Training in progress, epoch 0, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2751499016
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5283ab5dc1b04471ebdac9c69d4ddc3c9daadb1cd9659495feafd52deba15535
|
| 3 |
size 2751499016
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1397530085
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7343fb23d7b9ae991994bd95b71751b6fe75a7696f84ec85a1c9ef768b87760c
|
| 3 |
size 1397530085
|
last-checkpoint/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:a8e2011629d8bed3ef560fa11175cac55684c4e12a72634bb24abf767b6c7399
|
| 3 |
size 14645
|
last-checkpoint/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 sha256:589a29a7a9601ca691e8af1337d86f4213715f3591d526b3325012bd0d4260dd
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
"best_global_step": null,
|
| 3 |
-
"best_metric": 0.
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 50,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -174,6 +174,338 @@
|
|
| 174 |
"eval_samples_per_second": 11.342,
|
| 175 |
"eval_steps_per_second": 5.679,
|
| 176 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 177 |
}
|
| 178 |
],
|
| 179 |
"logging_steps": 5,
|
|
|
|
| 1 |
{
|
| 2 |
"best_global_step": null,
|
| 3 |
+
"best_metric": 0.024217354133725166,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.02294630564479119,
|
| 6 |
"eval_steps": 50,
|
| 7 |
+
"global_step": 150,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 174 |
"eval_samples_per_second": 11.342,
|
| 175 |
"eval_steps_per_second": 5.679,
|
| 176 |
"step": 50
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.008413645403090102,
|
| 180 |
+
"grad_norm": 0.0028622462414205074,
|
| 181 |
+
"learning_rate": 0.00014993216160618053,
|
| 182 |
+
"logits/chosen": 2.2500367164611816,
|
| 183 |
+
"logits/rejected": 0.9480382204055786,
|
| 184 |
+
"logps/chosen": -436.58172607421875,
|
| 185 |
+
"logps/rejected": -573.2215576171875,
|
| 186 |
+
"loss": 0.0001,
|
| 187 |
+
"rewards/accuracies": 1.0,
|
| 188 |
+
"rewards/chosen": -1.8464500904083252,
|
| 189 |
+
"rewards/margins": 13.733373641967773,
|
| 190 |
+
"rewards/rejected": -15.579824447631836,
|
| 191 |
+
"step": 55
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.009178522257916476,
|
| 195 |
+
"grad_norm": 0.00047387686208821833,
|
| 196 |
+
"learning_rate": 0.00016381476916230834,
|
| 197 |
+
"logits/chosen": 2.368880271911621,
|
| 198 |
+
"logits/rejected": 1.2400633096694946,
|
| 199 |
+
"logps/chosen": -460.0272521972656,
|
| 200 |
+
"logps/rejected": -477.02166748046875,
|
| 201 |
+
"loss": 0.0002,
|
| 202 |
+
"rewards/accuracies": 1.0,
|
| 203 |
+
"rewards/chosen": -1.5477186441421509,
|
| 204 |
+
"rewards/margins": 11.986092567443848,
|
| 205 |
+
"rewards/rejected": -13.53381061553955,
|
| 206 |
+
"step": 60
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.009943399112742848,
|
| 210 |
+
"grad_norm": 2.4312286768690683e-05,
|
| 211 |
+
"learning_rate": 0.00017769737671843617,
|
| 212 |
+
"logits/chosen": 2.1582725048065186,
|
| 213 |
+
"logits/rejected": 0.8890345692634583,
