Upload 6 files
Browse files- config.json +1 -0
- model.safetensors +1 -1
- trainer_state.json +410 -193
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"architectures": [
|
| 3 |
"VisionEncoderDecoderModel"
|
| 4 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "kms7530/ko-coco",
|
| 3 |
"architectures": [
|
| 4 |
"VisionEncoderDecoderModel"
|
| 5 |
],
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 901697272
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ae809b41c667fa9b9b0fcce2d610db8925c62d14621efb6401bc596ffaeace8
|
| 3 |
size 901697272
|
trainer_state.json
CHANGED
|
@@ -1,306 +1,523 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"eval_steps":
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"grad_norm": 0.
|
| 14 |
-
"learning_rate": 4.
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm": 0.
|
| 21 |
-
"learning_rate": 4.
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm": 0.
|
| 28 |
-
"learning_rate": 4.
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
-
"learning_rate": 4.
|
| 36 |
-
"loss": 0.
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
-
"learning_rate": 4.
|
| 43 |
-
"loss": 0.
|
| 44 |
-
"step": 500
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.16249593760155995,
|
| 48 |
-
"eval_gen_len": 19.964290067032298,
|
| 49 |
-
"eval_loss": 0.18788783252239227,
|
| 50 |
-
"eval_rouge1": 0.0068,
|
| 51 |
-
"eval_rouge2": 0.0,
|
| 52 |
-
"eval_rougeL": 0.0068,
|
| 53 |
-
"eval_rougeLsum": 0.0068,
|
| 54 |
-
"eval_runtime": 2174.226,
|
| 55 |
-
"eval_samples_per_second": 11.321,
|
| 56 |
-
"eval_steps_per_second": 0.354,
|
| 57 |
"step": 500
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"grad_norm": 0.
|
| 62 |
-
"learning_rate": 4.
|
| 63 |
-
"loss": 0.
|
| 64 |
"step": 600
|
| 65 |
},
|
| 66 |
{
|
| 67 |
-
"epoch": 0.
|
| 68 |
-
"grad_norm": 0.
|
| 69 |
-
"learning_rate":
|
| 70 |
-
"loss": 0.
|
| 71 |
"step": 700
|
| 72 |
},
|
| 73 |
{
|
| 74 |
-
"epoch": 0.
|
| 75 |
-
"grad_norm": 0.
|
| 76 |
-
"learning_rate":
|
| 77 |
-
"loss": 0.
|
| 78 |
"step": 800
|
| 79 |
},
|
| 80 |
{
|
| 81 |
-
"epoch": 0.
|
| 82 |
-
"grad_norm": 0.
|
| 83 |
-
"learning_rate":
|
| 84 |
-
"loss": 0.
|
| 85 |
"step": 900
|
| 86 |
},
|
| 87 |
{
|
| 88 |
-
"epoch": 0.
|
| 89 |
-
"grad_norm": 0.
|
| 90 |
-
"learning_rate":
|
| 91 |
-
"loss": 0.
|
| 92 |
"step": 1000
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
-
"eval_gen_len": 19.
|
| 97 |
-
"eval_loss": 0.
|
| 98 |
-
"eval_rouge1": 0.
|
| 99 |
-
"eval_rouge2": 0.
|
| 100 |
-
"eval_rougeL": 0.
|
| 101 |
-
"eval_rougeLsum": 0.
|
| 102 |
-
"eval_runtime":
|
| 103 |
-
"eval_samples_per_second": 11.
|
| 104 |
-
"eval_steps_per_second": 0.
|
| 105 |
"step": 1000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch": 0.
|
| 109 |
-
"grad_norm": 0.
|
| 110 |
-
"learning_rate":
|
| 111 |
-
"loss": 0.
|
| 112 |
"step": 1100
|
| 113 |
},
|
| 114 |
{
|
| 115 |
-
"epoch": 0.
|
| 116 |
-
"grad_norm": 0.
|
| 117 |
-
"learning_rate":
|
| 118 |
-
"loss": 0.
|
| 119 |
"step": 1200
|
| 120 |
},
|
| 121 |
{
|
| 122 |
-
"epoch": 0.
|
| 123 |
-
"grad_norm": 0.
