| 83 |
-
"normalized": true,
|
| 84 |
-
"rstrip": false,
|
| 85 |
-
"single_word": false
|
| 86 |
-
},
|
| 87 |
-
{
|
| 88 |
-
"content": "<extra_id_87>",
|
| 89 |
-
"lstrip": true,
|
| 90 |
-
"normalized": true,
|
| 91 |
-
"rstrip": false,
|
| 92 |
-
"single_word": false
|
| 93 |
-
},
|
| 94 |
-
{
|
| 95 |
-
"content": "<extra_id_86>",
|
| 96 |
-
"lstrip": true,
|
| 97 |
-
"normalized": true,
|
| 98 |
-
"rstrip": false,
|
| 99 |
-
"single_word": false
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"content": "<extra_id_85>",
|
| 103 |
-
"lstrip": true,
|
| 104 |
-
"normalized": true,
|
| 105 |
-
"rstrip": false,
|
| 106 |
-
"single_word": false
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"content": "<extra_id_84>",
|
| 110 |
-
"lstrip": true,
|
| 111 |
-
"normalized": true,
|
| 112 |
-
"rstrip": false,
|
| 113 |
-
"single_word": false
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"content": "<extra_id_83>",
|
| 117 |
-
"lstrip": true,
|
| 118 |
-
"normalized": true,
|
| 119 |
-
"rstrip": false,
|
| 120 |
-
"single_word": false
|
| 121 |
-
},
|
| 122 |
-
{
|
| 123 |
-
"content": "<extra_id_82>",
|
| 124 |
-
"lstrip": true,
|
| 125 |
-
"normalized": true,
|
| 126 |
-
"rstrip": false,
|
| 127 |
-
"single_word": false
|
| 128 |
-
},
|
| 129 |
-
{
|
| 130 |
-
"content": "<extra_id_81>",
|
| 131 |
-
"lstrip": true,
|
| 132 |
-
"normalized": true,
|
| 133 |
-
"rstrip": false,
|
| 134 |
-
"single_word": false
|
| 135 |
-
},
|
| 136 |
-
{
|
| 137 |
-
"content": "<extra_id_80>",
|
| 138 |
-
"lstrip": true,
|
| 139 |
-
"normalized": true,
|
| 140 |
-
"rstrip": false,
|
| 141 |
-
"single_word": false
|
| 142 |
-
},
|
| 143 |
-
{
|
| 144 |
-
"content": "<extra_id_79>",
|
| 145 |
-
"lstrip": true,
|
| 146 |
-
"normalized": true,
|
| 147 |
-
"rstrip": false,
|
| 148 |
-
"single_word": false
|
| 149 |
-
},
|
| 150 |
-
{
|
| 151 |
-
"content": "<extra_id_78>",
|
| 152 |
-
"lstrip": true,
|
| 153 |
-
"normalized": true,
|
| 154 |
-
"rstrip": false,
|
| 155 |
-
"single_word": false
|
| 156 |
-
},
|
| 157 |
-
{
|
| 158 |
-
"content": "<extra_id_77>",
|
| 159 |
-
"lstrip": true,
|
| 160 |
-
"normalized": true,
|
| 161 |
-
"rstrip": false,
|
| 162 |
-
"single_word": false
|
| 163 |
-
},
|
| 164 |
-
{
|
| 165 |
-
"content": "<extra_id_76>",
|
| 166 |
-
"lstrip": true,
|
| 167 |
-
"normalized": true,
|
| 168 |
-
"rstrip": false,
|
| 169 |
-
"single_word": false
|
| 170 |
-
},
|
| 171 |
-
{
|
| 172 |
-
"content": "<extra_id_75>",
|
| 173 |
-
"lstrip": true,
|
| 174 |
-
"normalized": true,
|
| 175 |
-
"rstrip": false,
|
| 176 |
-
"single_word": false
|
| 177 |
-
},
|
| 178 |
-
{
|
| 179 |
-
"content": "<extra_id_74>",
|
| 180 |
-
"lstrip": true,
|
| 181 |
-
"normalized": true,
|
| 182 |
-
"rstrip": false,
|
| 183 |
-
"single_word": false
|
| 184 |
-
},
|
| 185 |
-
{
|
| 186 |
-
"content": "<extra_id_73>",
|
| 187 |
-
"lstrip": true,
|
| 188 |
-
"normalized": true,
|
| 189 |
-
"rstrip": false,
|
| 190 |
-
"single_word": false
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"content": "<extra_id_72>",
|
| 194 |
-
"lstrip": true,
|
| 195 |
-
"normalized": true,
|
| 196 |
-
"rstrip": false,
|
| 197 |
-
"single_word": false
|
| 198 |
-
},
|
| 199 |
-
{
|
| 200 |
-
"content": "<extra_id_71>",
|
| 201 |
-
"lstrip": true,
|
| 202 |
-
"normalized": true,
|
| 203 |
-
"rstrip": false,
|
| 204 |
-
"single_word": false
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"content": "<extra_id_70>",
|
| 208 |
-
"lstrip": true,
|
| 209 |
-
"normalized": true,
|
| 210 |
-
"rstrip": false,
|
| 211 |
-
"single_word": false
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"content": "<extra_id_69>",
|
| 215 |
-
"lstrip": true,
|
| 216 |
-
"normalized": true,
|
| 217 |
-
"rstrip": false,
|
| 218 |
-
"single_word": false
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"content": "<extra_id_68>",
|
| 222 |
-
"lstrip": true,
|
| 223 |
-
"normalized": true,
|
| 224 |
-
"rstrip": false,
|
| 225 |
-
"single_word": false
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"content": "<extra_id_67>",
|
| 229 |
-
"lstrip": true,
|
| 230 |
-
"normalized": true,
|
| 231 |
-
"rstrip": false,
|
| 232 |
-
"single_word": false
|
| 233 |
-
},
|
| 234 |
-
{
|
| 235 |
-
"content": "<extra_id_66>",
|
| 236 |
-
"lstrip": true,
|
| 237 |
-
"normalized": true,
|
| 238 |
-
"rstrip": false,
|
| 239 |
-
"single_word": false
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"content": "<extra_id_65>",
|
| 243 |
-
"lstrip": true,
|
| 244 |
-
"normalized": true,
|
| 245 |
-
"rstrip": false,
|
| 246 |
-
"single_word": false
|
| 247 |
-
},
|
| 248 |
-
{
|
| 249 |
-
"content": "<extra_id_64>",
|
| 250 |
-
"lstrip": true,
|
| 251 |
-
"normalized": true,
|
| 252 |
-
"rstrip": false,
