Upload folder using huggingface_hub
Browse files
checkpoint-latest/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 62293080
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03190bb45309bbfedccad7959d6dd1878e0a3a2c49476630690e4b0b715ba571
|
| 3 |
size 62293080
|
checkpoint-latest/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 124642443
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08db5c721a857b9e376429cfbe60e90ff49062398fa0ce0ed683edf407f96d3c
|
| 3 |
size 124642443
|
checkpoint-latest/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b182b3dedf0ef16e42c7c650dc1dabae1c16baddbc2124f22d1cc0600b2e9222
|
| 3 |
size 14645
|
checkpoint-latest/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:47578247ee117cf4bc0a0c2af4d329dc3c61e90f1c21bdab88f7e138719d3e25
|
| 3 |
+
size 1383
|
checkpoint-latest/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6a5728e17488e0af5b9594b7a742e2aa414b404c43f45b9b8c470ef12744e2b
|
| 3 |
size 1465
|
checkpoint-latest/trainer_state.json
CHANGED
|
@@ -2,516 +2,84 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.021995424951610065,
|
| 14 |
-
"grad_norm": 0.
|
| 15 |
"learning_rate": 0.00029781805384480024,
|
| 16 |
-
"loss": 3.
|
| 17 |
"step": 125
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.04399084990322013,
|
| 21 |
-
"grad_norm": 0.
|
| 22 |
"learning_rate": 0.00029561851134963927,
|
| 23 |
-
"loss": 3.
|
| 24 |
"step": 250
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.06598627485483019,
|
| 28 |
-
"grad_norm": 0.
|
| 29 |
"learning_rate": 0.00029341896885447824,
|
| 30 |
-
"loss": 3.
|
| 31 |
"step": 375
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.08798169980644026,
|
| 35 |
-
"grad_norm": 0.
|
| 36 |
"learning_rate": 0.0002912194263593172,
|
| 37 |
-
"loss": 3.
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.08798169980644026,
|
| 42 |
-
"eval_loss":
|
| 43 |
-
"eval_runtime":
|
| 44 |
-
"eval_samples_per_second":
|
| 45 |
-
"eval_steps_per_second":
|
| 46 |
"step": 500
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.10997712475805033,
|
| 50 |
-
"grad_norm": 0.
|
| 51 |
"learning_rate": 0.00028901988386415624,
|
| 52 |
-
"loss": 3.
|
| 53 |
"step": 625
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.13197254970966038,
|
| 57 |
-
"grad_norm":
|
| 58 |
"learning_rate": 0.0002868203413689952,
|
| 59 |
-
"loss": 3.
|
| 60 |
"step": 750
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.15396797466127046,
|
| 64 |
-
"grad_norm": 0.
|
| 65 |
"learning_rate": 0.00028462079887383424,
|
| 66 |
-
"loss": 3.
|
| 67 |
"step": 875
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 0.17596339961288052,
|
| 71 |
-
"grad_norm": 0.
|
| 72 |
"learning_rate": 0.0002824212563786732,
|
| 73 |
-
"loss": 3.
|
| 74 |
"step": 1000
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.17596339961288052,
|
| 78 |
-
"eval_loss": 2.