|
| 124 |
-
"learning_rate":
|
| 125 |
-
"loss": 0.
|
| 126 |
"step": 1300
|
| 127 |
},
|
| 128 |
{
|
| 129 |
-
"epoch": 0.
|
| 130 |
-
"grad_norm": 0.
|
| 131 |
-
"learning_rate":
|
| 132 |
-
"loss": 0.
|
| 133 |
"step": 1400
|
| 134 |
},
|
| 135 |
{
|
| 136 |
-
"epoch": 0.
|
| 137 |
-
"grad_norm": 0.
|
| 138 |
-
"learning_rate":
|
| 139 |
-
"loss": 0.
|
| 140 |
"step": 1500
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch": 0.
|
| 144 |
-
"
|
| 145 |
-
"
|
| 146 |
-
"
|
| 147 |
-
"eval_rouge2": 0.0,
|
| 148 |
-
"eval_rougeL": 0.1828,
|
| 149 |
-
"eval_rougeLsum": 0.1851,
|
| 150 |
-
"eval_runtime": 2167.8807,
|
| 151 |
-
"eval_samples_per_second": 11.354,
|
| 152 |
-
"eval_steps_per_second": 0.355,
|
| 153 |
-
"step": 1500
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.5199870003249919,
|
| 157 |
-
"grad_norm": 0.2645615041255951,
|
| 158 |
-
"learning_rate": 2.4000649983750407e-05,
|
| 159 |
-
"loss": 0.1794,
|
| 160 |
"step": 1600
|
| 161 |
},
|
| 162 |
{
|
| 163 |
-
"epoch": 0.
|
| 164 |
-
"grad_norm": 0.
|
| 165 |
-
"learning_rate":
|
| 166 |
-
"loss": 0.
|
| 167 |
"step": 1700
|
| 168 |
},
|
| 169 |
{
|
| 170 |
-
"epoch": 0.
|
| 171 |
-
"grad_norm": 0.
|
| 172 |
-
"learning_rate":
|
| 173 |
-
"loss": 0.
|
| 174 |
"step": 1800
|
| 175 |
},
|
| 176 |
{
|
| 177 |
-
"epoch": 0.
|
| 178 |
-
"grad_norm": 0.
|
| 179 |
-
"learning_rate":
|
| 180 |
-
"loss": 0.
|
| 181 |
"step": 1900
|
| 182 |
},
|
| 183 |
{
|
| 184 |
-
"epoch": 0.
|
| 185 |
-
"grad_norm": 0.
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss": 0.
|
| 188 |
"step": 2000
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch": 0.
|
| 192 |
-
"eval_gen_len": 19.
|
| 193 |
-
"eval_loss": 0.
|
| 194 |
-
"eval_rouge1": 0.
|
| 195 |
"eval_rouge2": 0.0,
|
| 196 |
-
"eval_rougeL": 0.
|
| 197 |
-
"eval_rougeLsum": 0.
|
| 198 |
-
"eval_runtime":
|
| 199 |
-
"eval_samples_per_second": 11.
|
| 200 |
-
"eval_steps_per_second": 0.
|
| 201 |
"step": 2000
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch": 0.
|
| 205 |
-
"grad_norm": 0.
|
| 206 |
-
"learning_rate":
|
| 207 |
-
"loss": 0.
|
| 208 |
"step": 2100
|
| 209 |
},
|
| 210 |
{
|
| 211 |
-
"epoch": 0.
|
| 212 |
-
"grad_norm": 0.
|
| 213 |
-
"learning_rate":
|
| 214 |
-
"loss": 0.
|
| 215 |
"step": 2200
|
| 216 |
},
|
| 217 |
{
|
| 218 |
-
"epoch": 0.
|
| 219 |
-
"grad_norm": 0.
|
| 220 |
-
"learning_rate":
|
| 221 |
-
"loss": 0.
|
| 222 |
"step": 2300
|
| 223 |
},
|
| 224 |
{
|
| 225 |
-
"epoch": 0.
|
| 226 |
-
"grad_norm": 0.
|
| 227 |
-
"learning_rate":
|
| 228 |
-
"loss": 0.
|
| 229 |
"step": 2400
|
| 230 |
},
|
| 231 |
{
|
| 232 |
-
"epoch": 0.
