Delete trainer_state.json
Browse files- trainer_state.json +0 -1557
trainer_state.json
DELETED
|
@@ -1,1557 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": null,
|
| 3 |
-
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.0,
|
| 5 |
-
"eval_steps": 500,
|
| 6 |
-
"global_step": 501,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.001996007984031936,
|
| 13 |
-
"grad_norm": 16.492788975984936,
|
| 14 |
-
"learning_rate": 1.9607843137254902e-08,
|
| 15 |
-
"logits/chosen": 1.3011724948883057,
|
| 16 |
-
"logits/rejected": 1.3879833221435547,
|
| 17 |
-
"logps/chosen": -0.9583934545516968,
|
| 18 |
-
"logps/rejected": -0.908146858215332,
|
| 19 |
-
"loss": 0.6931,
|
| 20 |
-
"rewards/accuracies": 0.0,
|
| 21 |
-
"rewards/chosen": 0.0,
|
| 22 |
-
"rewards/margins": 0.0,
|
| 23 |
-
"rewards/rejected": 0.0,
|
| 24 |
-
"step": 1
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.00998003992015968,
|
| 28 |
-
"grad_norm": 18.616408754077472,
|
| 29 |
-
"learning_rate": 9.80392156862745e-08,
|
| 30 |
-
"logits/chosen": 1.230096697807312,
|
| 31 |
-
"logits/rejected": 1.2526068687438965,
|
| 32 |
-
"logps/chosen": -0.8566818833351135,
|
| 33 |
-
"logps/rejected": -0.9451415538787842,
|
| 34 |
-
"loss": 0.6931,
|
| 35 |
-
"rewards/accuracies": 0.34375,
|
| 36 |
-
"rewards/chosen": -0.00325668603181839,
|
| 37 |
-
"rewards/margins": -0.0027664820663630962,
|
| 38 |
-
"rewards/rejected": -0.0004902039654552937,
|
| 39 |
-
"step": 5
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 0.01996007984031936,
|
| 43 |
-
"grad_norm": 19.72943287901331,
|
| 44 |
-
"learning_rate": 1.96078431372549e-07,
|
| 45 |
-
"logits/chosen": 1.1732714176177979,
|
| 46 |
-
"logits/rejected": 1.2325665950775146,
|
| 47 |
-
"logps/chosen": -0.8571804165840149,
|
| 48 |
-
"logps/rejected": -0.9293950200080872,
|
| 49 |
-
"loss": 0.6935,
|
| 50 |
-
"rewards/accuracies": 0.5249999761581421,
|
| 51 |
-
"rewards/chosen": -0.0004458148032426834,
|
| 52 |
-
"rewards/margins": 0.00015245657414197922,
|
| 53 |
-
"rewards/rejected": -0.0005982713773846626,
|
| 54 |
-
"step": 10
|
| 55 |
-
},
|
| 56 |
-
{
|
| 57 |
-
"epoch": 0.029940119760479042,
|
| 58 |
-
"grad_norm": 15.989882260008397,
|
| 59 |
-
"learning_rate": 2.941176470588235e-07,
|
| 60 |
-
"logits/chosen": 1.1804090738296509,
|
| 61 |
-
"logits/rejected": 1.227670669555664,
|
| 62 |
-
"logps/chosen": -0.8887054324150085,
|
| 63 |
-
"logps/rejected": -0.9415630102157593,
|
| 64 |
-
"loss": 0.6942,
|
| 65 |
-
"rewards/accuracies": 0.4124999940395355,
|
| 66 |
-
"rewards/chosen": -0.005856434814631939,
|
| 67 |
-
"rewards/margins": -0.004528886638581753,
|
| 68 |
-
"rewards/rejected": -0.0013275481760501862,
|
| 69 |
-
"step": 15
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.03992015968063872,
|
| 73 |
-
"grad_norm": 21.617808491664096,
|
| 74 |
-
"learning_rate": 3.92156862745098e-07,
|
| 75 |
-
"logits/chosen": 1.2326724529266357,
|
| 76 |
-
"logits/rejected": 1.239334225654602,
|
| 77 |
-
"logps/chosen": -1.025779366493225,
|
| 78 |
-
"logps/rejected": -0.9815893173217773,
|
| 79 |
-
"loss": 0.6924,
|
| 80 |
-
"rewards/accuracies": 0.5249999761581421,
|
| 81 |
-
"rewards/chosen": -0.011453075334429741,
|
| 82 |
-
"rewards/margins": -0.006528028752654791,
|
| 83 |
-
"rewards/rejected": -0.004925046116113663,
|
| 84 |
-
"step": 20
|
| 85 |
-
},
|
| 86 |
-
{
|
| 87 |
-
"epoch": 0.0499001996007984,
|
| 88 |
-
"grad_norm": 14.85293949325317,
|
| 89 |
-
"learning_rate": 4.901960784313725e-07,
|
| 90 |
-
"logits/chosen": 1.1621646881103516,
|
| 91 |
-
"logits/rejected": 1.2202112674713135,
|
| 92 |
-
"logps/chosen": -0.9435383081436157,
|
| 93 |
-
"logps/rejected": -1.0095245838165283,
|
| 94 |
-
"loss": 0.6926,
|
| 95 |
-
"rewards/accuracies": 0.4749999940395355,
|
| 96 |
-
"rewards/chosen": -0.008930020034313202,
|
| 97 |
-
"rewards/margins": 0.011791445314884186,
|
| 98 |
-
"rewards/rejected": -0.020721465349197388,
|
| 99 |
-
"step": 25
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 0.059880239520958084,
|
| 103 |
-
"grad_norm": 18.28049213488445,
|
| 104 |
-
"learning_rate": 5.88235294117647e-07,
|
| 105 |
-
"logits/chosen": 1.2528660297393799,
|
| 106 |
-
"logits/rejected": 1.3036034107208252,
|
| 107 |
-
"logps/chosen": -0.9394888877868652,
|
| 108 |
-
"logps/rejected": -0.9887625575065613,
|
| 109 |
-
"loss": 0.6928,
|
| 110 |
-
"rewards/accuracies": 0.5,
|
| 111 |
-
"rewards/chosen": -0.020373398438096046,
|
| 112 |
-
"rewards/margins": -0.0037106983363628387,
|
| 113 |
-
"rewards/rejected": -0.016662701964378357,
|
| 114 |
-
"step": 30
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.06986027944111776,
|
| 118 |
-
"grad_norm": 15.999817261895291,
|
| 119 |
-
"learning_rate": 6.862745098039216e-07,
|
| 120 |
-
"logits/chosen": 1.1911170482635498,
|
| 121 |
-
"logits/rejected": 1.2343870401382446,
|
| 122 |
-
"logps/chosen": -0.9691197276115417,
|
| 123 |
-
"logps/rejected": -0.9998400807380676,
|
| 124 |
-
"loss": 0.6912,
|
| 125 |
-
"rewards/accuracies": 0.48750001192092896,
|
| 126 |
-
"rewards/chosen": -0.01592147909104824,
|
| 127 |
-
"rewards/margins": 0.0014309640973806381,
|
| 128 |
-
"rewards/rejected": -0.01735244318842888,
|
| 129 |
-
"step": 35
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.07984031936127745,
|
| 133 |
-
"grad_norm": 14.867532956796651,
|
| 134 |
-
"learning_rate": 7.84313725490196e-07,
|
| 135 |
-
"logits/chosen": 1.2250399589538574,
|
| 136 |
-
"logits/rejected": 1.2877371311187744,
|
| 137 |
-
"logps/chosen": -0.9882305860519409,
|
| 138 |
-
"logps/rejected": -1.013457179069519,
|
| 139 |
-
"loss": 0.6866,
|
| 140 |
-
"rewards/accuracies": 0.625,
|
| 141 |
-
"rewards/chosen": -0.014133378863334656,
|
| 142 |
-
"rewards/margins": 0.023214522749185562,
|
| 143 |
-
"rewards/rejected": -0.03734790161252022,
|
| 144 |
-
"step": 40
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 0.08982035928143713,
|
| 148 |
-
"grad_norm": 19.324457876267314,
|
| 149 |
-
"learning_rate": 8.823529411764705e-07,
|
| 150 |
-
"logits/chosen": 1.2738926410675049,
|
| 151 |
-
"logits/rejected": 1.2850474119186401,
|
| 152 |
-
"logps/chosen": -0.9675604701042175,
|
| 153 |
-
"logps/rejected": -1.0223758220672607,
|
| 154 |
-
"loss": 0.6804,
|
| 155 |
-
"rewards/accuracies": 0.6499999761581421,
|
| 156 |
-
"rewards/chosen": -0.010748568922281265,
|
| 157 |
-
"rewards/margins": 0.0401780791580677,
|
| 158 |
-
"rewards/rejected": -0.05092664435505867,
|
| 159 |
-
"step": 45
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.0998003992015968,
|
| 163 |
-
"grad_norm": 18.41965619240879,
|
| 164 |
-
"learning_rate": 9.80392156862745e-07,
|
| 165 |
-
"logits/chosen": 1.2459691762924194,
|
| 166 |
-
"logits/rejected": 1.3021132946014404,
|
| 167 |
-
"logps/chosen": -0.9164802432060242,
|
| 168 |
-
"logps/rejected": -1.013667345046997,
|
| 169 |
-
"loss": 0.6739,
|
| 170 |
-
"rewards/accuracies": 0.637499988079071,
|
| 171 |
-
"rewards/chosen": -0.026862669736146927,
|
| 172 |
-
"rewards/margins": 0.07785479724407196,
|
| 173 |
-
