Training in progress, epoch 0
Browse files- all_results.json +6 -6
- model.safetensors +1 -1
- runs/Sep04_17-57-52_93e6f912ca7d/events.out.tfevents.1725476123.93e6f912ca7d.1732.1 +3 -0
- runs/Sep04_19-03-43_93e6f912ca7d/events.out.tfevents.1725476631.93e6f912ca7d.1732.2 +3 -0
- train_results.json +6 -6
- trainer_state.json +550 -269
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 28.363636363636363,
|
| 3 |
+
"total_flos": 1.248846666821075e+18,
|
| 4 |
+
"train_loss": 0.9185584948613094,
|
| 5 |
+
"train_runtime": 3358.32,
|
| 6 |
+
"train_samples_per_second": 15.651,
|
| 7 |
+
"train_steps_per_second": 0.116
|
| 8 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 111317164
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4cdf28eb91d25a0607cd0c5e45fbb5c5811dd6a422b40f1b64b70f4a5a328977
|
| 3 |
size 111317164
|
runs/Sep04_17-57-52_93e6f912ca7d/events.out.tfevents.1725476123.93e6f912ca7d.1732.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce162503331238f6c8ee02fde0c40daead677bb56c46da821e7b757d3d1e3c0b
|
| 3 |
+
size 410
|
runs/Sep04_19-03-43_93e6f912ca7d/events.out.tfevents.1725476631.93e6f912ca7d.1732.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fe7eaf6ae676a8f6acc7c84c5f36ff7ff412a5a729b205444c4a7ce99b32085
|
| 3 |
+
size 6118
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second": 0.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 28.363636363636363,
|
| 3 |
+
"total_flos": 1.248846666821075e+18,
|
| 4 |
+
"train_loss": 0.9185584948613094,
|
| 5 |
+
"train_runtime": 3358.32,
|
| 6 |
+
"train_samples_per_second": 15.651,
|
| 7 |
+
"train_steps_per_second": 0.116
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,367 +1,648 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 1.
|
| 3 |
-
"best_model_checkpoint": "convnext-tiny-224-finetuned/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 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 |
-
"
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
"eval_logLoss": 1.6071832180023193,
|
| 17 |
-
"eval_loss": 1.6071830987930298,
|
| 18 |
-
"eval_runtime": 244.9974,
|
| 19 |
-
"eval_samples_per_second": 1.11,
|
| 20 |
-
"eval_steps_per_second": 0.037,
|
| 21 |
-
"step": 8
|
| 22 |
-
},
|
| 23 |
-
{
|
| 24 |
-
"epoch": 1.1764705882352942,
|
| 25 |
-
"grad_norm": 1.0974622964859009,
|
| 26 |
-
"learning_rate": 3.125e-05,
|
| 27 |
-
"loss": 1.6101,
|
| 28 |
"step": 10
|
| 29 |
},
|
| 30 |
{
|
| 31 |
-
"epoch":
|
| 32 |
"eval_accuracy": {
|
| 33 |
-
"accuracy": 0.
|
| 34 |
},
|
| 35 |
-
"eval_logLoss": 1.
|
| 36 |
-
"eval_loss": 1.
|
| 37 |
-
"eval_runtime":
|
| 38 |
-
"eval_samples_per_second":
|
| 39 |
-
"eval_steps_per_second":
|
| 40 |
-
"step":
|
| 41 |
-
},
|
| 42 |
-
{
|
| 43 |
-
"epoch":
|
| 44 |
-
"grad_norm": 1.
|
| 45 |
-
"learning_rate":
|
| 46 |
-
"loss": 1.
|
| 47 |
"step": 20
|
| 48 |
},
|
| 49 |
{
|
| 50 |
-
"epoch":
|
| 51 |
"eval_accuracy": {
|
| 52 |
-
"accuracy": 0.
|
| 53 |
},
|
| 54 |
-
"eval_logLoss": 1.
|
| 55 |
-
"eval_loss": 1.
|
| 56 |
-
"eval_runtime":
|
| 57 |
-
"eval_samples_per_second":
|
| 58 |
-
"eval_steps_per_second":
|
| 59 |
-
"step":
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch":
|
| 63 |
-
"grad_norm": 1.
|
| 64 |
-
"learning_rate":
|
| 65 |
-
"loss": 1.
|
| 66 |
"step": 30
|
| 67 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
{
|
| 69 |
"epoch": 4.0,
|
| 70 |
"eval_accuracy": {
|
| 71 |
-
"accuracy": 0.
|
| 72 |
},
|
| 73 |
-
"eval_logLoss": 1.
|
| 74 |
-
"eval_loss": 1.
|
| 75 |
-
"eval_runtime":
|
| 76 |
-
"eval_samples_per_second":
|
| 77 |
-
"eval_steps_per_second":
|
| 78 |
-
"step":
|
| 79 |
-
},
|
| 80 |
-
{
|
| 81 |
-
"epoch": 4.
|
| 82 |
-
"grad_norm": 1.
|
| 83 |
-
"learning_rate": 4.
|
| 84 |
-
"loss": 1.
|
| 85 |
-
"step":
|
| 86 |
},
|
| 87 |
{
|
| 88 |
-
"epoch": 4.
|
| 89 |
"eval_accuracy": {
|
| 90 |
-
"accuracy": 0.
|
| 91 |
},
|
| 92 |
-
"eval_logLoss": 1.
|
| 93 |
-
"eval_loss": 1.
|
| 94 |
-
"eval_runtime":
|
| 95 |
-
"eval_samples_per_second":
|
| 96 |
-
"eval_steps_per_second":
|
| 97 |
-
"step":
|
| 98 |
-
},
|
| 99 |
-
{
|
| 100 |
-
"epoch": 5.
|
| 101 |
-
"grad_norm": 1.