|
| 233 |
-
"grad_norm": 0.
|
| 234 |
-
"learning_rate":
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 2500
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch": 0.
|
| 240 |
-
"
|
| 241 |
-
"
|
| 242 |
-
"
|
| 243 |
-
"eval_rouge2": 0.0,
|
| 244 |
-
"eval_rougeL": 0.0027,
|
| 245 |
-
"eval_rougeLsum": 0.0027,
|
| 246 |
-
"eval_runtime": 2163.6252,
|
| 247 |
-
"eval_samples_per_second": 11.377,
|
| 248 |
-
"eval_steps_per_second": 0.356,
|
| 249 |
-
"step": 2500
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.8449788755281118,
|
| 253 |
-
"grad_norm": 0.23773913085460663,
|
| 254 |
-
"learning_rate": 7.75105622359441e-06,
|
| 255 |
-
"loss": 0.1733,
|
| 256 |
"step": 2600
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"epoch": 0.
|
| 260 |
-
"grad_norm": 0.
|
| 261 |
-
"learning_rate":
|
| 262 |
-
"loss": 0.
|
| 263 |
"step": 2700
|
| 264 |
},
|
| 265 |
{
|
| 266 |
-
"epoch": 0.
|
| 267 |
-
"grad_norm": 0.
|
| 268 |
-
"learning_rate":
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 2800
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch": 0.
|
| 274 |
-
"grad_norm": 0.
|
| 275 |
-
"learning_rate": 2.
|
| 276 |
-
"loss": 0.
|
| 277 |
"step": 2900
|
| 278 |
},
|
| 279 |
{
|
| 280 |
-
"epoch": 0.
|
| 281 |
-
"grad_norm": 0.
|
| 282 |
-
"learning_rate":
|
| 283 |
-
"loss": 0.
|
| 284 |
"step": 3000
|
| 285 |
},
|
| 286 |
{
|
| 287 |
-
"epoch": 0.
|
| 288 |
-
"eval_gen_len": 19.
|
| 289 |
-
"eval_loss": 0.
|
| 290 |
-
"eval_rouge1": 0.
|
| 291 |
"eval_rouge2": 0.0,
|
| 292 |
-
"eval_rougeL": 0.
|
| 293 |
-
"eval_rougeLsum": 0.
|
| 294 |
-
"eval_runtime":
|
| 295 |
-
"eval_samples_per_second": 11.
|
| 296 |
-
"eval_steps_per_second": 0.
|
| 297 |
"step": 3000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 298 |
}
|
| 299 |
],
|
| 300 |
"logging_steps": 100,
|
| 301 |
-
"max_steps":
|
| 302 |
"num_input_tokens_seen": 0,
|
| 303 |
-
"num_train_epochs":
|
| 304 |
"save_steps": 500,
|
| 305 |
"stateful_callbacks": {
|
| 306 |
"TrainerControl": {
|
|
@@ -314,7 +531,7 @@
|
|
| 314 |
"attributes": {}
|
| 315 |
}
|
| 316 |
},
|
| 317 |
-
"total_flos":
|
| 318 |
"train_batch_size": 32,
|
| 319 |
"trial_name": null,
|
| 320 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 1000,
|
| 6 |
+
"global_step": 6166,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.032435939020434644,
|
| 13 |
+
"grad_norm": 0.35351359844207764,
|
| 14 |
+
"learning_rate": 4.918910152448913e-05,
|
| 15 |
+
"loss": 0.1337,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.06487187804086929,
|
| 20 |
+
"grad_norm": 0.2617332935333252,
|
| 21 |
+
"learning_rate": 4.837820304897827e-05,
|
| 22 |
+
"loss": 0.1762,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.09730781706130393,
|
| 27 |
+
"grad_norm": 0.26761409640312195,
|
| 28 |
+
"learning_rate": 4.7567304573467406e-05,
|
| 29 |
+
"loss": 0.1729,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.12974375608173858,
|
| 34 |
+
"grad_norm": 0.2583613991737366,
|
| 35 |
+
"learning_rate": 4.675640609795654e-05,
|
| 36 |
+
"loss": 0.1728,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.1621796951021732,
|
| 41 |
+
"grad_norm": 0.24968299269676208,