|
| 79 |
-
"eval_runtime":
|
| 80 |
-
"eval_samples_per_second":
|
| 81 |
-
"eval_steps_per_second":
|
| 82 |
"step": 1000
|
| 83 |
-
},
|
| 84 |
-
{
|
| 85 |
-
"epoch": 0.1979588245644906,
|
| 86 |
-
"grad_norm": 0.8066490292549133,
|
| 87 |
-
"learning_rate": 0.0002802217138835122,
|
| 88 |
-
"loss": 3.5005,
|
| 89 |
-
"step": 1125
|
| 90 |
-
},
|
| 91 |
-
{
|
| 92 |
-
"epoch": 0.21995424951610065,
|
| 93 |
-
"grad_norm": 0.7767718434333801,
|
| 94 |
-
"learning_rate": 0.0002780221713883512,
|
| 95 |
-
"loss": 3.4749,
|
| 96 |
-
"step": 1250
|
| 97 |
-
},
|
| 98 |
-
{
|
| 99 |
-
"epoch": 0.2419496744677107,
|
| 100 |
-
"grad_norm": 0.7831096053123474,
|
| 101 |
-
"learning_rate": 0.0002758226288931902,
|
| 102 |
-
"loss": 3.4609,
|
| 103 |
-
"step": 1375
|
| 104 |
-
},
|
| 105 |
-
{
|
| 106 |
-
"epoch": 0.26394509941932076,
|
| 107 |
-
"grad_norm": 0.9593456387519836,
|
| 108 |
-
"learning_rate": 0.0002736230863980292,
|
| 109 |
-
"loss": 3.4383,
|
| 110 |
-
"step": 1500
|
| 111 |
-
},
|
| 112 |
-
{
|
| 113 |
-
"epoch": 0.26394509941932076,
|
| 114 |
-
"eval_loss": 2.885742664337158,
|
| 115 |
-
"eval_runtime": 91.283,
|
| 116 |
-
"eval_samples_per_second": 107.906,
|
| 117 |
-
"eval_steps_per_second": 4.152,
|
| 118 |
-
"step": 1500
|
| 119 |
-
},
|
| 120 |
-
{
|
| 121 |
-
"epoch": 0.28594052437093087,
|
| 122 |
-
"grad_norm": 0.7641668915748596,
|
| 123 |
-
"learning_rate": 0.0002714235439028682,
|
| 124 |
-
"loss": 3.4404,
|
| 125 |
-
"step": 1625
|
| 126 |
-
},
|
| 127 |
-
{
|
| 128 |
-
"epoch": 0.3079359493225409,
|
| 129 |
-
"grad_norm": 0.71437668800354,
|
| 130 |
-
"learning_rate": 0.00026922400140770716,
|
| 131 |
-
"loss": 3.4147,
|
| 132 |
-
"step": 1750
|
| 133 |
-
},
|
| 134 |
-
{
|
| 135 |
-
"epoch": 0.329931374274151,
|
| 136 |
-
"grad_norm": 0.7094811201095581,
|
| 137 |
-
"learning_rate": 0.0002670244589125462,
|
| 138 |
-
"loss": 3.4125,
|
| 139 |
-
"step": 1875
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 0.35192679922576103,
|
| 143 |
-
"grad_norm": 0.722710907459259,
|
| 144 |
-
"learning_rate": 0.00026482491641738516,
|
| 145 |
-
"loss": 3.3959,
|
| 146 |
-
"step": 2000
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.35192679922576103,
|
| 150 |
-
"eval_loss": 2.8463189601898193,
|
| 151 |
-
"eval_runtime": 91.3088,
|
| 152 |
-
"eval_samples_per_second": 107.876,
|
| 153 |
-
"eval_steps_per_second": 4.151,
|
| 154 |
-
"step": 2000
|
| 155 |
-
},
|
| 156 |
-
{
|
| 157 |
-
"epoch": 0.3739222241773711,
|
| 158 |
-
"grad_norm": 0.6693059802055359,
|
| 159 |
-
"learning_rate": 0.00026262537392222413,
|
| 160 |
-
"loss": 3.3749,
|
| 161 |
-
"step": 2125
|
| 162 |
-
},
|
| 163 |
-
{
|
| 164 |
-
"epoch": 0.3959176491289812,
|
| 165 |
-
"grad_norm": 0.8750757575035095,
|
| 166 |
-
"learning_rate": 0.00026042583142706316,
|
| 167 |
-
"loss": 3.3698,
|
| 168 |