|
| 214 |
+
"logps/chosen": -359.4410705566406,
|
| 215 |
+
"logps/rejected": -731.4852294921875,
|
| 216 |
+
"loss": 0.0,
|
| 217 |
+
"rewards/accuracies": 1.0,
|
| 218 |
+
"rewards/chosen": -1.1775275468826294,
|
| 219 |
+
"rewards/margins": 19.947010040283203,
|
| 220 |
+
"rewards/rejected": -21.12453842163086,
|
| 221 |
+
"step": 65
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.01070827596756922,
|
| 225 |
+
"grad_norm": 0.0019105683313682675,
|
| 226 |
+
"learning_rate": 0.00019157998427456397,
|
| 227 |
+
"logits/chosen": 1.9572877883911133,
|
| 228 |
+
"logits/rejected": 0.4613746106624603,
|
| 229 |
+
"logps/chosen": -433.6973571777344,
|
| 230 |
+
"logps/rejected": -575.4284057617188,
|
| 231 |
+
"loss": 0.0,
|
| 232 |
+
"rewards/accuracies": 1.0,
|
| 233 |
+
"rewards/chosen": -3.012183904647827,
|
| 234 |
+
"rewards/margins": 13.66297721862793,
|
| 235 |
+
"rewards/rejected": -16.675161361694336,
|
| 236 |
+
"step": 70
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.011473152822395595,
|
| 240 |
+
"grad_norm": 0.003828815184533596,
|
| 241 |
+
"learning_rate": 0.0002054625918306918,
|
| 242 |
+
"logits/chosen": NaN,
|
| 243 |
+
"logits/rejected": 0.8698433041572571,
|
| 244 |
+
"logps/chosen": -400.0428161621094,
|
| 245 |
+
"logps/rejected": -832.5679931640625,
|
| 246 |
+
"loss": 0.0693,
|
| 247 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 248 |
+
"rewards/chosen": -3.338275194168091,
|
| 249 |
+
"rewards/margins": 21.856613159179688,
|
| 250 |
+
"rewards/rejected": -25.194889068603516,
|
| 251 |
+
"step": 75
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.012238029677221967,
|
| 255 |
+
"grad_norm": 0.010469191707670689,
|
| 256 |
+
"learning_rate": 0.00021934519938681966,
|
| 257 |
+
"logits/chosen": 1.8671172857284546,
|
| 258 |
+
"logits/rejected": 0.7096843719482422,
|
| 259 |
+
"logps/chosen": -443.547119140625,
|
| 260 |
+
"logps/rejected": -688.80126953125,
|
| 261 |
+
"loss": 0.0,
|
| 262 |
+
"rewards/accuracies": 1.0,
|
| 263 |
+
"rewards/chosen": -3.6706607341766357,
|
| 264 |
+
"rewards/margins": 15.804478645324707,
|
| 265 |
+
"rewards/rejected": -19.47513771057129,
|
| 266 |
+
"step": 80
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.013002906532048341,
|
| 270 |
+
"grad_norm": 2.6392877771286294e-05,
|
| 271 |
+
"learning_rate": 0.00023322780694294746,
|
| 272 |
+
"logits/chosen": 1.9022146463394165,
|
| 273 |
+
"logits/rejected": 0.36020129919052124,
|
| 274 |
+
"logps/chosen": -396.5116271972656,
|
| 275 |
+
"logps/rejected": -583.2808837890625,
|
| 276 |
+
"loss": 0.0,
|
| 277 |
+
"rewards/accuracies": 1.0,
|
| 278 |
+
"rewards/chosen": -1.5602790117263794,
|
| 279 |
+
"rewards/margins": 15.830926895141602,
|
| 280 |
+
"rewards/rejected": -17.391204833984375,
|
| 281 |
+
"step": 85
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.013767783386874713,
|
| 285 |
+
"grad_norm": 1.727592433553582e-07,
|
| 286 |
+
"learning_rate": 0.0002471104144990753,
|
| 287 |
+
"logits/chosen": 1.5785582065582275,
|
| 288 |
+
