Training in progress, epoch 1, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2682482800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13d96965f916227845baa32615edd73ffb89f6906929907751c1d30b00563d96
|
| 3 |
size 2682482800
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5365108834
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a3162e8491462ae2471352df04c5e20159e90bc0bcb5af2ed19743b12a357a4a
|
| 3 |
size 5365108834
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15006
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7d16a323ec1c1c4d4bc21de4f5ba8f3d5aa610b99bd6f61b387ba35e64939e8
|
| 3 |
size 15006
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:acff3929c8e35c17a966d0166ffdaff0c7e8a7068b02c8a15d3a42c4f8762618
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,2073 +1,93 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 50,
|
| 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":
|
| 14 |
-
"learning_rate": 9.
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"eval_loss": 1.
|
| 21 |
-
"eval_runtime": 116.
|
| 22 |
-
"eval_samples_per_second": 41.
|
| 23 |
-
"eval_steps_per_second": 20.
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 9.
|
| 30 |
-
"loss": 1.
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"eval_loss": 1.
|
| 36 |
-
"eval_runtime": 116.
|
| 37 |
-
"eval_samples_per_second": 41.
|
| 38 |
-
"eval_steps_per_second": 20.
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"grad_norm":
|
| 44 |
-
"learning_rate": 9.
|
| 45 |
-
"loss": 1.
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 0.
|
| 50 |
-
"eval_loss": 1.
|
| 51 |
-
"eval_runtime": 116.
|
| 52 |
-
"eval_samples_per_second": 41.
|
| 53 |
-
"eval_steps_per_second": 20.
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
-
"epoch": 0.
|
| 58 |
-
"grad_norm":
|
| 59 |
-
"learning_rate": 9.
|
| 60 |
-
"loss": 1.
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
-
"epoch": 0.
|
| 65 |
-
"eval_loss": 1.
|
| 66 |
-
"eval_runtime": 116.
|
| 67 |
-
"eval_samples_per_second": 41.
|
| 68 |
-
"eval_steps_per_second": 20.
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"grad_norm":
|
| 74 |
-
"learning_rate": 9.
|
| 75 |
-
"loss": 1.
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
-
"epoch": 0.
|
| 80 |
-
"eval_loss": 1.
|
| 81 |
-
"eval_runtime": 116.
|
| 82 |
-
"eval_samples_per_second": 41.
|
| 83 |
-
"eval_steps_per_second": 20.
|
| 84 |
"step": 250
|
| 85 |
-
},
|
| 86 |
-
{
|
| 87 |
-
"epoch": 0.872885979268958,
|
| 88 |
-
"grad_norm": 67.02438354492188,
|
| 89 |
-
"learning_rate": 9.56268221574344e-05,
|
| 90 |
-
"loss": 1.2113,
|
| 91 |
-
"step": 300
|
| 92 |
-
},
|
| 93 |
-
{
|
| 94 |
-
"epoch": 0.872885979268958,
|
| 95 |
-
"eval_loss": 1.1557544469833374,
|
| 96 |
-
"eval_runtime": 116.6632,
|
| 97 |
-
"eval_samples_per_second": 41.367,
|
| 98 |
-
"eval_steps_per_second": 20.683,
|
| 99 |
-
"step": 300
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 1.0203673395162758,
|
| 103 |
-
"grad_norm": 118.28443145751953,
|
| 104 |
-
"learning_rate": 9.489795918367348e-05,
|
| 105 |
-
"loss": 1.1973,
|
| 106 |
-
"step": 350
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 1.0203673395162758,
|
| 110 |
-
"eval_loss": 1.1349806785583496,
|
| 111 |
-
"eval_runtime": 118.2776,
|
| 112 |
-
"eval_samples_per_second": 40.802,
|
| 113 |
-
"eval_steps_per_second": 20.401,
|
| 114 |
-
"step": 350
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 1.165848336061102,
|
| 118 |
-
"grad_norm": 71.47338104248047,
|
| 119 |
-
"learning_rate": 9.416909620991254e-05,
|
| 120 |
-
"loss": 1.159,
|
| 121 |
-
"step": 400
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 1.165848336061102,
|
| 125 |
-
"eval_loss": 1.1155215501785278,
|
| 126 |
-
"eval_runtime": 116.539,
|
| 127 |
-
"eval_samples_per_second": 41.411,
|
| 128 |
-
"eval_steps_per_second": 20.706,
|
| 129 |
-
"step": 400
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 1.3113293326059283,
|
| 133 |
-
"grad_norm": 49.708885192871094,
|
| 134 |
-
"learning_rate": 9.344023323615161e-05,
|
| 135 |
-
"loss": 1.1471,
|
| 136 |
-
"step": 450
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 1.3113293326059283,
|
| 140 |
-
"eval_loss": 1.0999969244003296,
|
| 141 |
-
"eval_runtime": 116.44,
|
| 142 |
-
"eval_samples_per_second": 41.446,
|
| 143 |
-
"eval_steps_per_second": 20.723,
|
| 144 |
-
"step": 450
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 1.4568103291507546,
|
| 148 |
-
"grad_norm": 59.27313995361328,
|
| 149 |
-
"learning_rate": 9.271137026239067e-05,
|
| 150 |
-
"loss": 1.1325,
|
| 151 |
-
"step": 500
|
| 152 |
-
},
|
| 153 |
-
{
|
| 154 |
-
"epoch": 1.4568103291507546,
|
| 155 |
-
"eval_loss": 1.0860569477081299,
|
| 156 |
-
"eval_runtime": 116.8329,
|
| 157 |
-
"eval_samples_per_second": 41.307,
|
| 158 |
-
"eval_steps_per_second": 20.653,
|
| 159 |
-
"step": 500
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 1.6022913256955809,
|
| 163 |
-
"grad_norm": 93.8194351196289,
|
| 164 |
-
"learning_rate": 9.198250728862974e-05,
|
| 165 |
-
"loss": 1.121,
|
| 166 |
-
"step": 550
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 1.6022913256955809,
|
| 170 |
-
"eval_loss": 1.0699989795684814,
|
| 171 |
-
"eval_runtime": 116.3351,
|
| 172 |
-
"eval_samples_per_second": 41.484,
|
| 173 |
-
"eval_steps_per_second": 20.742,
|
| 174 |
-
"step": 550
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 1.7477723222404073,
|
| 178 |
-
"grad_norm": 64.32953643798828,
|
| 179 |
-
"learning_rate": 9.125364431486881e-05,
|
| 180 |
-
"loss": 1.1071,
|
| 181 |
-
"step": 600
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 1.7477723222404073,
|
| 185 |
-
"eval_loss": 1.0575978755950928,
|
| 186 |
-
"eval_runtime": 116.4201,
|
| 187 |
-
"eval_samples_per_second": 41.453,
|
| 188 |
-
"eval_steps_per_second": 20.727,
|
| 189 |
-
"step": 600
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 1.8932533187852338,
|
| 193 |
-
"grad_norm": 55.454044342041016,
|
| 194 |
-
"learning_rate": 9.052478134110788e-05,
|
| 195 |
-
"loss": 1.092,
|
| 196 |
-
"step": 650
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 1.8932533187852338,
|
| 200 |
-
"eval_loss": 1.0517867803573608,
|
| 201 |
-
"eval_runtime": 116.5644,
|
| 202 |
-
"eval_samples_per_second": 41.402,
|
| 203 |
-
"eval_steps_per_second": 20.701,
|
| 204 |
-
"step": 650
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 2.0407346790325516,
|
| 208 |
-
"grad_norm": 96.50550842285156,
|
| 209 |
-
"learning_rate": 8.979591836734695e-05,
|
| 210 |
-
"loss": 1.0918,
|
| 211 |
-
"step": 700
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 2.0407346790325516,
|
| 215 |
-
"eval_loss": 1.0332311391830444,
|
| 216 |
-
"eval_runtime": 117.5493,
|
| 217 |
-
"eval_samples_per_second": 41.055,
|
| 218 |
-
"eval_steps_per_second": 20.528,
|
| 219 |
-
"step": 700
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 2.186215675577378,
|
| 223 |
-
"grad_norm": 83.90518188476562,
|
| 224 |
-
"learning_rate": 8.9067055393586e-05,
|
| 225 |
-
"loss": 1.06,
|
| 226 |
-
"step": 750
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 2.186215675577378,
|
| 230 |
-
"eval_loss": 1.020917534828186,
|
| 231 |
-
"eval_runtime": 116.365,
|
| 232 |
-
"eval_samples_per_second": 41.473,
|
| 233 |
-
"eval_steps_per_second": 20.736,
|
| 234 |
-
"step": 750
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 2.331696672122204,
|
| 238 |
-
"grad_norm": 59.09587860107422,
|
| 239 |
-
"learning_rate": 8.833819241982508e-05,
|
| 240 |
-
"loss": 1.0556,
|
| 241 |
-
"step": 800
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 2.331696672122204,
|
| 245 |
-
"eval_loss": 1.0105912685394287,
|
| 246 |
-
"eval_runtime": 116.7111,
|
| 247 |
-
"eval_samples_per_second": 41.35,
|
| 248 |
-
"eval_steps_per_second": 20.675,
|
| 249 |
-
"step": 800
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 2.4771776686670304,
|
| 253 |
-
"grad_norm": 41.396785736083984,
|
| 254 |
-
"learning_rate": 8.760932944606415e-05,
|
| 255 |
-
"loss": 1.046,
|
| 256 |
-
"step": 850
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 2.4771776686670304,
|
| 260 |
-
"eval_loss": 1.002454161643982,
|
| 261 |
-
"eval_runtime": 116.4653,
|
| 262 |
-
"eval_samples_per_second": 41.437,
|
| 263 |
-
"eval_steps_per_second": 20.719,
|
| 264 |
-
"step": 850
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 2.6226586652118566,
|
| 268 |
-
"grad_norm": 49.805667877197266,
|
| 269 |
-
"learning_rate": 8.688046647230322e-05,
|
| 270 |
-
"loss": 1.036,
|
| 271 |
-
"step": 900
|
| 272 |
-
},
|
| 273 |
-
{
|
| 274 |
-
"epoch": 2.6226586652118566,
|
| 275 |
-
"eval_loss": 0.9919629096984863,
|
| 276 |
-
"eval_runtime": 116.6035,
|
| 277 |
-
"eval_samples_per_second": 41.388,
|
| 278 |
-
"eval_steps_per_second": 20.694,
|
| 279 |
-
"step": 900
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 2.768139661756683,
|
| 283 |
-
"grad_norm": 39.228797912597656,
|
| 284 |
-
"learning_rate": 8.615160349854227e-05,
|
| 285 |