"rewards/rejected": -0.10471747070550919,
|
| 174 |
-
"step": 50
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 0.10978043912175649,
|
| 178 |
-
"grad_norm": 16.897558680765105,
|
| 179 |
-
"learning_rate": 9.99805057520177e-07,
|
| 180 |
-
"logits/chosen": 1.1725871562957764,
|
| 181 |
-
"logits/rejected": 1.173532247543335,
|
| 182 |
-
"logps/chosen": -0.9711742401123047,
|
| 183 |
-
"logps/rejected": -0.9595904350280762,
|
| 184 |
-
"loss": 0.6738,
|
| 185 |
-
"rewards/accuracies": 0.737500011920929,
|
| 186 |
-
"rewards/chosen": -0.03100317157804966,
|
| 187 |
-
"rewards/margins": 0.11643954366445541,
|
| 188 |
-
"rewards/rejected": -0.14744271337985992,
|
| 189 |
-
"step": 55
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.11976047904191617,
|
| 193 |
-
"grad_norm": 14.265898399249132,
|
| 194 |
-
"learning_rate": 9.990133642141357e-07,
|
| 195 |
-
"logits/chosen": 1.1897612810134888,
|
| 196 |
-
"logits/rejected": 1.2000809907913208,
|
| 197 |
-
"logps/chosen": -0.9207404255867004,
|
| 198 |
-
"logps/rejected": -0.9565631747245789,
|
| 199 |
-
"loss": 0.6693,
|
| 200 |
-
"rewards/accuracies": 0.6000000238418579,
|
| 201 |
-
"rewards/chosen": -0.06663166731595993,
|
| 202 |
-
"rewards/margins": 0.18790733814239502,
|
| 203 |
-
"rewards/rejected": -0.25453901290893555,
|
| 204 |
-
"step": 60
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 0.12974051896207583,
|
| 208 |
-
"grad_norm": 19.11690615448299,
|
| 209 |
-
"learning_rate": 9.976136999909155e-07,
|
| 210 |
-
"logits/chosen": 1.261150598526001,
|
| 211 |
-
"logits/rejected": 1.2706950902938843,
|
| 212 |
-
"logps/chosen": -0.9832024574279785,
|
| 213 |
-
"logps/rejected": -0.9762998819351196,
|
| 214 |
-
"loss": 0.6736,
|
| 215 |
-
"rewards/accuracies": 0.675000011920929,
|
| 216 |
-
"rewards/chosen": -0.08731752634048462,
|
| 217 |
-
"rewards/margins": 0.037145014852285385,
|
| 218 |
-
"rewards/rejected": -0.1244625449180603,
|
| 219 |
-
"step": 65
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.13972055888223553,
|
| 223 |
-
"grad_norm": 16.308947856088572,
|
| 224 |
-
"learning_rate": 9.956077701257707e-07,
|
| 225 |
-
"logits/chosen": 1.1949059963226318,
|
| 226 |
-
"logits/rejected": 1.2373930215835571,
|
| 227 |
-
"logps/chosen": -0.923254132270813,
|
| 228 |
-
"logps/rejected": -0.9537776708602905,
|
| 229 |
-
"loss": 0.6588,
|
| 230 |
-
"rewards/accuracies": 0.800000011920929,
|
| 231 |
-
"rewards/chosen": -0.0689881294965744,
|
| 232 |
-
"rewards/margins": 0.0771615132689476,
|
| 233 |
-
"rewards/rejected": -0.1461496502161026,
|
| 234 |
-
"step": 70
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.1497005988023952,
|
| 238 |
-
"grad_norm": 18.31597399561215,
|
| 239 |
-
"learning_rate": 9.929980185352525e-07,
|
| 240 |
-
"logits/chosen": 1.1303855180740356,
|
| 241 |
-
"logits/rejected": 1.166472315788269,
|
| 242 |
-
"logps/chosen": -1.0079195499420166,
|
| 243 |
-
"logps/rejected": -1.0075769424438477,
|
| 244 |
-
"loss": 0.6486,
|
| 245 |
-
"rewards/accuracies": 0.7124999761581421,
|
| 246 |
-
"rewards/chosen": -0.026683980599045753,
|
| 247 |
-
"rewards/margins": 0.0837903693318367,
|
| 248 |
-
"rewards/rejected": -0.1104743480682373,
|
| 249 |
-
"step": 75
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.1596806387225549,
|
| 253 |
-
"grad_norm": 16.142069213484454,
|
| 254 |
-
"learning_rate": 9.89787624799672e-07,
|
| 255 |
-
"logits/chosen": 1.1199300289154053,
|
| 256 |
-
"logits/rejected": 1.1778942346572876,
|
| 257 |
-
"logps/chosen": -0.9652664065361023,
|
| 258 |
-
"logps/rejected": -1.0307899713516235,
|
| 259 |
-
"loss": 0.6496,
|
| 260 |
-
"rewards/accuracies": 0.737500011920929,
|
| 261 |
-
"rewards/chosen": -0.0005544237792491913,
|
| 262 |
-
"rewards/margins": 0.10419625043869019,
|
| 263 |
-
"rewards/rejected": -0.10475067049264908,
|
| 264 |
-
"step": 80
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 0.16966067864271456,
|
| 268 |
-
"grad_norm": 15.306215682029483,
|
| 269 |
-
"learning_rate": 9.859805002892731e-07,
|
| 270 |
-
"logits/chosen": 1.1645634174346924,
|
| 271 |
-
"logits/rejected": 1.1519969701766968,
|
| 272 |
-
"logps/chosen": -0.9917243719100952,
|
| 273 |
-
"logps/rejected": -1.060416340827942,
|
| 274 |
-
"loss": 0.661,
|
| 275 |
-
"rewards/accuracies": 0.675000011920929,
|
| 276 |
-
"rewards/chosen": -0.09171438962221146,
|
| 277 |
-
"rewards/margins": 0.20642364025115967,
|
| 278 |
-
"rewards/rejected": -0.2981380820274353,
|
| 279 |
-
"step": 85
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 0.17964071856287425,
|
| 283 |
-
"grad_norm": 15.254186655042064,
|
| 284 |
-
"learning_rate": 9.81581283398829e-07,
|
| 285 |
-
"logits/chosen": 1.103563904762268,
|
| 286 |
-
"logits/rejected": 1.1300289630889893,
|
| 287 |
-
"logps/chosen": -0.969647228717804,
|
| 288 |
-
"logps/rejected": -0.9916777610778809,
|
| 289 |
-
"loss": 0.6337,
|
| 290 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 291 |
-
"rewards/chosen": 0.013281228020787239,
|
| 292 |
-
"rewards/margins": 0.1793346405029297,
|
| 293 |
-
"rewards/rejected": -0.1660534292459488,
|
| 294 |
-
"step": 90
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 0.18962075848303392,
|
| 298 |
-
"grad_norm": 15.282160453652757,
|
| 299 |
-
"learning_rate": 9.765953338964734e-07,
|
| 300 |
-
"logits/chosen": 1.0921884775161743,
|
| 301 |
-
"logits/rejected": 1.1226933002471924,
|
| 302 |
-
"logps/chosen": -1.0730102062225342,
|
| 303 |
-
"logps/rejected": -1.1024601459503174,
|
| 304 |
-
"loss": 0.6385,
|
| 305 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 306 |
-
"rewards/chosen": 0.0058188484981656075,
|
| 307 |
-
"rewards/margins": 0.21344614028930664,
|
| 308 |
-
"rewards/rejected": -0.2076272964477539,
|
| 309 |
-
"step": 95
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 0.1996007984031936,
|
| 313 |
-
"grad_norm": 21.61748232589367,
|
| 314 |
-
"learning_rate": 9.710287263936483e-07,
|
| 315 |
-
"logits/chosen": 1.0795061588287354,
|
| 316 |
-
"logits/rejected": 1.0822367668151855,
|
| 317 |
-
"logps/chosen": -1.0821864604949951,
|
| 318 |
-
"logps/rejected": -1.1186153888702393,
|
| 319 |
-
"loss": 0.621,
|
| 320 |
-
"rewards/accuracies": 0.6875,
|
| 321 |
-
"rewards/chosen": 0.07217474281787872,
|
| 322 |
-
"rewards/margins": 0.3310089707374573,
|
| 323 |
-
"rewards/rejected": -0.25883427262306213,
|
| 324 |
-
"step": 100
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.20958083832335328,
|
| 328 |
-
"grad_norm": 23.129340339628143,
|
| 329 |
-
"learning_rate": 9.648882429441256e-07,
|
| 330 |
-
"logits/chosen": 1.0758105516433716,
|
| 331 |
-
"logits/rejected": 1.0959206819534302,
|
| 332 |
-
"logps/chosen": -0.9559111595153809,
|
| 333 |
-
"logps/rejected": -0.9702944755554199,
|
| 334 |
-
"loss": 0.6278,
|
| 335 |
-
"rewards/accuracies": 0.7124999761581421,
|
| 336 |
-
"rewards/chosen": -0.027358528226614,
|
| 337 |
-
"rewards/margins": 0.17237558960914612,
|
| 338 |
-
"rewards/rejected": -0.1997341364622116,
|
| 339 |
-
"step": 105
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.21956087824351297,