|
| 42 |
+
"learning_rate": 4.594550762244567e-05,
|
| 43 |
+
"loss": 0.1721,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.19461563412260785,
|
| 48 |
+
"grad_norm": 0.252204567193985,
|
| 49 |
+
"learning_rate": 4.513460914693481e-05,
|
| 50 |
+
"loss": 0.1729,
|
| 51 |
"step": 600
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.2270515731430425,
|
| 55 |
+
"grad_norm": 0.25790390372276306,
|
| 56 |
+
"learning_rate": 4.432371067142394e-05,
|
| 57 |
+
"loss": 0.1716,
|
| 58 |
"step": 700
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.25948751216347715,
|
| 62 |
+
"grad_norm": 0.2649690508842468,
|
| 63 |
+
"learning_rate": 4.351281219591307e-05,
|
| 64 |
+
"loss": 0.1726,
|
| 65 |
"step": 800
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.2919234511839118,
|
| 69 |
+
"grad_norm": 0.2653878927230835,
|
| 70 |
+
"learning_rate": 4.270191372040221e-05,
|
| 71 |
+
"loss": 0.172,
|
| 72 |
"step": 900
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.3243593902043464,
|
| 76 |
+
"grad_norm": 0.2365722954273224,
|
| 77 |
+
"learning_rate": 4.1891015244891344e-05,
|
| 78 |
+
"loss": 0.1734,
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.3243593902043464,
|
| 83 |
+
"eval_gen_len": 19.97732987265796,
|
| 84 |
+
"eval_loss": 0.17071112990379333,
|
| 85 |
+
"eval_rouge1": 0.146,
|
| 86 |
+
"eval_rouge2": 0.0,
|
| 87 |
+
"eval_rougeL": 0.146,
|
| 88 |
+
"eval_rougeLsum": 0.1477,
|
| 89 |
+
"eval_runtime": 2204.8687,
|
| 90 |
+
"eval_samples_per_second": 11.183,
|
| 91 |
+
"eval_steps_per_second": 0.35,
|
| 92 |
"step": 1000
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.3567953292247811,
|
| 96 |
+
"grad_norm": 0.2285241037607193,
|
| 97 |
+
"learning_rate": 4.108011676938048e-05,
|
| 98 |
+
"loss": 0.1721,
|
| 99 |
"step": 1100
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.3892312682452157,
|
| 103 |
+
"grad_norm": 0.2493802011013031,
|
| 104 |
+
"learning_rate": 4.026921829386961e-05,
|
| 105 |
+
"loss": 0.1694,
|
| 106 |
"step": 1200
|
| 107 |
},
|
| 108 |
{
|
| 109 |
+
"epoch": 0.4216672072656503,
|
| 110 |
+
"grad_norm": 0.25236544013023376,
|
| 111 |
+
"learning_rate": 3.945831981835874e-05,
|
| 112 |
+
"loss": 0.1698,
|
| 113 |
"step": 1300
|
| 114 |
},
|
| 115 |
{
|
| 116 |
+
"epoch": 0.454103146286085,
|
| 117 |
+
"grad_norm": 0.243851900100708,
|
| 118 |
+
"learning_rate": 3.864742134284788e-05,
|
| 119 |
+
"loss": 0.1683,
|
| 120 |
"step": 1400
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"epoch": 0.4865390853065196,
|
| 124 |
+
"grad_norm": 0.24125412106513977,
|
| 125 |
+
"learning_rate": 3.783652286733701e-05,
|
| 126 |
+
"loss": 0.1694,
|
| 127 |
"step": 1500
|
| 128 |
},
|
| 129 |
{
|
| 130 |
+
"epoch": 0.5189750243269543,
|
| 131 |
+
"grad_norm": 0.25099146366119385,
|
| 132 |
+
"learning_rate": 3.7025624391826144e-05,
|
| 133 |
+
"loss": 0.1669,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 134 |
"step": 1600
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.5514109633473889,
|
| 138 |
+
"grad_norm": 0.23592032492160797,
|
| 139 |
+
"learning_rate": 3.621472591631528e-05,
|
| 140 |
+
"loss": 0.1668,
|
| 141 |
"step": 1700