|
| 253 |
-
"single_word": false
|
| 254 |
-
},
|
| 255 |
-
{
|
| 256 |
-
"content": "<extra_id_63>",
|
| 257 |
-
"lstrip": true,
|
| 258 |
-
"normalized": true,
|
| 259 |
-
"rstrip": false,
|
| 260 |
-
"single_word": false
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"content": "<extra_id_62>",
|
| 264 |
-
"lstrip": true,
|
| 265 |
-
"normalized": true,
|
| 266 |
-
"rstrip": false,
|
| 267 |
-
"single_word": false
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"content": "<extra_id_61>",
|
| 271 |
-
"lstrip": true,
|
| 272 |
-
"normalized": true,
|
| 273 |
-
"rstrip": false,
|
| 274 |
-
"single_word": false
|
| 275 |
-
},
|
| 276 |
-
{
|
| 277 |
-
"content": "<extra_id_60>",
|
| 278 |
-
"lstrip": true,
|
| 279 |
-
"normalized": true,
|
| 280 |
-
"rstrip": false,
|
| 281 |
-
"single_word": false
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"content": "<extra_id_59>",
|
| 285 |
-
"lstrip": true,
|
| 286 |
-
"normalized": true,
|
| 287 |
-
"rstrip": false,
|
| 288 |
-
"single_word": false
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"content": "<extra_id_58>",
|
| 292 |
-
"lstrip": true,
|
| 293 |
-
"normalized": true,
|
| 294 |
-
"rstrip": false,
|
| 295 |
-
"single_word": false
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"content": "<extra_id_57>",
|
| 299 |
-
"lstrip": true,
|
| 300 |
-
"normalized": true,
|
| 301 |
-
"rstrip": false,
|
| 302 |
-
"single_word": false
|
| 303 |
-
},
|
| 304 |
-
{
|
| 305 |
-
"content": "<extra_id_56>",
|
| 306 |
-
"lstrip": true,
|
| 307 |
-
"normalized": true,
|
| 308 |
-
"rstrip": false,
|
| 309 |
-
"single_word": false
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"content": "<extra_id_55>",
|
| 313 |
-
"lstrip": true,
|
| 314 |
-
"normalized": true,
|
| 315 |
-
"rstrip": false,
|
| 316 |
-
"single_word": false
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"content": "<extra_id_54>",
|
| 320 |
-
"lstrip": true,
|
| 321 |
-
"normalized": true,
|
| 322 |
-
"rstrip": false,
|
| 323 |
-
"single_word": false
|
| 324 |
-
},
|
| 325 |
-
{
|
| 326 |
-
"content": "<extra_id_53>",
|
| 327 |
-
"lstrip": true,
|
| 328 |
-
"normalized": true,
|
| 329 |
-
"rstrip": false,
|
| 330 |
-
"single_word": false
|
| 331 |
-
},
|
| 332 |
-
{
|
| 333 |
-
"content": "<extra_id_52>",
|
| 334 |
-
"lstrip": true,
|
| 335 |
-
"normalized": true,
|
| 336 |
-
"rstrip": false,
|
| 337 |
-
"single_word": false
|
| 338 |
-
},
|
| 339 |
-
{
|
| 340 |
-
"content": "<extra_id_51>",
|
| 341 |
-
"lstrip": true,
|
| 342 |
-
"normalized": true,
|
| 343 |
-
"rstrip": false,
|
| 344 |
-
"single_word": false
|
| 345 |
-
},
|
| 346 |
-
{
|
| 347 |
-
"content": "<extra_id_50>",
|
| 348 |
-
"lstrip": true,
|
| 349 |
-
"normalized": true,
|
| 350 |
-
"rstrip": false,
|
| 351 |
-
"single_word": false
|
| 352 |
-
},
|
| 353 |
-
{
|
| 354 |
-
"content": "<extra_id_49>",
|
| 355 |
-
"lstrip": true,
|
| 356 |
-
"normalized": true,
|
| 357 |
-
"rstrip": false,
|
| 358 |
-
"single_word": false
|
| 359 |
-
},
|
| 360 |
-
{
|
| 361 |
-
"content": "<extra_id_48>",
|
| 362 |
-
"lstrip": true,
|
| 363 |
-
"normalized": true,
|
| 364 |
-
"rstrip": false,
|
| 365 |
-
"single_word": false
|
| 366 |
-
},
|
| 367 |
-
{
|
| 368 |
-
"content": "<extra_id_47>",
|
| 369 |
-
"lstrip": true,
|
| 370 |
-
"normalized": true,
|
| 371 |
-
"rstrip": false,
|
| 372 |
-
"single_word": false
|
| 373 |
-
},
|
| 374 |
-
{
|
| 375 |
-
"content": "<extra_id_46>",
|
| 376 |
-
"lstrip": true,
|
| 377 |
-
"normalized": true,
|
| 378 |
-
"rstrip": false,
|
| 379 |
-
"single_word": false
|
| 380 |
-
},
|
| 381 |
-
{
|
| 382 |
-
"content": "<extra_id_45>",
|
| 383 |
-
"lstrip": true,
|
| 384 |
-
"normalized": true,
|
| 385 |
-
"rstrip": false,
|
| 386 |
-
"single_word": false
|
| 387 |
-
},
|
| 388 |
-
{
|
| 389 |
-
"content": "<extra_id_44>",
|
| 390 |
-
"lstrip": true,
|
| 391 |
-
"normalized": true,
|
| 392 |
-
"rstrip": false,
|
| 393 |
-
"single_word": false
|
| 394 |
-
},
|
| 395 |
-
{
|
| 396 |
-
"content": "<extra_id_43>",
|
| 397 |
-
"lstrip": true,
|
| 398 |
-
"normalized": true,
|
| 399 |
-
"rstrip": false,
|
| 400 |
-
"single_word": false
|
| 401 |
-
},
|
| 402 |
-
{
|
| 403 |
-
"content": "<extra_id_42>",
|
| 404 |
-
"lstrip": true,
|
| 405 |
-
"normalized": true,
|
| 406 |
-
"rstrip": false,
|
| 407 |
-
"single_word": false
|
| 408 |
-
},
|
| 409 |
-
{
|
| 410 |
-
"content": "<extra_id_41>",
|
| 411 |
-
"lstrip": true,
|
| 412 |
-
"normalized": true,
|
| 413 |
-
"rstrip": false,
|
| 414 |
-
"single_word": false
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"content": "<extra_id_40>",
|
| 418 |
-
"lstrip": true,
|
| 419 |
-
"normalized": true,
|
| 420 |
-
"rstrip": false,
|
| 421 |
-
"single_word": false
|
| 422 |
-
},