|
| 343 |
-
"grad_norm": 15.179993551058994,
|
| 344 |
-
"learning_rate": 9.581813647811197e-07,
|
| 345 |
-
"logits/chosen": 1.094158411026001,
|
| 346 |
-
"logits/rejected": 1.114650011062622,
|
| 347 |
-
"logps/chosen": -0.9534612894058228,
|
| 348 |
-
"logps/rejected": -0.9674752354621887,
|
| 349 |
-
"loss": 0.6124,
|
| 350 |
-
"rewards/accuracies": 0.75,
|
| 351 |
-
"rewards/chosen": 0.058080364018678665,
|
| 352 |
-
"rewards/margins": 0.30136996507644653,
|
| 353 |
-
"rewards/rejected": -0.24328958988189697,
|
| 354 |
-
"step": 110
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 0.22954091816367264,
|
| 358 |
-
"grad_norm": 13.705886523650493,
|
| 359 |
-
"learning_rate": 9.509162632025569e-07,
|
| 360 |
-
"logits/chosen": 1.0607202053070068,
|
| 361 |
-
"logits/rejected": 1.045424461364746,
|
| 362 |
-
"logps/chosen": -1.1245949268341064,
|
| 363 |
-
"logps/rejected": -1.1509120464324951,
|
| 364 |
-
"loss": 0.6167,
|
| 365 |
-
"rewards/accuracies": 0.675000011920929,
|
| 366 |
-
"rewards/chosen": -0.07293939590454102,
|
| 367 |
-
"rewards/margins": 0.17659255862236023,
|
| 368 |
-
"rewards/rejected": -0.24953195452690125,
|
| 369 |
-
"step": 115
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 0.23952095808383234,
|
| 373 |
-
"grad_norm": 27.63745397695,
|
| 374 |
-
"learning_rate": 9.431017896156073e-07,
|
| 375 |
-
"logits/chosen": 1.0884307622909546,
|
| 376 |
-
"logits/rejected": 1.101423978805542,
|
| 377 |
-
"logps/chosen": -0.9731513857841492,
|
| 378 |
-
"logps/rejected": -1.023944616317749,
|
| 379 |
-
"loss": 0.6061,
|
| 380 |
-
"rewards/accuracies": 0.6625000238418579,
|
| 381 |
-
"rewards/chosen": 0.055571239441633224,
|
| 382 |
-
"rewards/margins": 0.36377543210983276,
|
| 383 |
-
"rewards/rejected": -0.3082042336463928,
|
| 384 |
-
"step": 120
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 0.249500998003992,
|
| 388 |
-
"grad_norm": 14.89222449165367,
|
| 389 |
-
"learning_rate": 9.347474647526095e-07,
|
| 390 |
-
"logits/chosen": 1.0270758867263794,
|
| 391 |
-
"logits/rejected": 1.061629056930542,
|
| 392 |
-
"logps/chosen": -0.9520384073257446,
|
| 393 |
-
"logps/rejected": -0.9609392285346985,
|
| 394 |
-
"loss": 0.5982,
|
| 395 |
-
"rewards/accuracies": 0.75,
|
| 396 |
-
"rewards/chosen": -0.03607760742306709,
|
| 397 |
-
"rewards/margins": 0.205668643116951,
|
| 398 |
-
"rewards/rejected": -0.24174626171588898,
|
| 399 |
-
"step": 125
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 0.25948103792415167,
|
| 403 |
-
"grad_norm": 16.598481366896863,
|
| 404 |
-
"learning_rate": 9.258634670715237e-07,
|
| 405 |
-
"logits/chosen": 0.9625661969184875,
|
| 406 |
-
"logits/rejected": 0.9923114776611328,
|
| 407 |
-
"logps/chosen": -0.9781230092048645,
|
| 408 |
-
"logps/rejected": -1.0380465984344482,
|
| 409 |
-
"loss": 0.6033,
|
| 410 |
-
"rewards/accuracies": 0.8374999761581421,
|
| 411 |
-
"rewards/chosen": -0.08731977641582489,
|
| 412 |
-
"rewards/margins": 0.32325008511543274,
|
| 413 |
-
"rewards/rejected": -0.41056984663009644,
|
| 414 |
-
"step": 130
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 0.2694610778443114,
|
| 418 |
-
"grad_norm": 17.332548482467526,
|
| 419 |
-
"learning_rate": 9.164606203550497e-07,
|
| 420 |
-
"logits/chosen": 1.0289146900177002,
|
| 421 |
-
"logits/rejected": 1.0656378269195557,
|
| 422 |
-
"logps/chosen": -0.898147463798523,
|
| 423 |
-
"logps/rejected": -0.949881911277771,
|
| 424 |
-
"loss": 0.602,
|
| 425 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 426 |
-
"rewards/chosen": -0.01896672695875168,
|
| 427 |
-
"rewards/margins": 0.33230799436569214,
|
| 428 |
-
"rewards/rejected": -0.351274698972702,
|
| 429 |
-
"step": 135
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.27944111776447106,
|
| 433 |
-
"grad_norm": 16.198930205544908,
|
| 434 |
-
"learning_rate": 9.065503805235137e-07,
|
| 435 |
-
"logits/chosen": 1.0169068574905396,
|
| 436 |
-
"logits/rejected": 1.0443648099899292,
|
| 437 |
-
"logps/chosen": -0.9837234616279602,
|
| 438 |
-
"logps/rejected": -1.015891432762146,
|
| 439 |
-
"loss": 0.5962,
|
| 440 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 441 |
-
"rewards/chosen": -0.11362324655056,
|
| 442 |
-
"rewards/margins": 0.29118040204048157,
|
| 443 |
-
"rewards/rejected": -0.40480366349220276,
|
| 444 |
-
"step": 140
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.2894211576846307,
|
| 448 |
-
"grad_norm": 14.26896893631501,
|
| 449 |
-
"learning_rate": 8.961448216775953e-07,
|
| 450 |
-
"logits/chosen": 0.9983326196670532,
|
| 451 |
-
"logits/rejected": 0.984410285949707,
|
| 452 |
-
"logps/chosen": -1.0478496551513672,
|
| 453 |
-
"logps/rejected": -1.0458667278289795,
|
| 454 |
-
"loss": 0.6106,
|
| 455 |
-
"rewards/accuracies": 0.6625000238418579,
|
| 456 |
-
"rewards/chosen": -0.384859174489975,
|
| 457 |
-
"rewards/margins": 0.23550406098365784,
|
| 458 |
-
"rewards/rejected": -0.6203632354736328,
|
| 459 |
-
"step": 145
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 0.2994011976047904,
|
| 463 |
-
"grad_norm": 17.114721290662626,
|
| 464 |
-
"learning_rate": 8.852566213878946e-07,
|
| 465 |
-
"logits/chosen": 0.9434808492660522,
|
| 466 |
-
"logits/rejected": 0.9709761738777161,
|
| 467 |
-
"logps/chosen": -0.946380615234375,
|
| 468 |
-
"logps/rejected": -1.011891484260559,
|
| 469 |
-
"loss": 0.6051,
|
| 470 |
-
"rewards/accuracies": 0.8125,
|
| 471 |
-
"rewards/chosen": -0.10636933147907257,
|
| 472 |
-
"rewards/margins": 0.38454684615135193,
|
| 473 |
-
"rewards/rejected": -0.4909161627292633,
|
| 474 |
-
"step": 150
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 0.3093812375249501,
|
| 478 |
-
"grad_norm": 12.532368784046184,
|
| 479 |
-
"learning_rate": 8.73899045249266e-07,
|
| 480 |
-
"logits/chosen": 0.9879257082939148,
|
| 481 |
-
"logits/rejected": 1.0284427404403687,
|
| 482 |
-
"logps/chosen": -0.9887377619743347,
|
| 483 |
-
"logps/rejected": -1.0425301790237427,
|
| 484 |
-
"loss": 0.5785,
|
| 485 |
-
"rewards/accuracies": 0.7124999761581421,
|
| 486 |
-
"rewards/chosen": -0.18314163386821747,
|
| 487 |
-
"rewards/margins": 0.3873257339000702,
|
| 488 |
-
"rewards/rejected": -0.5704673528671265,
|
| 489 |
-
"step": 155
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 0.3193612774451098,
|
| 493 |
-
"grad_norm": 20.653306879612042,
|
| 494 |
-
"learning_rate": 8.620859307187338e-07,
|
| 495 |
-
"logits/chosen": 0.9365812540054321,
|
| 496 |
-
"logits/rejected": 0.9702633619308472,
|
| 497 |
-
"logps/chosen": -0.9792872667312622,
|
| 498 |
-
"logps/rejected": -1.0119134187698364,
|
| 499 |
-
"loss": 0.5915,
|
| 500 |
-
"rewards/accuracies": 0.7124999761581421,
|
| 501 |
-
"rewards/chosen": -0.28814950585365295,
|
| 502 |
-
"rewards/margins": 0.3159632086753845,
|
| 503 |
-
"rewards/rejected": -0.6041126847267151,
|
| 504 |
-
"step": 160
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 0.32934131736526945,
|
| 508 |
-
"grad_norm": 46.72961396087624,
|
| 509 |
-
"learning_rate": 8.498316702566826e-07,
|