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 1.
|
| 104 |
-
"step":
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 108 |
"eval_accuracy": {
|
| 109 |
-
"accuracy": 0.
|
| 110 |
},
|
| 111 |
-
"eval_logLoss": 1.
|
| 112 |
-
"eval_loss": 1.
|
| 113 |
-
"eval_runtime":
|
| 114 |
-
"eval_samples_per_second":
|
| 115 |
-
"eval_steps_per_second":
|
| 116 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch": 6.
|
| 120 |
"eval_accuracy": {
|
| 121 |
-
"accuracy": 0.
|
| 122 |
},
|
| 123 |
-
"eval_logLoss": 1.
|
| 124 |
-
"eval_loss": 1.
|
| 125 |
-
"eval_runtime":
|
| 126 |
-
"eval_samples_per_second":
|
| 127 |
-
"eval_steps_per_second":
|
| 128 |
-
"step":
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 7.
|
| 132 |
-
"grad_norm":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 1.
|
| 135 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 8.0,
|
| 139 |
"eval_accuracy": {
|
| 140 |
-
"accuracy": 0.
|
| 141 |
},
|
| 142 |
-
"eval_logLoss": 1.
|
| 143 |
-
"eval_loss": 1.
|
| 144 |
-
"eval_runtime":
|
| 145 |
-
"eval_samples_per_second":
|
| 146 |
-
"eval_steps_per_second":
|
| 147 |
-
"step":
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch": 8.
|
| 151 |
-
"grad_norm":
|
| 152 |
-
"learning_rate": 3.
|
| 153 |
-
"loss": 1.
|
| 154 |
-
"step":
|
| 155 |
},
|
| 156 |
{
|
| 157 |
-
"epoch": 8.
|
| 158 |
"eval_accuracy": {
|
| 159 |
-
"accuracy": 0.
|
| 160 |
},
|
| 161 |
-
"eval_logLoss": 1.
|
| 162 |
-
"eval_loss": 1.
|
| 163 |
-
"eval_runtime":
|
| 164 |
-
"eval_samples_per_second":
|
| 165 |
-
"eval_steps_per_second":
|
| 166 |
-
"step":
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 9.
|
| 170 |
-
"grad_norm":
|
| 171 |
-
"learning_rate":
|
| 172 |
-
"loss": 1.
|
| 173 |
-
"step":
|
| 174 |
},
|
| 175 |
{
|
| 176 |
-
"epoch":
|
| 177 |
"eval_accuracy": {
|
| 178 |
-
"accuracy": 0.
|
| 179 |
},
|
| 180 |
-
"eval_logLoss": 1.
|
| 181 |
-
"eval_loss": 1.
|
| 182 |
-
"eval_runtime":
|
| 183 |
-
"eval_samples_per_second":
|
| 184 |
-
"eval_steps_per_second":
|
| 185 |
-
"step":
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 10.
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss":
|
| 192 |
-
"step":
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch": 10.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 196 |
"eval_accuracy": {
|
| 197 |
-
"accuracy": 0.
|
| 198 |
},
|
| 199 |
-
"eval_logLoss": 1.
|
| 200 |
-
"eval_loss": 1.
|
| 201 |
-
"eval_runtime":
|
| 202 |
-
"eval_samples_per_second":
|
| 203 |
-
"eval_steps_per_second":
|
| 204 |
-
"step":
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 11.
|
| 208 |
-
"grad_norm":
|
| 209 |
-
"learning_rate":
|
| 210 |
-
"loss":
|
| 211 |
-
"step":
|
| 212 |
},
|
| 213 |
{
|
| 214 |
"epoch": 12.0,
|
| 215 |
"eval_accuracy": {
|
| 216 |
-
"accuracy": 0.
|
| 217 |
},
|
| 218 |
-
"eval_logLoss": 1.
|
| 219 |
-
"eval_loss": 1.
|
| 220 |
-
"eval_runtime":
|
| 221 |
-
"eval_samples_per_second":
|
| 222 |
-
"eval_steps_per_second":
|
| 223 |
-
"step":
|
| 224 |
-
},
|
| 225 |
-
{
|
| 226 |
-
"epoch": 12.
|
| 227 |
-
"grad_norm":
|
| 228 |
-
"learning_rate":
|
| 229 |
-
"loss":
|
| 230 |
-
"step":
|
| 231 |
},
|
| 232 |
{
|
| 233 |
-
"epoch": 12.
|
| 234 |
"eval_accuracy": {
|
| 235 |
-
"accuracy": 0.
|
| 236 |
},
|
| 237 |
-
"eval_logLoss": 1.
|
| 238 |
-
"eval_loss": 1.
|
| 239 |
-
"eval_runtime":
|
| 240 |
-
"eval_samples_per_second":
|
| 241 |
-
"eval_steps_per_second":
|
| 242 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 243 |
},
|
| 244 |
{
|
| 245 |
-
"epoch":
|
| 246 |
"eval_accuracy": {
|
| 247 |
-
"accuracy": 0.
|
| 248 |
},
|
| 249 |
-
"eval_logLoss": 1.
|
| 250 |
-
"eval_loss": 1.
|
| 251 |
-
"eval_runtime":
|
| 252 |
-
"eval_samples_per_second":
|
| 253 |
-
"eval_steps_per_second":
|
| 254 |
-
"step":
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 14.