-
"step": 2250
|
| 169 |
-
},
|
| 170 |
-
{
|
| 171 |
-
"epoch": 0.41791307408059125,
|
| 172 |
-
"grad_norm": 0.8450427651405334,
|
| 173 |
-
"learning_rate": 0.00025822628893190213,
|
| 174 |
-
"loss": 3.3556,
|
| 175 |
-
"step": 2375
|
| 176 |
-
},
|
| 177 |
-
{
|
| 178 |
-
"epoch": 0.4399084990322013,
|
| 179 |
-
"grad_norm": 0.660953938961029,
|
| 180 |
-
"learning_rate": 0.00025602674643674116,
|
| 181 |
-
"loss": 3.3503,
|
| 182 |
-
"step": 2500
|
| 183 |
-
},
|
| 184 |
-
{
|
| 185 |
-
"epoch": 0.4399084990322013,
|
| 186 |
-
"eval_loss": 2.8230228424072266,
|
| 187 |
-
"eval_runtime": 91.2675,
|
| 188 |
-
"eval_samples_per_second": 107.924,
|
| 189 |
-
"eval_steps_per_second": 4.153,
|
| 190 |
-
"step": 2500
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"epoch": 0.46190392398381136,
|
| 194 |
-
"grad_norm": 0.7277446389198303,
|
| 195 |
-
"learning_rate": 0.00025382720394158013,
|
| 196 |
-
"loss": 3.358,
|
| 197 |
-
"step": 2625
|
| 198 |
-
},
|
| 199 |
-
{
|
| 200 |
-
"epoch": 0.4838993489354214,
|
| 201 |
-
"grad_norm": 0.7908076047897339,
|
| 202 |
-
"learning_rate": 0.0002516276614464191,
|
| 203 |
-
"loss": 3.3433,
|
| 204 |
-
"step": 2750
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 0.5058947738870315,
|
| 208 |
-
"grad_norm": 0.847932755947113,
|
| 209 |
-
"learning_rate": 0.00024942811895125813,
|
| 210 |
-
"loss": 3.3363,
|
| 211 |
-
"step": 2875
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.5278901988386415,
|
| 215 |
-
"grad_norm": 0.7486276626586914,
|
| 216 |
-
"learning_rate": 0.0002472285764560971,
|
| 217 |
-
"loss": 3.3197,
|
| 218 |
-
"step": 3000
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"epoch": 0.5278901988386415,
|
| 222 |
-
"eval_loss": 2.7938477993011475,
|
| 223 |
-
"eval_runtime": 91.2355,
|
| 224 |
-
"eval_samples_per_second": 107.962,
|
| 225 |
-
"eval_steps_per_second": 4.154,
|
| 226 |
-
"step": 3000
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.5498856237902516,
|
| 230 |
-
"grad_norm": 0.7471196055412292,
|
| 231 |
-
"learning_rate": 0.0002450290339609361,
|
| 232 |
-
"loss": 3.3313,
|
| 233 |
-
"step": 3125
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.5718810487418617,
|
| 237 |
-
"grad_norm": 0.7734588384628296,
|
| 238 |
-
"learning_rate": 0.00024282949146577507,
|
| 239 |
-
"loss": 3.3004,
|
| 240 |
-
"step": 3250
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.5938764736934717,
|
| 244 |
-
"grad_norm": 0.7516855597496033,
|
| 245 |
-
"learning_rate": 0.0002406299489706141,
|
| 246 |
-
"loss": 3.304,
|
| 247 |
-
"step": 3375
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.6158718986450818,
|
| 251 |
-
"grad_norm": 0.7859501242637634,
|
| 252 |
-
"learning_rate": 0.0002384304064754531,
|
| 253 |
-
"loss": 3.2958,
|
| 254 |
-
"step": 3500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.6158718986450818,