-
"loss": 1.0333,
|
| 286 |
-
"step": 950
|
| 287 |
-
},
|
| 288 |
-
{
|
| 289 |
-
"epoch": 2.768139661756683,
|
| 290 |
-
"eval_loss": 0.9831019639968872,
|
| 291 |
-
"eval_runtime": 116.7019,
|
| 292 |
-
"eval_samples_per_second": 41.353,
|
| 293 |
-
"eval_steps_per_second": 20.677,
|
| 294 |
-
"step": 950
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 2.913620658301509,
|
| 298 |
-
"grad_norm": 85.81388854980469,
|
| 299 |
-
"learning_rate": 8.542274052478134e-05,
|
| 300 |
-
"loss": 1.0256,
|
| 301 |
-
"step": 1000
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 2.913620658301509,
|
| 305 |
-
"eval_loss": 0.976150631904602,
|
| 306 |
-
"eval_runtime": 116.6678,
|
| 307 |
-
"eval_samples_per_second": 41.365,
|
| 308 |
-
"eval_steps_per_second": 20.683,
|
| 309 |
-
"step": 1000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 3.061102018548827,
|
| 313 |
-
"grad_norm": 51.54584503173828,
|
| 314 |
-
"learning_rate": 8.469387755102041e-05,
|
| 315 |
-
"loss": 1.014,
|
| 316 |
-
"step": 1050
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"epoch": 3.061102018548827,
|
| 320 |
-
"eval_loss": 0.9659164547920227,
|
| 321 |
-
"eval_runtime": 116.6584,
|
| 322 |
-
"eval_samples_per_second": 41.369,
|
| 323 |
-
"eval_steps_per_second": 20.684,
|
| 324 |
-
"step": 1050
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 3.2065830150936536,
|
| 328 |
-
"grad_norm": 65.59361267089844,
|
| 329 |
-
"learning_rate": 8.396501457725948e-05,
|
| 330 |
-
"loss": 0.9976,
|
| 331 |
-
"step": 1100
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 3.2065830150936536,
|
| 335 |
-
"eval_loss": 0.9607768058776855,
|
| 336 |
-
"eval_runtime": 116.7925,
|
| 337 |
-
"eval_samples_per_second": 41.321,
|
| 338 |
-
"eval_steps_per_second": 20.661,
|
| 339 |
-
"step": 1100
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 3.35206401163848,
|
| 343 |
-
"grad_norm": 45.97025680541992,
|
| 344 |
-
"learning_rate": 8.323615160349854e-05,
|
| 345 |
-
"loss": 0.9917,
|
| 346 |
-
"step": 1150
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 3.35206401163848,
|
| 350 |
-
"eval_loss": 0.9531689882278442,
|
| 351 |
-
"eval_runtime": 116.6276,
|
| 352 |
-
"eval_samples_per_second": 41.38,
|
| 353 |
-
"eval_steps_per_second": 20.69,
|
| 354 |
-
"step": 1150
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 3.497545008183306,
|
| 358 |
-
"grad_norm": 92.67556762695312,
|
| 359 |
-
"learning_rate": 8.250728862973761e-05,
|
| 360 |
-
"loss": 0.9858,
|
| 361 |
-
"step": 1200
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 3.497545008183306,
|
| 365 |
-
"eval_loss": 0.9395309090614319,
|
| 366 |
-
"eval_runtime": 116.5916,
|
| 367 |
-
"eval_samples_per_second": 41.392,
|
| 368 |
-
"eval_steps_per_second": 20.696,
|
| 369 |
-
"step": 1200
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 3.6430260047281324,
|
| 373 |
-
"grad_norm": 69.46937561035156,
|
| 374 |
-
"learning_rate": 8.177842565597668e-05,
|
| 375 |
-
"loss": 0.9727,
|
| 376 |
-
"step": 1250
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 3.6430260047281324,
|
| 380 |
-
"eval_loss": 0.9338119029998779,
|
| 381 |
-
"eval_runtime": 116.657,
|
| 382 |
-
"eval_samples_per_second": 41.369,
|
| 383 |
-
"eval_steps_per_second": 20.685,
|
| 384 |
-
"step": 1250
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 3.7885070012729587,
|
| 388 |
-
"grad_norm": 49.612693786621094,
|
| 389 |
-
"learning_rate": 8.104956268221575e-05,
|
| 390 |
-
"loss": 0.9714,
|
| 391 |
-
"step": 1300
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 3.7885070012729587,
|
| 395 |
-
"eval_loss": 0.9243752956390381,
|
| 396 |
-
"eval_runtime": 116.916,
|
| 397 |
-
"eval_samples_per_second": 41.277,
|
| 398 |
-
"eval_steps_per_second": 20.639,
|
| 399 |
-
"step": 1300
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 3.933987997817785,
|
| 403 |
-
"grad_norm": 45.8222541809082,
|
| 404 |
-
"learning_rate": 8.032069970845482e-05,
|
| 405 |
-
"loss": 0.9661,
|
| 406 |
-
"step": 1350
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 3.933987997817785,
|
| 410 |
-
"eval_loss": 0.9216246604919434,
|
| 411 |
-
"eval_runtime": 116.5963,
|
| 412 |
-
"eval_samples_per_second": 41.391,
|
| 413 |
-
"eval_steps_per_second": 20.695,
|
| 414 |
-
"step": 1350
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 4.081469358065103,
|
| 418 |
-
"grad_norm": 53.19688034057617,
|
| 419 |
-
"learning_rate": 7.959183673469388e-05,
|
| 420 |
-
"loss": 0.9632,
|
| 421 |
-
"step": 1400
|
| 422 |
-
},
|
| 423 |
-
{
|
| 424 |
-
"epoch": 4.081469358065103,
|
| 425 |
-
"eval_loss": 0.9175893068313599,
|
| 426 |
-
"eval_runtime": 116.3144,
|
| 427 |
-
"eval_samples_per_second": 41.491,
|
| 428 |
-
"eval_steps_per_second": 20.745,
|
| 429 |
-
"step": 1400
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 4.226950354609929,
|
| 433 |
-
"grad_norm": 62.0031623840332,
|
| 434 |
-
"learning_rate": 7.886297376093295e-05,
|
| 435 |
-
"loss": 0.9455,
|
| 436 |
-
"step": 1450
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 4.226950354609929,
|
| 440 |
-
"eval_loss": 0.909070611000061,
|
| 441 |
-
"eval_runtime": 116.6224,
|
| 442 |
-
"eval_samples_per_second": 41.381,
|
| 443 |
-
"eval_steps_per_second": 20.691,
|
| 444 |
-
"step": 1450
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 4.372431351154756,
|
| 448 |
-
"grad_norm": 91.25264739990234,
|
| 449 |
-
"learning_rate": 7.8134110787172e-05,
|
| 450 |
-
"loss": 0.9364,
|
| 451 |
-
"step": 1500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 4.372431351154756,
|
| 455 |
-
"eval_loss": 0.9035021066665649,
|
| 456 |
-
"eval_runtime": 116.658,
|
| 457 |
-
"eval_samples_per_second": 41.369,
|
| 458 |
-
"eval_steps_per_second": 20.684,
|
| 459 |
-
"step": 1500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 4.517912347699582,
|
| 463 |
-
"grad_norm": 56.04409408569336,
|
| 464 |
-
"learning_rate": 7.740524781341109e-05,
|
| 465 |
-
"loss": 0.9328,
|
| 466 |
-
"step": 1550
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 4.517912347699582,
|
| 470 |
-
"eval_loss": 0.8970184922218323,
|
| 471 |
-
"eval_runtime": 116.4662,
|
| 472 |
-
"eval_samples_per_second": 41.437,
|
| 473 |
-
"eval_steps_per_second": 20.718,
|
| 474 |
-
"step": 1550
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 4.663393344244408,
|
| 478 |
-
"grad_norm": 49.968868255615234,
|
| 479 |
-
"learning_rate": 7.667638483965015e-05,
|
| 480 |
-
"loss": 0.9376,
|
| 481 |
-
"step": 1600
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 4.663393344244408,
|
| 485 |
-
"eval_loss": 0.8949723839759827,
|
| 486 |
-
"eval_runtime": 116.5953,
|
| 487 |
-
"eval_samples_per_second": 41.391,
|
| 488 |
-
"eval_steps_per_second": 20.696,
|
| 489 |
-
"step": 1600
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 4.8088743407892345,
|
| 493 |
-
"grad_norm": 70.03372192382812,
|
| 494 |
-
"learning_rate": 7.594752186588922e-05,
|
| 495 |
-
"loss": 0.9234,
|
| 496 |
-
"step": 1650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 4.8088743407892345,
|
| 500 |
-
"eval_loss": 0.8886305689811707,
|
| 501 |
-
"eval_runtime": 116.6405,
|
| 502 |
-
"eval_samples_per_second": 41.375,
|
| 503 |
-
"eval_steps_per_second": 20.688,
|
| 504 |
-
"step": 1650
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 4.954355337334061,
|
| 508 |
-
"grad_norm": 66.96846771240234,
|
| 509 |
-
"learning_rate": 7.521865889212827e-05,
|
| 510 |
-
"loss": 0.9264,
|
| 511 |
-
"step": 1700
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 4.954355337334061,
|
| 515 |
-
"eval_loss": 0.8805985450744629,
|
| 516 |
-
"eval_runtime": 116.6258,
|
| 517 |
-
"eval_samples_per_second": 41.38,
|
| 518 |
-
"eval_steps_per_second": 20.69,
|
| 519 |
-
"step": 1700
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 5.101836697581378,
|
| 523 |
-
"grad_norm": 79.30951690673828,
|
| 524 |
-
"learning_rate": 7.448979591836736e-05,
|
| 525 |
-
"loss": 0.9154,
|
| 526 |
-
"step": 1750
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 5.101836697581378,
|
| 530 |
-
"eval_loss": 0.8716277480125427,
|
| 531 |
-
"eval_runtime": 116.5039,
|
| 532 |
-
"eval_samples_per_second": 41.424,
|
| 533 |
-
"eval_steps_per_second": 20.712,
|
| 534 |
-
"step": 1750
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 5.247317694126205,
|
| 538 |
-
"grad_norm": 42.0803337097168,
|
| 539 |
-
"learning_rate": 7.376093294460641e-05,
|
| 540 |
-
"loss": 0.9036,
|
| 541 |
-
"step": 1800
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 5.247317694126205,
|
| 545 |
-
"eval_loss": 0.8666579127311707,
|
| 546 |
-
"eval_runtime": 116.7061,
|
| 547 |
-
"eval_samples_per_second": 41.352,
|
| 548 |
-
"eval_steps_per_second": 20.676,
|
| 549 |
-
"step": 1800
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 5.3927986906710315,
|
| 553 |
-
"grad_norm": 48.07575988769531,
|
| 554 |
-
"learning_rate": 7.303206997084548e-05,
|
| 555 |
-
"loss": 0.9033,
|
| 556 |
-
"step": 1850
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 5.3927986906710315,