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.5838469023678236,
|
| 145 |
+
"grad_norm": 0.25207582116127014,
|
| 146 |
+
"learning_rate": 3.540382744080442e-05,
|
| 147 |
+
"loss": 0.1678,
|
| 148 |
"step": 1800
|
| 149 |
},
|
| 150 |
{
|
| 151 |
+
"epoch": 0.6162828413882582,
|
| 152 |
+
"grad_norm": 0.2521522045135498,
|
| 153 |
+
"learning_rate": 3.459292896529355e-05,
|
| 154 |
+
"loss": 0.166,
|
| 155 |
"step": 1900
|
| 156 |
},
|
| 157 |
{
|
| 158 |
+
"epoch": 0.6487187804086928,
|
| 159 |
+
"grad_norm": 0.2497226446866989,
|
| 160 |
+
"learning_rate": 3.378203048978268e-05,
|
| 161 |
+
"loss": 0.169,
|
| 162 |
"step": 2000
|
| 163 |
},
|
| 164 |
{
|
| 165 |
+
"epoch": 0.6487187804086928,
|
| 166 |
+
"eval_gen_len": 19.993389569308135,
|
| 167 |
+
"eval_loss": 0.1660359501838684,
|
| 168 |
+
"eval_rouge1": 0.1899,
|
| 169 |
"eval_rouge2": 0.0,
|
| 170 |
+
"eval_rougeL": 0.1899,
|
| 171 |
+
"eval_rougeLsum": 0.1913,
|
| 172 |
+
"eval_runtime": 2195.0671,
|
| 173 |
+
"eval_samples_per_second": 11.233,
|
| 174 |
+
"eval_steps_per_second": 0.351,
|
| 175 |
"step": 2000
|
| 176 |
},
|
| 177 |
{
|
| 178 |
+
"epoch": 0.6811547194291274,
|
| 179 |
+
"grad_norm": 0.25921228528022766,
|
| 180 |
+
"learning_rate": 3.2971132014271815e-05,
|
| 181 |
+
"loss": 0.1663,
|
| 182 |
"step": 2100
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.7135906584495622,
|
| 186 |
+
"grad_norm": 0.22675295174121857,
|
| 187 |
+
"learning_rate": 3.2160233538760945e-05,
|
| 188 |
+
"loss": 0.1661,
|
| 189 |
"step": 2200
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 0.7460265974699968,
|
| 193 |
+
"grad_norm": 0.253683477640152,
|
| 194 |
+
"learning_rate": 3.134933506325008e-05,
|
| 195 |
+
"loss": 0.1661,
|
| 196 |
"step": 2300
|
| 197 |
},
|
| 198 |
{
|
| 199 |
+
"epoch": 0.7784625364904314,
|
| 200 |
+
"grad_norm": 0.23699304461479187,
|
| 201 |
+
"learning_rate": 3.053843658773922e-05,
|
| 202 |
+
"loss": 0.1664,
|
| 203 |
"step": 2400
|
| 204 |
},
|
| 205 |
{
|
| 206 |
+
"epoch": 0.810898475510866,
|
| 207 |
+
"grad_norm": 0.21723595261573792,
|
| 208 |
+
"learning_rate": 2.972753811222835e-05,
|
| 209 |
+
"loss": 0.1628,
|
| 210 |
"step": 2500
|
| 211 |
},
|
| 212 |
{
|
| 213 |
+
"epoch": 0.8433344145313006,
|
| 214 |
+
"grad_norm": 0.2416258603334427,
|
| 215 |
+
"learning_rate": 2.8916639636717486e-05,
|
| 216 |
+
"loss": 0.1684,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 217 |
"step": 2600
|
| 218 |
},
|
| 219 |
{
|
| 220 |
+
"epoch": 0.8757703535517353,
|
| 221 |
+
"grad_norm": 0.24152766168117523,
|
| 222 |
+
"learning_rate": 2.8105741161206616e-05,
|
| 223 |
+
"loss": 0.1655,
|
| 224 |
"step": 2700
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"epoch": 0.90820629257217,
|
| 228 |
+
"grad_norm": 0.23950466513633728,
|
| 229 |
+
"learning_rate": 2.7294842685695753e-05,
|
| 230 |
+
"loss": 0.1662,
|
| 231 |
"step": 2800
|
| 232 |
},
|
| 233 |
{
|
| 234 |
+
"epoch": 0.9406422315926046,
|
| 235 |
+
"grad_norm": 0.2509647011756897,
|
| 236 |
+
"learning_rate": 2.6483944210184886e-05,
|
| 237 |
+
"loss": 0.1642,
|
| 238 |
"step": 2900
|
| 239 |
},
|
| 240 |
{
|
| 241 |
+
"epoch": 0.9730781706130393,
|
| 242 |