|
| 423 |
-
{
|
| 424 |
-
"content": "<extra_id_39>",
|
| 425 |
-
"lstrip": true,
|
| 426 |
-
"normalized": true,
|
| 427 |
-
"rstrip": false,
|
| 428 |
-
"single_word": false
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"content": "<extra_id_38>",
|
| 432 |
-
"lstrip": true,
|
| 433 |
-
"normalized": true,
|
| 434 |
-
"rstrip": false,
|
| 435 |
-
"single_word": false
|
| 436 |
-
},
|
| 437 |
-
{
|
| 438 |
-
"content": "<extra_id_37>",
|
| 439 |
-
"lstrip": true,
|
| 440 |
-
"normalized": true,
|
| 441 |
-
"rstrip": false,
|
| 442 |
-
"single_word": false
|
| 443 |
-
},
|
| 444 |
-
{
|
| 445 |
-
"content": "<extra_id_36>",
|
| 446 |
-
"lstrip": true,
|
| 447 |
-
"normalized": true,
|
| 448 |
-
"rstrip": false,
|
| 449 |
-
"single_word": false
|
| 450 |
-
},
|
| 451 |
-
{
|
| 452 |
-
"content": "<extra_id_35>",
|
| 453 |
-
"lstrip": true,
|
| 454 |
-
"normalized": true,
|
| 455 |
-
"rstrip": false,
|
| 456 |
-
"single_word": false
|
| 457 |
-
},
|
| 458 |
-
{
|
| 459 |
-
"content": "<extra_id_34>",
|
| 460 |
-
"lstrip": true,
|
| 461 |
-
"normalized": true,
|
| 462 |
-
"rstrip": false,
|
| 463 |
-
"single_word": false
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"content": "<extra_id_33>",
|
| 467 |
-
"lstrip": true,
|
| 468 |
-
"normalized": true,
|
| 469 |
-
"rstrip": false,
|
| 470 |
-
"single_word": false
|
| 471 |
-
},
|
| 472 |
-
{
|
| 473 |
-
"content": "<extra_id_32>",
|
| 474 |
-
"lstrip": true,
|
| 475 |
-
"normalized": true,
|
| 476 |
-
"rstrip": false,
|
| 477 |
-
"single_word": false
|
| 478 |
-
},
|
| 479 |
-
{
|
| 480 |
-
"content": "<extra_id_31>",
|
| 481 |
-
"lstrip": true,
|
| 482 |
-
"normalized": true,
|
| 483 |
-
"rstrip": false,
|
| 484 |
-
"single_word": false
|
| 485 |
-
},
|
| 486 |
-
{
|
| 487 |
-
"content": "<extra_id_30>",
|
| 488 |
-
"lstrip": true,
|
| 489 |
-
"normalized": true,
|
| 490 |
-
"rstrip": false,
|
| 491 |
-
"single_word": false
|
| 492 |
-
},
|
| 493 |
-
{
|
| 494 |
-
"content": "<extra_id_29>",
|
| 495 |
-
"lstrip": true,
|
| 496 |
-
"normalized": true,
|
| 497 |
-
"rstrip": false,
|
| 498 |
-
"single_word": false
|
| 499 |
-
},
|
| 500 |
-
{
|
| 501 |
-
"content": "<extra_id_28>",
|
| 502 |
-
"lstrip": true,
|
| 503 |
-
"normalized": true,
|
| 504 |
-
"rstrip": false,
|
| 505 |
-
"single_word": false
|
| 506 |
-
},
|
| 507 |
-
{
|
| 508 |
-
"content": "<extra_id_27>",
|
| 509 |
-
"lstrip": true,
|
| 510 |
-
"normalized": true,
|
| 511 |
-
"rstrip": false,
|
| 512 |
-
"single_word": false
|
| 513 |
-
},
|
| 514 |
-
{
|
| 515 |
-
"content": "<extra_id_26>",
|
| 516 |
-
"lstrip": true,
|
| 517 |
-
"normalized": true,
|
| 518 |
-
"rstrip": false,
|
| 519 |
-
"single_word": false
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"content": "<extra_id_25>",
|
| 523 |
-
"lstrip": true,
|
| 524 |
-
"normalized": true,
|
| 525 |
-
"rstrip": false,
|
| 526 |
-
"single_word": false
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"content": "<extra_id_24>",
|
| 530 |
-
"lstrip": true,
|
| 531 |
-
"normalized": true,
|
| 532 |
-
"rstrip": false,
|
| 533 |
-
"single_word": false
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"content": "<extra_id_23>",
|
| 537 |
-
"lstrip": true,
|
| 538 |
-
"normalized": true,
|
| 539 |
-
"rstrip": false,
|
| 540 |
-
"single_word": false
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"content": "<extra_id_22>",
|
| 544 |
-
"lstrip": true,
|
| 545 |
-
"normalized": true,
|
| 546 |
-
"rstrip": false,
|
| 547 |
-
"single_word": false
|
| 548 |
-
},
|
| 549 |
-
{
|
| 550 |
-
"content": "<extra_id_21>",
|
| 551 |
-
"lstrip": true,
|
| 552 |
-
"normalized": true,
|
| 553 |
-
"rstrip": false,
|
| 554 |
-
"single_word": false
|
| 555 |
-
},
|
| 556 |
-
{
|
| 557 |
-
"content": "<extra_id_20>",
|
| 558 |
-
"lstrip": true,
|
| 559 |
-
"normalized": true,
|
| 560 |
-
"rstrip": false,
|
| 561 |
-
"single_word": false
|
| 562 |
-
},
|
| 563 |
-
{
|
| 564 |
-
"content": "<extra_id_19>",
|
| 565 |
-
"lstrip": true,
|
| 566 |
-
"normalized": true,
|
| 567 |
-
"rstrip": false,
|
| 568 |
-
"single_word": false
|
| 569 |
-
},
|
| 570 |
-
{
|
| 571 |
-
"content": "<extra_id_18>",
|
| 572 |
-
"lstrip": true,
|
| 573 |
-
"normalized": true,
|
| 574 |
-
"rstrip": false,
|
| 575 |
-
"single_word": false
|
| 576 |
-
},
|
| 577 |
-
{
|
| 578 |
-
"content": "<extra_id_17>",
|
| 579 |
-
"lstrip": true,
|
| 580 |
-
"normalized": true,
|
| 581 |
-
"rstrip": false,
|
| 582 |
-
"single_word": false
|
| 583 |
-
},
|
| 584 |
-
{
|
| 585 |
-
"content": "<extra_id_16>",
|
| 586 |
-
"lstrip": true,
|
| 587 |
-
"normalized": true,
|
| 588 |
-
"rstrip": false,
|
| 589 |
-
"single_word": false
|
| 590 |
-
},
|
| 591 |
-
{
|
| 592 |
-
"content": "<extra_id_15>",
|
| 593 |
-
"lstrip": true,