|
| 560 |
-
"eval_loss": 0.8641985058784485,
|
| 561 |
-
"eval_runtime": 116.5215,
|
| 562 |
-
"eval_samples_per_second": 41.417,
|
| 563 |
-
"eval_steps_per_second": 20.709,
|
| 564 |
-
"step": 1850
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 5.538279687215858,
|
| 568 |
-
"grad_norm": 75.88548278808594,
|
| 569 |
-
"learning_rate": 7.230320699708455e-05,
|
| 570 |
-
"loss": 0.896,
|
| 571 |
-
"step": 1900
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 5.538279687215858,
|
| 575 |
-
"eval_loss": 0.858232319355011,
|
| 576 |
-
"eval_runtime": 116.5164,
|
| 577 |
-
"eval_samples_per_second": 41.419,
|
| 578 |
-
"eval_steps_per_second": 20.71,
|
| 579 |
-
"step": 1900
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 5.683760683760684,
|
| 583 |
-
"grad_norm": 56.688232421875,
|
| 584 |
-
"learning_rate": 7.157434402332361e-05,
|
| 585 |
-
"loss": 0.8861,
|
| 586 |
-
"step": 1950
|
| 587 |
-
},
|
| 588 |
-
{
|
| 589 |
-
"epoch": 5.683760683760684,
|
| 590 |
-
"eval_loss": 0.8528502583503723,
|
| 591 |
-
"eval_runtime": 116.6212,
|
| 592 |
-
"eval_samples_per_second": 41.382,
|
| 593 |
-
"eval_steps_per_second": 20.691,
|
| 594 |
-
"step": 1950
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 5.82924168030551,
|
| 598 |
-
"grad_norm": 72.85212707519531,
|
| 599 |
-
"learning_rate": 7.08454810495627e-05,
|
| 600 |
-
"loss": 0.8876,
|
| 601 |
-
"step": 2000
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 5.82924168030551,
|
| 605 |
-
"eval_loss": 0.8465431928634644,
|
| 606 |
-
"eval_runtime": 116.5542,
|
| 607 |
-
"eval_samples_per_second": 41.406,
|
| 608 |
-
"eval_steps_per_second": 20.703,
|
| 609 |
-
"step": 2000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 5.9747226768503365,
|
| 613 |
-
"grad_norm": 86.02615356445312,
|
| 614 |
-
"learning_rate": 7.011661807580175e-05,
|
| 615 |
-
"loss": 0.8888,
|
| 616 |
-
"step": 2050
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 5.9747226768503365,
|
| 620 |
-
"eval_loss": 0.8466458916664124,
|
| 621 |
-
"eval_runtime": 116.5437,
|
| 622 |
-
"eval_samples_per_second": 41.409,
|
| 623 |
-
"eval_steps_per_second": 20.705,
|
| 624 |
-
"step": 2050
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 6.122204037097654,
|
| 628 |
-
"grad_norm": 58.913875579833984,
|
| 629 |
-
"learning_rate": 6.938775510204082e-05,
|
| 630 |
-
"loss": 0.8781,
|
| 631 |
-
"step": 2100
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 6.122204037097654,
|
| 635 |
-
"eval_loss": 0.8386573791503906,
|
| 636 |
-
"eval_runtime": 116.5363,
|
| 637 |
-
"eval_samples_per_second": 41.412,
|
| 638 |
-
"eval_steps_per_second": 20.706,
|
| 639 |
-
"step": 2100
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 6.26768503364248,
|
| 643 |
-
"grad_norm": 60.61414337158203,
|
| 644 |
-
"learning_rate": 6.865889212827988e-05,
|
| 645 |
-
"loss": 0.8687,
|
| 646 |
-
"step": 2150
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 6.26768503364248,
|
| 650 |
-
"eval_loss": 0.8333680033683777,
|
| 651 |
-
"eval_runtime": 116.5626,
|
| 652 |
-
"eval_samples_per_second": 41.403,
|
| 653 |
-
"eval_steps_per_second": 20.701,
|
| 654 |
-
"step": 2150
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 6.413166030187307,
|
| 658 |
-
"grad_norm": 53.10786437988281,
|
| 659 |
-
"learning_rate": 6.793002915451895e-05,
|
| 660 |
-
"loss": 0.8615,
|
| 661 |
-
"step": 2200
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 6.413166030187307,
|
| 665 |
-
"eval_loss": 0.8312305808067322,
|
| 666 |
-
"eval_runtime": 116.6202,
|
| 667 |
-
"eval_samples_per_second": 41.382,
|
| 668 |
-
"eval_steps_per_second": 20.691,
|
| 669 |
-
"step": 2200
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 6.5586470267321335,
|
| 673 |
-
"grad_norm": 95.9214859008789,
|
| 674 |
-
"learning_rate": 6.720116618075802e-05,
|
| 675 |
-
"loss": 0.8689,
|
| 676 |
-
"step": 2250
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 6.5586470267321335,
|
| 680 |
-
"eval_loss": 0.8315210938453674,
|
| 681 |
-
"eval_runtime": 116.7326,
|
| 682 |
-
"eval_samples_per_second": 41.342,
|
| 683 |
-
"eval_steps_per_second": 20.671,
|
| 684 |
-
"step": 2250
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 6.70412802327696,
|
| 688 |
-
"grad_norm": 98.01502990722656,
|
| 689 |
-
"learning_rate": 6.647230320699709e-05,
|
| 690 |
-
"loss": 0.8596,
|
| 691 |
-
"step": 2300
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 6.70412802327696,
|
| 695 |
-
"eval_loss": 0.8249507546424866,
|
| 696 |
-
"eval_runtime": 116.5909,
|
| 697 |
-
"eval_samples_per_second": 41.393,
|
| 698 |
-
"eval_steps_per_second": 20.696,
|
| 699 |
-
"step": 2300
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 6.849609019821786,
|
| 703 |
-
"grad_norm": 42.227508544921875,
|
| 704 |
-
"learning_rate": 6.574344023323615e-05,
|
| 705 |
-
"loss": 0.8546,
|
| 706 |
-
"step": 2350
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 6.849609019821786,
|
| 710 |
-
"eval_loss": 0.8198081254959106,
|
| 711 |
-
"eval_runtime": 116.65,
|
| 712 |
-
"eval_samples_per_second": 41.372,
|
| 713 |
-
"eval_steps_per_second": 20.686,
|
| 714 |
-
"step": 2350
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 6.995090016366612,
|
| 718 |
-
"grad_norm": 67.66838073730469,
|
| 719 |
-
"learning_rate": 6.501457725947522e-05,
|
| 720 |
-
"loss": 0.8589,
|
| 721 |
-
"step": 2400
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 6.995090016366612,
|
| 725 |
-
"eval_loss": 0.8165889978408813,
|
| 726 |
-
"eval_runtime": 116.6484,
|
| 727 |
-
"eval_samples_per_second": 41.372,
|
| 728 |
-
"eval_steps_per_second": 20.686,
|
| 729 |
-
"step": 2400
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 7.14257137661393,
|
| 733 |
-
"grad_norm": 49.362857818603516,
|
| 734 |
-
"learning_rate": 6.428571428571429e-05,
|
| 735 |
-
"loss": 0.8429,
|
| 736 |
-
"step": 2450
|
| 737 |
-
},
|
| 738 |
-
{
|
| 739 |
-
"epoch": 7.14257137661393,
|
| 740 |
-
"eval_loss": 0.8078887462615967,
|
| 741 |
-
"eval_runtime": 116.7726,
|
| 742 |
-
"eval_samples_per_second": 41.328,
|
| 743 |
-
"eval_steps_per_second": 20.664,
|
| 744 |
-
"step": 2450
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 7.288052373158756,
|
| 748 |
-
"grad_norm": 73.06404113769531,
|
| 749 |
-
"learning_rate": 6.355685131195336e-05,
|
| 750 |
-
"loss": 0.8374,
|
| 751 |
-
"step": 2500
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 7.288052373158756,
|
| 755 |
-
"eval_loss": 0.8106158375740051,
|
| 756 |
-
"eval_runtime": 116.3346,
|
| 757 |
-
"eval_samples_per_second": 41.484,
|
| 758 |
-
"eval_steps_per_second": 20.742,
|
| 759 |
-
"step": 2500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 7.433533369703582,
|
| 763 |
-
"grad_norm": 72.56304931640625,
|
| 764 |
-
"learning_rate": 6.282798833819243e-05,
|
| 765 |
-
"loss": 0.8371,
|
| 766 |
-
"step": 2550
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 7.433533369703582,
|
| 770 |
-
"eval_loss": 0.8068947792053223,
|
| 771 |
-
"eval_runtime": 116.5087,
|
| 772 |
-
"eval_samples_per_second": 41.422,
|
| 773 |
-
"eval_steps_per_second": 20.711,
|
| 774 |
-
"step": 2550
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 7.579014366248408,
|
| 778 |
-
"grad_norm": 62.97400665283203,
|
| 779 |
-
"learning_rate": 6.209912536443149e-05,
|
| 780 |
-
"loss": 0.8323,
|
| 781 |
-
"step": 2600
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 7.579014366248408,
|
| 785 |
-
"eval_loss": 0.804251492023468,
|
| 786 |
-
"eval_runtime": 116.5526,
|
| 787 |
-
"eval_samples_per_second": 41.406,
|
| 788 |
-
"eval_steps_per_second": 20.703,
|
| 789 |
-
"step": 2600
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 7.724495362793235,
|
| 793 |
-
"grad_norm": 46.578704833984375,
|
| 794 |
-
"learning_rate": 6.137026239067056e-05,
|
| 795 |
-
"loss": 0.8381,
|
| 796 |
-
"step": 2650
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 7.724495362793235,
|
| 800 |
-
"eval_loss": 0.7970871925354004,
|
| 801 |
-
"eval_runtime": 116.4414,
|
| 802 |
-
"eval_samples_per_second": 41.446,
|
| 803 |
-
"eval_steps_per_second": 20.723,
|
| 804 |
-
"step": 2650
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 7.869976359338062,
|
| 808 |
-
"grad_norm": 63.2580680847168,
|
| 809 |
-
"learning_rate": 6.0641399416909626e-05,
|
| 810 |
-
"loss": 0.833,
|
| 811 |
-
"step": 2700
|
| 812 |
-
},
|
| 813 |
-
{
|
| 814 |
-
"epoch": 7.869976359338062,
|
| 815 |
-
"eval_loss": 0.7932024002075195,
|
| 816 |
-
"eval_runtime": 116.5379,
|
| 817 |
-
"eval_samples_per_second": 41.411,
|
| 818 |
-
"eval_steps_per_second": 20.706,
|
| 819 |
-
"step": 2700
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 8.01745771958538,
|
| 823 |
-
"grad_norm": 81.85997009277344,
|
| 824 |
-
"learning_rate": 5.991253644314869e-05,
|
| 825 |
-
"loss": 0.8295,
|
| 826 |
-
"step": 2750
|
| 827 |
-
},
|
| 828 |
-
{
|
| 829 |
-
"epoch": 8.01745771958538,
|
| 830 |
-
"eval_loss": 0.7912312746047974,
|
| 831 |
-
"eval_runtime": 117.8188,
|
| 832 |
-