|
| 258 |
-
"eval_loss": 2.7763619422912598,
|
| 259 |
-
"eval_runtime": 91.3446,
|
| 260 |
-
"eval_samples_per_second": 107.833,
|
| 261 |
-
"eval_steps_per_second": 4.149,
|
| 262 |
-
"step": 3500
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.6378673235966918,
|
| 266 |
-
"grad_norm": 0.754370927810669,
|
| 267 |
-
"learning_rate": 0.00023623086398029207,
|
| 268 |
-
"loss": 3.2671,
|
| 269 |
-
"step": 3625
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.659862748548302,
|
| 273 |
-
"grad_norm": 0.8835521936416626,
|
| 274 |
-
"learning_rate": 0.00023403132148513107,
|
| 275 |
-
"loss": 3.284,
|
| 276 |
-
"step": 3750
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.6818581734999121,
|
| 280 |
-
"grad_norm": 0.7852393984794617,
|
| 281 |
-
"learning_rate": 0.00023183177898997007,
|
| 282 |
-
"loss": 3.2879,
|
| 283 |
-
"step": 3875
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.7038535984515221,
|
| 287 |
-
"grad_norm": 0.6869296431541443,
|
| 288 |
-
"learning_rate": 0.00022963223649480905,
|
| 289 |
-
"loss": 3.2702,
|
| 290 |
-
"step": 4000
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.7038535984515221,
|
| 294 |
-
"eval_loss": 2.7806477546691895,
|
| 295 |
-
"eval_runtime": 91.4522,
|
| 296 |
-
"eval_samples_per_second": 107.707,
|
| 297 |
-
"eval_steps_per_second": 4.144,
|
| 298 |
-
"step": 4000
|
| 299 |
-
},
|
| 300 |
-
{
|
| 301 |
-
"epoch": 0.7258490234031322,
|
| 302 |
-
"grad_norm": 1.0874834060668945,
|
| 303 |
-
"learning_rate": 0.00022743269399964805,
|
| 304 |
-
"loss": 3.254,
|
| 305 |
-
"step": 4125
|
| 306 |
-
},
|
| 307 |
-
{
|
| 308 |
-
"epoch": 0.7478444483547422,
|
| 309 |
-
"grad_norm": 0.6965116262435913,
|
| 310 |
-
"learning_rate": 0.00022523315150448702,
|
| 311 |
-
"loss": 3.2525,
|
| 312 |
-
"step": 4250
|
| 313 |
-
},
|
| 314 |
-
{
|
| 315 |
-
"epoch": 0.7698398733063523,
|
| 316 |
-
"grad_norm": 0.6829759478569031,
|
| 317 |
-
"learning_rate": 0.00022303360900932605,
|
| 318 |
-
"loss": 3.2502,
|
| 319 |
-
"step": 4375
|
| 320 |
-
},
|
| 321 |
-
{
|
| 322 |
-
"epoch": 0.7918352982579624,
|
| 323 |
-
"grad_norm": 0.7037951946258545,
|
| 324 |
-
"learning_rate": 0.00022083406651416505,
|
| 325 |
-
"loss": 3.2419,
|
| 326 |
-
"step": 4500
|
| 327 |
-
},
|
| 328 |
-
{
|
| 329 |
-
"epoch": 0.7918352982579624,
|
| 330 |
-
"eval_loss": 2.7353272438049316,
|
| 331 |
-
"eval_runtime": 90.7352,
|
| 332 |
-
"eval_samples_per_second": 108.558,
|
| 333 |
-
"eval_steps_per_second": 4.177,
|
| 334 |
-
"step": 4500
|
| 335 |
-
},
|
| 336 |
-
{
|
| 337 |
-
"epoch": 0.8138307232095724,
|
| 338 |
-
"grad_norm": 0.9032262563705444,
|
| 339 |
-
"learning_rate": 0.00021863452401900402,
|
| 340 |
-
"loss": 3.2426,
|
| 341 |
-
"step": 4625
|
| 342 |
-
},
|
| 343 |
-
{
|
| 344 |
-
"epoch": 0.8358261481611825,
|