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss":
|
| 261 |
-
"step":
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch": 14.
|
| 265 |
"eval_accuracy": {
|
| 266 |
-
"accuracy": 0.
|
| 267 |
},
|
| 268 |
-
"eval_logLoss": 1.
|
| 269 |
-
"eval_loss": 1.
|
| 270 |
-
"eval_runtime":
|
| 271 |
-
"eval_samples_per_second":
|
| 272 |
-
"eval_steps_per_second":
|
| 273 |
-
"step":
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 15.
|
| 277 |
-
"grad_norm":
|
| 278 |
-
"learning_rate":
|
| 279 |
-
"loss":
|
| 280 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 281 |
},
|
| 282 |
{
|
| 283 |
"epoch": 16.0,
|
| 284 |
"eval_accuracy": {
|
| 285 |
-
"accuracy": 0.
|
| 286 |
},
|
| 287 |
-
"eval_logLoss": 1.
|
| 288 |
-
"eval_loss": 1.
|
| 289 |
-
"eval_runtime":
|
| 290 |
-
"eval_samples_per_second":
|
| 291 |
-
"eval_steps_per_second":
|
| 292 |
-
"step":
|
| 293 |
-
},
|
| 294 |
-
{
|
| 295 |
-
"epoch": 16.
|
| 296 |
-
"grad_norm":
|
| 297 |
-
"learning_rate":
|
| 298 |
-
"loss":
|
| 299 |
-
"step":
|
| 300 |
},
|
| 301 |
{
|
| 302 |
-
"epoch": 16.
|
| 303 |
"eval_accuracy": {
|
| 304 |
-
"accuracy": 0.
|
| 305 |
},
|
| 306 |
-
"eval_logLoss": 1.
|
| 307 |
-
"eval_loss": 1.
|
| 308 |
-
"eval_runtime":
|
| 309 |
-
"eval_samples_per_second":
|
| 310 |
-
"eval_steps_per_second":
|
| 311 |
-
"step":
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 17.
|
| 315 |
-
"grad_norm":
|
| 316 |
-
"learning_rate":
|
| 317 |
-
"loss":
|
| 318 |
-
"step":
|
| 319 |
},
|
| 320 |
{
|
| 321 |
-
"epoch":
|
| 322 |
"eval_accuracy": {
|
| 323 |
-
"accuracy": 0.
|
| 324 |
},
|
| 325 |
-
"eval_logLoss": 1.
|
| 326 |
-
"eval_loss": 1.
|
| 327 |
-
"eval_runtime":
|
| 328 |
-
"eval_samples_per_second":
|
| 329 |
-
"eval_steps_per_second":
|
| 330 |
-
"step":
|
| 331 |
},
|
| 332 |
{
|
| 333 |
-
"epoch": 18.
|
| 334 |
-
"grad_norm":
|
| 335 |
-
"learning_rate":
|
| 336 |
-
"loss": 0.
|
| 337 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 338 |
},
|
| 339 |
{
|
| 340 |
-
"epoch": 18.
|
| 341 |
"eval_accuracy": {
|
| 342 |
-
"accuracy": 0.
|
| 343 |
},
|
| 344 |
-
"eval_logLoss": 1.
|
| 345 |
-
"eval_loss": 1.
|
| 346 |
-
"eval_runtime":
|
| 347 |
-
"eval_samples_per_second":
|
| 348 |
-
"eval_steps_per_second":
|
| 349 |
-
"step":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 350 |
},
|
| 351 |
{
|
| 352 |
-
"epoch":
|
| 353 |
-
"
|
| 354 |
-
"
|
| 355 |
-
"
|
| 356 |
-
"
|
| 357 |
-
|
| 358 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 359 |
}
|
| 360 |
],
|
| 361 |
"logging_steps": 10,
|
| 362 |
-
"max_steps":
|
| 363 |
"num_input_tokens_seen": 0,
|
| 364 |
-
"num_train_epochs":
|
| 365 |
"save_steps": 500,
|
| 366 |
"stateful_callbacks": {
|
| 367 |
"TrainerControl": {
|
|
@@ -375,7 +656,7 @@
|
|
| 375 |
"attributes": {}
|
| 376 |
}
|
| 377 |
},
|
| 378 |
-
"total_flos":
|
| 379 |
"train_batch_size": 32,
|
| 380 |
"trial_name": null,
|
| 381 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.0196388959884644,
|
| 3 |
+
"best_model_checkpoint": "convnext-tiny-224-finetuned/checkpoint-261",
|
| 4 |
+
"epoch": 28.363636363636363,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 390,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.7272727272727273,
|
| 13 |
+
"grad_norm": 0.9955764412879944,
|
| 14 |
+
"learning_rate": 1.282051282051282e-05,
|
| 15 |
+
"loss": 1.6135,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.9454545454545454,
|
| 20 |
"eval_accuracy": {
|
| 21 |
+
"accuracy": 0.22323462414578588
|
| 22 |
},
|
| 23 |
+
"eval_logLoss": 1.5881173610687256,
|
| 24 |
+
"eval_loss": 1.588117241859436,
|
| 25 |
+
"eval_runtime": 405.8824,
|
| 26 |
+
"eval_samples_per_second": 1.082,
|
| 27 |
+
"eval_steps_per_second": 0.034,
|
| 28 |
+
"step": 13
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 1.4545454545454546,
|
| 32 |
+
"grad_norm": 1.0506945848464966,
|
| 33 |
+
"learning_rate": 2.564102564102564e-05,
|
| 34 |
+
"loss": 1.5823,
|
| 35 |
"step": 20
|
| 36 |
},
|
| 37 |
{
|
| 38 |
+
"epoch": 1.9636363636363636,