"eval_samples_per_second": 40.961,
|
| 833 |
-
"eval_steps_per_second": 20.481,
|
| 834 |
-
"step": 2750
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 8.162938716130206,
|
| 838 |
-
"grad_norm": 87.41828918457031,
|
| 839 |
-
"learning_rate": 5.918367346938776e-05,
|
| 840 |
-
"loss": 0.8146,
|
| 841 |
-
"step": 2800
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 8.162938716130206,
|
| 845 |
-
"eval_loss": 0.7906960248947144,
|
| 846 |
-
"eval_runtime": 116.6653,
|
| 847 |
-
"eval_samples_per_second": 41.366,
|
| 848 |
-
"eval_steps_per_second": 20.683,
|
| 849 |
-
"step": 2800
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 8.308419712675033,
|
| 853 |
-
"grad_norm": 50.575706481933594,
|
| 854 |
-
"learning_rate": 5.845481049562682e-05,
|
| 855 |
-
"loss": 0.8104,
|
| 856 |
-
"step": 2850
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 8.308419712675033,
|
| 860 |
-
"eval_loss": 0.7849209308624268,
|
| 861 |
-
"eval_runtime": 116.6593,
|
| 862 |
-
"eval_samples_per_second": 41.368,
|
| 863 |
-
"eval_steps_per_second": 20.684,
|
| 864 |
-
"step": 2850
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 8.453900709219859,
|
| 868 |
-
"grad_norm": 55.96903991699219,
|
| 869 |
-
"learning_rate": 5.77259475218659e-05,
|
| 870 |
-
"loss": 0.8135,
|
| 871 |
-
"step": 2900
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 8.453900709219859,
|
| 875 |
-
"eval_loss": 0.7816675305366516,
|
| 876 |
-
"eval_runtime": 116.6099,
|
| 877 |
-
"eval_samples_per_second": 41.386,
|
| 878 |
-
"eval_steps_per_second": 20.693,
|
| 879 |
-
"step": 2900
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 8.599381705764685,
|
| 883 |
-
"grad_norm": 46.01133346557617,
|
| 884 |
-
"learning_rate": 5.6997084548104964e-05,
|
| 885 |
-
"loss": 0.8061,
|
| 886 |
-
"step": 2950
|
| 887 |
-
},
|
| 888 |
-
{
|
| 889 |
-
"epoch": 8.599381705764685,
|
| 890 |
-
"eval_loss": 0.7791692614555359,
|
| 891 |
-
"eval_runtime": 116.4849,
|
| 892 |
-
"eval_samples_per_second": 41.43,
|
| 893 |
-
"eval_steps_per_second": 20.715,
|
| 894 |
-
"step": 2950
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 8.744862702309511,
|
| 898 |
-
"grad_norm": 57.846012115478516,
|
| 899 |
-
"learning_rate": 5.626822157434403e-05,
|
| 900 |
-
"loss": 0.8114,
|
| 901 |
-
"step": 3000
|
| 902 |
-
},
|
| 903 |
-
{
|
| 904 |
-
"epoch": 8.744862702309511,
|
| 905 |
-
"eval_loss": 0.77262282371521,
|
| 906 |
-
"eval_runtime": 116.493,
|
| 907 |
-
"eval_samples_per_second": 41.427,
|
| 908 |
-
"eval_steps_per_second": 20.714,
|
| 909 |
-
"step": 3000
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 8.890343698854338,
|
| 913 |
-
"grad_norm": 86.41826629638672,
|
| 914 |
-
"learning_rate": 5.553935860058309e-05,
|
| 915 |
-
"loss": 0.8061,
|
| 916 |
-
"step": 3050
|
| 917 |
-
},
|
| 918 |
-
{
|
| 919 |
-
"epoch": 8.890343698854338,
|
| 920 |
-
"eval_loss": 0.7756626605987549,
|
| 921 |
-
"eval_runtime": 116.3032,
|
| 922 |
-
"eval_samples_per_second": 41.495,
|
| 923 |
-
"eval_steps_per_second": 20.747,
|
| 924 |
-
"step": 3050
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 9.037825059101655,
|
| 928 |
-
"grad_norm": 50.474700927734375,
|
| 929 |
-
"learning_rate": 5.4810495626822155e-05,
|
| 930 |
-
"loss": 0.8096,
|
| 931 |
-
"step": 3100
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 9.037825059101655,
|
| 935 |
-
"eval_loss": 0.7688556909561157,
|
| 936 |
-
"eval_runtime": 118.1452,
|
| 937 |
-
"eval_samples_per_second": 40.848,
|
| 938 |
-
"eval_steps_per_second": 20.424,
|
| 939 |
-
"step": 3100
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 9.183306055646481,
|
| 943 |
-
"grad_norm": 62.775550842285156,
|
| 944 |
-
"learning_rate": 5.408163265306123e-05,
|
| 945 |
-
"loss": 0.7868,
|
| 946 |
-
"step": 3150
|
| 947 |
-
},
|
| 948 |
-
{
|
| 949 |
-
"epoch": 9.183306055646481,
|
| 950 |
-
"eval_loss": 0.7667975425720215,
|
| 951 |
-
"eval_runtime": 116.4923,
|
| 952 |
-
"eval_samples_per_second": 41.428,
|
| 953 |
-
"eval_steps_per_second": 20.714,
|
| 954 |
-
"step": 3150
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 9.328787052191307,
|
| 958 |
-
"grad_norm": 39.37485122680664,
|
| 959 |
-
"learning_rate": 5.3352769679300295e-05,
|
| 960 |
-
"loss": 0.7921,
|
| 961 |
-
"step": 3200
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 9.328787052191307,
|
| 965 |
-
"eval_loss": 0.7623693346977234,
|
| 966 |
-
"eval_runtime": 116.502,
|
| 967 |
-
"eval_samples_per_second": 41.424,
|
| 968 |
-
"eval_steps_per_second": 20.712,
|
| 969 |
-
"step": 3200
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 9.474268048736134,
|
| 973 |
-
"grad_norm": 52.29838943481445,
|
| 974 |
-
"learning_rate": 5.262390670553936e-05,
|
| 975 |
-
"loss": 0.7899,
|
| 976 |
-
"step": 3250
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 9.474268048736134,
|
| 980 |
-
"eval_loss": 0.7582478523254395,
|
| 981 |
-
"eval_runtime": 116.5025,
|
| 982 |
-
"eval_samples_per_second": 41.424,
|
| 983 |
-
"eval_steps_per_second": 20.712,
|
| 984 |
-
"step": 3250
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 9.61974904528096,
|
| 988 |
-
"grad_norm": 58.95834732055664,
|
| 989 |
-
"learning_rate": 5.189504373177842e-05,
|
| 990 |
-
"loss": 0.7869,
|
| 991 |
-
"step": 3300
|
| 992 |
-
},
|
| 993 |
-
{
|
| 994 |
-
"epoch": 9.61974904528096,
|
| 995 |
-
"eval_loss": 0.760308563709259,
|
| 996 |
-
"eval_runtime": 116.4675,
|
| 997 |
-
"eval_samples_per_second": 41.436,
|
| 998 |
-
"eval_steps_per_second": 20.718,
|
| 999 |
-
"step": 3300
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 9.765230041825786,
|
| 1003 |
-
"grad_norm": 58.980167388916016,
|
| 1004 |
-
"learning_rate": 5.116618075801749e-05,
|
| 1005 |
-
"loss": 0.7834,
|
| 1006 |
-
"step": 3350
|
| 1007 |
-
},
|
| 1008 |
-
{
|
| 1009 |
-
"epoch": 9.765230041825786,
|
| 1010 |
-
"eval_loss": 0.7558280229568481,
|
| 1011 |
-
"eval_runtime": 116.6353,
|
| 1012 |
-
"eval_samples_per_second": 41.377,
|
| 1013 |
-
"eval_steps_per_second": 20.688,
|
| 1014 |
-
"step": 3350
|
| 1015 |
-
},
|
| 1016 |
-
{
|
| 1017 |
-
"epoch": 9.910711038370613,
|
| 1018 |
-
"grad_norm": 47.75642013549805,
|
| 1019 |
-
"learning_rate": 5.043731778425656e-05,
|
| 1020 |
-
"loss": 0.789,
|
| 1021 |
-
"step": 3400
|
| 1022 |
-
},
|
| 1023 |
-
{
|
| 1024 |
-
"epoch": 9.910711038370613,
|
| 1025 |
-
"eval_loss": 0.7518753409385681,
|
| 1026 |
-
"eval_runtime": 116.5888,
|
| 1027 |
-
"eval_samples_per_second": 41.393,
|
| 1028 |
-
"eval_steps_per_second": 20.697,
|
| 1029 |
-
"step": 3400
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 10.05819239861793,
|
| 1033 |
-
"grad_norm": 73.24003601074219,
|
| 1034 |
-
"learning_rate": 4.970845481049563e-05,
|
| 1035 |
-
"loss": 0.7873,
|
| 1036 |
-
"step": 3450
|
| 1037 |
-
},
|
| 1038 |
-
{
|
| 1039 |
-
"epoch": 10.05819239861793,
|
| 1040 |
-
"eval_loss": 0.7501546144485474,
|
| 1041 |
-
"eval_runtime": 117.2672,
|
| 1042 |
-
"eval_samples_per_second": 41.154,
|
| 1043 |
-
"eval_steps_per_second": 20.577,
|
| 1044 |
-
"step": 3450
|
| 1045 |
-
},
|
| 1046 |
-
{
|
| 1047 |
-
"epoch": 10.203673395162756,
|
| 1048 |
-
"grad_norm": 64.36954498291016,
|
| 1049 |
-
"learning_rate": 4.89795918367347e-05,
|
| 1050 |
-
"loss": 0.7681,
|
| 1051 |
-
"step": 3500
|
| 1052 |
-
},
|
| 1053 |
-
{
|
| 1054 |
-
"epoch": 10.203673395162756,
|
| 1055 |
-
"eval_loss": 0.7464452385902405,
|
| 1056 |
-
"eval_runtime": 116.6327,
|
| 1057 |
-
"eval_samples_per_second": 41.378,
|
| 1058 |
-
"eval_steps_per_second": 20.689,
|
| 1059 |
-
"step": 3500
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 10.349154391707582,
|
| 1063 |
-
"grad_norm": 53.44997024536133,
|
| 1064 |
-
"learning_rate": 4.825072886297377e-05,
|
| 1065 |
-
"loss": 0.7657,
|
| 1066 |
-
"step": 3550
|
| 1067 |
-
},
|
| 1068 |
-
{
|
| 1069 |
-
"epoch": 10.349154391707582,
|
| 1070 |
-
"eval_loss": 0.7468724250793457,
|
| 1071 |
-
"eval_runtime": 116.3252,
|
| 1072 |
-
"eval_samples_per_second": 41.487,
|
| 1073 |
-
"eval_steps_per_second": 20.744,
|
| 1074 |
-
"step": 3550
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 10.49463538825241,
|
| 1078 |
-
"grad_norm": 54.961727142333984,
|
| 1079 |
-
"learning_rate": 4.752186588921283e-05,
|
| 1080 |
-
"loss": 0.7746,
|
| 1081 |
-
"step": 3600
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 10.49463538825241,
|
| 1085 |
-
"eval_loss": 0.7428314685821533,
|
| 1086 |
-
"eval_runtime": 116.5215,
|
| 1087 |
-
"eval_samples_per_second": 41.417,
|
| 1088 |
-
"eval_steps_per_second": 20.709,
|
| 1089 |
-
"step": 3600
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 10.640116384797237,
|
| 1093 |
-
"grad_norm": 72.78599548339844,
|
| 1094 |
-
"learning_rate": 4.6793002915451894e-05,
|
| 1095 |
-
"loss": 0.7666,
|
| 1096 |
-
"step": 3650
|
| 1097 |
-
},
|
| 1098 |
-
{
|
| 1099 |
-
"epoch": 10.640116384797237,
|
| 1100 |
-
"eval_loss": 0.7381331324577332,
|
| 1101 |
-
"eval_runtime": 116.4741,
|
| 1102 |
-
"eval_samples_per_second": 41.434,
|
| 1103 |
-