| 510 |
-
"logits/chosen": 1.0088741779327393,
|
| 511 |
-
"logits/rejected": 1.0628801584243774,
|
| 512 |
-
"logps/chosen": -1.0191477537155151,
|
| 513 |
-
"logps/rejected": -1.0589340925216675,
|
| 514 |
-
"loss": 0.5784,
|
| 515 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 516 |
-
"rewards/chosen": -0.43470463156700134,
|
| 517 |
-
"rewards/margins": 0.3699150085449219,
|
| 518 |
-
"rewards/rejected": -0.8046196103096008,
|
| 519 |
-
"step": 165
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.3393213572854291,
|
| 523 |
-
"grad_norm": 17.875828780004216,
|
| 524 |
-
"learning_rate": 8.371511937918617e-07,
|
| 525 |
-
"logits/chosen": 0.8878538012504578,
|
| 526 |
-
"logits/rejected": 0.9223688840866089,
|
| 527 |
-
"logps/chosen": -0.9879400134086609,
|
| 528 |
-
"logps/rejected": -1.029184103012085,
|
| 529 |
-
"loss": 0.6033,
|
| 530 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 531 |
-
"rewards/chosen": -0.3266729712486267,
|
| 532 |
-
"rewards/margins": 0.5894866585731506,
|
| 533 |
-
"rewards/rejected": -0.9161597490310669,
|
| 534 |
-
"step": 170
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.34930139720558884,
|
| 538 |
-
"grad_norm": 16.753564996079323,
|
| 539 |
-
"learning_rate": 8.240599505315654e-07,
|
| 540 |
-
"logits/chosen": 0.8993169665336609,
|
| 541 |
-
"logits/rejected": 0.931827187538147,
|
| 542 |
-
"logps/chosen": -1.0300993919372559,
|
| 543 |
-
"logps/rejected": -1.104100227355957,
|
| 544 |
-
"loss": 0.5719,
|
| 545 |
-
"rewards/accuracies": 0.762499988079071,
|
| 546 |
-
"rewards/chosen": -0.26795631647109985,
|
| 547 |
-
"rewards/margins": 0.4897838234901428,
|
| 548 |
-
"rewards/rejected": -0.7577401399612427,
|
| 549 |
-
"step": 175
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.3592814371257485,
|
| 553 |
-
"grad_norm": 37.583807172799176,
|
| 554 |
-
"learning_rate": 8.105738901391551e-07,
|
| 555 |
-
"logits/chosen": 0.9503185153007507,
|
| 556 |
-
"logits/rejected": 0.9612258076667786,
|
| 557 |
-
"logps/chosen": -0.9984599947929382,
|
| 558 |
-
"logps/rejected": -1.0410314798355103,
|
| 559 |
-
"loss": 0.5749,
|
| 560 |
-
"rewards/accuracies": 0.800000011920929,
|
| 561 |
-
"rewards/chosen": -0.3058738708496094,
|
| 562 |
-
"rewards/margins": 0.4171958565711975,
|
| 563 |
-
"rewards/rejected": -0.7230697274208069,
|
| 564 |
-
"step": 180
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 0.36926147704590817,
|
| 568 |
-
"grad_norm": 31.25023262450056,
|
| 569 |
-
"learning_rate": 7.967094433018508e-07,
|
| 570 |
-
"logits/chosen": 0.8538301587104797,
|
| 571 |
-
"logits/rejected": 0.8765438199043274,
|
| 572 |
-
"logps/chosen": -1.0639922618865967,
|
| 573 |
-
"logps/rejected": -1.1690170764923096,
|
| 574 |
-
"loss": 0.5628,
|
| 575 |
-
"rewards/accuracies": 0.75,
|
| 576 |
-
"rewards/chosen": -0.31266123056411743,
|
| 577 |
-
"rewards/margins": 0.6013716459274292,
|
| 578 |
-
"rewards/rejected": -0.9140329360961914,
|
| 579 |
-
"step": 185
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 0.37924151696606784,
|
| 583 |
-
"grad_norm": 21.697159728289254,
|
| 584 |
-
"learning_rate": 7.82483501712469e-07,
|
| 585 |
-
"logits/chosen": 0.8337196111679077,
|
| 586 |
-
"logits/rejected": 0.8647937774658203,
|
| 587 |
-
"logps/chosen": -1.0461093187332153,
|
| 588 |
-
"logps/rejected": -1.0427898168563843,
|
| 589 |
-
"loss": 0.5701,
|
| 590 |
-
"rewards/accuracies": 0.737500011920929,
|
| 591 |
-
"rewards/chosen": -0.48062148690223694,
|
| 592 |
-
"rewards/margins": 0.4651462435722351,
|
| 593 |
-
"rewards/rejected": -0.9457677602767944,
|
| 594 |
-
"step": 190
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 0.38922155688622756,
|
| 598 |
-
"grad_norm": 13.452601846054984,
|
| 599 |
-
"learning_rate": 7.679133974894982e-07,
|
| 600 |
-
"logits/chosen": 0.8574188351631165,
|
| 601 |
-
"logits/rejected": 0.8771845102310181,
|
| 602 |
-
"logps/chosen": -1.0106494426727295,
|
| 603 |
-
"logps/rejected": -1.1300721168518066,
|
| 604 |
-
"loss": 0.5649,
|
| 605 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 606 |
-
"rewards/chosen": -0.5489331483840942,
|
| 607 |
-
"rewards/margins": 0.6363680958747864,
|
| 608 |
-
"rewards/rejected": -1.1853011846542358,
|
| 609 |
-
"step": 195
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 0.3992015968063872,
|
| 613 |
-
"grad_norm": 14.045999687736995,
|
| 614 |
-
"learning_rate": 7.530168820605818e-07,
|
| 615 |
-
"logits/chosen": 0.8429155349731445,
|
| 616 |
-
"logits/rejected": 0.8586641550064087,
|
| 617 |
-
"logps/chosen": -1.1243480443954468,
|
| 618 |
-
"logps/rejected": -1.126988172531128,
|
| 619 |
-
"loss": 0.5556,
|
| 620 |
-
"rewards/accuracies": 0.699999988079071,
|
| 621 |
-
"rewards/chosen": -0.611405074596405,
|
| 622 |
-
"rewards/margins": 0.5212133526802063,
|
| 623 |
-
"rewards/rejected": -1.1326184272766113,
|
| 624 |
-
"step": 200
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 0.4091816367265469,
|
| 628 |
-
"grad_norm": 11.17840118621038,
|
| 629 |
-
"learning_rate": 7.378121045351377e-07,
|
| 630 |
-
"logits/chosen": 0.8805230855941772,
|
| 631 |
-
"logits/rejected": 0.8616092801094055,
|
| 632 |
-
"logps/chosen": -1.025155782699585,
|
| 633 |
-
"logps/rejected": -1.1043505668640137,
|
| 634 |
-
"loss": 0.5274,
|
| 635 |
-
"rewards/accuracies": 0.75,
|
| 636 |
-
"rewards/chosen": -0.5864971280097961,
|
| 637 |
-
"rewards/margins": 0.6259642839431763,
|
| 638 |
-
"rewards/rejected": -1.2124613523483276,
|
| 639 |
-
"step": 205
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.41916167664670656,
|
| 643 |
-
"grad_norm": 17.02714464481709,
|
| 644 |
-
"learning_rate": 7.223175895924637e-07,
|
| 645 |
-
"logits/chosen": 0.8006958961486816,
|
| 646 |
-
"logits/rejected": 0.8346986770629883,
|
| 647 |
-
"logps/chosen": -1.0029652118682861,
|
| 648 |
-
"logps/rejected": -1.0653495788574219,
|
| 649 |
-
"loss": 0.5437,
|
| 650 |
-
"rewards/accuracies": 0.800000011920929,
|
| 651 |
-
"rewards/chosen": -0.5938572287559509,
|
| 652 |
-
"rewards/margins": 0.5961542725563049,
|
| 653 |
-
"rewards/rejected": -1.1900113821029663,
|
| 654 |
-
"step": 210
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.4291417165668663,
|
| 658 |
-
"grad_norm": 12.708513730083673,
|
| 659 |
-
"learning_rate": 7.065522149122709e-07,
|
| 660 |
-
"logits/chosen": 0.7250524759292603,
|
| 661 |
-
"logits/rejected": 0.7448651790618896,
|
| 662 |
-
"logps/chosen": -1.0977203845977783,
|
| 663 |
-
"logps/rejected": -1.1671373844146729,
|
| 664 |
-
"loss": 0.5296,
|
| 665 |
-
"rewards/accuracies": 0.8374999761581421,
|
| 666 |
-
"rewards/chosen": -0.8822553753852844,
|
| 667 |
-
"rewards/margins": 0.5623547434806824,
|
| 668 |
-
"rewards/rejected": -1.4446099996566772,
|
| 669 |
-
"step": 215
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 0.43912175648702595,
|
| 673 |
-
"grad_norm": 23.02316300872982,
|
| 674 |
-
"learning_rate": 6.905351881751371e-07,
|
| 675 |
-
"logits/chosen": 0.7448990345001221,
|
| 676 |
-
"logits/rejected": 0.759792685508728,
|
| 677 |
-