+
"grad_norm": 0.24778147041797638,
|
| 243 |
+
"learning_rate": 2.5673045734674023e-05,
|
| 244 |
+
"loss": 0.1628,
|
| 245 |
"step": 3000
|
| 246 |
},
|
| 247 |
{
|
| 248 |
+
"epoch": 0.9730781706130393,
|
| 249 |
+
"eval_gen_len": 19.997323383891636,
|
| 250 |
+
"eval_loss": 0.16280025243759155,
|
| 251 |
+
"eval_rouge1": 0.0561,
|
| 252 |
"eval_rouge2": 0.0,
|
| 253 |
+
"eval_rougeL": 0.0561,
|
| 254 |
+
"eval_rougeLsum": 0.0547,
|
| 255 |
+
"eval_runtime": 2183.6993,
|
| 256 |
+
"eval_samples_per_second": 11.292,
|
| 257 |
+
"eval_steps_per_second": 0.353,
|
| 258 |
"step": 3000
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 1.0055141096334739,
|
| 262 |
+
"grad_norm": 0.22274711728096008,
|
| 263 |
+
"learning_rate": 2.4862147259163153e-05,
|
| 264 |
+
"loss": 0.1629,
|
| 265 |
+
"step": 3100
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 1.0379500486539086,
|
| 269 |
+
"grad_norm": 0.23203332722187042,
|
| 270 |
+
"learning_rate": 2.405124878365229e-05,
|
| 271 |
+
"loss": 0.1556,
|
| 272 |
+
"step": 3200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.0703859876743431,
|
| 276 |
+
"grad_norm": 0.24074462056159973,
|
| 277 |
+
"learning_rate": 2.3240350308141423e-05,
|
| 278 |
+
"loss": 0.1548,
|
| 279 |
+
"step": 3300
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.1028219266947779,
|
| 283 |
+
"grad_norm": 0.23876875638961792,
|
| 284 |
+
"learning_rate": 2.2429451832630553e-05,
|
| 285 |
+
"loss": 0.1535,
|
| 286 |
+
"step": 3400
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 1.1352578657152124,
|
| 290 |
+
"grad_norm": 0.23867164552211761,
|
| 291 |
+
"learning_rate": 2.161855335711969e-05,
|
| 292 |
+
"loss": 0.1528,
|
| 293 |
+
"step": 3500
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 1.167693804735647,
|
| 297 |
+
"grad_norm": 0.2626864016056061,
|
| 298 |
+
"learning_rate": 2.0807654881608824e-05,
|
| 299 |
+
"loss": 0.1526,
|
| 300 |
+
"step": 3600
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 1.2001297437560818,
|
| 304 |
+
"grad_norm": 0.26715072989463806,
|
| 305 |
+
"learning_rate": 1.9996756406097957e-05,
|
| 306 |
+
"loss": 0.1526,
|
| 307 |
+
"step": 3700
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 1.2325656827765163,
|
| 311 |
+
"grad_norm": 0.2496395856142044,
|
| 312 |
+
"learning_rate": 1.9185857930587094e-05,
|
| 313 |
+
"loss": 0.1518,
|
| 314 |
+
"step": 3800
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.265001621796951,
|
| 318 |
+
"grad_norm": 0.24588952958583832,
|
| 319 |
+
"learning_rate": 1.8374959455076227e-05,
|
| 320 |
+
"loss": 0.1519,
|
| 321 |
+
"step": 3900
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.2974375608173856,
|
| 325 |
+
"grad_norm": 0.23679636418819427,
|
| 326 |
+
"learning_rate": 1.7564060979565357e-05,
|
| 327 |
+
"loss": 0.1504,
|
| 328 |
+
"step": 4000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 1.2974375608173856,
|
| 332 |
+
"eval_gen_len": 19.98292643361181,
|
| 333 |
+
"eval_loss": 0.16194650530815125,
|
| 334 |
+
"eval_rouge1": 0.1229,
|
| 335 |
+
"eval_rouge2": 0.0041,
|
| 336 |
+
"eval_rougeL": 0.123,
|
| 337 |
+
"eval_rougeLsum": 0.1233,
|
| 338 |
+
"eval_runtime": 2185.3639,