"eval_steps_per_second": 20.717,
|
| 1104 |
-
"step": 3650
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 10.785597381342063,
|
| 1108 |
-
"grad_norm": 57.89088821411133,
|
| 1109 |
-
"learning_rate": 4.6064139941690965e-05,
|
| 1110 |
-
"loss": 0.7685,
|
| 1111 |
-
"step": 3700
|
| 1112 |
-
},
|
| 1113 |
-
{
|
| 1114 |
-
"epoch": 10.785597381342063,
|
| 1115 |
-
"eval_loss": 0.7335684299468994,
|
| 1116 |
-
"eval_runtime": 116.9024,
|
| 1117 |
-
"eval_samples_per_second": 41.282,
|
| 1118 |
-
"eval_steps_per_second": 20.641,
|
| 1119 |
-
"step": 3700
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 10.93107837788689,
|
| 1123 |
-
"grad_norm": 63.70279312133789,
|
| 1124 |
-
"learning_rate": 4.533527696793003e-05,
|
| 1125 |
-
"loss": 0.7666,
|
| 1126 |
-
"step": 3750
|
| 1127 |
-
},
|
| 1128 |
-
{
|
| 1129 |
-
"epoch": 10.93107837788689,
|
| 1130 |
-
"eval_loss": 0.7345423102378845,
|
| 1131 |
-
"eval_runtime": 116.4317,
|
| 1132 |
-
"eval_samples_per_second": 41.449,
|
| 1133 |
-
"eval_steps_per_second": 20.725,
|
| 1134 |
-
"step": 3750
|
| 1135 |
-
},
|
| 1136 |
-
{
|
| 1137 |
-
"epoch": 11.078559738134206,
|
| 1138 |
-
"grad_norm": 61.34590530395508,
|
| 1139 |
-
"learning_rate": 4.46064139941691e-05,
|
| 1140 |
-
"loss": 0.7639,
|
| 1141 |
-
"step": 3800
|
| 1142 |
-
},
|
| 1143 |
-
{
|
| 1144 |
-
"epoch": 11.078559738134206,
|
| 1145 |
-
"eval_loss": 0.731596052646637,
|
| 1146 |
-
"eval_runtime": 116.4502,
|
| 1147 |
-
"eval_samples_per_second": 41.443,
|
| 1148 |
-
"eval_steps_per_second": 20.721,
|
| 1149 |
-
"step": 3800
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 11.224040734679033,
|
| 1153 |
-
"grad_norm": 74.86040496826172,
|
| 1154 |
-
"learning_rate": 4.387755102040816e-05,
|
| 1155 |
-
"loss": 0.7515,
|
| 1156 |
-
"step": 3850
|
| 1157 |
-
},
|
| 1158 |
-
{
|
| 1159 |
-
"epoch": 11.224040734679033,
|
| 1160 |
-
"eval_loss": 0.7297601699829102,
|
| 1161 |
-
"eval_runtime": 116.5358,
|
| 1162 |
-
"eval_samples_per_second": 41.412,
|
| 1163 |
-
"eval_steps_per_second": 20.706,
|
| 1164 |
-
"step": 3850
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 11.369521731223859,
|
| 1168 |
-
"grad_norm": 53.91130828857422,
|
| 1169 |
-
"learning_rate": 4.314868804664723e-05,
|
| 1170 |
-
"loss": 0.7567,
|
| 1171 |
-
"step": 3900
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 11.369521731223859,
|
| 1175 |
-
"eval_loss": 0.7294139266014099,
|
| 1176 |
-
"eval_runtime": 116.7175,
|
| 1177 |
-
"eval_samples_per_second": 41.348,
|
| 1178 |
-
"eval_steps_per_second": 20.674,
|
| 1179 |
-
"step": 3900
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 11.515002727768685,
|
| 1183 |
-
"grad_norm": 63.51275634765625,
|
| 1184 |
-
"learning_rate": 4.2419825072886296e-05,
|
| 1185 |
-
"loss": 0.7519,
|
| 1186 |
-
"step": 3950
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 11.515002727768685,
|
| 1190 |
-
"eval_loss": 0.7229241728782654,
|
| 1191 |
-
"eval_runtime": 116.7452,
|
| 1192 |
-
"eval_samples_per_second": 41.338,
|
| 1193 |
-
"eval_steps_per_second": 20.669,
|
| 1194 |
-
"step": 3950
|
| 1195 |
-
},
|
| 1196 |
-
{
|
| 1197 |
-
"epoch": 11.660483724313512,
|
| 1198 |
-
"grad_norm": 44.026119232177734,
|
| 1199 |
-
"learning_rate": 4.1690962099125366e-05,
|
| 1200 |
-
"loss": 0.7494,
|
| 1201 |
-
"step": 4000
|
| 1202 |
-
},
|
| 1203 |
-
{
|
| 1204 |
-
"epoch": 11.660483724313512,
|
| 1205 |
-
"eval_loss": 0.7239974141120911,
|
| 1206 |
-
"eval_runtime": 116.4117,
|
| 1207 |
-
"eval_samples_per_second": 41.456,
|
| 1208 |
-
"eval_steps_per_second": 20.728,
|
| 1209 |
-
"step": 4000
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 11.805964720858338,
|
| 1213 |
-
"grad_norm": 54.68580627441406,
|
| 1214 |
-
"learning_rate": 4.0962099125364436e-05,
|
| 1215 |
-
"loss": 0.7452,
|
| 1216 |
-
"step": 4050
|
| 1217 |
-
},
|
| 1218 |
-
{
|
| 1219 |
-
"epoch": 11.805964720858338,
|
| 1220 |
-
"eval_loss": 0.7212454080581665,
|
| 1221 |
-
"eval_runtime": 116.4854,
|
| 1222 |
-
"eval_samples_per_second": 41.43,
|
| 1223 |
-
"eval_steps_per_second": 20.715,
|
| 1224 |
-
"step": 4050
|
| 1225 |
-
},
|
| 1226 |
-
{
|
| 1227 |
-
"epoch": 11.951445717403164,
|
| 1228 |
-
"grad_norm": 68.30030822753906,
|
| 1229 |
-
"learning_rate": 4.02332361516035e-05,
|
| 1230 |
-
"loss": 0.7498,
|
| 1231 |
-
"step": 4100
|
| 1232 |
-
},
|
| 1233 |
-
{
|
| 1234 |
-
"epoch": 11.951445717403164,
|
| 1235 |
-
"eval_loss": 0.7175871133804321,
|
| 1236 |
-
"eval_runtime": 116.3899,
|
| 1237 |
-
"eval_samples_per_second": 41.464,
|
| 1238 |
-
"eval_steps_per_second": 20.732,
|
| 1239 |
-
"step": 4100
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 12.098927077650481,
|
| 1243 |
-
"grad_norm": 61.96623611450195,
|
| 1244 |
-
"learning_rate": 3.950437317784257e-05,
|
| 1245 |
-
"loss": 0.7457,
|
| 1246 |
-
"step": 4150
|
| 1247 |
-
},
|
| 1248 |
-
{
|
| 1249 |
-
"epoch": 12.098927077650481,
|
| 1250 |
-
"eval_loss": 0.7167355418205261,
|
| 1251 |
-
"eval_runtime": 116.6266,
|
| 1252 |
-
"eval_samples_per_second": 41.38,
|
| 1253 |
-
"eval_steps_per_second": 20.69,
|
| 1254 |
-
"step": 4150
|
| 1255 |
-
},
|
| 1256 |
-
{
|
| 1257 |
-
"epoch": 12.244408074195308,
|
| 1258 |
-
"grad_norm": 112.12747955322266,
|
| 1259 |
-
"learning_rate": 3.8775510204081634e-05,
|
| 1260 |
-
"loss": 0.7356,
|
| 1261 |
-
"step": 4200
|
| 1262 |
-
},
|
| 1263 |
-
{
|
| 1264 |
-
"epoch": 12.244408074195308,
|
| 1265 |
-
"eval_loss": 0.7129220962524414,
|
| 1266 |
-
"eval_runtime": 116.6342,
|
| 1267 |
-
"eval_samples_per_second": 41.377,
|
| 1268 |
-
"eval_steps_per_second": 20.689,
|
| 1269 |
-
"step": 4200
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 12.389889070740134,
|
| 1273 |
-
"grad_norm": 51.88462829589844,
|
| 1274 |
-
"learning_rate": 3.8046647230320704e-05,
|
| 1275 |
-
"loss": 0.7361,
|
| 1276 |
-
"step": 4250
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 12.389889070740134,
|
| 1280 |
-
"eval_loss": 0.7131578326225281,
|
| 1281 |
-
"eval_runtime": 116.4086,
|
| 1282 |
-
"eval_samples_per_second": 41.457,
|
| 1283 |
-
"eval_steps_per_second": 20.729,
|
| 1284 |
-
"step": 4250
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 12.53537006728496,
|
| 1288 |
-
"grad_norm": 65.30543518066406,
|
| 1289 |
-
"learning_rate": 3.731778425655977e-05,
|
| 1290 |
-
"loss": 0.7323,
|
| 1291 |
-
"step": 4300
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 12.53537006728496,
|
| 1295 |
-
"eval_loss": 0.7133215665817261,
|
| 1296 |
-
"eval_runtime": 116.5949,
|
| 1297 |
-
"eval_samples_per_second": 41.391,
|
| 1298 |
-
"eval_steps_per_second": 20.696,
|
| 1299 |
-
"step": 4300
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 12.680851063829786,
|
| 1303 |
-
"grad_norm": 59.78929901123047,
|
| 1304 |
-
"learning_rate": 3.658892128279884e-05,
|
| 1305 |
-
"loss": 0.7358,
|
| 1306 |
-
"step": 4350
|
| 1307 |
-
},
|
| 1308 |
-
{
|
| 1309 |
-
"epoch": 12.680851063829786,
|
| 1310 |
-
"eval_loss": 0.7077481746673584,
|
| 1311 |
-
"eval_runtime": 116.19,
|
| 1312 |
-
"eval_samples_per_second": 41.535,
|
| 1313 |
-
"eval_steps_per_second": 20.768,
|
| 1314 |
-
"step": 4350
|
| 1315 |
-
},
|
| 1316 |
-
{
|
| 1317 |
-
"epoch": 12.826332060374614,
|
| 1318 |
-
"grad_norm": 43.919498443603516,
|
| 1319 |
-
"learning_rate": 3.58600583090379e-05,
|
| 1320 |
-
"loss": 0.7359,
|
| 1321 |
-
"step": 4400
|
| 1322 |
-
},
|
| 1323 |
-
{
|
| 1324 |
-
"epoch": 12.826332060374614,
|
| 1325 |
-
"eval_loss": 0.7035172581672668,
|
| 1326 |
-
"eval_runtime": 116.6188,
|
| 1327 |
-
"eval_samples_per_second": 41.383,
|
| 1328 |
-
"eval_steps_per_second": 20.691,
|
| 1329 |
-
"step": 4400
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 12.97181305691944,
|
| 1333 |
-
"grad_norm": 48.08681869506836,
|
| 1334 |
-
"learning_rate": 3.5131195335276965e-05,
|
| 1335 |
-
"loss": 0.7392,
|
| 1336 |
-
"step": 4450
|
| 1337 |
-
},
|
| 1338 |
-
{
|
| 1339 |
-
"epoch": 12.97181305691944,
|
| 1340 |
-
"eval_loss": 0.7051976323127747,
|
| 1341 |
-
"eval_runtime": 116.393,
|
| 1342 |
-
"eval_samples_per_second": 41.463,
|
| 1343 |
-
"eval_steps_per_second": 20.731,
|
| 1344 |
-
"step": 4450
|
| 1345 |
-
},
|
| 1346 |
-
{
|
| 1347 |
-
"epoch": 13.119294417166758,
|
| 1348 |
-
"grad_norm": 64.73822021484375,
|
| 1349 |
-
"learning_rate": 3.4402332361516035e-05,
|
| 1350 |
-
"loss": 0.7291,
|
| 1351 |
-
"step": 4500
|
| 1352 |
-
},
|
| 1353 |
-
{
|
| 1354 |
-
"epoch": 13.119294417166758,
|
| 1355 |
-
"eval_loss": 0.705399751663208,
|
| 1356 |
-
"eval_runtime": 116.3819,
|
| 1357 |
-
"eval_samples_per_second": 41.467,
|
| 1358 |
-
"eval_steps_per_second": 20.733,
|
| 1359 |
-
"step": 4500
|
| 1360 |
-
},
|
| 1361 |
-
{
|
| 1362 |
-
"epoch": 13.264775413711584,
|
| 1363 |
-
"grad_norm": 48.628440856933594,
|
| 1364 |
-
"learning_rate": 3.36734693877551e-05,
|
| 1365 |
-
"loss": 0.726,
|
| 1366 |
-
"step": 4550
|
| 1367 |
-
},
|
| 1368 |
-
{
|
| 1369 |
-
"epoch": 13.264775413711584,
|
| 1370 |
-
"eval_loss": 0.6991727352142334,
|
| 1371 |
-