"logps/chosen": -1.0590717792510986,
|
| 678 |
-
"logps/rejected": -1.1264311075210571,
|
| 679 |
-
"loss": 0.5411,
|
| 680 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 681 |
-
"rewards/chosen": -0.7196047306060791,
|
| 682 |
-
"rewards/margins": 0.6396269798278809,
|
| 683 |
-
"rewards/rejected": -1.35923171043396,
|
| 684 |
-
"step": 220
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 0.4491017964071856,
|
| 688 |
-
"grad_norm": 15.317013674505024,
|
| 689 |
-
"learning_rate": 6.742860236609076e-07,
|
| 690 |
-
"logits/chosen": 0.7387903928756714,
|
| 691 |
-
"logits/rejected": 0.7482835054397583,
|
| 692 |
-
"logps/chosen": -1.0746328830718994,
|
| 693 |
-
"logps/rejected": -1.0935697555541992,
|
| 694 |
-
"loss": 0.5362,
|
| 695 |
-
"rewards/accuracies": 0.75,
|
| 696 |
-
"rewards/chosen": -0.91168612241745,
|
| 697 |
-
"rewards/margins": 0.43347805738449097,
|
| 698 |
-
"rewards/rejected": -1.3451642990112305,
|
| 699 |
-
"step": 225
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 0.4590818363273453,
|
| 703 |
-
"grad_norm": 20.979897267224555,
|
| 704 |
-
"learning_rate": 6.578245184735512e-07,
|
| 705 |
-
"logits/chosen": 0.7530170679092407,
|
| 706 |
-
"logits/rejected": 0.7971817851066589,
|
| 707 |
-
"logps/chosen": -1.0877180099487305,
|
| 708 |
-
"logps/rejected": -1.1133688688278198,
|
| 709 |
-
"loss": 0.5297,
|
| 710 |
-
"rewards/accuracies": 0.762499988079071,
|
| 711 |
-
"rewards/chosen": -0.7731893062591553,
|
| 712 |
-
"rewards/margins": 0.6359795331954956,
|
| 713 |
-
"rewards/rejected": -1.4091689586639404,
|
| 714 |
-
"step": 230
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 0.469061876247505,
|
| 718 |
-
"grad_norm": 23.654443057551042,
|
| 719 |
-
"learning_rate": 6.411707284214383e-07,
|
| 720 |
-
"logits/chosen": 0.7936985492706299,
|
| 721 |
-
"logits/rejected": 0.8071283102035522,
|
| 722 |
-
"logps/chosen": -1.1007978916168213,
|
| 723 |
-
"logps/rejected": -1.175083041191101,
|
| 724 |
-
"loss": 0.5189,
|
| 725 |
-
"rewards/accuracies": 0.75,
|
| 726 |
-
"rewards/chosen": -0.8833421468734741,
|
| 727 |
-
"rewards/margins": 0.6522087454795837,
|
| 728 |
-
"rewards/rejected": -1.535550832748413,
|
| 729 |
-
"step": 235
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 0.47904191616766467,
|
| 733 |
-
"grad_norm": 17.617163550183303,
|
| 734 |
-
"learning_rate": 6.243449435824276e-07,
|
| 735 |
-
"logits/chosen": 0.7208539247512817,
|
| 736 |
-
"logits/rejected": 0.7419721484184265,
|
| 737 |
-
"logps/chosen": -0.9637038111686707,
|
| 738 |
-
"logps/rejected": -1.0924652814865112,
|
| 739 |
-
"loss": 0.5239,
|
| 740 |
-
"rewards/accuracies": 0.75,
|
| 741 |
-
"rewards/chosen": -0.823961079120636,
|
| 742 |
-
"rewards/margins": 0.9264790415763855,
|
| 743 |
-
"rewards/rejected": -1.750440239906311,
|
| 744 |
-
"step": 240
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.48902195608782434,
|
| 748 |
-
"grad_norm": 23.429183199427385,
|
| 749 |
-
"learning_rate": 6.073676635835316e-07,
|
| 750 |
-
"logits/chosen": 0.7787541747093201,
|
| 751 |
-
"logits/rejected": 0.8052628636360168,
|
| 752 |
-
"logps/chosen": -1.0102448463439941,
|
| 753 |
-
"logps/rejected": -1.2205474376678467,
|
| 754 |
-
"loss": 0.5087,
|
| 755 |
-
"rewards/accuracies": 0.8125,
|
| 756 |
-
"rewards/chosen": -0.7600202560424805,
|
| 757 |
-
"rewards/margins": 1.2399606704711914,
|
| 758 |
-
"rewards/rejected": -1.9999806880950928,
|
| 759 |
-
"step": 245
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.499001996007984,
|
| 763 |
-
"grad_norm": 15.897882119964894,
|
| 764 |
-
"learning_rate": 5.9025957262528e-07,
|
| 765 |
-
"logits/chosen": 0.674079179763794,
|
| 766 |
-
"logits/rejected": 0.6560205221176147,
|
| 767 |
-
"logps/chosen": -1.0422561168670654,
|
| 768 |
-
"logps/rejected": -1.15647292137146,
|
| 769 |
-
"loss": 0.5129,
|
| 770 |
-
"rewards/accuracies": 0.8374999761581421,
|
| 771 |
-
"rewards/chosen": -0.9762313961982727,
|
| 772 |
-
"rewards/margins": 0.9589719772338867,
|
| 773 |
-
"rewards/rejected": -1.9352035522460938,
|
| 774 |
-
"step": 250
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 0.5089820359281437,
|
| 778 |
-
"grad_norm": 23.926407573258526,
|
| 779 |
-
"learning_rate": 5.730415142812058e-07,
|
| 780 |
-
"logits/chosen": 0.6435045003890991,
|
| 781 |
-
"logits/rejected": 0.6540420055389404,
|
| 782 |
-
"logps/chosen": -1.1076724529266357,
|
| 783 |
-
"logps/rejected": -1.2272051572799683,
|
| 784 |
-
"loss": 0.5207,
|
| 785 |
-
"rewards/accuracies": 0.824999988079071,
|
| 786 |
-
"rewards/chosen": -1.1593824625015259,
|
| 787 |
-
"rewards/margins": 0.9408347010612488,
|
| 788 |
-
"rewards/rejected": -2.10021710395813,
|
| 789 |
-
"step": 255
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 0.5189620758483033,
|
| 793 |
-
"grad_norm": 25.873872774153636,
|
| 794 |
-
"learning_rate": 5.557344661031627e-07,
|
| 795 |
-
"logits/chosen": 0.6078227758407593,
|
| 796 |
-
"logits/rejected": 0.6234728097915649,
|
| 797 |
-
"logps/chosen": -1.1247365474700928,
|
| 798 |
-
"logps/rejected": -1.1961596012115479,
|
| 799 |
-
"loss": 0.5558,
|
| 800 |
-
"rewards/accuracies": 0.7124999761581421,
|
| 801 |
-
"rewards/chosen": -1.4092304706573486,
|
| 802 |
-
"rewards/margins": 0.7921189665794373,
|
| 803 |
-
"rewards/rejected": -2.2013492584228516,
|
| 804 |
-
"step": 260
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 0.5289421157684631,
|
| 808 |
-
"grad_norm": 14.32961070454234,
|
| 809 |
-
"learning_rate": 5.383595140634093e-07,
|
| 810 |
-
"logits/chosen": 0.634459912776947,
|
| 811 |
-
"logits/rejected": 0.6405037641525269,
|
| 812 |
-
"logps/chosen": -1.054018259048462,
|
| 813 |
-
"logps/rejected": -1.1594079732894897,
|
| 814 |
-
"loss": 0.5108,
|
| 815 |
-
"rewards/accuracies": 0.699999988079071,
|
| 816 |
-
"rewards/chosen": -1.3187201023101807,
|
| 817 |
-
"rewards/margins": 0.7670294046401978,
|
| 818 |
-
"rewards/rejected": -2.085749626159668,
|
| 819 |
-
"step": 265
|
| 820 |
-
},
|
| 821 |
-
{
|
| 822 |
-
"epoch": 0.5389221556886228,
|
| 823 |
-
"grad_norm": 12.765239103380393,
|
| 824 |
-
"learning_rate": 5.209378268645997e-07,
|
| 825 |
-
"logits/chosen": 0.6027650237083435,
|
| 826 |
-
"logits/rejected": 0.6287568211555481,
|
| 827 |
-
"logps/chosen": -1.008739948272705,
|
| 828 |
-
"logps/rejected": -1.1450139284133911,
|
| 829 |
-
"loss": 0.5318,
|
| 830 |
-
"rewards/accuracies": 0.75,
|
| 831 |
-
"rewards/chosen": -1.2432196140289307,
|
| 832 |
-
"rewards/margins": 0.6896325349807739,
|
| 833 |
-
"rewards/rejected": -1.932852029800415,
|
| 834 |
-
"step": 270
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 0.5489021956087824,
|
| 838 |
-
"grad_norm": 17.632252940280708,
|
| 839 |
-
"learning_rate": 5.034906301489807e-07,
|
| 840 |
-
"logits/chosen": 0.6728609204292297,
|
| 841 |
-
"logits/rejected": 0.6657983064651489,
|
| 842 |
-
"logps/chosen": -1.1084893941879272,
|
| 843 |
-
"logps/rejected": -1.182081937789917,
|
| 844 |
-
"loss": 0.5413,
|
| 845 |
-
"rewards/accuracies": 0.762499988079071,
|
| 846 |
-