|
| 39 |
"eval_accuracy": {
|
| 40 |
+
"accuracy": 0.35990888382687924
|
| 41 |
},
|
| 42 |
+
"eval_logLoss": 1.5302075147628784,
|
| 43 |
+
"eval_loss": 1.5302073955535889,
|
| 44 |
+
"eval_runtime": 6.3692,
|
| 45 |
+
"eval_samples_per_second": 68.926,
|
| 46 |
+
"eval_steps_per_second": 2.198,
|
| 47 |
+
"step": 27
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 2.1818181818181817,
|
| 51 |
+
"grad_norm": 1.089643120765686,
|
| 52 |
+
"learning_rate": 3.846153846153846e-05,
|
| 53 |
+
"loss": 1.5485,
|
| 54 |
"step": 30
|
| 55 |
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 2.909090909090909,
|
| 58 |
+
"grad_norm": 1.260988712310791,
|
| 59 |
+
"learning_rate": 4.985754985754986e-05,
|
| 60 |
+
"loss": 1.4988,
|
| 61 |
+
"step": 40
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 2.981818181818182,
|
| 65 |
+
"eval_accuracy": {
|
| 66 |
+
"accuracy": 0.4874715261958998
|
| 67 |
+
},
|
| 68 |
+
"eval_logLoss": 1.4480493068695068,
|
| 69 |
+
"eval_loss": 1.4480493068695068,
|
| 70 |
+
"eval_runtime": 6.4414,
|
| 71 |
+
"eval_samples_per_second": 68.153,
|
| 72 |
+
"eval_steps_per_second": 2.173,
|
| 73 |
+
"step": 41
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 3.6363636363636362,
|
| 77 |
+
"grad_norm": 1.314483642578125,
|
| 78 |
+
"learning_rate": 4.8433048433048433e-05,
|
| 79 |
+
"loss": 1.4303,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
{
|
| 83 |
"epoch": 4.0,
|
| 84 |
"eval_accuracy": {
|
| 85 |
+
"accuracy": 0.5034168564920274
|
| 86 |
},
|
| 87 |
+
"eval_logLoss": 1.3424228429794312,
|
| 88 |
+
"eval_loss": 1.3424230813980103,
|
| 89 |
+
"eval_runtime": 6.3233,
|
| 90 |
+
"eval_samples_per_second": 69.426,
|
| 91 |
+
"eval_steps_per_second": 2.214,
|
| 92 |
+
"step": 55
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.363636363636363,
|
| 96 |
+
"grad_norm": 1.4022053480148315,
|
| 97 |
+
"learning_rate": 4.700854700854701e-05,
|
| 98 |
+
"loss": 1.3653,
|
| 99 |
+
"step": 60
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 4.945454545454545,
|
| 103 |
"eval_accuracy": {
|
| 104 |
+
"accuracy": 0.5239179954441914
|
| 105 |
},
|
| 106 |
+
"eval_logLoss": 1.2543689012527466,
|
| 107 |
+
"eval_loss": 1.2543689012527466,
|
| 108 |
+
"eval_runtime": 6.3884,
|
| 109 |
+
"eval_samples_per_second": 68.718,
|
| 110 |
+
"eval_steps_per_second": 2.191,
|
| 111 |
+
"step": 68
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 5.090909090909091,
|
| 115 |
+
"grad_norm": 1.9723025560379028,
|
| 116 |
+
"learning_rate": 4.558404558404559e-05,
|
| 117 |
+
"loss": 1.2829,
|
| 118 |
+
"step": 70
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 5.818181818181818,
|
| 122 |
+
"grad_norm": 1.580014944076538,
|
| 123 |
+
"learning_rate": 4.415954415954416e-05,
|
| 124 |
+
"loss": 1.2232,
|
| 125 |
+
"step": 80
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 5.963636363636364,
|
| 129 |
"eval_accuracy": {
|
| 130 |
+
"accuracy": 0.5239179954441914
|
| 131 |
},
|
| 132 |
+
"eval_logLoss": 1.1867280006408691,
|
| 133 |
+
"eval_loss": 1.1867281198501587,
|
| 134 |
+
"eval_runtime": 6.3056,
|
| 135 |
+
"eval_samples_per_second": 69.621,
|
| 136 |
+
"eval_steps_per_second": 2.22,
|
| 137 |
+
"step": 82
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 6.545454545454545,
|
| 141 |
+
"grad_norm": 1.6402294635772705,
|
| 142 |
+
"learning_rate": 4.2735042735042735e-05,
|
| 143 |
+
"loss": 1.1734,
|
| 144 |
+
"step": 90
|
| 145 |
},
|
| 146 |
{
|
| 147 |
+
"epoch": 6.9818181818181815,
|
| 148 |
"eval_accuracy": {
|
| 149 |
+
"accuracy": 0.5466970387243736
|
| 150 |
},
|
| 151 |
+
"eval_logLoss": 1.1329878568649292,
|
| 152 |
+
"eval_loss": 1.1329878568649292,
|
| 153 |
+
"eval_runtime": 6.5296,
|
| 154 |
+
"eval_samples_per_second": 67.232,
|
| 155 |
+
"eval_steps_per_second": 2.144,
|
| 156 |
+
"step": 96
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 7.2727272727272725,
|
| 160 |
+
"grad_norm": 2.135455846786499,
|
| 161 |
+
"learning_rate": 4.131054131054131e-05,
|
| 162 |
+