|
| 594 |
-
"normalized": true,
|
| 595 |
-
"rstrip": false,
|
| 596 |
-
"single_word": false
|
| 597 |
-
},
|
| 598 |
-
{
|
| 599 |
-
"content": "<extra_id_14>",
|
| 600 |
-
"lstrip": true,
|
| 601 |
-
"normalized": true,
|
| 602 |
-
"rstrip": false,
|
| 603 |
-
"single_word": false
|
| 604 |
-
},
|
| 605 |
-
{
|
| 606 |
-
"content": "<extra_id_13>",
|
| 607 |
-
"lstrip": true,
|
| 608 |
-
"normalized": true,
|
| 609 |
-
"rstrip": false,
|
| 610 |
-
"single_word": false
|
| 611 |
-
},
|
| 612 |
-
{
|
| 613 |
-
"content": "<extra_id_12>",
|
| 614 |
-
"lstrip": true,
|
| 615 |
-
"normalized": true,
|
| 616 |
-
"rstrip": false,
|
| 617 |
-
"single_word": false
|
| 618 |
-
},
|
| 619 |
-
{
|
| 620 |
-
"content": "<extra_id_11>",
|
| 621 |
-
"lstrip": true,
|
| 622 |
-
"normalized": true,
|
| 623 |
-
"rstrip": false,
|
| 624 |
-
"single_word": false
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"content": "<extra_id_10>",
|
| 628 |
-
"lstrip": true,
|
| 629 |
-
"normalized": true,
|
| 630 |
-
"rstrip": false,
|
| 631 |
-
"single_word": false
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"content": "<extra_id_9>",
|
| 635 |
-
"lstrip": true,
|
| 636 |
-
"normalized": true,
|
| 637 |
-
"rstrip": false,
|
| 638 |
-
"single_word": false
|
| 639 |
-
},
|
| 640 |
-
{
|
| 641 |
-
"content": "<extra_id_8>",
|
| 642 |
-
"lstrip": true,
|
| 643 |
-
"normalized": true,
|
| 644 |
-
"rstrip": false,
|
| 645 |
-
"single_word": false
|
| 646 |
-
},
|
| 647 |
-
{
|
| 648 |
-
"content": "<extra_id_7>",
|
| 649 |
-
"lstrip": true,
|
| 650 |
-
"normalized": true,
|
| 651 |
-
"rstrip": false,
|
| 652 |
-
"single_word": false
|
| 653 |
-
},
|
| 654 |
-
{
|
| 655 |
-
"content": "<extra_id_6>",
|
| 656 |
-
"lstrip": true,
|
| 657 |
-
"normalized": true,
|
| 658 |
-
"rstrip": false,
|
| 659 |
-
"single_word": false
|
| 660 |
-
},
|
| 661 |
-
{
|
| 662 |
-
"content": "<extra_id_5>",
|
| 663 |
-
"lstrip": true,
|
| 664 |
-
"normalized": true,
|
| 665 |
-
"rstrip": false,
|
| 666 |
-
"single_word": false
|
| 667 |
-
},
|
| 668 |
-
{
|
| 669 |
-
"content": "<extra_id_4>",
|
| 670 |
-
"lstrip": true,
|
| 671 |
-
"normalized": true,
|
| 672 |
-
"rstrip": false,
|
| 673 |
-
"single_word": false
|
| 674 |
-
},
|
| 675 |
-
{
|
| 676 |
-
"content": "<extra_id_3>",
|
| 677 |
-
"lstrip": true,
|
| 678 |
-
"normalized": true,
|
| 679 |
-
"rstrip": false,
|
| 680 |
-
"single_word": false
|
| 681 |
-
},
|
| 682 |
-
{
|
| 683 |
-
"content": "<extra_id_2>",
|
| 684 |
-
"lstrip": true,
|
| 685 |
-
"normalized": true,
|
| 686 |
-
"rstrip": false,
|
| 687 |
-
"single_word": false
|
| 688 |
-
},
|
| 689 |
-
{
|
| 690 |
-
"content": "<extra_id_1>",
|
| 691 |
-
"lstrip": true,
|
| 692 |
-
"normalized": true,
|
| 693 |
-
"rstrip": false,
|
| 694 |
-
"single_word": false
|
| 695 |
-
},
|
| 696 |
-
{
|
| 697 |
-
"content": "<extra_id_0>",
|
| 698 |
-
"lstrip": true,
|
| 699 |
-
"normalized": true,
|
| 700 |
-
"rstrip": false,
|
| 701 |
-
"single_word": false
|
| 702 |
-
}
|
| 703 |
-
],
|
| 704 |
-
"bos_token": {
|
| 705 |
-
"content": "<s>",
|
| 706 |
-
"lstrip": false,
|
| 707 |
-
"normalized": true,
|
| 708 |
-
"rstrip": false,
|
| 709 |
-
"single_word": false
|
| 710 |
-
},
|
| 711 |
-
"cls_token": {
|
| 712 |
-
"content": "<s>",
|
| 713 |
-
"lstrip": false,
|
| 714 |
-
"normalized": true,
|
| 715 |
-
"rstrip": false,
|
| 716 |
-
"single_word": false
|
| 717 |
-
},
|
| 718 |
-
"eos_token": {
|
| 719 |
-
"content": "</s>",
|
| 720 |
-
"lstrip": false,
|
| 721 |
-
"normalized": true,
|
| 722 |
-
"rstrip": false,
|
| 723 |
-
"single_word": false
|
| 724 |
-
},
|
| 725 |
-
"mask_token": {
|
| 726 |
-
"content": "<mask>",
|
| 727 |
-
"lstrip": true,
|
| 728 |
-
"normalized": true,
|
| 729 |
-
"rstrip": false,
|
| 730 |
-
"single_word": false
|
| 731 |
-
},
|
| 732 |
-
"pad_token": {
|
| 733 |
-
"content": "<pad>",
|
| 734 |
-
"lstrip": false,
|
| 735 |
-
"normalized": true,
|
| 736 |
-
"rstrip": false,
|
| 737 |
-
"single_word": false
|
| 738 |
-
},
|
| 739 |
-
"sep_token": {
|
| 740 |
-
"content": "</s>",
|
| 741 |
-
"lstrip": false,
|
| 742 |
-
"normalized": true,
|
| 743 |
-
"rstrip": false,
|
| 744 |
-
"single_word": false
|
| 745 |
-
},
|
| 746 |
-
"unk_token": {
|
| 747 |
-
"content": "<unk>",
|
| 748 |
-
"lstrip": false,
|
| 749 |
-
"normalized": true,
|
| 750 |
-
"rstrip": false,
|
| 751 |
-
"single_word": false
|
| 752 |
-
}
|
| 753 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/tokenizer.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|
last-checkpoint/tokenizer_config.json
DELETED
|
@@ -1,64 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"add_prefix_space": false,
|
| 3 |
-
"bos_token": {
|
| 4 |
-
"__type": "AddedToken",
|
| 5 |
-
"content": "<s>",