|
| 339 |
+
"eval_samples_per_second": 11.283,
|
| 340 |
+
"eval_steps_per_second": 0.353,
|
| 341 |
+
"step": 4000
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.3298734998378203,
|
| 345 |
+
"grad_norm": 0.24863706529140472,
|
| 346 |
+
"learning_rate": 1.6753162504054494e-05,
|
| 347 |
+
"loss": 0.1511,
|
| 348 |
+
"step": 4100
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.3623094388582548,
|
| 352 |
+
"grad_norm": 0.2532482445240021,
|
| 353 |
+
"learning_rate": 1.5942264028543628e-05,
|
| 354 |
+
"loss": 0.1503,
|
| 355 |
+
"step": 4200
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 1.3947453778786896,
|
| 359 |
+
"grad_norm": 0.2682236135005951,
|
| 360 |
+
"learning_rate": 1.5131365553032761e-05,
|
| 361 |
+
"loss": 0.1529,
|
| 362 |
+
"step": 4300
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.4271813168991243,
|
| 366 |
+
"grad_norm": 0.23846983909606934,
|
| 367 |
+
"learning_rate": 1.4320467077521896e-05,
|
| 368 |
+
"loss": 0.1514,
|
| 369 |
+
"step": 4400
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.4596172559195588,
|
| 373 |
+
"grad_norm": 0.2428126037120819,
|
| 374 |
+
"learning_rate": 1.350956860201103e-05,
|
| 375 |
+
"loss": 0.1486,
|
| 376 |
+
"step": 4500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.4920531949399936,
|
| 380 |
+
"grad_norm": 0.25097399950027466,
|
| 381 |
+
"learning_rate": 1.2698670126500161e-05,
|
| 382 |
+
"loss": 0.15,
|
| 383 |
+
"step": 4600
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.524489133960428,
|
| 387 |
+
"grad_norm": 0.2583732008934021,
|
| 388 |
+
"learning_rate": 1.1887771650989297e-05,
|
| 389 |
+
"loss": 0.1528,
|
| 390 |
+
"step": 4700
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.5569250729808628,
|
| 394 |
+
"grad_norm": 0.23872007429599762,
|
| 395 |
+
"learning_rate": 1.107687317547843e-05,
|
| 396 |
+
"loss": 0.1486,
|
| 397 |
+
"step": 4800
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.5893610120012975,
|
| 401 |
+
"grad_norm": 0.2473708838224411,
|
| 402 |
+
"learning_rate": 1.0265974699967565e-05,
|
| 403 |
+
"loss": 0.1493,
|
| 404 |
+
"step": 4900
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.621796951021732,
|
| 408 |
+
"grad_norm": 0.24233025312423706,
|
| 409 |
+
"learning_rate": 9.455076224456699e-06,
|
| 410 |
+
"loss": 0.1505,
|
| 411 |
+
"step": 5000
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.621796951021732,
|
| 415 |
+
"eval_gen_len": 19.982034228242355,
|
| 416 |
+
"eval_loss": 0.16026677191257477,
|
| 417 |
+
"eval_rouge1": 0.0744,
|
| 418 |
+
"eval_rouge2": 0.0041,
|
| 419 |
+
"eval_rougeL": 0.0744,
|
| 420 |
+
"eval_rougeLsum": 0.0744,
|
| 421 |
+
"eval_runtime": 2176.6648,
|
| 422 |
+
"eval_samples_per_second": 11.328,
|
| 423 |
+
"eval_steps_per_second": 0.354,
|
| 424 |
+
"step": 5000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.6542328900421666,
|
| 428 |
+
"grad_norm": 0.25053468346595764,
|
| 429 |
+
"learning_rate": 8.644177748945832e-06,
|
| 430 |
+
"loss": 0.1508,
|
| 431 |
+
"step": 5100
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.6866688290626013,
|
| 435 |
+
"grad_norm": 0.24528223276138306,
|
| 436 |
+