"rewards/chosen": -1.4515244960784912,
|
| 847 |
-
"rewards/margins": 0.8624111413955688,
|
| 848 |
-
"rewards/rejected": -2.3139355182647705,
|
| 849 |
-
"step": 275
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 0.5588822355289421,
|
| 853 |
-
"grad_norm": 22.324779245169193,
|
| 854 |
-
"learning_rate": 4.860391806382156e-07,
|
| 855 |
-
"logits/chosen": 0.6076717972755432,
|
| 856 |
-
"logits/rejected": 0.6105688810348511,
|
| 857 |
-
"logps/chosen": -1.0900273323059082,
|
| 858 |
-
"logps/rejected": -1.1191543340682983,
|
| 859 |
-
"loss": 0.512,
|
| 860 |
-
"rewards/accuracies": 0.862500011920929,
|
| 861 |
-
"rewards/chosen": -1.21976900100708,
|
| 862 |
-
"rewards/margins": 0.7310382723808289,
|
| 863 |
-
"rewards/rejected": -1.9508073329925537,
|
| 864 |
-
"step": 280
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 0.5688622754491018,
|
| 868 |
-
"grad_norm": 18.31410229389494,
|
| 869 |
-
"learning_rate": 4.686047402353433e-07,
|
| 870 |
-
"logits/chosen": 0.6535500288009644,
|
| 871 |
-
"logits/rejected": 0.6188856959342957,
|
| 872 |
-
"logps/chosen": -1.0886776447296143,
|
| 873 |
-
"logps/rejected": -1.1758826971054077,
|
| 874 |
-
"loss": 0.4871,
|
| 875 |
-
"rewards/accuracies": 0.8125,
|
| 876 |
-
"rewards/chosen": -1.381490707397461,
|
| 877 |
-
"rewards/margins": 0.8018454313278198,
|
| 878 |
-
"rewards/rejected": -2.1833362579345703,
|
| 879 |
-
"step": 285
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 0.5788423153692615,
|
| 883 |
-
"grad_norm": 28.66415855277427,
|
| 884 |
-
"learning_rate": 4.512085501204253e-07,
|
| 885 |
-
"logits/chosen": 0.7206050157546997,
|
| 886 |
-
"logits/rejected": 0.7634114623069763,
|
| 887 |
-
"logps/chosen": -1.116304636001587,
|
| 888 |
-
"logps/rejected": -1.1659752130508423,
|
| 889 |
-
"loss": 0.553,
|
| 890 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 891 |
-
"rewards/chosen": -1.4132583141326904,
|
| 892 |
-
"rewards/margins": 0.657631516456604,
|
| 893 |
-
"rewards/rejected": -2.070889711380005,
|
| 894 |
-
"step": 290
|
| 895 |
-
},
|
| 896 |
-
{
|
| 897 |
-
"epoch": 0.5888223552894212,
|
| 898 |
-
"grad_norm": 17.845472804681062,
|
| 899 |
-
"learning_rate": 4.338718048714387e-07,
|
| 900 |
-
"logits/chosen": 0.6159842610359192,
|
| 901 |
-
"logits/rejected": 0.6279784440994263,
|
| 902 |
-
"logps/chosen": -1.0876820087432861,
|
| 903 |
-
"logps/rejected": -1.1220331192016602,
|
| 904 |
-
"loss": 0.5134,
|
| 905 |
-
"rewards/accuracies": 0.8374999761581421,
|
| 906 |
-
"rewards/chosen": -1.233598232269287,
|
| 907 |
-
"rewards/margins": 0.5510263442993164,
|
| 908 |
-
"rewards/rejected": -1.784624695777893,
|
| 909 |
-
"step": 295
|
| 910 |
-
},
|
| 911 |
-
{
|
| 912 |
-
"epoch": 0.5988023952095808,
|
| 913 |
-
"grad_norm": 22.693877555997958,
|
| 914 |
-
"learning_rate": 4.166156266419489e-07,
|
| 915 |
-
"logits/chosen": 0.6069675087928772,
|
| 916 |
-
"logits/rejected": 0.6444044709205627,
|
| 917 |
-
"logps/chosen": -1.0315536260604858,
|
| 918 |
-
"logps/rejected": -1.1916632652282715,
|
| 919 |
-
"loss": 0.5015,
|
| 920 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 921 |
-
"rewards/chosen": -1.4063842296600342,
|
| 922 |
-
"rewards/margins": 0.8028200268745422,
|
| 923 |
-
"rewards/rejected": -2.2092041969299316,
|
| 924 |
-
"step": 300
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 0.6087824351297405,
|
| 928 |
-
"grad_norm": 29.176362707547757,
|
| 929 |
-
"learning_rate": 3.9946103942701775e-07,
|
| 930 |
-
"logits/chosen": 0.6731375455856323,
|
| 931 |
-
"logits/rejected": 0.6558948755264282,
|
| 932 |
-
"logps/chosen": -1.0202916860580444,
|
| 933 |
-
"logps/rejected": -1.1923191547393799,
|
| 934 |
-
"loss": 0.4922,
|
| 935 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 936 |
-
"rewards/chosen": -1.134216547012329,
|
| 937 |
-
"rewards/margins": 0.8526105880737305,
|
| 938 |
-
"rewards/rejected": -1.9868271350860596,
|
| 939 |
-
"step": 305
|
| 940 |
-
},
|
| 941 |
-
{
|
| 942 |
-
"epoch": 0.6187624750499002,
|
| 943 |
-
"grad_norm": 16.68482157863775,
|
| 944 |
-
"learning_rate": 3.8242894344870495e-07,
|
| 945 |
-
"logits/chosen": 0.6814749836921692,
|
| 946 |
-
"logits/rejected": 0.6544811129570007,
|
| 947 |
-
"logps/chosen": -1.1450976133346558,
|
| 948 |
-
"logps/rejected": -1.2438080310821533,
|
| 949 |
-
"loss": 0.518,
|
| 950 |
-
"rewards/accuracies": 0.875,
|
| 951 |
-
"rewards/chosen": -1.6444076299667358,
|
| 952 |
-
"rewards/margins": 0.7449675798416138,
|
| 953 |
-
"rewards/rejected": -2.3893752098083496,
|
| 954 |
-
"step": 310
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 0.6287425149700598,
|
| 958 |
-
"grad_norm": 24.235832517472502,
|
| 959 |
-
"learning_rate": 3.6554008969236715e-07,
|
| 960 |
-
"logits/chosen": 0.6473463773727417,
|
| 961 |
-
"logits/rejected": 0.6526838541030884,
|
| 962 |
-
"logps/chosen": -1.1155353784561157,
|
| 963 |
-
"logps/rejected": -1.2412601709365845,
|
| 964 |
-
"loss": 0.5036,
|
| 965 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 966 |
-
"rewards/chosen": -1.5284216403961182,
|
| 967 |
-
"rewards/margins": 0.7738613486289978,
|
| 968 |
-
"rewards/rejected": -2.3022828102111816,
|
| 969 |
-
"step": 315
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 0.6387225548902196,
|
| 973 |
-
"grad_norm": 14.637858980431309,
|
| 974 |
-
"learning_rate": 3.488150546247778e-07,
|
| 975 |
-
"logits/chosen": 0.641592264175415,
|
| 976 |
-
"logits/rejected": 0.6488014459609985,
|
| 977 |
-
"logps/chosen": -1.0486619472503662,
|
| 978 |
-
"logps/rejected": -1.117785930633545,
|
| 979 |
-
"loss": 0.5076,
|
| 980 |
-
"rewards/accuracies": 0.800000011920929,
|
| 981 |
-
"rewards/chosen": -1.3681632280349731,
|
| 982 |
-
"rewards/margins": 0.7144045829772949,
|
| 983 |
-
"rewards/rejected": -2.0825679302215576,
|
| 984 |
-
"step": 320
|
| 985 |
-
},
|
| 986 |
-
{
|
| 987 |
-
"epoch": 0.6487025948103793,
|
| 988 |
-
"grad_norm": 14.611695255214977,
|
| 989 |
-
"learning_rate": 3.3227421512487255e-07,
|
| 990 |
-
"logits/chosen": 0.6020836234092712,
|
| 991 |
-
"logits/rejected": 0.5903235673904419,
|
| 992 |
-
"logps/chosen": -1.0694186687469482,
|
| 993 |
-
"logps/rejected": -1.1450960636138916,
|
| 994 |
-
"loss": 0.4988,
|
| 995 |
-
"rewards/accuracies": 0.75,
|
| 996 |
-
"rewards/chosen": -1.495102047920227,
|
| 997 |
-
"rewards/margins": 0.6063732504844666,
|
| 998 |
-
"rewards/rejected": -2.101475238800049,
|
| 999 |
-
"step": 325
|
| 1000 |
-
},
|
| 1001 |
-
{
|
| 1002 |
-
"epoch": 0.6586826347305389,
|
| 1003 |
-
"grad_norm": 53.83501877641933,
|
| 1004 |
-
"learning_rate": 3.15937723657661e-07,
|
| 1005 |
-
"logits/chosen": 0.5937372446060181,
|
| 1006 |
-
"logits/rejected": 0.5983537435531616,
|
| 1007 |
-
"logps/chosen": -1.0911985635757446,
|
| 1008 |
-
"logps/rejected": -1.1970782279968262,
|
| 1009 |
-
"loss": 0.5292,
|
| 1010 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 1011 |
-
"rewards/chosen": -1.5829055309295654,
|
| 1012 |
-
"rewards/margins": 0.6139482259750366,
|
| 1013 |
-
"rewards/rejected": -2.1968536376953125,
|
| 1014 |