"logits/rejected": 0.47088107466697693,
|
| 289 |
+
"logps/chosen": -400.20953369140625,
|
| 290 |
+
"logps/rejected": -640.7406616210938,
|
| 291 |
+
"loss": 0.0,
|
| 292 |
+
"rewards/accuracies": 1.0,
|
| 293 |
+
"rewards/chosen": -0.2463744431734085,
|
| 294 |
+
"rewards/margins": 18.521793365478516,
|
| 295 |
+
"rewards/rejected": -18.768169403076172,
|
| 296 |
+
"step": 90
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.014532660241701086,
|
| 300 |
+
"grad_norm": 5.100153430248611e-05,
|
| 301 |
+
"learning_rate": 0.0002609930220552031,
|
| 302 |
+
"logits/chosen": 1.551195502281189,
|
| 303 |
+
"logits/rejected": 0.6250472664833069,
|
| 304 |
+
"logps/chosen": -400.8873596191406,
|
| 305 |
+
"logps/rejected": -630.4702758789062,
|
| 306 |
+
"loss": 0.0,
|
| 307 |
+
"rewards/accuracies": 1.0,
|
| 308 |
+
"rewards/chosen": -0.10455240309238434,
|
| 309 |
+
"rewards/margins": 17.793054580688477,
|
| 310 |
+
"rewards/rejected": -17.897607803344727,
|
| 311 |
+
"step": 95
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.01529753709652746,
|
| 315 |
+
"grad_norm": 1.2168807472789922e-07,
|
| 316 |
+
"learning_rate": 0.00027487562961133095,
|
| 317 |
+
"logits/chosen": 2.179185390472412,
|
| 318 |
+
"logits/rejected": 0.9079009890556335,
|
| 319 |
+
"logps/chosen": -295.3583984375,
|
| 320 |
+
"logps/rejected": -859.0484619140625,
|
| 321 |
+
"loss": 0.0,
|
| 322 |
+
"rewards/accuracies": 1.0,
|
| 323 |
+
"rewards/chosen": 0.7290161848068237,
|
| 324 |
+
"rewards/margins": 25.750385284423828,
|
| 325 |
+
"rewards/rejected": -25.02136993408203,
|
| 326 |
+
"step": 100
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.01529753709652746,
|
| 330 |
+
"eval_logits/chosen": NaN,
|
| 331 |
+
"eval_logits/rejected": 0.5695675611495972,
|
| 332 |
+
"eval_logps/chosen": -330.794189453125,
|
| 333 |
+
"eval_logps/rejected": -718.6740112304688,
|
| 334 |
+
"eval_loss": 0.02476140484213829,
|
| 335 |
+
"eval_rewards/accuracies": 0.9811594486236572,
|
| 336 |
+
"eval_rewards/chosen": 0.7904098033905029,
|
| 337 |
+
"eval_rewards/margins": 22.044944763183594,
|
| 338 |
+
"eval_rewards/rejected": -21.254533767700195,
|
| 339 |
+
"eval_runtime": 62.4706,
|
| 340 |
+
"eval_samples_per_second": 11.029,
|
| 341 |
+
"eval_steps_per_second": 5.523,
|
| 342 |
+
"step": 100
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.016062413951353834,
|
| 346 |
+
"grad_norm": 8.499349557489211e-10,
|
| 347 |
+
"learning_rate": 0.0002776520392843145,
|
| 348 |
+
"logits/chosen": 2.1345207691192627,
|
| 349 |
+
"logits/rejected": 0.47184300422668457,
|
| 350 |
+
"logps/chosen": -315.34930419921875,
|
| 351 |
+
"logps/rejected": -695.0374145507812,
|
| 352 |
+
"loss": 0.0,
|
| 353 |
+
"rewards/accuracies": 1.0,
|
| 354 |
+
"rewards/chosen": 0.8640382885932922,
|
| 355 |
+
"rewards/margins": 21.742961883544922,
|
| 356 |
+
"rewards/rejected": -20.878923416137695,
|
| 357 |
+
"step": 105
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.016827290806180204,
|
| 361 |
+
"grad_norm": 3.2272314456349704e-06,
|
| 362 |
+
"learning_rate": 0.0002776515849417652,