| 345 |
-
"grad_norm": 0.6463780403137207,
|
| 346 |
-
"learning_rate": 0.00021643498152384302,
|
| 347 |
-
"loss": 3.2408,
|
| 348 |
-
"step": 4750
|
| 349 |
-
},
|
| 350 |
-
{
|
| 351 |
-
"epoch": 0.8578215731127925,
|
| 352 |
-
"grad_norm": 0.8066023588180542,
|
| 353 |
-
"learning_rate": 0.00021423543902868202,
|
| 354 |
-
"loss": 3.2178,
|
| 355 |
-
"step": 4875
|
| 356 |
-
},
|
| 357 |
-
{
|
| 358 |
-
"epoch": 0.8798169980644026,
|
| 359 |
-
"grad_norm": 0.7497230172157288,
|
| 360 |
-
"learning_rate": 0.000212035896533521,
|
| 361 |
-
"loss": 3.2268,
|
| 362 |
-
"step": 5000
|
| 363 |
-
},
|
| 364 |
-
{
|
| 365 |
-
"epoch": 0.8798169980644026,
|
| 366 |
-
"eval_loss": 2.721640110015869,
|
| 367 |
-
"eval_runtime": 90.4581,
|
| 368 |
-
"eval_samples_per_second": 108.89,
|
| 369 |
-
"eval_steps_per_second": 4.19,
|
| 370 |
-
"step": 5000
|
| 371 |
-
},
|
| 372 |
-
{
|
| 373 |
-
"epoch": 0.9018124230160127,
|
| 374 |
-
"grad_norm": 0.6645826101303101,
|
| 375 |
-
"learning_rate": 0.00020983635403836,
|
| 376 |
-
"loss": 3.1946,
|
| 377 |
-
"step": 5125
|
| 378 |
-
},
|
| 379 |
-
{
|
| 380 |
-
"epoch": 0.9238078479676227,
|
| 381 |
-
"grad_norm": 0.7062675952911377,
|
| 382 |
-
"learning_rate": 0.00020763681154319902,
|
| 383 |
-
"loss": 3.2228,
|
| 384 |
-
"step": 5250
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 0.9458032729192328,
|
| 388 |
-
"grad_norm": 0.7374680638313293,
|
| 389 |
-
"learning_rate": 0.000205437269048038,
|
| 390 |
-
"loss": 3.1905,
|
| 391 |
-
"step": 5375
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 0.9677986978708428,
|
| 395 |
-
"grad_norm": 0.853108286857605,
|
| 396 |
-
"learning_rate": 0.000203237726552877,
|
| 397 |
-
"loss": 3.2011,
|
| 398 |
-
"step": 5500
|
| 399 |
-
},
|
| 400 |
-
{
|
| 401 |
-
"epoch": 0.9677986978708428,
|
| 402 |
-
"eval_loss": 2.714653491973877,
|
| 403 |
-
"eval_runtime": 90.854,
|
| 404 |
-
"eval_samples_per_second": 108.416,
|
| 405 |
-
"eval_steps_per_second": 4.172,
|
| 406 |
-
"step": 5500
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 0.9897941228224529,
|
| 410 |
-
"grad_norm": 0.6738927364349365,
|
| 411 |
-
"learning_rate": 0.000201038184057716,
|
| 412 |
-
"loss": 3.1888,
|
| 413 |
-
"step": 5625
|
| 414 |
-
},
|
| 415 |
-
{
|
| 416 |
-
"epoch": 1.011789547774063,
|
| 417 |
-
"grad_norm": 0.7420991659164429,
|
| 418 |
-
"learning_rate": 0.00019883864156255496,
|
| 419 |
-
"loss": 3.1835,
|
| 420 |
-
"step": 5750
|
| 421 |
-
},
|
| 422 |
-
{
|
| 423 |
-
"epoch": 1.0337849727256732,
|
| 424 |
-
"grad_norm": 0.8556201457977295,
|
| 425 |
-
"learning_rate": 0.00019663909906739396,
|
| 426 |
-
"loss": 3.1885,
|
| 427 |
-
"step": 5875
|
| 428 |
-
},
|
| 429 |
-
{
|
| 430 |
-
"epoch": 1.055780397677283,
|
| 431 |
-
"grad_norm": 0.7332561016082764,
|
| 432 |
-
"learning_rate": 0.00019443955657223294,