"eval_runtime": 116.4083,
|
| 1372 |
-
"eval_samples_per_second": 41.458,
|
| 1373 |
-
"eval_steps_per_second": 20.729,
|
| 1374 |
-
"step": 4550
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 13.41025641025641,
|
| 1378 |
-
"grad_norm": 42.37076187133789,
|
| 1379 |
-
"learning_rate": 3.294460641399417e-05,
|
| 1380 |
-
"loss": 0.7257,
|
| 1381 |
-
"step": 4600
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 13.41025641025641,
|
| 1385 |
-
"eval_loss": 0.6997016668319702,
|
| 1386 |
-
"eval_runtime": 116.654,
|
| 1387 |
-
"eval_samples_per_second": 41.37,
|
| 1388 |
-
"eval_steps_per_second": 20.685,
|
| 1389 |
-
"step": 4600
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 13.555737406801237,
|
| 1393 |
-
"grad_norm": 54.22138977050781,
|
| 1394 |
-
"learning_rate": 3.221574344023324e-05,
|
| 1395 |
-
"loss": 0.721,
|
| 1396 |
-
"step": 4650
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 13.555737406801237,
|
| 1400 |
-
"eval_loss": 0.6972126960754395,
|
| 1401 |
-
"eval_runtime": 116.6938,
|
| 1402 |
-
"eval_samples_per_second": 41.356,
|
| 1403 |
-
"eval_steps_per_second": 20.678,
|
| 1404 |
-
"step": 4650
|
| 1405 |
-
},
|
| 1406 |
-
{
|
| 1407 |
-
"epoch": 13.701218403346063,
|
| 1408 |
-
"grad_norm": 70.08407592773438,
|
| 1409 |
-
"learning_rate": 3.148688046647231e-05,
|
| 1410 |
-
"loss": 0.7219,
|
| 1411 |
-
"step": 4700
|
| 1412 |
-
},
|
| 1413 |
-
{
|
| 1414 |
-
"epoch": 13.701218403346063,
|
| 1415 |
-
"eval_loss": 0.697705864906311,
|
| 1416 |
-
"eval_runtime": 116.8913,
|
| 1417 |
-
"eval_samples_per_second": 41.286,
|
| 1418 |
-
"eval_steps_per_second": 20.643,
|
| 1419 |
-
"step": 4700
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 13.84669939989089,
|
| 1423 |
-
"grad_norm": 59.16844177246094,
|
| 1424 |
-
"learning_rate": 3.0758017492711373e-05,
|
| 1425 |
-
"loss": 0.7206,
|
| 1426 |
-
"step": 4750
|
| 1427 |
-
},
|
| 1428 |
-
{
|
| 1429 |
-
"epoch": 13.84669939989089,
|
| 1430 |
-
"eval_loss": 0.6945058107376099,
|
| 1431 |
-
"eval_runtime": 117.2096,
|
| 1432 |
-
"eval_samples_per_second": 41.174,
|
| 1433 |
-
"eval_steps_per_second": 20.587,
|
| 1434 |
-
"step": 4750
|
| 1435 |
-
},
|
| 1436 |
-
{
|
| 1437 |
-
"epoch": 13.992180396435716,
|
| 1438 |
-
"grad_norm": 70.3475112915039,
|
| 1439 |
-
"learning_rate": 3.0029154518950437e-05,
|
| 1440 |
-
"loss": 0.7173,
|
| 1441 |
-
"step": 4800
|
| 1442 |
-
},
|
| 1443 |
-
{
|
| 1444 |
-
"epoch": 13.992180396435716,
|
| 1445 |
-
"eval_loss": 0.692059338092804,
|
| 1446 |
-
"eval_runtime": 116.7938,
|
| 1447 |
-
"eval_samples_per_second": 41.321,
|
| 1448 |
-
"eval_steps_per_second": 20.66,
|
| 1449 |
-
"step": 4800
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 14.139661756683033,
|
| 1453 |
-
"grad_norm": 49.7451286315918,
|
| 1454 |
-
"learning_rate": 2.9300291545189507e-05,
|
| 1455 |
-
"loss": 0.7159,
|
| 1456 |
-
"step": 4850
|
| 1457 |
-
},
|
| 1458 |
-
{
|
| 1459 |
-
"epoch": 14.139661756683033,
|
| 1460 |
-
"eval_loss": 0.6897387504577637,
|
| 1461 |
-
"eval_runtime": 116.4679,
|
| 1462 |
-
"eval_samples_per_second": 41.436,
|
| 1463 |
-
"eval_steps_per_second": 20.718,
|
| 1464 |
-
"step": 4850
|
| 1465 |
-
},
|
| 1466 |
-
{
|
| 1467 |
-
"epoch": 14.28514275322786,
|
| 1468 |
-
"grad_norm": 43.889793395996094,
|
| 1469 |
-
"learning_rate": 2.857142857142857e-05,
|
| 1470 |
-
"loss": 0.7094,
|
| 1471 |
-
"step": 4900
|
| 1472 |
-
},
|
| 1473 |
-
{
|
| 1474 |
-
"epoch": 14.28514275322786,
|
| 1475 |
-
"eval_loss": 0.690019428730011,
|
| 1476 |
-
"eval_runtime": 116.7069,
|
| 1477 |
-
"eval_samples_per_second": 41.351,
|
| 1478 |
-
"eval_steps_per_second": 20.676,
|
| 1479 |
-
"step": 4900
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 14.430623749772685,
|
| 1483 |
-
"grad_norm": 48.33151626586914,
|
| 1484 |
-
"learning_rate": 2.784256559766764e-05,
|
| 1485 |
-
"loss": 0.7059,
|
| 1486 |
-
"step": 4950
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 14.430623749772685,
|
| 1490 |
-
"eval_loss": 0.6871985793113708,
|
| 1491 |
-
"eval_runtime": 116.4941,
|
| 1492 |
-
"eval_samples_per_second": 41.427,
|
| 1493 |
-
"eval_steps_per_second": 20.713,
|
| 1494 |
-
"step": 4950
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 14.576104746317512,
|
| 1498 |
-
"grad_norm": 53.56142807006836,
|
| 1499 |
-
"learning_rate": 2.7113702623906705e-05,
|
| 1500 |
-
"loss": 0.7124,
|
| 1501 |
-
"step": 5000
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 14.576104746317512,
|
| 1505 |
-
"eval_loss": 0.6843300461769104,
|
| 1506 |
-
"eval_runtime": 116.6779,
|
| 1507 |
-
"eval_samples_per_second": 41.362,
|
| 1508 |
-
"eval_steps_per_second": 20.681,
|
| 1509 |
-
"step": 5000
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 14.721585742862338,
|
| 1513 |
-
"grad_norm": 46.541114807128906,
|
| 1514 |
-
"learning_rate": 2.6384839650145775e-05,
|
| 1515 |
-
"loss": 0.708,
|
| 1516 |
-
"step": 5050
|
| 1517 |
-
},
|
| 1518 |
-
{
|
| 1519 |
-
"epoch": 14.721585742862338,
|
| 1520 |
-
"eval_loss": 0.683451235294342,
|
| 1521 |
-
"eval_runtime": 116.8285,
|
| 1522 |
-
"eval_samples_per_second": 41.308,
|
| 1523 |
-
"eval_steps_per_second": 20.654,
|
| 1524 |
-
"step": 5050
|
| 1525 |
-
},
|
| 1526 |
-
{
|
| 1527 |
-
"epoch": 14.867066739407164,
|
| 1528 |
-
"grad_norm": 47.15498352050781,
|
| 1529 |
-
"learning_rate": 2.5655976676384842e-05,
|
| 1530 |
-
"loss": 0.7062,
|
| 1531 |
-
"step": 5100
|
| 1532 |
-
},
|
| 1533 |
-
{
|
| 1534 |
-
"epoch": 14.867066739407164,
|
| 1535 |
-
"eval_loss": 0.6834617853164673,
|
| 1536 |
-
"eval_runtime": 116.6798,
|
| 1537 |
-
"eval_samples_per_second": 41.361,
|
| 1538 |
-
"eval_steps_per_second": 20.681,
|
| 1539 |
-
"step": 5100
|
| 1540 |
-
},
|
| 1541 |
-
{
|
| 1542 |
-
"epoch": 15.014548099654483,
|
| 1543 |
-
"grad_norm": 63.07834243774414,
|
| 1544 |
-
"learning_rate": 2.492711370262391e-05,
|
| 1545 |
-
"loss": 0.7085,
|
| 1546 |
-
"step": 5150
|
| 1547 |
-
},
|
| 1548 |
-
{
|
| 1549 |
-
"epoch": 15.014548099654483,
|
| 1550 |
-
"eval_loss": 0.6828727722167969,
|
| 1551 |
-
"eval_runtime": 117.6601,
|
| 1552 |
-
"eval_samples_per_second": 41.016,
|
| 1553 |
-
"eval_steps_per_second": 20.508,
|
| 1554 |
-
"step": 5150
|
| 1555 |
-
},
|
| 1556 |
-
{
|
| 1557 |
-
"epoch": 15.16002909619931,
|
| 1558 |
-
"grad_norm": 47.388702392578125,
|
| 1559 |
-
"learning_rate": 2.4198250728862976e-05,
|
| 1560 |
-
"loss": 0.6936,
|
| 1561 |
-
"step": 5200
|
| 1562 |
-
},
|
| 1563 |
-
{
|
| 1564 |
-
"epoch": 15.16002909619931,
|
| 1565 |
-
"eval_loss": 0.6810610294342041,
|
| 1566 |
-
"eval_runtime": 116.7247,
|
| 1567 |
-
"eval_samples_per_second": 41.345,
|
| 1568 |
-
"eval_steps_per_second": 20.673,
|
| 1569 |
-
"step": 5200
|
| 1570 |
-
},
|
| 1571 |
-
{
|
| 1572 |
-
"epoch": 15.305510092744136,
|
| 1573 |
-
"grad_norm": 52.29343032836914,
|
| 1574 |
-
"learning_rate": 2.3469387755102043e-05,
|
| 1575 |
-
"loss": 0.6988,
|
| 1576 |
-
"step": 5250
|
| 1577 |
-
},
|
| 1578 |
-
{
|
| 1579 |
-
"epoch": 15.305510092744136,
|
| 1580 |
-
"eval_loss": 0.6771013736724854,
|
| 1581 |
-
"eval_runtime": 116.5346,
|
| 1582 |
-
"eval_samples_per_second": 41.413,
|
| 1583 |
-
"eval_steps_per_second": 20.706,
|
| 1584 |
-
"step": 5250
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 15.450991089288962,
|
| 1588 |
-
"grad_norm": 52.001590728759766,
|
| 1589 |
-
"learning_rate": 2.2740524781341106e-05,
|
| 1590 |
-
"loss": 0.6969,
|
| 1591 |
-
"step": 5300
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 15.450991089288962,
|
| 1595 |
-
"eval_loss": 0.6796217560768127,
|
| 1596 |
-
"eval_runtime": 116.7294,
|
| 1597 |
-
"eval_samples_per_second": 41.343,
|
| 1598 |
-
"eval_steps_per_second": 20.672,
|
| 1599 |
-
"step": 5300
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 15.596472085833788,
|
| 1603 |
-
"grad_norm": 64.73352813720703,
|
| 1604 |
-
"learning_rate": 2.2011661807580177e-05,
|
| 1605 |
-
"loss": 0.6955,
|
| 1606 |
-
"step": 5350
|
| 1607 |
-
},
|
| 1608 |
-
{
|
| 1609 |
-
"epoch": 15.596472085833788,
|
| 1610 |
-
"eval_loss": 0.6759679317474365,
|
| 1611 |
-
"eval_runtime": 116.3825,
|
| 1612 |
-
"eval_samples_per_second": 41.467,
|
| 1613 |
-
"eval_steps_per_second": 20.733,
|
| 1614 |
-
"step": 5350
|
| 1615 |
-
},
|
| 1616 |
-
{
|
| 1617 |
-
"epoch": 15.741953082378615,
|
| 1618 |
-
"grad_norm": 44.738739013671875,
|
| 1619 |
-
"learning_rate": 2.1282798833819244e-05,
|
| 1620 |
-
"loss": 0.6985,
|
| 1621 |
-
"step": 5400
|
| 1622 |
-
},
|
| 1623 |
-
{
|
| 1624 |
-
"epoch": 15.741953082378615,
|
| 1625 |
-
"eval_loss": 0.6740979552268982,
|
| 1626 |
-
"eval_runtime": 116.6554,
|
| 1627 |
-
"eval_samples_per_second": 41.37,
|
| 1628 |
-
"eval_steps_per_second": 20.685,
|
| 1629 |
-
"step": 5400
|
| 1630 |
-
},
|
| 1631 |
-
{
|
| 1632 |
-
"epoch": 15.887434078923441,
|
| 1633 |
-
"grad_norm": 97.4302749633789,
|
| 1634 |
-
"learning_rate": 2.055393586005831e-05,
|
| 1635 |
-
"loss": 0.6959,
|
| 1636 |
-
"step": 5450
|
| 1637 |
-
},
|
| 1638 |
-
{
|
| 1639 |
-