"loss": 1.146,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 8.0,
|
| 167 |
+
"grad_norm": 2.3969764709472656,
|
| 168 |
+
"learning_rate": 3.988603988603989e-05,
|
| 169 |
+
"loss": 1.0747,
|
| 170 |
+
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 8.0,
|
| 174 |
"eval_accuracy": {
|
| 175 |
+
"accuracy": 0.5626423690205011
|
| 176 |
},
|
| 177 |
+
"eval_logLoss": 1.1196777820587158,
|
| 178 |
+
"eval_loss": 1.1196777820587158,
|
| 179 |
+
"eval_runtime": 6.221,
|
| 180 |
+
"eval_samples_per_second": 70.567,
|
| 181 |
+
"eval_steps_per_second": 2.25,
|
| 182 |
+
"step": 110
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 8.727272727272727,
|
| 186 |
+
"grad_norm": 2.5098071098327637,
|
| 187 |
+
"learning_rate": 3.846153846153846e-05,
|
| 188 |
+
"loss": 1.0405,
|
| 189 |
+
"step": 120
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 8.945454545454545,
|
| 193 |
"eval_accuracy": {
|
| 194 |
+
"accuracy": 0.5535307517084282
|
| 195 |
},
|
| 196 |
+
"eval_logLoss": 1.0871223211288452,
|
| 197 |
+
"eval_loss": 1.0871223211288452,
|
| 198 |
+
"eval_runtime": 6.225,
|
| 199 |
+
"eval_samples_per_second": 70.522,
|
| 200 |
+
"eval_steps_per_second": 2.249,
|
| 201 |
+
"step": 123
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 9.454545454545455,
|
| 205 |
+
"grad_norm": 2.5964548587799072,
|
| 206 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 207 |
+
"loss": 1.0313,
|
| 208 |
+
"step": 130
|
| 209 |
},
|
| 210 |
{
|
| 211 |
+
"epoch": 9.963636363636363,
|
| 212 |
"eval_accuracy": {
|
| 213 |
+
"accuracy": 0.5671981776765376
|
| 214 |
},
|
| 215 |
+
"eval_logLoss": 1.0899914503097534,
|
| 216 |
+
"eval_loss": 1.0899913311004639,
|
| 217 |
+
"eval_runtime": 6.2747,
|
| 218 |
+
"eval_samples_per_second": 69.964,
|
| 219 |
+
"eval_steps_per_second": 2.231,
|
| 220 |
+
"step": 137
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 10.181818181818182,
|
| 224 |
+
"grad_norm": 3.431312084197998,
|
| 225 |
+
"learning_rate": 3.561253561253561e-05,
|
| 226 |
+
"loss": 0.9867,
|
| 227 |
+
"step": 140
|
| 228 |
},
|
| 229 |
{
|
| 230 |
+
"epoch": 10.909090909090908,
|
| 231 |
+
"grad_norm": 1.9337528944015503,
|
| 232 |
+
"learning_rate": 3.418803418803419e-05,
|
| 233 |
+
"loss": 0.959,
|
| 234 |
+
"step": 150
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 10.981818181818182,
|
| 238 |
"eval_accuracy": {
|
| 239 |
+
"accuracy": 0.5603644646924829
|
| 240 |
},
|
| 241 |
+
"eval_logLoss": 1.0766180753707886,
|
| 242 |
+
"eval_loss": 1.0766180753707886,
|
| 243 |
+
"eval_runtime": 6.3132,
|
| 244 |
+
"eval_samples_per_second": 69.537,
|
| 245 |
+
"eval_steps_per_second": 2.218,
|
| 246 |
+
"step": 151
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 11.636363636363637,
|
| 250 |
+
"grad_norm": 2.0123398303985596,
|
| 251 |
+
"learning_rate": 3.2763532763532764e-05,
|
| 252 |
+
"loss": 0.9314,
|
| 253 |
+
"step": 160
|
| 254 |
},
|
| 255 |
{
|
| 256 |
"epoch": 12.0,
|
| 257 |
"eval_accuracy": {
|
| 258 |
+
"accuracy": 0.5603644646924829
|
| 259 |
},
|
| 260 |
+
"eval_logLoss": 1.0608227252960205,
|
| 261 |
+
"eval_loss": 1.0608227252960205,
|
| 262 |
+
"eval_runtime": 6.2643,
|
| 263 |
+
"eval_samples_per_second": 70.079,
|
| 264 |
+
"eval_steps_per_second": 2.235,
|
| 265 |
+
"step": 165
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 12.363636363636363,
|
| 269 |
+
"grad_norm": 2.423022508621216,
|
| 270 |
+
"learning_rate": 3.133903133903134e-05,
|
| 271 |
+
"loss": 0.9102,
|
| 272 |
+
"step": 170
|
| 273 |
},
|
| 274 |
{
|
| 275 |
+
"epoch": 12.945454545454545,
|
| 276 |
"eval_accuracy": {
|
| 277 |
+
"accuracy": 0.5649202733485194
|
| 278 |
},
|
| 279 |
+
"eval_logLoss": 1.0388399362564087,
|
| 280 |
+
"eval_loss": 1.0388399362564087,
|
| 281 |
+
"eval_runtime": 6.2456,
|
| 282 |
+
"eval_samples_per_second": 70.289,
|
| 283 |
+
"eval_steps_per_second": 2.242,
|
| 284 |
+
"step": 178
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 13.090909090909092,
|
| 288 |
+
"grad_norm": 2.3961944580078125,
|