|
| 6 |
-
"lstrip": false,
|
| 7 |
-
"normalized": true,
|
| 8 |
-
"rstrip": false,
|
| 9 |
-
"single_word": false
|
| 10 |
-
},
|
| 11 |
-
"cls_token": {
|
| 12 |
-
"__type": "AddedToken",
|
| 13 |
-
"content": "<s>",
|
| 14 |
-
"lstrip": false,
|
| 15 |
-
"normalized": true,
|
| 16 |
-
"rstrip": false,
|
| 17 |
-
"single_word": false
|
| 18 |
-
},
|
| 19 |
-
"eos_token": {
|
| 20 |
-
"__type": "AddedToken",
|
| 21 |
-
"content": "</s>",
|
| 22 |
-
"lstrip": false,
|
| 23 |
-
"normalized": true,
|
| 24 |
-
"rstrip": false,
|
| 25 |
-
"single_word": false
|
| 26 |
-
},
|
| 27 |
-
"errors": "replace",
|
| 28 |
-
"mask_token": {
|
| 29 |
-
"__type": "AddedToken",
|
| 30 |
-
"content": "<mask>",
|
| 31 |
-
"lstrip": true,
|
| 32 |
-
"normalized": true,
|
| 33 |
-
"rstrip": false,
|
| 34 |
-
"single_word": false
|
| 35 |
-
},
|
| 36 |
-
"model_max_length": 512,
|
| 37 |
-
"pad_token": {
|
| 38 |
-
"__type": "AddedToken",
|
| 39 |
-
"content": "<pad>",
|
| 40 |
-
"lstrip": false,
|
| 41 |
-
"normalized": true,
|
| 42 |
-
"rstrip": false,
|
| 43 |
-
"single_word": false
|
| 44 |
-
},
|
| 45 |
-
"sep_token": {
|
| 46 |
-
"__type": "AddedToken",
|
| 47 |
-
"content": "</s>",
|
| 48 |
-
"lstrip": false,
|
| 49 |
-
"normalized": true,
|
| 50 |
-
"rstrip": false,
|
| 51 |
-
"single_word": false
|
| 52 |
-
},
|
| 53 |
-
"special_tokens_map_file": "/home/s5431786/.cache/huggingface/hub/models--Salesforce--codet5-small/snapshots/a642dc934e5475185369d09ac07091dfe72a31fc/special_tokens_map.json",
|
| 54 |
-
"tokenizer_class": "RobertaTokenizer",
|
| 55 |
-
"trim_offsets": true,
|
| 56 |
-
"unk_token": {
|
| 57 |
-
"__type": "AddedToken",
|
| 58 |
-
"content": "<unk>",
|
| 59 |
-
"lstrip": false,
|
| 60 |
-
"normalized": true,
|
| 61 |
-
"rstrip": false,
|
| 62 |
-
"single_word": false
|
| 63 |
-
}
|
| 64 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/trainer_state.json
DELETED
|
@@ -1,476 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": 0.7621653079986572,
|
| 3 |
-
"best_model_checkpoint": "/home2/s5431786/jb-internship/results/Salesforce-codet5-small-CodeXGLUE-CONCODE-adamw/checkpoint-9000",
|
| 4 |
-
"epoch": 3.68,
|
| 5 |
-
"global_step": 11500,
|
| 6 |
-
"is_hyper_param_search": false,
|
| 7 |
-
"is_local_process_zero": true,
|
| 8 |
-
"is_world_process_zero": true,
|
| 9 |
-
"log_history": [
|
| 10 |
-
{
|
| 11 |
-
"epoch": 0.16,
|
| 12 |
-
"learning_rate": 3.166986564299425e-05,
|
| 13 |
-
"loss": 2.3935,
|
| 14 |
-
"step": 500
|
| 15 |
-
},
|
| 16 |
-
{
|
| 17 |
-
"epoch": 0.16,
|
| 18 |
-
"eval_bleu": 0.11268710945019392,
|
| 19 |
-
"eval_exact_match": 0.129,
|
| 20 |
-
"eval_loss": 0.9724392294883728,
|
| 21 |
-
"eval_rouge1": 0.5286362208212969,
|
| 22 |
-
"eval_rouge2": 0.3465696747006132,
|
| 23 |
-
"eval_rougeL": 0.5098110686689669,
|
| 24 |
-
"eval_rougeLsum": 0.5153482514973529,
|
| 25 |
-
"eval_runtime": 27.4689,
|
| 26 |
-
"eval_samples_per_second": 72.81,
|
| 27 |
-
"eval_steps_per_second": 2.294,
|
| 28 |
-
"step": 500
|
| 29 |
-
},
|
| 30 |
-
{
|
| 31 |
-
"epoch": 0.32,
|
| 32 |
-
"learning_rate": 6.3659628918746e-05,
|
| 33 |
-
"loss": 0.8984,
|
| 34 |
-
"step": 1000
|
| 35 |
-
},
|
| 36 |
-
{
|
| 37 |
-
"epoch": 0.32,
|
| 38 |
-
"eval_bleu": 0.11998407865663271,
|
| 39 |
-
"eval_exact_match": 0.138,
|
| 40 |
-
"eval_loss": 0.8919487595558167,
|
| 41 |
-
"eval_rouge1": 0.546341326360675,
|
| 42 |
-
"eval_rouge2": 0.3714077684410343,
|
| 43 |
-
"eval_rougeL": 0.5285236062967356,
|
| 44 |
-
"eval_rougeLsum": 0.535261365301597,
|
| 45 |
-
"eval_runtime": 26.8866,
|
| 46 |
-
"eval_samples_per_second": 74.386,
|
| 47 |
-
"eval_steps_per_second": 2.343,
|
| 48 |
-
"step": 1000
|
| 49 |
-
},
|
| 50 |
-
{
|
| 51 |
-
"epoch": 0.48,
|
| 52 |
-
"learning_rate": 9.564939219449776e-05,
|
| 53 |
-
"loss": 0.8121,
|
| 54 |
-
"step": 1500
|
| 55 |
-
},
|
| 56 |
-
{
|
| 57 |
-
"epoch": 0.48,
|
| 58 |
-
"eval_bleu": 0.11580804646751057,
|
| 59 |
-
"eval_exact_match": 0.1455,
|
| 60 |
-
"eval_loss": 0.8583337664604187,
|
| 61 |
-
"eval_rouge1": 0.5528659122019346,
|
| 62 |
-
"eval_rouge2": 0.3786855675064588,
|
| 63 |
-
"eval_rougeL": 0.535002514875875,
|
| 64 |
-
"eval_rougeLsum": 0.5425756750433103,
|
| 65 |
-
"eval_runtime": 27.5302,
|
| 66 |
-
"eval_samples_per_second": 72.647,
|
| 67 |
-
"eval_steps_per_second": 2.288,
|
| 68 |
-
"step": 1500
|
| 69 |
-
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 0.64,
|
| 72 |
-
"learning_rate": 9.854481759692796e-05,
|
| 73 |
-
"loss": 0.7598,
|
| 74 |
-
"step": 2000
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 0.64,
|
| 78 |
-
"eval_bleu": 0.11973654523564387,
|
| 79 |
-
"eval_exact_match": 0.1485,
|
| 80 |
-
"eval_loss": 0.8436598181724548,
|
| 81 |
-
"eval_rouge1": 0.5541164072038078,