"epoch": 15.887434078923441,
|
| 1640 |
-
"eval_loss": 0.6725562810897827,
|
| 1641 |
-
"eval_runtime": 116.2903,
|
| 1642 |
-
"eval_samples_per_second": 41.5,
|
| 1643 |
-
"eval_steps_per_second": 20.75,
|
| 1644 |
-
"step": 5450
|
| 1645 |
-
},
|
| 1646 |
-
{
|
| 1647 |
-
"epoch": 16.03491543917076,
|
| 1648 |
-
"grad_norm": 56.59613037109375,
|
| 1649 |
-
"learning_rate": 1.9825072886297377e-05,
|
| 1650 |
-
"loss": 0.7036,
|
| 1651 |
-
"step": 5500
|
| 1652 |
-
},
|
| 1653 |
-
{
|
| 1654 |
-
"epoch": 16.03491543917076,
|
| 1655 |
-
"eval_loss": 0.6698302626609802,
|
| 1656 |
-
"eval_runtime": 118.2947,
|
| 1657 |
-
"eval_samples_per_second": 40.796,
|
| 1658 |
-
"eval_steps_per_second": 20.398,
|
| 1659 |
-
"step": 5500
|
| 1660 |
-
},
|
| 1661 |
-
{
|
| 1662 |
-
"epoch": 16.180396435715586,
|
| 1663 |
-
"grad_norm": 105.1061019897461,
|
| 1664 |
-
"learning_rate": 1.9096209912536444e-05,
|
| 1665 |
-
"loss": 0.6882,
|
| 1666 |
-
"step": 5550
|
| 1667 |
-
},
|
| 1668 |
-
{
|
| 1669 |
-
"epoch": 16.180396435715586,
|
| 1670 |
-
"eval_loss": 0.6712303161621094,
|
| 1671 |
-
"eval_runtime": 116.6934,
|
| 1672 |
-
"eval_samples_per_second": 41.356,
|
| 1673 |
-
"eval_steps_per_second": 20.678,
|
| 1674 |
-
"step": 5550
|
| 1675 |
-
},
|
| 1676 |
-
{
|
| 1677 |
-
"epoch": 16.325877432260413,
|
| 1678 |
-
"grad_norm": 48.417171478271484,
|
| 1679 |
-
"learning_rate": 1.836734693877551e-05,
|
| 1680 |
-
"loss": 0.6877,
|
| 1681 |
-
"step": 5600
|
| 1682 |
-
},
|
| 1683 |
-
{
|
| 1684 |
-
"epoch": 16.325877432260413,
|
| 1685 |
-
"eval_loss": 0.6700485944747925,
|
| 1686 |
-
"eval_runtime": 117.069,
|
| 1687 |
-
"eval_samples_per_second": 41.224,
|
| 1688 |
-
"eval_steps_per_second": 20.612,
|
| 1689 |
-
"step": 5600
|
| 1690 |
-
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 16.47135842880524,
|
| 1693 |
-
"grad_norm": 71.32566833496094,
|
| 1694 |
-
"learning_rate": 1.7638483965014578e-05,
|
| 1695 |
-
"loss": 0.6912,
|
| 1696 |
-
"step": 5650
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 16.47135842880524,
|
| 1700 |
-
"eval_loss": 0.6679624915122986,
|
| 1701 |
-
"eval_runtime": 116.8492,
|
| 1702 |
-
"eval_samples_per_second": 41.301,
|
| 1703 |
-
"eval_steps_per_second": 20.651,
|
| 1704 |
-
"step": 5650
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 16.616839425350065,
|
| 1708 |
-
"grad_norm": 50.87607955932617,
|
| 1709 |
-
"learning_rate": 1.6909620991253645e-05,
|
| 1710 |
-
"loss": 0.6874,
|
| 1711 |
-
"step": 5700
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 16.616839425350065,
|
| 1715 |
-
"eval_loss": 0.6632840633392334,
|
| 1716 |
-
"eval_runtime": 116.778,
|
| 1717 |
-
"eval_samples_per_second": 41.326,
|
| 1718 |
-
"eval_steps_per_second": 20.663,
|
| 1719 |
-
"step": 5700
|
| 1720 |
-
},
|
| 1721 |
-
{
|
| 1722 |
-
"epoch": 16.76232042189489,
|
| 1723 |
-
"grad_norm": 62.6202392578125,
|
| 1724 |
-
"learning_rate": 1.6180758017492712e-05,
|
| 1725 |
-
"loss": 0.685,
|
| 1726 |
-
"step": 5750
|
| 1727 |
-
},
|
| 1728 |
-
{
|
| 1729 |
-
"epoch": 16.76232042189489,
|
| 1730 |
-
"eval_loss": 0.6643268465995789,
|
| 1731 |
-
"eval_runtime": 116.7182,
|
| 1732 |
-
"eval_samples_per_second": 41.347,
|
| 1733 |
-
"eval_steps_per_second": 20.674,
|
| 1734 |
-
"step": 5750
|
| 1735 |
-
},
|
| 1736 |
-
{
|
| 1737 |
-
"epoch": 16.907801418439718,
|
| 1738 |
-
"grad_norm": 57.434730529785156,
|
| 1739 |
-
"learning_rate": 1.545189504373178e-05,
|
| 1740 |
-
"loss": 0.6861,
|
| 1741 |
-
"step": 5800
|
| 1742 |
-
},
|
| 1743 |
-
{
|
| 1744 |
-
"epoch": 16.907801418439718,
|
| 1745 |
-
"eval_loss": 0.663483738899231,
|
| 1746 |
-
"eval_runtime": 116.4065,
|
| 1747 |
-
"eval_samples_per_second": 41.458,
|
| 1748 |
-
"eval_steps_per_second": 20.729,
|
| 1749 |
-
"step": 5800
|
| 1750 |
-
},
|
| 1751 |
-
{
|
| 1752 |
-
"epoch": 17.055282778687033,
|
| 1753 |
-
"grad_norm": 69.1272201538086,
|
| 1754 |
-
"learning_rate": 1.4723032069970846e-05,
|
| 1755 |
-
"loss": 0.686,
|
| 1756 |
-
"step": 5850
|
| 1757 |
-
},
|
| 1758 |
-
{
|
| 1759 |
-
"epoch": 17.055282778687033,
|
| 1760 |
-
"eval_loss": 0.662264883518219,
|
| 1761 |
-
"eval_runtime": 116.5098,
|
| 1762 |
-
"eval_samples_per_second": 41.421,
|
| 1763 |
-
"eval_steps_per_second": 20.711,
|
| 1764 |
-
"step": 5850
|
| 1765 |
-
},
|
| 1766 |
-
{
|
| 1767 |
-
"epoch": 17.20076377523186,
|
| 1768 |
-
"grad_norm": 46.41282653808594,
|
| 1769 |
-
"learning_rate": 1.3994169096209913e-05,
|
| 1770 |
-
"loss": 0.678,
|
| 1771 |
-
"step": 5900
|
| 1772 |
-
},
|
| 1773 |
-
{
|
| 1774 |
-
"epoch": 17.20076377523186,
|
| 1775 |
-
"eval_loss": 0.6592727899551392,
|
| 1776 |
-
"eval_runtime": 116.359,
|
| 1777 |
-
"eval_samples_per_second": 41.475,
|
| 1778 |
-
"eval_steps_per_second": 20.738,
|
| 1779 |
-
"step": 5900
|
| 1780 |
-
},
|
| 1781 |
-
{
|
| 1782 |
-
"epoch": 17.346244771776686,
|
| 1783 |
-
"grad_norm": 69.46429443359375,
|
| 1784 |
-
"learning_rate": 1.3265306122448982e-05,
|
| 1785 |
-
"loss": 0.6799,
|
| 1786 |
-
"step": 5950
|
| 1787 |
-
},
|
| 1788 |
-
{
|
| 1789 |
-
"epoch": 17.346244771776686,
|
| 1790 |
-
"eval_loss": 0.6556416153907776,
|
| 1791 |
-
"eval_runtime": 116.5416,
|
| 1792 |
-
"eval_samples_per_second": 41.41,
|
| 1793 |
-
"eval_steps_per_second": 20.705,
|
| 1794 |
-
"step": 5950
|
| 1795 |
-
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 17.491725768321512,
|
| 1798 |
-
"grad_norm": 53.518436431884766,
|
| 1799 |
-
"learning_rate": 1.2536443148688048e-05,
|
| 1800 |
-
"loss": 0.6783,
|
| 1801 |
-
"step": 6000
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 17.491725768321512,
|
| 1805 |
-
"eval_loss": 0.6580107808113098,
|
| 1806 |
-
"eval_runtime": 116.1219,
|
| 1807 |
-
"eval_samples_per_second": 41.56,
|
| 1808 |
-
"eval_steps_per_second": 20.78,
|
| 1809 |
-
"step": 6000
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 17.637206764866338,
|
| 1813 |
-
"grad_norm": 57.20518493652344,
|
| 1814 |
-
"learning_rate": 1.1807580174927114e-05,
|
| 1815 |
-
"loss": 0.6784,
|
| 1816 |
-
"step": 6050
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 17.637206764866338,
|
| 1820 |
-
"eval_loss": 0.6599807143211365,
|
| 1821 |
-
"eval_runtime": 116.2815,
|
| 1822 |
-
"eval_samples_per_second": 41.503,
|
| 1823 |
-
"eval_steps_per_second": 20.751,
|
| 1824 |
-
"step": 6050
|
| 1825 |
-
},
|
| 1826 |
-
{
|
| 1827 |
-
"epoch": 17.782687761411164,
|
| 1828 |
-
"grad_norm": 49.93397521972656,
|
| 1829 |
-
"learning_rate": 1.1078717201166182e-05,
|
| 1830 |
-
"loss": 0.676,
|
| 1831 |
-
"step": 6100
|
| 1832 |
-
},
|
| 1833 |
-
{
|
| 1834 |
-
"epoch": 17.782687761411164,
|
| 1835 |
-
"eval_loss": 0.6560738682746887,
|
| 1836 |
-
"eval_runtime": 116.7097,
|
| 1837 |
-
"eval_samples_per_second": 41.35,
|
| 1838 |
-
"eval_steps_per_second": 20.675,
|
| 1839 |
-
"step": 6100
|
| 1840 |
-
},
|
| 1841 |
-
{
|
| 1842 |
-
"epoch": 17.92816875795599,
|
| 1843 |
-
"grad_norm": 48.3265495300293,
|
| 1844 |
-
"learning_rate": 1.0349854227405248e-05,
|
| 1845 |
-
"loss": 0.6778,
|
| 1846 |
-
"step": 6150
|
| 1847 |
-
},
|
| 1848 |
-
{
|
| 1849 |
-
"epoch": 17.92816875795599,
|
| 1850 |
-
"eval_loss": 0.6563200950622559,
|
| 1851 |
-
"eval_runtime": 116.4229,
|
| 1852 |
-
"eval_samples_per_second": 41.452,
|
| 1853 |
-
"eval_steps_per_second": 20.726,
|
| 1854 |
-
"step": 6150
|
| 1855 |
-
},
|
| 1856 |
-
{
|
| 1857 |
-
"epoch": 18.07565011820331,
|
| 1858 |
-
"grad_norm": 74.32230377197266,
|
| 1859 |
-
"learning_rate": 9.620991253644314e-06,
|
| 1860 |
-
"loss": 0.6763,
|
| 1861 |
-
"step": 6200
|
| 1862 |
-
},
|
| 1863 |
-
{
|
| 1864 |
-
"epoch": 18.07565011820331,
|
| 1865 |
-
"eval_loss": 0.6560296416282654,
|
| 1866 |
-
"eval_runtime": 116.3284,
|
| 1867 |
-
"eval_samples_per_second": 41.486,
|
| 1868 |
-
"eval_steps_per_second": 20.743,
|
| 1869 |
-
"step": 6200
|
| 1870 |
-
},
|
| 1871 |
-
{
|
| 1872 |
-
"epoch": 18.221131114748136,
|
| 1873 |
-
"grad_norm": 62.084503173828125,
|
| 1874 |
-
"learning_rate": 8.892128279883383e-06,
|
| 1875 |
-
"loss": 0.6667,
|
| 1876 |
-
"step": 6250
|
| 1877 |
-
},
|
| 1878 |
-
{
|
| 1879 |
-
"epoch": 18.221131114748136,
|
| 1880 |
-
"eval_loss": 0.6540623903274536,
|
| 1881 |
-
"eval_runtime": 116.3245,
|
| 1882 |
-
"eval_samples_per_second": 41.487,
|
| 1883 |
-
"eval_steps_per_second": 20.744,
|
| 1884 |
-
"step": 6250
|
| 1885 |
-
},
|
| 1886 |
-
{
|
| 1887 |
-
"epoch": 18.366612111292962,
|
| 1888 |
-
"grad_norm": 43.38812255859375,
|
| 1889 |
-
"learning_rate": 8.163265306122448e-06,
|
| 1890 |
-
"loss": 0.6761,
|
| 1891 |
-
"step": 6300
|
| 1892 |
-
},
|
| 1893 |
-
{
|
| 1894 |
-
"epoch": 18.366612111292962,
|
| 1895 |
-
"eval_loss": 0.6527947187423706,
|
| 1896 |
-
"eval_runtime": 116.5368,
|
| 1897 |
-
"eval_samples_per_second": 41.412,
|
| 1898 |
-
"eval_steps_per_second": 20.706,
|
| 1899 |
-
"step": 6300
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 18.51209310783779,
|
| 1903 |
-
"grad_norm": 81.97037506103516,
|
| 1904 |
-
"learning_rate": 7.434402332361516e-06,
|
| 1905 |
-