|
| 433 |
-
"loss": 3.1691,
|
| 434 |
-
"step": 6000
|
| 435 |
-
},
|
| 436 |
-
{
|
| 437 |
-
"epoch": 1.055780397677283,
|
| 438 |
-
"eval_loss": 2.692098617553711,
|
| 439 |
-
"eval_runtime": 90.3724,
|
| 440 |
-
"eval_samples_per_second": 108.993,
|
| 441 |
-
"eval_steps_per_second": 4.194,
|
| 442 |
-
"step": 6000
|
| 443 |
-
},
|
| 444 |
-
{
|
| 445 |
-
"epoch": 1.0777758226288932,
|
| 446 |
-
"grad_norm": 0.8108923435211182,
|
| 447 |
-
"learning_rate": 0.00019224001407707196,
|
| 448 |
-
"loss": 3.1627,
|
| 449 |
-
"step": 6125
|
| 450 |
-
},
|
| 451 |
-
{
|
| 452 |
-
"epoch": 1.0997712475805033,
|
| 453 |
-
"grad_norm": 0.8860889077186584,
|
| 454 |
-
"learning_rate": 0.00019004047158191096,
|
| 455 |
-
"loss": 3.1703,
|
| 456 |
-
"step": 6250
|
| 457 |
-
},
|
| 458 |
-
{
|
| 459 |
-
"epoch": 1.1217666725321134,
|
| 460 |
-
"grad_norm": 0.6584394574165344,
|
| 461 |
-
"learning_rate": 0.00018784092908674993,
|
| 462 |
-
"loss": 3.1704,
|
| 463 |
-
"step": 6375
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"epoch": 1.1437620974837235,
|
| 467 |
-
"grad_norm": 0.7717780470848083,
|
| 468 |
-
"learning_rate": 0.00018564138659158893,
|
| 469 |
-
"loss": 3.1712,
|
| 470 |
-
"step": 6500
|
| 471 |
-
},
|
| 472 |
-
{
|
| 473 |
-
"epoch": 1.1437620974837235,
|
| 474 |
-
"eval_loss": 2.6807186603546143,
|
| 475 |
-
"eval_runtime": 90.3407,
|
| 476 |
-
"eval_samples_per_second": 109.032,
|
| 477 |
-
"eval_steps_per_second": 4.195,
|
| 478 |
-
"step": 6500
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 1.1657575224353334,
|
| 482 |
-
"grad_norm": 0.666558563709259,
|
| 483 |
-
"learning_rate": 0.00018344184409642793,
|
| 484 |
-
"loss": 3.1564,
|
| 485 |
-
"step": 6625
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 1.1877529473869435,
|
| 489 |
-
"grad_norm": 0.8080345392227173,
|
| 490 |
-
"learning_rate": 0.0001812423016012669,
|
| 491 |
-
"loss": 3.1725,
|
| 492 |
-
"step": 6750
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 1.2097483723385536,
|
| 496 |
-
"grad_norm": 0.7759010791778564,
|
| 497 |
-
"learning_rate": 0.0001790427591061059,
|
| 498 |
-
"loss": 3.1469,
|
| 499 |
-
"step": 6875
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 1.2317437972901637,
|
| 503 |
-
"grad_norm": 0.7258739471435547,
|
| 504 |
-
"learning_rate": 0.00017684321661094488,
|
| 505 |
-
"loss": 3.147,
|
| 506 |
-
"step": 7000
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 1.2317437972901637,
|
| 510 |
-
"eval_loss": 2.674105405807495,
|
| 511 |
-
"eval_runtime": 90.5221,
|
| 512 |
-
"eval_samples_per_second": 108.813,
|
| 513 |
-
"eval_steps_per_second": 4.187,
|
| 514 |
-
"step": 7000
|
| 515 |
}
|
| 516 |
],
|
| 517 |
"logging_steps": 125,
|
|
@@ -531,7 +99,7 @@
|
|
| 531 |
"attributes": {}
|
| 532 |
}
|
| 533 |
},
|
| 534 |
-
"total_flos":
|
| 535 |