|
| 82 |
-
"eval_rouge2": 0.3813328094993438,
|
| 83 |
-
"eval_rougeL": 0.5354722356320805,
|
| 84 |
-
"eval_rougeLsum": 0.5432424855167899,
|
| 85 |
-
"eval_runtime": 26.6962,
|
| 86 |
-
"eval_samples_per_second": 74.917,
|
| 87 |
-
"eval_steps_per_second": 2.36,
|
| 88 |
-
"step": 2000
|
| 89 |
-
},
|
| 90 |
-
{
|
| 91 |
-
"epoch": 0.8,
|
| 92 |
-
"learning_rate": 9.686057870448345e-05,
|
| 93 |
-
"loss": 0.7289,
|
| 94 |
-
"step": 2500
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.8,
|
| 98 |
-
"eval_bleu": 0.12215125938884584,
|
| 99 |
-
"eval_exact_match": 0.158,
|
| 100 |
-
"eval_loss": 0.818922758102417,
|
| 101 |
-
"eval_rouge1": 0.5596647043092989,
|
| 102 |
-
"eval_rouge2": 0.390646741130832,
|
| 103 |
-
"eval_rougeL": 0.5415781749603358,
|
| 104 |
-
"eval_rougeLsum": 0.5501483448643409,
|
| 105 |
-
"eval_runtime": 26.4885,
|
| 106 |
-
"eval_samples_per_second": 75.504,
|
| 107 |
-
"eval_steps_per_second": 2.378,
|
| 108 |
-
"step": 2500
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.96,
|
| 112 |
-
"learning_rate": 9.517633981203894e-05,
|
| 113 |
-
"loss": 0.7053,
|
| 114 |
-
"step": 3000
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.96,
|
| 118 |
-
"eval_bleu": 0.12223620743911237,
|
| 119 |
-
"eval_exact_match": 0.161,
|
| 120 |
-
"eval_loss": 0.8145434856414795,
|
| 121 |
-
"eval_rouge1": 0.5571603682618643,
|
| 122 |
-
"eval_rouge2": 0.3887993919896527,
|
| 123 |
-
"eval_rougeL": 0.5392381596924516,
|
| 124 |
-
"eval_rougeLsum": 0.5469429464366221,
|
| 125 |
-
"eval_runtime": 27.992,
|
| 126 |
-
"eval_samples_per_second": 71.449,
|
| 127 |
-
"eval_steps_per_second": 2.251,
|
| 128 |
-
"step": 3000
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 1.12,
|
| 132 |
-
"learning_rate": 9.349210091959444e-05,
|
| 133 |
-
"loss": 0.6544,
|
| 134 |
-
"step": 3500
|
| 135 |
-
},
|
| 136 |
-
{
|
| 137 |
-
"epoch": 1.12,
|
| 138 |
-
"eval_bleu": 0.1260404864365832,
|
| 139 |
-
"eval_exact_match": 0.1565,
|
| 140 |
-
"eval_loss": 0.7981612682342529,
|
| 141 |
-
"eval_rouge1": 0.5606467979991658,
|
| 142 |
-
"eval_rouge2": 0.3920123242296646,
|
| 143 |
-
"eval_rougeL": 0.5435919709288143,
|
| 144 |
-
"eval_rougeLsum": 0.5516785996846396,
|
| 145 |
-
"eval_runtime": 29.5652,
|
| 146 |
-
"eval_samples_per_second": 67.647,
|
| 147 |
-
"eval_steps_per_second": 2.131,
|
| 148 |
-
"step": 3500
|
| 149 |
-
},
|
| 150 |
-
{
|
| 151 |
-
"epoch": 1.28,
|
| 152 |
-
"learning_rate": 9.180786202714994e-05,
|
| 153 |
-
"loss": 0.6334,
|
| 154 |
-
"step": 4000
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 1.28,
|
| 158 |
-
"eval_bleu": 0.12839687243556736,
|
| 159 |
-
"eval_exact_match": 0.1585,
|
| 160 |
-
"eval_loss": 0.7973963618278503,
|
| 161 |
-
"eval_rouge1": 0.5633046028358244,
|
| 162 |
-
"eval_rouge2": 0.3905965243856279,
|
| 163 |
-
"eval_rougeL": 0.5447842726270915,
|
| 164 |
-
"eval_rougeLsum": 0.5529333478079217,
|
| 165 |
-
"eval_runtime": 26.3748,
|
| 166 |
-
"eval_samples_per_second": 75.83,
|
| 167 |
-
"eval_steps_per_second": 2.389,
|
| 168 |
-
"step": 4000
|
| 169 |
-
},
|
| 170 |
-
{
|
| 171 |
-
"epoch": 1.44,
|
| 172 |
-
"learning_rate": 9.012362313470542e-05,
|
| 173 |
-
"loss": 0.6236,
|
| 174 |
-
"step": 4500
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 1.44,
|
| 178 |
-
"eval_bleu": 0.12748409364928656,
|
| 179 |
-
"eval_exact_match": 0.163,
|
| 180 |
-
"eval_loss": 0.7942736744880676,
|
| 181 |
-
"eval_rouge1": 0.563881072938117,
|
| 182 |
-
"eval_rouge2": 0.3930844670404538,
|
| 183 |
-
"eval_rougeL": 0.5455192493710137,
|
| 184 |
-
"eval_rougeLsum": 0.5541560433164279,
|
| 185 |
-
"eval_runtime": 26.3891,
|
| 186 |
-
"eval_samples_per_second": 75.789,
|
| 187 |
-
"eval_steps_per_second": 2.387,
|
| 188 |
-
"step": 4500
|
| 189 |
-
},
|
| 190 |
-
{
|
| 191 |
-
"epoch": 1.6,
|
| 192 |
-
"learning_rate": 8.843938424226093e-05,
|
| 193 |
-
"loss": 0.6221,
|
| 194 |
-
"step": 5000
|
| 195 |
-
},
|
| 196 |
-
{
|
| 197 |
-
"epoch": 1.6,
|
| 198 |
-
"eval_bleu": 0.13095132849637034,
|
| 199 |
-
"eval_exact_match": 0.1655,
|
| 200 |
-
"eval_loss": 0.782397449016571,
|
| 201 |
-
"eval_rouge1": 0.571774543234057,
|
| 202 |
-
"eval_rouge2": 0.40112070812038864,
|
| 203 |
-
"eval_rougeL": 0.5536602781518731,
|
| 204 |
-
"eval_rougeLsum": 0.5621325917699274,
|
| 205 |
-
"eval_runtime": 27.1882,
|
| 206 |
-
"eval_samples_per_second": 73.561,
|
| 207 |
-
"eval_steps_per_second": 2.317,
|
| 208 |
-
"step": 5000
|
| 209 |
-
},
|
| 210 |
-
{
|
| 211 |
-
"epoch": 1.76,
|
| 212 |
-
"learning_rate": 8.675514534981643e-05,
|
| 213 |
-
"loss": 0.608,
|
| 214 |
-
"step": 5500
|
| 215 |
-
},
|
| 216 |
-
{
|
| 217 |
-
"epoch": 1.76,
|
| 218 |
-
"eval_bleu": 0.13144704649097816,
|
| 219 |
-
"eval_exact_match": 0.163,