"loss": 0.6708,
|
| 1906 |
-
"step": 6350
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 18.51209310783779,
|
| 1910 |
-
"eval_loss": 0.656480073928833,
|
| 1911 |
-
"eval_runtime": 116.5086,
|
| 1912 |
-
"eval_samples_per_second": 41.422,
|
| 1913 |
-
"eval_steps_per_second": 20.711,
|
| 1914 |
-
"step": 6350
|
| 1915 |
-
},
|
| 1916 |
-
{
|
| 1917 |
-
"epoch": 18.657574104382615,
|
| 1918 |
-
"grad_norm": 60.97893524169922,
|
| 1919 |
-
"learning_rate": 6.705539358600584e-06,
|
| 1920 |
-
"loss": 0.6704,
|
| 1921 |
-
"step": 6400
|
| 1922 |
-
},
|
| 1923 |
-
{
|
| 1924 |
-
"epoch": 18.657574104382615,
|
| 1925 |
-
"eval_loss": 0.6558669209480286,
|
| 1926 |
-
"eval_runtime": 116.735,
|
| 1927 |
-
"eval_samples_per_second": 41.342,
|
| 1928 |
-
"eval_steps_per_second": 20.671,
|
| 1929 |
-
"step": 6400
|
| 1930 |
-
},
|
| 1931 |
-
{
|
| 1932 |
-
"epoch": 18.80305510092744,
|
| 1933 |
-
"grad_norm": 75.70346069335938,
|
| 1934 |
-
"learning_rate": 5.97667638483965e-06,
|
| 1935 |
-
"loss": 0.6695,
|
| 1936 |
-
"step": 6450
|
| 1937 |
-
},
|
| 1938 |
-
{
|
| 1939 |
-
"epoch": 18.80305510092744,
|
| 1940 |
-
"eval_loss": 0.6515429615974426,
|
| 1941 |
-
"eval_runtime": 116.3122,
|
| 1942 |
-
"eval_samples_per_second": 41.492,
|
| 1943 |
-
"eval_steps_per_second": 20.746,
|
| 1944 |
-
"step": 6450
|
| 1945 |
-
},
|
| 1946 |
-
{
|
| 1947 |
-
"epoch": 18.948536097472267,
|
| 1948 |
-
"grad_norm": 49.48427963256836,
|
| 1949 |
-
"learning_rate": 5.247813411078718e-06,
|
| 1950 |
-
"loss": 0.6716,
|
| 1951 |
-
"step": 6500
|
| 1952 |
-
},
|
| 1953 |
-
{
|
| 1954 |
-
"epoch": 18.948536097472267,
|
| 1955 |
-
"eval_loss": 0.6520217061042786,
|
| 1956 |
-
"eval_runtime": 116.8293,
|
| 1957 |
-
"eval_samples_per_second": 41.308,
|
| 1958 |
-
"eval_steps_per_second": 20.654,
|
| 1959 |
-
"step": 6500
|
| 1960 |
-
},
|
| 1961 |
-
{
|
| 1962 |
-
"epoch": 19.096017457719586,
|
| 1963 |
-
"grad_norm": 58.43903732299805,
|
| 1964 |
-
"learning_rate": 4.518950437317785e-06,
|
| 1965 |
-
"loss": 0.6733,
|
| 1966 |
-
"step": 6550
|
| 1967 |
-
},
|
| 1968 |
-
{
|
| 1969 |
-
"epoch": 19.096017457719586,
|
| 1970 |
-
"eval_loss": 0.6509958505630493,
|
| 1971 |
-
"eval_runtime": 116.5229,
|
| 1972 |
-
"eval_samples_per_second": 41.417,
|
| 1973 |
-
"eval_steps_per_second": 20.708,
|
| 1974 |
-
"step": 6550
|
| 1975 |
-
},
|
| 1976 |
-
{
|
| 1977 |
-
"epoch": 19.241498454264413,
|
| 1978 |
-
"grad_norm": 52.80235290527344,
|
| 1979 |
-
"learning_rate": 3.7900874635568516e-06,
|
| 1980 |
-
"loss": 0.666,
|
| 1981 |
-
"step": 6600
|
| 1982 |
-
},
|
| 1983 |
-
{
|
| 1984 |
-
"epoch": 19.241498454264413,
|
| 1985 |
-
"eval_loss": 0.6503157615661621,
|
| 1986 |
-
"eval_runtime": 116.552,
|
| 1987 |
-
"eval_samples_per_second": 41.406,
|
| 1988 |
-
"eval_steps_per_second": 20.703,
|
| 1989 |
-
"step": 6600
|
| 1990 |
-
},
|
| 1991 |
-
{
|
| 1992 |
-
"epoch": 19.38697945080924,
|
| 1993 |
-
"grad_norm": 45.26133346557617,
|
| 1994 |
-
"learning_rate": 3.0612244897959185e-06,
|
| 1995 |
-
"loss": 0.6627,
|
| 1996 |
-
"step": 6650
|
| 1997 |
-
},
|
| 1998 |
-
{
|
| 1999 |
-
"epoch": 19.38697945080924,
|
| 2000 |
-
"eval_loss": 0.6515105366706848,
|
| 2001 |
-
"eval_runtime": 116.6286,
|
| 2002 |
-
"eval_samples_per_second": 41.379,
|
| 2003 |
-
"eval_steps_per_second": 20.69,
|
| 2004 |
-
"step": 6650
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 19.532460447354065,
|
| 2008 |
-
"grad_norm": 55.74667739868164,
|
| 2009 |
-
"learning_rate": 2.3323615160349855e-06,
|
| 2010 |
-
"loss": 0.6683,
|
| 2011 |
-
"step": 6700
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 19.532460447354065,
|
| 2015 |
-
"eval_loss": 0.6518427133560181,
|
| 2016 |
-
"eval_runtime": 116.5738,
|
| 2017 |
-
"eval_samples_per_second": 41.399,
|
| 2018 |
-
"eval_steps_per_second": 20.699,
|
| 2019 |
-
"step": 6700
|
| 2020 |
-
},
|
| 2021 |
-
{
|
| 2022 |
-
"epoch": 19.67794144389889,
|
| 2023 |
-
"grad_norm": 58.34127426147461,
|
| 2024 |
-
"learning_rate": 1.6034985422740526e-06,
|
| 2025 |
-
"loss": 0.6638,
|
| 2026 |
-
"step": 6750
|
| 2027 |
-
},
|
| 2028 |
-
{
|
| 2029 |
-
"epoch": 19.67794144389889,
|
| 2030 |
-
"eval_loss": 0.6493617296218872,
|
| 2031 |
-
"eval_runtime": 116.3645,
|
| 2032 |
-
"eval_samples_per_second": 41.473,
|
| 2033 |
-
"eval_steps_per_second": 20.737,
|
| 2034 |
-
"step": 6750
|
| 2035 |
-
},
|
| 2036 |
-
{
|
| 2037 |
-
"epoch": 19.823422440443718,
|
| 2038 |
-
"grad_norm": 58.03067398071289,
|
| 2039 |
-
"learning_rate": 8.746355685131196e-07,
|
| 2040 |
-
"loss": 0.6682,
|
| 2041 |
-
"step": 6800
|
| 2042 |
-
},
|
| 2043 |
-
{
|
| 2044 |
-
"epoch": 19.823422440443718,
|
| 2045 |
-
"eval_loss": 0.6479066014289856,
|
| 2046 |
-
"eval_runtime": 116.4677,
|
| 2047 |
-
"eval_samples_per_second": 41.436,
|
| 2048 |
-
"eval_steps_per_second": 20.718,
|
| 2049 |
-
"step": 6800
|
| 2050 |
-
},
|
| 2051 |
-
{
|
| 2052 |
-
"epoch": 19.968903436988544,
|
| 2053 |
-
"grad_norm": 50.071128845214844,
|
| 2054 |
-
"learning_rate": 1.457725947521866e-07,
|
| 2055 |
-
"loss": 0.6677,
|
| 2056 |
-
"step": 6850
|
| 2057 |
-
},
|
| 2058 |
-
{
|
| 2059 |
-
"epoch": 19.968903436988544,
|
| 2060 |
-
"eval_loss": 0.6467408537864685,
|
| 2061 |
-
"eval_runtime": 116.3823,
|
| 2062 |
-
"eval_samples_per_second": 41.467,
|
| 2063 |
-
"eval_steps_per_second": 20.733,
|
| 2064 |
-
"step": 6850
|
| 2065 |
}
|
| 2066 |
],
|
| 2067 |
"logging_steps": 50,
|
| 2068 |
-
"max_steps":
|
| 2069 |
"num_input_tokens_seen": 0,
|
| 2070 |
-
"num_train_epochs":
|
| 2071 |
"save_steps": 500,
|
| 2072 |
"stateful_callbacks": {
|
| 2073 |
"TrainerControl": {
|
|
@@ -2076,12 +96,12 @@
|
|
| 2076 |
"should_evaluate": false,
|
| 2077 |
"should_log": false,
|
| 2078 |
"should_save": true,
|
| 2079 |
-
"should_training_stop":
|
| 2080 |
},
|
| 2081 |
"attributes": {}
|
| 2082 |
}
|
| 2083 |
},
|
| 2084 |
-
"total_flos":
|
| 2085 |
"train_batch_size": 2,
|
| 2086 |
"trial_name": null,
|
| 2087 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 273,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.18380241240666284,
|
| 13 |
+
"grad_norm": 96.95903778076172,
|
| 14 |
+
"learning_rate": 9.816176470588235e-05,
|
| 15 |
+
"loss": 1.4053,
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.18380241240666284,
|
| 20 |
+
"eval_loss": 1.357384204864502,
|
| 21 |
+
"eval_runtime": 116.0223,
|
| 22 |
+
"eval_samples_per_second": 41.595,
|
| 23 |
+
"eval_steps_per_second": 20.798,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.3676048248133257,
|
| 28 |
+
"grad_norm": 62.13881301879883,
|
| 29 |
+
"learning_rate": 9.632352941176472e-05,
|
| 30 |
+
"loss": 1.3129,
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.3676048248133257,
|
| 35 |
+
"eval_loss": 1.3342629671096802,
|
| 36 |
+
"eval_runtime": 116.0756,
|
| 37 |
+
"eval_samples_per_second": 41.576,
|
| 38 |
+
"eval_steps_per_second": 20.788,
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.5514072372199885,
|
| 43 |
+
"grad_norm": 94.01299285888672,
|
| 44 |
+
"learning_rate": 9.448529411764707e-05,
|
| 45 |
+
"loss": 1.2775,
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
+
"epoch": 0.5514072372199885,
|
| 50 |
+
"eval_loss": 1.315340280532837,
|
| 51 |
+
"eval_runtime": 116.2655,
|
| 52 |
+
"eval_samples_per_second": 41.508,
|
| 53 |
+
"eval_steps_per_second": 20.754,
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
+
"epoch": 0.7352096496266514,
|
| 58 |
+
"grad_norm": 106.19926452636719,
|
| 59 |
+
"learning_rate": 9.264705882352942e-05,
|
| 60 |
+
"loss": 1.2405,
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 0.7352096496266514,
|
| 65 |
+
"eval_loss": 1.3019272089004517,
|
| 66 |
+
"eval_runtime": 116.2502,
|
| 67 |
+
"eval_samples_per_second": 41.514,
|
| 68 |
+
"eval_steps_per_second": 20.757,
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.9190120620333142,
|
| 73 |
+
"grad_norm": 65.5792465209961,
|
| 74 |
+
"learning_rate": 9.080882352941177e-05,
|
| 75 |
+
"loss": 1.2157,
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
+
"epoch": 0.9190120620333142,
|
| 80 |
+
"eval_loss": 1.290924310684204,
|
| 81 |
+
"eval_runtime": 116.1975,
|
| 82 |
+
"eval_samples_per_second": 41.533,
|
| 83 |
+
"eval_steps_per_second": 20.766,
|
| 84 |
"step": 250
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 85 |
}
|
| 86 |
],
|
| 87 |
"logging_steps": 50,
|
| 88 |
+
"max_steps": 2720,
|
| 89 |
"num_input_tokens_seen": 0,
|
| 90 |
+
"num_train_epochs": 10,
|
| 91 |
"save_steps": 500,
|
| 92 |
"stateful_callbacks": {
|
| 93 |
"TrainerControl": {
|
|
|
|
| 96 |
"should_evaluate": false,
|
| 97 |
"should_log": false,
|
| 98 |
"should_save": true,
|
| 99 |
+
"should_training_stop": false
|
| 100 |
},
|
| 101 |
"attributes": {}
|
| 102 |
}
|
| 103 |
},
|
| 104 |
+
"total_flos": 7.014086072913101e+16,
|
| 105 |
"train_batch_size": 2,
|
| 106 |
"trial_name": null,
|
| 107 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba3ba4c9440deaddd402fb37b83a2039f6da0a755f6ddc398099f7184d4f6b4e
|
| 3 |
size 5368
|