"train_batch_size": 26,
|
| 536 |
"trial_name": null,
|
| 537 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.17596339961288052,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
"epoch": 0.021995424951610065,
|
| 14 |
+
"grad_norm": 0.7317402958869934,
|
| 15 |
"learning_rate": 0.00029781805384480024,
|
| 16 |
+
"loss": 3.1594,
|
| 17 |
"step": 125
|
| 18 |
},
|
| 19 |
{
|
| 20 |
"epoch": 0.04399084990322013,
|
| 21 |
+
"grad_norm": 0.7342692613601685,
|
| 22 |
"learning_rate": 0.00029561851134963927,
|
| 23 |
+
"loss": 3.1518,
|
| 24 |
"step": 250
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.06598627485483019,
|
| 28 |
+
"grad_norm": 0.6049332022666931,
|
| 29 |
"learning_rate": 0.00029341896885447824,
|
| 30 |
+
"loss": 3.1479,
|
| 31 |
"step": 375
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.08798169980644026,
|
| 35 |
+
"grad_norm": 0.7177291512489319,
|
| 36 |
"learning_rate": 0.0002912194263593172,
|
| 37 |
+
"loss": 3.1586,
|
| 38 |
"step": 500
|
| 39 |
},
|
| 40 |
{
|
| 41 |
"epoch": 0.08798169980644026,
|
| 42 |
+
"eval_loss": 2.6734766960144043,
|
| 43 |
+
"eval_runtime": 59.2675,
|
| 44 |
+
"eval_samples_per_second": 166.196,
|
| 45 |
+
"eval_steps_per_second": 6.395,
|
| 46 |
"step": 500
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.10997712475805033,
|
| 50 |
+
"grad_norm": 0.7241224050521851,
|
| 51 |
"learning_rate": 0.00028901988386415624,
|
| 52 |
+
"loss": 3.1257,
|
| 53 |
"step": 625
|
| 54 |
},
|
| 55 |
{
|
| 56 |
"epoch": 0.13197254970966038,
|
| 57 |
+
"grad_norm": 1.1103954315185547,
|
| 58 |
"learning_rate": 0.0002868203413689952,
|
| 59 |
+
"loss": 3.1179,
|
| 60 |
"step": 750
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"epoch": 0.15396797466127046,
|
| 64 |
+
"grad_norm": 0.7277866005897522,
|
| 65 |
"learning_rate": 0.00028462079887383424,
|
| 66 |
+
"loss": 3.1451,
|
| 67 |
"step": 875
|
| 68 |
},
|
| 69 |
{
|
| 70 |
"epoch": 0.17596339961288052,
|
| 71 |
+
"grad_norm": 0.7038848400115967,
|
| 72 |
"learning_rate": 0.0002824212563786732,
|
| 73 |
+
"loss": 3.1223,
|
| 74 |
"step": 1000
|
| 75 |
},
|
| 76 |
{
|
| 77 |
"epoch": 0.17596339961288052,
|
| 78 |
+
"eval_loss": 2.664580821990967,
|
| 79 |
+
"eval_runtime": 59.2093,
|
| 80 |
+
"eval_samples_per_second": 166.359,
|
| 81 |
+
"eval_steps_per_second": 6.401,
|
| 82 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 83 |
}
|
| 84 |
],
|
| 85 |
"logging_steps": 125,
|
|
|
|
| 99 |
"attributes": {}
|
| 100 |
}
|
| 101 |
},
|
| 102 |
+
"total_flos": 586712875008000.0,
|
| 103 |
"train_batch_size": 26,
|
| 104 |
"trial_name": null,
|
| 105 |
"trial_params": null
|
checkpoint-latest/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6033
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0bef956c560dcdc3bac6492fdc576c6dff35538e184d8d9734adbb5fe3c9b01e
|
| 3 |
size 6033
|