-
"step": 330
|
| 1015 |
-
},
|
| 1016 |
-
{
|
| 1017 |
-
"epoch": 0.6686626746506986,
|
| 1018 |
-
"grad_norm": 22.263382020354737,
|
| 1019 |
-
"learning_rate": 2.9982548372155256e-07,
|
| 1020 |
-
"logits/chosen": 0.5615344047546387,
|
| 1021 |
-
"logits/rejected": 0.5818489789962769,
|
| 1022 |
-
"logps/chosen": -1.0336883068084717,
|
| 1023 |
-
"logps/rejected": -1.1681592464447021,
|
| 1024 |
-
"loss": 0.5036,
|
| 1025 |
-
"rewards/accuracies": 0.824999988079071,
|
| 1026 |
-
"rewards/chosen": -1.2369213104248047,
|
| 1027 |
-
"rewards/margins": 1.0152759552001953,
|
| 1028 |
-
"rewards/rejected": -2.252197265625,
|
| 1029 |
-
"step": 335
|
| 1030 |
-
},
|
| 1031 |
-
{
|
| 1032 |
-
"epoch": 0.6786427145708582,
|
| 1033 |
-
"grad_norm": 28.648845585620847,
|
| 1034 |
-
"learning_rate": 2.8395712559900874e-07,
|
| 1035 |
-
"logits/chosen": 0.5865073204040527,
|
| 1036 |
-
"logits/rejected": 0.5887445211410522,
|
| 1037 |
-
"logps/chosen": -1.2165284156799316,
|
| 1038 |
-
"logps/rejected": -1.285627007484436,
|
| 1039 |
-
"loss": 0.5287,
|
| 1040 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 1041 |
-
"rewards/chosen": -1.876123070716858,
|
| 1042 |
-
"rewards/margins": 0.8445480465888977,
|
| 1043 |
-
"rewards/rejected": -2.7206711769104004,
|
| 1044 |
-
"step": 340
|
| 1045 |
-
},
|
| 1046 |
-
{
|
| 1047 |
-
"epoch": 0.688622754491018,
|
| 1048 |
-
"grad_norm": 27.7967265314739,
|
| 1049 |
-
"learning_rate": 2.683519824400692e-07,
|
| 1050 |
-
"logits/chosen": 0.6274746656417847,
|
| 1051 |
-
"logits/rejected": 0.6499799489974976,
|
| 1052 |
-
"logps/chosen": -1.1229360103607178,
|
| 1053 |
-
"logps/rejected": -1.2630244493484497,
|
| 1054 |
-
"loss": 0.5228,
|
| 1055 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 1056 |
-
"rewards/chosen": -1.6735212802886963,
|
| 1057 |
-
"rewards/margins": 0.8694586753845215,
|
| 1058 |
-
"rewards/rejected": -2.5429797172546387,
|
| 1059 |
-
"step": 345
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 0.6986027944111777,
|
| 1063 |
-
"grad_norm": 22.50466227883414,
|
| 1064 |
-
"learning_rate": 2.530290667078846e-07,
|
| 1065 |
-
"logits/chosen": 0.620844841003418,
|
| 1066 |
-
"logits/rejected": 0.6195460557937622,
|
| 1067 |
-
"logps/chosen": -1.1165497303009033,
|
| 1068 |
-
"logps/rejected": -1.1850183010101318,
|
| 1069 |
-
"loss": 0.5019,
|
| 1070 |
-
"rewards/accuracies": 0.824999988079071,
|
| 1071 |
-
"rewards/chosen": -1.5181663036346436,
|
| 1072 |
-
"rewards/margins": 0.7825000286102295,
|
| 1073 |
-
"rewards/rejected": -2.300666332244873,
|
| 1074 |
-
"step": 350
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.7085828343313373,
|
| 1078 |
-
"grad_norm": 21.26239260918309,
|
| 1079 |
-
"learning_rate": 2.380070470149605e-07,
|
| 1080 |
-
"logits/chosen": 0.6597515940666199,
|
| 1081 |
-
"logits/rejected": 0.6680246591567993,
|
| 1082 |
-
"logps/chosen": -1.1044973134994507,
|
| 1083 |
-
"logps/rejected": -1.2354698181152344,
|
| 1084 |
-
"loss": 0.4914,
|
| 1085 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 1086 |
-
"rewards/chosen": -1.4524997472763062,
|
| 1087 |
-
"rewards/margins": 1.0394423007965088,
|
| 1088 |
-
"rewards/rejected": -2.4919424057006836,
|
| 1089 |
-
"step": 355
|
| 1090 |
-
},
|
| 1091 |
-
{
|
| 1092 |
-
"epoch": 0.718562874251497,
|
| 1093 |
-
"grad_norm": 30.986440285036473,
|
| 1094 |
-
"learning_rate": 2.23304225378328e-07,
|
| 1095 |
-
"logits/chosen": 0.5663326978683472,
|
| 1096 |
-
"logits/rejected": 0.5595775842666626,
|
| 1097 |
-
"logps/chosen": -1.2192121744155884,
|
| 1098 |
-
"logps/rejected": -1.2986266613006592,
|
| 1099 |
-
"loss": 0.4991,
|
| 1100 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 1101 |
-
"rewards/chosen": -1.6858431100845337,
|
| 1102 |
-
"rewards/margins": 1.0494027137756348,
|
| 1103 |
-
"rewards/rejected": -2.735245704650879,
|
| 1104 |
-
"step": 360
|
| 1105 |
-
},
|
| 1106 |
-
{
|
| 1107 |
-
"epoch": 0.7285429141716567,
|
| 1108 |
-
"grad_norm": 31.154865644231116,
|
| 1109 |
-
"learning_rate": 2.0893851492135532e-07,
|
| 1110 |
-
"logits/chosen": 0.5782791376113892,
|
| 1111 |
-
"logits/rejected": 0.5733534097671509,
|
| 1112 |
-
"logps/chosen": -1.108449935913086,
|
| 1113 |
-
"logps/rejected": -1.2320600748062134,
|
| 1114 |
-
"loss": 0.4969,
|
| 1115 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 1116 |
-
"rewards/chosen": -1.4567809104919434,
|
| 1117 |
-
"rewards/margins": 1.0569483041763306,
|
| 1118 |
-
"rewards/rejected": -2.5137288570404053,
|
| 1119 |
-
"step": 365
|
| 1120 |
-
},
|
| 1121 |
-
{
|
| 1122 |
-
"epoch": 0.7385229540918163,
|
| 1123 |
-
"grad_norm": 25.665361579684205,
|
| 1124 |
-
"learning_rate": 1.9492741804936618e-07,
|
| 1125 |
-
"logits/chosen": 0.5590307116508484,
|
| 1126 |
-
"logits/rejected": 0.5289372205734253,
|
| 1127 |
-
"logps/chosen": -1.135460615158081,
|
| 1128 |
-
"logps/rejected": -1.1982572078704834,
|
| 1129 |
-
"loss": 0.5123,
|
| 1130 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 1131 |
-
"rewards/chosen": -1.5751354694366455,
|
| 1132 |
-
"rewards/margins": 0.7708719968795776,
|
| 1133 |
-
"rewards/rejected": -2.346007823944092,
|
| 1134 |
-
"step": 370
|
| 1135 |
-
},
|
| 1136 |
-
{
|
| 1137 |
-
"epoch": 0.7485029940119761,
|
| 1138 |
-
"grad_norm": 13.466456049238714,
|
| 1139 |
-
"learning_rate": 1.812880051256551e-07,
|
| 1140 |
-
"logits/chosen": 0.5437551736831665,
|
| 1141 |
-
"logits/rejected": 0.540005087852478,
|
| 1142 |
-
"logps/chosen": -1.1057159900665283,
|
| 1143 |
-
"logps/rejected": -1.1711170673370361,
|
| 1144 |
-
"loss": 0.4956,
|
| 1145 |
-
"rewards/accuracies": 0.6875,
|
| 1146 |
-
"rewards/chosen": -1.607832670211792,
|
| 1147 |
-
"rewards/margins": 0.5974467396736145,
|
| 1148 |
-
"rewards/rejected": -2.205279588699341,
|
| 1149 |
-
"step": 375
|
| 1150 |
-
},
|
| 1151 |
-
{
|
| 1152 |
-
"epoch": 0.7584830339321357,
|
| 1153 |
-
"grad_norm": 22.830693166932168,
|
| 1154 |
-
"learning_rate": 1.6803689367387918e-07,
|
| 1155 |
-
"logits/chosen": 0.5521084666252136,
|
| 1156 |
-
"logits/rejected": 0.563881516456604,
|
| 1157 |
-
"logps/chosen": -1.0188974142074585,
|
| 1158 |
-
"logps/rejected": -1.0934759378433228,
|
| 1159 |
-
"loss": 0.5364,
|
| 1160 |
-
"rewards/accuracies": 0.6625000238418579,
|
| 1161 |
-
"rewards/chosen": -1.6915432214736938,
|
| 1162 |
-
"rewards/margins": 0.41655927896499634,
|
| 1163 |
-
"rewards/rejected": -2.108102560043335,
|
| 1164 |
-
"step": 380
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 0.7684630738522954,
|
| 1168 |
-
"grad_norm": 16.717527209109356,
|
| 1169 |
-
"learning_rate": 1.551902281321651e-07,
|
| 1170 |
-
"logits/chosen": 0.6741904020309448,
|
| 1171 |
-
"logits/rejected": 0.6558393836021423,
|
| 1172 |
-
"logps/chosen": -1.153136134147644,
|
| 1173 |
-
"logps/rejected": -1.2626793384552002,
|
| 1174 |
-
"loss": 0.4885,
|
| 1175 |
-
"rewards/accuracies": 0.8374999761581421,
|
| 1176 |
-
"rewards/chosen": -1.4042468070983887,
|
| 1177 |
-
"rewards/margins": 1.1682214736938477,
|
| 1178 |
-
"rewards/rejected": -2.5724682807922363,