| 289 |
+
"learning_rate": 2.9914529914529915e-05,
|
| 290 |
+
"loss": 0.8825,
|
| 291 |
+
"step": 180
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 13.818181818181818,
|
| 295 |
+
"grad_norm": 2.1838817596435547,
|
| 296 |
+
"learning_rate": 2.8490028490028492e-05,
|
| 297 |
+
"loss": 0.8437,
|
| 298 |
+
"step": 190
|
| 299 |
},
|
| 300 |
{
|
| 301 |
+
"epoch": 13.963636363636363,
|
| 302 |
"eval_accuracy": {
|
| 303 |
+
"accuracy": 0.5785876993166287
|
| 304 |
},
|
| 305 |
+
"eval_logLoss": 1.0331932306289673,
|
| 306 |
+
"eval_loss": 1.0331931114196777,
|
| 307 |
+
"eval_runtime": 6.3139,
|
| 308 |
+
"eval_samples_per_second": 69.529,
|
| 309 |
+
"eval_steps_per_second": 2.217,
|
| 310 |
+
"step": 192
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 14.545454545454545,
|
| 314 |
+
"grad_norm": 2.22560453414917,
|
| 315 |
+
"learning_rate": 2.706552706552707e-05,
|
| 316 |
+
"loss": 0.8234,
|
| 317 |
+
"step": 200
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 14.981818181818182,
|
| 321 |
"eval_accuracy": {
|
| 322 |
+
"accuracy": 0.5763097949886105
|
| 323 |
},
|
| 324 |
+
"eval_logLoss": 1.030157446861267,
|
| 325 |
+
"eval_loss": 1.0301573276519775,
|
| 326 |
+
"eval_runtime": 6.2242,
|
| 327 |
+
"eval_samples_per_second": 70.531,
|
| 328 |
+
"eval_steps_per_second": 2.249,
|
| 329 |
+
"step": 206
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 15.272727272727273,
|
| 333 |
+
"grad_norm": 2.565896987915039,
|
| 334 |
+
"learning_rate": 2.564102564102564e-05,
|
| 335 |
+
"loss": 0.7888,
|
| 336 |
+
"step": 210
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 16.0,
|
| 340 |
+
"grad_norm": 4.106865882873535,
|
| 341 |
+
"learning_rate": 2.4216524216524217e-05,
|
| 342 |
+
"loss": 0.7883,
|
| 343 |
+
"step": 220
|
| 344 |
},
|
| 345 |
{
|
| 346 |
"epoch": 16.0,
|
| 347 |
"eval_accuracy": {
|
| 348 |
+
"accuracy": 0.5649202733485194
|
| 349 |
},
|
| 350 |
+
"eval_logLoss": 1.0276356935501099,
|
| 351 |
+
"eval_loss": 1.0276355743408203,
|
| 352 |
+
"eval_runtime": 6.221,
|
| 353 |
+
"eval_samples_per_second": 70.568,
|
| 354 |
+
"eval_steps_per_second": 2.25,
|
| 355 |
+
"step": 220
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 16.727272727272727,
|
| 359 |
+
"grad_norm": 2.70666766166687,
|
| 360 |
+
"learning_rate": 2.2792022792022794e-05,
|
| 361 |
+
"loss": 0.7364,
|
| 362 |
+
"step": 230
|
| 363 |
},
|
| 364 |
{
|
| 365 |
+
"epoch": 16.945454545454545,
|
| 366 |
"eval_accuracy": {
|
| 367 |
+
"accuracy": 0.5649202733485194
|
| 368 |
},
|
| 369 |
+
"eval_logLoss": 1.0277541875839233,
|
| 370 |
+
"eval_loss": 1.0277544260025024,
|
| 371 |
+
"eval_runtime": 6.2998,
|
| 372 |
+
"eval_samples_per_second": 69.685,
|
| 373 |
+
"eval_steps_per_second": 2.222,
|
| 374 |
+
"step": 233
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 17.454545454545453,
|
| 378 |
+
"grad_norm": 3.2285561561584473,
|
| 379 |
+
"learning_rate": 2.1367521367521368e-05,
|
| 380 |
+
"loss": 0.7561,
|
| 381 |
+
"step": 240
|
| 382 |
},
|
| 383 |
{
|
| 384 |
+
"epoch": 17.963636363636365,
|
| 385 |
"eval_accuracy": {
|
| 386 |
+
"accuracy": 0.5649202733485194
|
| 387 |
},
|
| 388 |
+
"eval_logLoss": 1.0257948637008667,
|
| 389 |
+
"eval_loss": 1.0257947444915771,
|
| 390 |
+
"eval_runtime": 6.223,
|
| 391 |
+
"eval_samples_per_second": 70.544,
|
| 392 |
+
"eval_steps_per_second": 2.25,
|
| 393 |
+
"step": 247
|
| 394 |
},
|
| 395 |
{
|
| 396 |
+
"epoch": 18.181818181818183,
|
| 397 |
+
"grad_norm": 3.2319438457489014,
|
| 398 |
+
"learning_rate": 1.9943019943019945e-05,
|
| 399 |
+
"loss": 0.7162,
|
| 400 |
+
"step": 250
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 18.90909090909091,
|
| 404 |
+
"grad_norm": 3.1322107315063477,
|
| 405 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 406 |
+
"loss": 0.7062,
|
| 407 |
+
"step": 260
|
| 408 |
},
|
| 409 |
{
|
| 410 |
+
"epoch": 18.98181818181818,
|
| 411 |
"eval_accuracy": {
|
| 412 |
+
"accuracy": 0.5694760820045558
|
| 413 |
},
|
| 414 |
+
"eval_logLoss": 1.019639015197754,