|
| 363 |
+
"logits/chosen": 2.066007614135742,
|
| 364 |
+
"logits/rejected": 0.7821122407913208,
|
| 365 |
+
"logps/chosen": -273.3198547363281,
|
| 366 |
+
"logps/rejected": -622.60205078125,
|
| 367 |
+
"loss": 0.0,
|
| 368 |
+
"rewards/accuracies": 1.0,
|
| 369 |
+
"rewards/chosen": 1.230340600013733,
|
| 370 |
+
"rewards/margins": 17.653335571289062,
|
| 371 |
+
"rewards/rejected": -16.42299461364746,
|
| 372 |
+
"step": 110
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.017592167661006578,
|
| 376 |
+
"grad_norm": 2.764949513789361e-08,
|
| 377 |
+
"learning_rate": 0.00027765078110616136,
|
| 378 |
+
"logits/chosen": 2.0481059551239014,
|
| 379 |
+
"logits/rejected": 0.06188444048166275,
|
| 380 |
+
"logps/chosen": -310.6258239746094,
|
| 381 |
+
"logps/rejected": -710.1292724609375,
|
| 382 |
+
"loss": 0.0,
|
| 383 |
+
"rewards/accuracies": 1.0,
|
| 384 |
+
"rewards/chosen": 1.083168387413025,
|
| 385 |
+
"rewards/margins": 23.025583267211914,
|
| 386 |
+
"rewards/rejected": -21.942415237426758,
|
| 387 |
+
"step": 115
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.018357044515832952,
|
| 391 |
+
"grad_norm": 1.2643829450098565e-06,
|
| 392 |
+
"learning_rate": 0.00027764962777952664,
|
| 393 |
+
"logits/chosen": 2.1907763481140137,
|
| 394 |
+
"logits/rejected": 1.0400992631912231,
|
| 395 |
+
"logps/chosen": -307.0046081542969,
|
| 396 |
+
"logps/rejected": -755.6473999023438,
|
| 397 |
+
"loss": 0.0,
|
| 398 |
+
"rewards/accuracies": 1.0,
|
| 399 |
+
"rewards/chosen": 1.624768853187561,
|
| 400 |
+
"rewards/margins": 22.524160385131836,
|
| 401 |
+
"rewards/rejected": -20.89939308166504,
|
| 402 |
+
"step": 120
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.019121921370659323,
|
| 406 |
+
"grad_norm": 9.230769360685231e-13,
|
| 407 |
+
"learning_rate": 0.0002776481249647646,
|
| 408 |
+
"logits/chosen": 2.0665574073791504,
|
| 409 |
+
"logits/rejected": 0.20953090488910675,
|
| 410 |
+
"logps/chosen": -278.8282775878906,
|
| 411 |
+
"logps/rejected": -763.1021728515625,
|
| 412 |
+
"loss": 0.0,
|
| 413 |
+
"rewards/accuracies": 1.0,
|
| 414 |
+
"rewards/chosen": 0.6925491094589233,
|
| 415 |
+
"rewards/margins": 24.283218383789062,
|
| 416 |
+
"rewards/rejected": -23.590667724609375,
|
| 417 |
+
"step": 125
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.019886798225485697,
|
| 421 |
+
"grad_norm": 2.1791736344312085e-06,
|
| 422 |
+
"learning_rate": 0.00027764627266565854,
|
| 423 |
+
"logits/chosen": 2.2263553142547607,
|
| 424 |
+
"logits/rejected": 0.7493933439254761,
|
| 425 |
+
"logps/chosen": -355.59344482421875,
|
| 426 |
+
"logps/rejected": -662.4058837890625,
|
| 427 |
+
"loss": 0.0,
|
| 428 |
+
"rewards/accuracies": 1.0,
|
| 429 |
+
"rewards/chosen": 1.3247101306915283,
|
| 430 |
+
"rewards/margins": 20.672582626342773,
|
| 431 |
+
"rewards/rejected": -19.34787368774414,
|
| 432 |
+
"step": 130
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.02065167508031207,
|
| 436 |
+
"grad_norm": 1.3962798220745753e-05,
|
| 437 |
+