|
| 220 |
-
"eval_loss": 0.7792329788208008,
|
| 221 |
-
"eval_rouge1": 0.5663801985181786,
|
| 222 |
-
"eval_rouge2": 0.3996756504619762,
|
| 223 |
-
"eval_rougeL": 0.5490225477397204,
|
| 224 |
-
"eval_rougeLsum": 0.556696752293008,
|
| 225 |
-
"eval_runtime": 26.3826,
|
| 226 |
-
"eval_samples_per_second": 75.808,
|
| 227 |
-
"eval_steps_per_second": 2.388,
|
| 228 |
-
"step": 5500
|
| 229 |
-
},
|
| 230 |
-
{
|
| 231 |
-
"epoch": 1.92,
|
| 232 |
-
"learning_rate": 8.507090645737192e-05,
|
| 233 |
-
"loss": 0.5956,
|
| 234 |
-
"step": 6000
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 1.92,
|
| 238 |
-
"eval_bleu": 0.129401969984806,
|
| 239 |
-
"eval_exact_match": 0.1605,
|
| 240 |
-
"eval_loss": 0.7785006165504456,
|
| 241 |
-
"eval_rouge1": 0.5640650180352025,
|
| 242 |
-
"eval_rouge2": 0.39806305672301234,
|
| 243 |
-
"eval_rougeL": 0.5469935550994057,
|
| 244 |
-
"eval_rougeLsum": 0.554556710156975,
|
| 245 |
-
"eval_runtime": 27.3051,
|
| 246 |
-
"eval_samples_per_second": 73.246,
|
| 247 |
-
"eval_steps_per_second": 2.307,
|
| 248 |
-
"step": 6000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 2.08,
|
| 252 |
-
"learning_rate": 8.338666756492741e-05,
|
| 253 |
-
"loss": 0.5701,
|
| 254 |
-
"step": 6500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 2.08,
|
| 258 |
-
"eval_bleu": 0.13363680452673993,
|
| 259 |
-
"eval_exact_match": 0.157,
|
| 260 |
-
"eval_loss": 0.7799766063690186,
|
| 261 |
-
"eval_rouge1": 0.5672994124585793,
|
| 262 |
-
"eval_rouge2": 0.39553178231486497,
|
| 263 |
-
"eval_rougeL": 0.548936720126607,
|
| 264 |
-
"eval_rougeLsum": 0.5568480771592814,
|
| 265 |
-
"eval_runtime": 26.8025,
|
| 266 |
-
"eval_samples_per_second": 74.62,
|
| 267 |
-
"eval_steps_per_second": 2.351,
|
| 268 |
-
"step": 6500
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 2.24,
|
| 272 |
-
"learning_rate": 8.170242867248291e-05,
|
| 273 |
-
"loss": 0.5378,
|
| 274 |
-
"step": 7000
|
| 275 |
-
},
|
| 276 |
-
{
|
| 277 |
-
"epoch": 2.24,
|
| 278 |
-
"eval_bleu": 0.13079743156246348,
|
| 279 |
-
"eval_exact_match": 0.1655,
|
| 280 |
-
"eval_loss": 0.7720439434051514,
|
| 281 |
-
"eval_rouge1": 0.5685604278995763,
|
| 282 |
-
"eval_rouge2": 0.3999533201987808,
|
| 283 |
-
"eval_rougeL": 0.5504327783527796,
|
| 284 |
-
"eval_rougeLsum": 0.5582442604826536,
|
| 285 |
-
"eval_runtime": 26.7858,
|
| 286 |
-
"eval_samples_per_second": 74.666,
|
| 287 |
-
"eval_steps_per_second": 2.352,
|
| 288 |
-
"step": 7000
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"epoch": 2.4,
|
| 292 |
-
"learning_rate": 8.00181897800384e-05,
|
| 293 |
-
"loss": 0.541,
|
| 294 |
-
"step": 7500
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 2.4,
|
| 298 |
-
"eval_bleu": 0.1312928889201776,
|
| 299 |
-
"eval_exact_match": 0.1625,
|
| 300 |
-
"eval_loss": 0.7708559632301331,
|
| 301 |
-
"eval_rouge1": 0.5698993331552187,
|
| 302 |
-
"eval_rouge2": 0.39836218536255497,
|
| 303 |
-
"eval_rougeL": 0.5510518000592355,
|
| 304 |
-
"eval_rougeLsum": 0.5590054429199334,
|
| 305 |
-
"eval_runtime": 26.5491,
|
| 306 |
-
"eval_samples_per_second": 75.332,
|
| 307 |
-
"eval_steps_per_second": 2.373,
|
| 308 |
-
"step": 7500
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"epoch": 2.56,
|
| 312 |
-
"learning_rate": 7.83339508875939e-05,
|
| 313 |
-
"loss": 0.5359,
|
| 314 |
-
"step": 8000
|
| 315 |
-
},
|
| 316 |
-
{
|
| 317 |
-
"epoch": 2.56,
|
| 318 |
-
"eval_bleu": 0.133238582586381,
|
| 319 |
-
"eval_exact_match": 0.164,
|
| 320 |
-
"eval_loss": 0.767312228679657,
|
| 321 |
-
"eval_rouge1": 0.5696523576197161,
|
| 322 |
-
"eval_rouge2": 0.40229167377006236,
|
| 323 |
-
"eval_rougeL": 0.552121263102242,
|
| 324 |
-
"eval_rougeLsum": 0.5600576350734027,
|
| 325 |
-
"eval_runtime": 26.7037,
|
| 326 |
-
"eval_samples_per_second": 74.896,
|
| 327 |
-
"eval_steps_per_second": 2.359,
|
| 328 |
-
"step": 8000
|
| 329 |
-
},
|
| 330 |
-
{
|
| 331 |
-
"epoch": 2.72,
|
| 332 |
-
"learning_rate": 7.664971199514939e-05,
|
| 333 |
-
"loss": 0.5322,
|
| 334 |
-
"step": 8500
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"epoch": 2.72,
|
| 338 |
-
"eval_bleu": 0.13498312180179964,
|
| 339 |
-
"eval_exact_match": 0.1665,
|
| 340 |
-
"eval_loss": 0.7641937732696533,
|
| 341 |
-
"eval_rouge1": 0.5707696930577069,
|
| 342 |
-
"eval_rouge2": 0.403346138073659,
|
| 343 |
-
"eval_rougeL": 0.5527280072901636,
|
| 344 |
-
"eval_rougeLsum": 0.5605579795104842,
|
| 345 |
-
"eval_runtime": 26.4882,
|
| 346 |
-
"eval_samples_per_second": 75.505,
|
| 347 |
-
"eval_steps_per_second": 2.378,
|
| 348 |
-
"step": 8500
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 2.88,
|
| 352 |
-
"learning_rate": 7.496547310270489e-05,
|
| 353 |
-
"loss": 0.5387,