|
| 415 |
+
"eval_loss": 1.0196388959884644,
|
| 416 |
+
"eval_runtime": 6.1923,
|
| 417 |
+
"eval_samples_per_second": 70.894,
|
| 418 |
+
"eval_steps_per_second": 2.261,
|
| 419 |
+
"step": 261
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 19.636363636363637,
|
| 423 |
+
"grad_norm": 2.8413918018341064,
|
| 424 |
+
"learning_rate": 1.7094017094017095e-05,
|
| 425 |
+
"loss": 0.6897,
|
| 426 |
+
"step": 270
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 20.0,
|
| 430 |
+
"eval_accuracy": {
|
| 431 |
+
"accuracy": 0.5558086560364465
|
| 432 |
+
},
|
| 433 |
+
"eval_logLoss": 1.0308281183242798,
|
| 434 |
+
"eval_loss": 1.0308281183242798,
|
| 435 |
+
"eval_runtime": 6.1943,
|
| 436 |
+
"eval_samples_per_second": 70.871,
|
| 437 |
+
"eval_steps_per_second": 2.26,
|
| 438 |
+
"step": 275
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 20.363636363636363,
|
| 442 |
+
"grad_norm": 2.9819412231445312,
|
| 443 |
+
"learning_rate": 1.566951566951567e-05,
|
| 444 |
+
"loss": 0.6511,
|
| 445 |
+
"step": 280
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 20.945454545454545,
|
| 449 |
+
"eval_accuracy": {
|
| 450 |
+
"accuracy": 0.5626423690205011
|
| 451 |
+
},
|
| 452 |
+
"eval_logLoss": 1.0247496366500854,
|
| 453 |
+
"eval_loss": 1.0247496366500854,
|
| 454 |
+
"eval_runtime": 6.2512,
|
| 455 |
+
"eval_samples_per_second": 70.226,
|
| 456 |
+
"eval_steps_per_second": 2.24,
|
| 457 |
+
"step": 288
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 21.09090909090909,
|
| 461 |
+
"grad_norm": 4.308482646942139,
|
| 462 |
+
"learning_rate": 1.4245014245014246e-05,
|
| 463 |
+
"loss": 0.6456,
|
| 464 |
+
"step": 290
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 21.818181818181817,
|
| 468 |
+
"grad_norm": 3.1177256107330322,
|
| 469 |
+
"learning_rate": 1.282051282051282e-05,
|
| 470 |
+
"loss": 0.6338,
|
| 471 |
+
"step": 300
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 21.963636363636365,
|
| 475 |
+
"eval_accuracy": {
|
| 476 |
+
"accuracy": 0.5603644646924829
|
| 477 |
+
},
|
| 478 |
+
"eval_logLoss": 1.0310317277908325,
|
| 479 |
+
"eval_loss": 1.031031608581543,
|
| 480 |
+
"eval_runtime": 6.1732,
|
| 481 |
+
"eval_samples_per_second": 71.114,
|
| 482 |
+
"eval_steps_per_second": 2.268,
|
| 483 |
+
"step": 302
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 22.545454545454547,
|
| 487 |
+
"grad_norm": 2.941620111465454,
|
| 488 |
+
"learning_rate": 1.1396011396011397e-05,
|
| 489 |
+
"loss": 0.619,
|
| 490 |
+
"step": 310
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 22.98181818181818,
|
| 494 |
+
"eval_accuracy": {
|
| 495 |
+
"accuracy": 0.5671981776765376
|
| 496 |
+
},
|
| 497 |
+
"eval_logLoss": 1.0257664918899536,
|
| 498 |
+
"eval_loss": 1.025766372680664,
|
| 499 |
+
"eval_runtime": 6.1895,
|
| 500 |
+
"eval_samples_per_second": 70.927,
|
| 501 |
+
"eval_steps_per_second": 2.262,
|
| 502 |
+
"step": 316
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 23.272727272727273,
|
| 506 |
+
"grad_norm": 3.034752368927002,
|
| 507 |
+
"learning_rate": 9.971509971509972e-06,
|
| 508 |
+
"loss": 0.6187,
|
| 509 |
+
"step": 320
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 24.0,
|
| 513 |
+
"grad_norm": 3.183673620223999,
|
| 514 |
+
"learning_rate": 8.547008547008548e-06,
|
| 515 |
+
"loss": 0.6008,
|
| 516 |
+
"step": 330
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 24.0,
|
| 520 |
+
"eval_accuracy": {
|
| 521 |
+
"accuracy": 0.5626423690205011
|
| 522 |
+
},
|
| 523 |
+
"eval_logLoss": 1.0298672914505005,
|
| 524 |
+
"eval_loss": 1.029867172241211,
|
| 525 |
+
"eval_runtime": 6.3628,
|
| 526 |
+
"eval_samples_per_second": 68.995,
|
| 527 |
+
"eval_steps_per_second": 2.2,
|
| 528 |
+
"step": 330
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 24.727272727272727,
|
| 532 |
+
"grad_norm": 2.810622215270996,
|
| 533 |
+
"learning_rate": 7.122507122507123e-06,
|
| 534 |
+
"loss": 0.601,
|
| 535 |
+
"step": 340
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 24.945454545454545,