"learning_rate": 0.00027764407088687153,
|
| 438 |
+
"logits/chosen": 2.0784218311309814,
|
| 439 |
+
"logits/rejected": 0.6179172396659851,
|
| 440 |
+
"logps/chosen": -327.31329345703125,
|
| 441 |
+
"logps/rejected": -678.7785034179688,
|
| 442 |
+
"loss": 0.0,
|
| 443 |
+
"rewards/accuracies": 1.0,
|
| 444 |
+
"rewards/chosen": 1.206441044807434,
|
| 445 |
+
"rewards/margins": 20.837465286254883,
|
| 446 |
+
"rewards/rejected": -19.631023406982422,
|
| 447 |
+
"step": 135
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.02141655193513844,
|
| 451 |
+
"grad_norm": 5.620458409794082e-07,
|
| 452 |
+
"learning_rate": 0.0002776415196339467,
|
| 453 |
+
"logits/chosen": 1.9481678009033203,
|
| 454 |
+
"logits/rejected": 0.5887425541877747,
|
| 455 |
+
"logps/chosen": -309.68682861328125,
|
| 456 |
+
"logps/rejected": -587.2570190429688,
|
| 457 |
+
"loss": 0.0,
|
| 458 |
+
"rewards/accuracies": 1.0,
|
| 459 |
+
"rewards/chosen": 0.8636436462402344,
|
| 460 |
+
"rewards/margins": 18.32933807373047,
|
| 461 |
+
"rewards/rejected": -17.465694427490234,
|
| 462 |
+
"step": 140
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.022181428789964815,
|
| 466 |
+
"grad_norm": 1.5234973034239374e-06,
|
| 467 |
+
"learning_rate": 0.00027763861891330675,
|
| 468 |
+
"logits/chosen": 1.926863670349121,
|
| 469 |
+
"logits/rejected": 0.5701156854629517,
|
| 470 |
+
"logps/chosen": -344.00811767578125,
|
| 471 |
+
"logps/rejected": -644.8851318359375,
|
| 472 |
+
"loss": 0.0,
|
| 473 |
+
"rewards/accuracies": 1.0,
|
| 474 |
+
"rewards/chosen": 1.6971412897109985,
|
| 475 |
+
"rewards/margins": 20.91071891784668,
|
| 476 |
+
"rewards/rejected": -19.213577270507812,
|
| 477 |
+
"step": 145
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.02294630564479119,
|
| 481 |
+
"grad_norm": 3.310283773316769e-06,
|
| 482 |
+
"learning_rate": 0.0002776353687322543,
|
| 483 |
+
"logits/chosen": 1.9721519947052002,
|
| 484 |
+
"logits/rejected": 0.6008085012435913,
|
| 485 |
+
"logps/chosen": -346.241455078125,
|
| 486 |
+
"logps/rejected": -690.4124755859375,
|
| 487 |
+
"loss": 0.0,
|
| 488 |
+
"rewards/accuracies": 1.0,
|
| 489 |
+
"rewards/chosen": 0.7422004342079163,
|
| 490 |
+
"rewards/margins": 21.550968170166016,
|
| 491 |
+
"rewards/rejected": -20.808767318725586,
|
| 492 |
+
"step": 150
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.02294630564479119,
|
| 496 |
+
"eval_logits/chosen": NaN,
|
| 497 |
+
"eval_logits/rejected": 0.5568678379058838,
|
| 498 |
+
"eval_logps/chosen": -325.7964782714844,
|
| 499 |
+
"eval_logps/rejected": -757.2901000976562,
|
| 500 |
+
"eval_loss": 0.024217354133725166,
|
| 501 |
+
"eval_rewards/accuracies": 0.9811594486236572,
|
| 502 |
+
"eval_rewards/chosen": 0.9729923605918884,
|
| 503 |
+
"eval_rewards/margins": 23.638286590576172,
|
| 504 |
+
"eval_rewards/rejected": -22.665294647216797,
|
| 505 |
+
"eval_runtime": 63.3584,
|
| 506 |
+
"eval_samples_per_second": 10.875,
|
| 507 |
+
"eval_steps_per_second": 5.445,
|
| 508 |
+
"step": 150
|
| 509 |
}
|
| 510 |
],
|
| 511 |
"logging_steps": 5,
|