"learning_rate": 7.833279273434967e-06,
|
| 437 |
+
"loss": 0.1499,
|
| 438 |
+
"step": 5200
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.719104768083036,
|
| 442 |
+
"grad_norm": 0.2794703543186188,
|
| 443 |
+
"learning_rate": 7.0223807979241e-06,
|
| 444 |
+
"loss": 0.1476,
|
| 445 |
+
"step": 5300
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.7515407071034708,
|
| 449 |
+
"grad_norm": 0.257614403963089,
|
| 450 |
+
"learning_rate": 6.211482322413234e-06,
|
| 451 |
+
"loss": 0.1498,
|
| 452 |
+
"step": 5400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.7839766461239053,
|
| 456 |
+
"grad_norm": 0.24895663559436798,
|
| 457 |
+
"learning_rate": 5.400583846902368e-06,
|
| 458 |
+
"loss": 0.1485,
|
| 459 |
+
"step": 5500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.8164125851443398,
|
| 463 |
+
"grad_norm": 0.26096346974372864,
|
| 464 |
+
"learning_rate": 4.589685371391502e-06,
|
| 465 |
+
"loss": 0.1486,
|
| 466 |
+
"step": 5600
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.8488485241647745,
|
| 470 |
+
"grad_norm": 0.23756231367588043,
|
| 471 |
+
"learning_rate": 3.7787868958806357e-06,
|
| 472 |
+
"loss": 0.1506,
|
| 473 |
+
"step": 5700
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.8812844631852093,
|
| 477 |
+
"grad_norm": 0.2549228072166443,
|
| 478 |
+
"learning_rate": 2.9678884203697696e-06,
|
| 479 |
+
"loss": 0.1483,
|
| 480 |
+
"step": 5800
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.913720402205644,
|
| 484 |
+
"grad_norm": 0.24819372594356537,
|
| 485 |
+
"learning_rate": 2.156989944858904e-06,
|
| 486 |
+
"loss": 0.149,
|
| 487 |
+
"step": 5900
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.9461563412260785,
|
| 491 |
+
"grad_norm": 0.23372448980808258,
|
| 492 |
+
"learning_rate": 1.3460914693480378e-06,
|
| 493 |
+
"loss": 0.1506,
|
| 494 |
+
"step": 6000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.9461563412260785,
|
| 498 |
+
"eval_gen_len": 19.98361586503366,
|
| 499 |
+
"eval_loss": 0.15918129682540894,
|
| 500 |
+
"eval_rouge1": 0.073,
|
| 501 |
+
"eval_rouge2": 0.002,
|
| 502 |
+
"eval_rougeL": 0.073,
|
| 503 |
+
"eval_rougeLsum": 0.073,
|
| 504 |
+
"eval_runtime": 2174.8737,
|
| 505 |
+
"eval_samples_per_second": 11.338,
|
| 506 |
+
"eval_steps_per_second": 0.355,
|
| 507 |
+
"step": 6000
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.978592280246513,
|
| 511 |
+
"grad_norm": 0.25701549649238586,
|
| 512 |
+
"learning_rate": 5.351929938371715e-07,
|
| 513 |
+
"loss": 0.15,
|
| 514 |
+
"step": 6100
|
| 515 |
}
|
| 516 |
],
|
| 517 |
"logging_steps": 100,
|
| 518 |
+
"max_steps": 6166,
|
| 519 |
"num_input_tokens_seen": 0,
|
| 520 |
+
"num_train_epochs": 2,
|
| 521 |
"save_steps": 500,
|
| 522 |
"stateful_callbacks": {
|
| 523 |
"TrainerControl": {
|
|
|
|
| 531 |
"attributes": {}
|
| 532 |
}
|
| 533 |
},
|
| 534 |
+
"total_flos": 3.5709139077411373e+19,
|
| 535 |
"train_batch_size": 32,
|
| 536 |
"trial_name": null,
|
| 537 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5304
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5d585a2c4f69235113a2ad29e36aef3be9f972906a8fd4eddd987849af1a5a8
|
| 3 |
size 5304
|