|
| 539 |
+
"eval_accuracy": {
|
| 540 |
+
"accuracy": 0.5671981776765376
|
| 541 |
+
},
|
| 542 |
+
"eval_logLoss": 1.0329437255859375,
|
| 543 |
+
"eval_loss": 1.0329439640045166,
|
| 544 |
+
"eval_runtime": 6.2697,
|
| 545 |
+
"eval_samples_per_second": 70.019,
|
| 546 |
+
"eval_steps_per_second": 2.233,
|
| 547 |
+
"step": 343
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 25.454545454545453,
|
| 551 |
+
"grad_norm": 4.746636867523193,
|
| 552 |
+
"learning_rate": 5.6980056980056985e-06,
|
| 553 |
+
"loss": 0.595,
|
| 554 |
+
"step": 350
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 25.963636363636365,
|
| 558 |
+
"eval_accuracy": {
|
| 559 |
+
"accuracy": 0.5694760820045558
|
| 560 |
+
},
|
| 561 |
+
"eval_logLoss": 1.0277411937713623,
|
| 562 |
+
"eval_loss": 1.0277411937713623,
|
| 563 |
+
"eval_runtime": 6.3782,
|
| 564 |
+
"eval_samples_per_second": 68.828,
|
| 565 |
+
"eval_steps_per_second": 2.195,
|
| 566 |
+
"step": 357
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 26.181818181818183,
|
| 570 |
+
"grad_norm": 3.6217236518859863,
|
| 571 |
+
"learning_rate": 4.273504273504274e-06,
|
| 572 |
+
"loss": 0.5713,
|
| 573 |
+
"step": 360
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 26.90909090909091,
|
| 577 |
+
"grad_norm": 4.607295989990234,
|
| 578 |
+
"learning_rate": 2.8490028490028492e-06,
|
| 579 |
+
"loss": 0.598,
|
| 580 |
+
"step": 370
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 26.98181818181818,
|
| 584 |
+
"eval_accuracy": {
|
| 585 |
+
"accuracy": 0.5671981776765376
|
| 586 |
+
},
|
| 587 |
+
"eval_logLoss": 1.0288209915161133,
|
| 588 |
+
"eval_loss": 1.0288212299346924,
|
| 589 |
+
"eval_runtime": 6.3179,
|
| 590 |
+
"eval_samples_per_second": 69.485,
|
| 591 |
+
"eval_steps_per_second": 2.216,
|
| 592 |
+
"step": 371
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 27.636363636363637,
|
| 596 |
+
"grad_norm": 3.8291075229644775,
|
| 597 |
+
"learning_rate": 1.4245014245014246e-06,
|
| 598 |
+
"loss": 0.5771,
|
| 599 |
+
"step": 380
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 28.0,
|
| 603 |
+
"eval_accuracy": {
|
| 604 |
+
"accuracy": 0.5603644646924829
|
| 605 |
+
},
|
| 606 |
+
"eval_logLoss": 1.031100869178772,
|
| 607 |
+
"eval_loss": 1.0311009883880615,
|
| 608 |
+
"eval_runtime": 6.2861,
|
| 609 |
+
"eval_samples_per_second": 69.836,
|
| 610 |
+
"eval_steps_per_second": 2.227,
|
| 611 |
+
"step": 385
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 28.363636363636363,
|
| 615 |
+
"grad_norm": 3.588949203491211,
|
| 616 |
+
"learning_rate": 0.0,
|
| 617 |
+
"loss": 0.5829,
|
| 618 |
+
"step": 390
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 28.363636363636363,
|
| 622 |
+
"eval_accuracy": {
|
| 623 |
+
"accuracy": 0.5626423690205011
|
| 624 |
+
},
|
| 625 |
+
"eval_logLoss": 1.0311228036880493,
|
| 626 |
+
"eval_loss": 1.0311226844787598,
|
| 627 |
+
"eval_runtime": 6.7799,
|
| 628 |
+
"eval_samples_per_second": 64.75,
|
| 629 |
+
"eval_steps_per_second": 2.065,
|
| 630 |
+
"step": 390
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 28.363636363636363,
|
| 634 |
+
"step": 390,
|
| 635 |
+
"total_flos": 1.248846666821075e+18,
|
| 636 |
+
"train_loss": 0.9185584948613094,
|
| 637 |
+
"train_runtime": 3358.32,
|
| 638 |
+
"train_samples_per_second": 15.651,
|
| 639 |
+
"train_steps_per_second": 0.116
|
| 640 |
}
|
| 641 |
],
|
| 642 |
"logging_steps": 10,
|
| 643 |
+
"max_steps": 390,
|
| 644 |
"num_input_tokens_seen": 0,
|
| 645 |
+
"num_train_epochs": 30,
|
| 646 |
"save_steps": 500,
|
| 647 |
"stateful_callbacks": {
|
| 648 |
"TrainerControl": {
|
|
|
|
| 656 |
"attributes": {}
|
| 657 |
}
|
| 658 |
},
|
| 659 |
+
"total_flos": 1.248846666821075e+18,
|
| 660 |
"train_batch_size": 32,
|
| 661 |
"trial_name": null,
|
| 662 |
"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 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd6e90f1620446f81e54f3779005a5ba165bf79bc20b0a5b3ab3e30f8560836a
|
| 3 |
size 5240
|