|
| 354 |
-
"step": 9000
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 2.88,
|
| 358 |
-
"eval_bleu": 0.13416837075728422,
|
| 359 |
-
"eval_exact_match": 0.159,
|
| 360 |
-
"eval_loss": 0.7621653079986572,
|
| 361 |
-
"eval_rouge1": 0.5671758743776842,
|
| 362 |
-
"eval_rouge2": 0.3987572074969532,
|
| 363 |
-
"eval_rougeL": 0.5499787838213916,
|
| 364 |
-
"eval_rougeLsum": 0.55728689293535,
|
| 365 |
-
"eval_runtime": 26.7851,
|
| 366 |
-
"eval_samples_per_second": 74.668,
|
| 367 |
-
"eval_steps_per_second": 2.352,
|
| 368 |
-
"step": 9000
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 3.04,
|
| 372 |
-
"learning_rate": 7.328123421026039e-05,
|
| 373 |
-
"loss": 0.514,
|
| 374 |
-
"step": 9500
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 3.04,
|
| 378 |
-
"eval_bleu": 0.13519274318497243,
|
| 379 |
-
"eval_exact_match": 0.166,
|
| 380 |
-
"eval_loss": 0.7700380682945251,
|
| 381 |
-
"eval_rouge1": 0.572246331088752,
|
| 382 |
-
"eval_rouge2": 0.40519052168186803,
|
| 383 |
-
"eval_rougeL": 0.5545858464982438,
|
| 384 |
-
"eval_rougeLsum": 0.5618015934184992,
|
| 385 |
-
"eval_runtime": 26.4898,
|
| 386 |
-
"eval_samples_per_second": 75.501,
|
| 387 |
-
"eval_steps_per_second": 2.378,
|
| 388 |
-
"step": 9500
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 3.2,
|
| 392 |
-
"learning_rate": 7.159699531781589e-05,
|
| 393 |
-
"loss": 0.4895,
|
| 394 |
-
"step": 10000
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 3.2,
|
| 398 |
-
"eval_bleu": 0.13587785975036965,
|
| 399 |
-
"eval_exact_match": 0.1615,
|
| 400 |
-
"eval_loss": 0.7675954699516296,
|
| 401 |
-
"eval_rouge1": 0.5695980848985347,
|
| 402 |
-
"eval_rouge2": 0.40156815801776047,
|
| 403 |
-
"eval_rougeL": 0.551645493171788,
|
| 404 |
-
"eval_rougeLsum": 0.5591021610099983,
|
| 405 |
-
"eval_runtime": 28.9264,
|
| 406 |
-
"eval_samples_per_second": 69.141,
|
| 407 |
-
"eval_steps_per_second": 2.178,
|
| 408 |
-
"step": 10000
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 3.36,
|
| 412 |
-
"learning_rate": 6.991275642537137e-05,
|
| 413 |
-
"loss": 0.4827,
|
| 414 |
-
"step": 10500
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 3.36,
|
| 418 |
-
"eval_bleu": 0.1366806488179042,
|
| 419 |
-
"eval_exact_match": 0.162,
|
| 420 |
-
"eval_loss": 0.7664822936058044,
|
| 421 |
-
"eval_rouge1": 0.5756384334726231,
|
| 422 |
-
"eval_rouge2": 0.4071922211902318,
|
| 423 |
-
"eval_rougeL": 0.557732929347675,
|
| 424 |
-
"eval_rougeLsum": 0.5655550288614574,
|
| 425 |
-
"eval_runtime": 28.1804,
|
| 426 |
-
"eval_samples_per_second": 70.971,
|
| 427 |
-
"eval_steps_per_second": 2.236,
|
| 428 |
-
"step": 10500
|
| 429 |
-
},
|
| 430 |
-
{
|
| 431 |
-
"epoch": 3.52,
|
| 432 |
-
"learning_rate": 6.822851753292687e-05,
|
| 433 |
-
"loss": 0.4814,
|
| 434 |
-
"step": 11000
|
| 435 |
-
},
|
| 436 |
-
{
|
| 437 |
-
"epoch": 3.52,
|
| 438 |
-
"eval_bleu": 0.13337542765383828,
|
| 439 |
-
"eval_exact_match": 0.1605,
|
| 440 |
-
"eval_loss": 0.7700090408325195,
|
| 441 |
-
"eval_rouge1": 0.5708999247293218,
|
| 442 |
-
"eval_rouge2": 0.4026343245543057,
|
| 443 |
-
"eval_rougeL": 0.5528368884976037,
|
| 444 |
-
"eval_rougeLsum": 0.5605321832570869,
|
| 445 |
-
"eval_runtime": 27.6804,
|
| 446 |
-
"eval_samples_per_second": 72.253,
|
| 447 |
-
"eval_steps_per_second": 2.276,
|
| 448 |
-
"step": 11000
|
| 449 |
-
},
|
| 450 |
-
{
|
| 451 |
-
"epoch": 3.68,
|
| 452 |
-
"learning_rate": 6.654427864048237e-05,
|
| 453 |
-
"loss": 0.4847,
|
| 454 |
-
"step": 11500
|
| 455 |
-
},
|
| 456 |
-
{
|
| 457 |
-
"epoch": 3.68,
|
| 458 |
-
"eval_bleu": 0.13346084775685726,
|
| 459 |
-
"eval_exact_match": 0.163,
|
| 460 |
-
"eval_loss": 0.7665703296661377,
|
| 461 |
-
"eval_rouge1": 0.5716177390806905,
|
| 462 |
-
"eval_rouge2": 0.4045876918243749,
|
| 463 |
-
"eval_rougeL": 0.553575747159084,
|
| 464 |
-
"eval_rougeLsum": 0.561417118599018,
|
| 465 |
-
"eval_runtime": 26.5079,
|
| 466 |
-
"eval_samples_per_second": 75.449,
|
| 467 |
-
"eval_steps_per_second": 2.377,
|
| 468 |
-
"step": 11500
|
| 469 |
-
}
|
| 470 |
-
],
|
| 471 |
-
"max_steps": 31250,
|
| 472 |
-
"num_train_epochs": 10,
|
| 473 |
-
"total_flos": 4.890379750578586e+16,
|
| 474 |
-
"trial_name": null,
|
| 475 |
-
"trial_params": null
|
| 476 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/training_args.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:7a29fb8466ad38b6da30b44f49a9f1c4c5fb10cf12d549ef6042b665fc563462
|
| 3 |
-
size 3887
|
|
|
|
|
|
|
|
|
|
|
|
last-checkpoint/vocab.json
DELETED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 242012923
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:236ac2a99de8898e9adfad014e34dc559fd10b5c876168a4128b74cab112d650
|
| 3 |
size 242012923
|