|
| 1179 |
-
"step": 385
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.7784431137724551,
|
| 1183 |
-
"grad_norm": 21.35977175588003,
|
| 1184 |
-
"learning_rate": 1.4276366018359842e-07,
|
| 1185 |
-
"logits/chosen": 0.5667208433151245,
|
| 1186 |
-
"logits/rejected": 0.5196897387504578,
|
| 1187 |
-
"logps/chosen": -1.1266697645187378,
|
| 1188 |
-
"logps/rejected": -1.2171615362167358,
|
| 1189 |
-
"loss": 0.5164,
|
| 1190 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 1191 |
-
"rewards/chosen": -1.5344607830047607,
|
| 1192 |
-
"rewards/margins": 0.9302822947502136,
|
| 1193 |
-
"rewards/rejected": -2.464743137359619,
|
| 1194 |
-
"step": 390
|
| 1195 |
-
},
|
| 1196 |
-
{
|
| 1197 |
-
"epoch": 0.7884231536926147,
|
| 1198 |
-
"grad_norm": 21.011004494402407,
|
| 1199 |
-
"learning_rate": 1.3077232968705805e-07,
|
| 1200 |
-
"logits/chosen": 0.581541121006012,
|
| 1201 |
-
"logits/rejected": 0.5735315084457397,
|
| 1202 |
-
"logps/chosen": -1.1142101287841797,
|
| 1203 |
-
"logps/rejected": -1.204474687576294,
|
| 1204 |
-
"loss": 0.4938,
|
| 1205 |
-
"rewards/accuracies": 0.8125,
|
| 1206 |
-
"rewards/chosen": -1.4655622243881226,
|
| 1207 |
-
"rewards/margins": 0.7967506647109985,
|
| 1208 |
-
"rewards/rejected": -2.262312650680542,
|
| 1209 |
-
"step": 395
|
| 1210 |
-
},
|
| 1211 |
-
{
|
| 1212 |
-
"epoch": 0.7984031936127745,
|
| 1213 |
-
"grad_norm": 21.73473549415687,
|
| 1214 |
-
"learning_rate": 1.192308462316317e-07,
|
| 1215 |
-
"logits/chosen": 0.49174603819847107,
|
| 1216 |
-
"logits/rejected": 0.5455900430679321,
|
| 1217 |
-
"logps/chosen": -1.1133276224136353,
|
| 1218 |
-
"logps/rejected": -1.2326544523239136,
|
| 1219 |
-
"loss": 0.5169,
|
| 1220 |
-
"rewards/accuracies": 0.862500011920929,
|
| 1221 |
-
"rewards/chosen": -1.5614674091339111,
|
| 1222 |
-
"rewards/margins": 0.7484425902366638,
|
| 1223 |
-
"rewards/rejected": -2.309910297393799,
|
| 1224 |
-
"step": 400
|
| 1225 |
-
},
|
| 1226 |
-
{
|
| 1227 |
-
"epoch": 0.8083832335329342,
|
| 1228 |
-
"grad_norm": 18.213898305037567,
|
| 1229 |
-
"learning_rate": 1.0815327133708013e-07,
|
| 1230 |
-
"logits/chosen": 0.5780737400054932,
|
| 1231 |
-
"logits/rejected": 0.6099402904510498,
|
| 1232 |
-
"logps/chosen": -1.0353389978408813,
|
| 1233 |
-
"logps/rejected": -1.1462222337722778,
|
| 1234 |
-
"loss": 0.5117,
|
| 1235 |
-
"rewards/accuracies": 0.800000011920929,
|
| 1236 |
-
"rewards/chosen": -1.678412675857544,
|
| 1237 |
-
"rewards/margins": 0.7521916031837463,
|
| 1238 |
-
"rewards/rejected": -2.4306042194366455,
|
| 1239 |
-
"step": 405
|
| 1240 |
-
},
|
| 1241 |
-
{
|
| 1242 |
-
"epoch": 0.8183632734530938,
|
| 1243 |
-
"grad_norm": 14.585406444054382,
|
| 1244 |
-
"learning_rate": 9.755310132204297e-08,
|
| 1245 |
-
"logits/chosen": 0.5654971599578857,
|
| 1246 |
-
"logits/rejected": 0.5592461824417114,
|
| 1247 |
-
"logps/chosen": -1.0622889995574951,
|
| 1248 |
-
"logps/rejected": -1.2071908712387085,
|
| 1249 |
-
"loss": 0.5324,
|
| 1250 |
-
"rewards/accuracies": 0.75,
|
| 1251 |
-
"rewards/chosen": -1.5269814729690552,
|
| 1252 |
-
"rewards/margins": 1.1922402381896973,
|
| 1253 |
-
"rewards/rejected": -2.719221591949463,
|
| 1254 |
-
"step": 410
|
| 1255 |
-
},
|
| 1256 |
-
{
|
| 1257 |
-
"epoch": 0.8283433133732535,
|
| 1258 |
-
"grad_norm": 23.842758561266724,
|
| 1259 |
-
"learning_rate": 8.744325086085247e-08,
|
| 1260 |
-
"logits/chosen": 0.5762808322906494,
|
| 1261 |
-
"logits/rejected": 0.5668230652809143,
|
| 1262 |
-
"logps/chosen": -1.1735143661499023,
|
| 1263 |
-
"logps/rejected": -1.2483824491500854,
|
| 1264 |
-
"loss": 0.504,
|
| 1265 |
-
"rewards/accuracies": 0.800000011920929,
|
| 1266 |
-
"rewards/chosen": -1.7859411239624023,
|
| 1267 |
-
"rewards/margins": 0.8882120847702026,
|
| 1268 |
-
"rewards/rejected": -2.6741533279418945,
|
| 1269 |
-
"step": 415
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.8383233532934131,
|
| 1273 |
-
"grad_norm": 41.62435891120011,
|
| 1274 |
-
"learning_rate": 7.783603724899257e-08,
|
| 1275 |
-
"logits/chosen": 0.5172094106674194,
|
| 1276 |
-
"logits/rejected": 0.5516868829727173,
|
| 1277 |
-
"logps/chosen": -1.145578145980835,
|
| 1278 |
-
"logps/rejected": -1.248124599456787,
|
| 1279 |
-
"loss": 0.5313,
|
| 1280 |
-
"rewards/accuracies": 0.737500011920929,
|
| 1281 |
-
"rewards/chosen": -1.742923378944397,
|
| 1282 |
-
"rewards/margins": 0.8347417712211609,
|
| 1283 |
-
"rewards/rejected": -2.577665090560913,
|
| 1284 |
-
"step": 420
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 0.8483033932135728,
|
| 1288 |
-
"grad_norm": 26.49206528554136,
|
| 1289 |
-
"learning_rate": 6.874316539637126e-08,
|
| 1290 |
-
"logits/chosen": 0.5917887091636658,
|
| 1291 |
-
"logits/rejected": 0.6128644943237305,
|
| 1292 |
-
"logps/chosen": -1.0141854286193848,
|
| 1293 |
-
"logps/rejected": -1.1528804302215576,
|
| 1294 |
-
"loss": 0.4854,
|
| 1295 |
-
"rewards/accuracies": 0.824999988079071,
|
| 1296 |
-
"rewards/chosen": -1.2433063983917236,
|
| 1297 |
-
"rewards/margins": 0.826849639415741,
|
| 1298 |
-
"rewards/rejected": -2.0701558589935303,
|
| 1299 |
-
"step": 425
|
| 1300 |
-
},
|
| 1301 |
-
{
|
| 1302 |
-
"epoch": 0.8582834331337326,
|
| 1303 |
-
"grad_norm": 14.052261790037898,
|
| 1304 |
-
"learning_rate": 6.017571356669182e-08,
|
| 1305 |
-
"logits/chosen": 0.5848813652992249,
|
| 1306 |
-
"logits/rejected": 0.5729445219039917,
|
| 1307 |
-
"logps/chosen": -1.1584407091140747,
|
| 1308 |
-
"logps/rejected": -1.304329752922058,
|
| 1309 |
-
"loss": 0.5097,
|
| 1310 |
-
"rewards/accuracies": 0.7749999761581421,
|
| 1311 |
-
"rewards/chosen": -1.7920862436294556,
|
| 1312 |
-
"rewards/margins": 0.670229971408844,
|
| 1313 |
-
"rewards/rejected": -2.4623162746429443,
|
| 1314 |
-
"step": 430
|
| 1315 |
-
},
|
| 1316 |
-
{
|
| 1317 |
-
"epoch": 0.8682634730538922,
|
| 1318 |
-
"grad_norm": 18.538952507210343,
|
| 1319 |
-
"learning_rate": 5.2144119880293544e-08,
|
| 1320 |
-
"logits/chosen": 0.5895706415176392,
|
| 1321 |
-
"logits/rejected": 0.5678045749664307,
|
| 1322 |
-
"logps/chosen": -1.075687050819397,
|
| 1323 |
-
"logps/rejected": -1.2755447626113892,
|
| 1324 |
-
"loss": 0.5187,
|
| 1325 |
-
"rewards/accuracies": 0.824999988079071,
|
| 1326 |
-
"rewards/chosen": -1.3742207288742065,
|
| 1327 |
-
"rewards/margins": 1.1585171222686768,
|
| 1328 |
-
"rewards/rejected": -2.532737970352173,
|
| 1329 |
-
"step": 435
|
| 1330 |
-
},
|
| 1331 |
-
{
|
| 1332 |
-
"epoch": 0.8782435129740519,
|
| 1333 |
-
"grad_norm": 17.852951128940067,
|
| 1334 |
-
"learning_rate": 4.465816959691149e-08,
|
| 1335 |
-
"logits/chosen": 0.5652838945388794,
|
| 1336 |
-
"logits/rejected": 0.6052907705307007,
|
| 1337 |
-
"logps/chosen": -1.2037353515625,
|
| 1338 |
-
"logps/rejected": -1.3117984533309937,
|
| 1339 |
-
"loss": 0.5009,
|
| 1340 |
-
"rewards/accuracies": 0.887499988079071,
|
| 1341 |
-
"rewards/chosen": -1.7123981714248657,
|
| 1342 |
-
"rewards/margins": 1.019058108329773,
|
| 1343 |
-
"rewards/rejected": -2.7314560413360596,
|
| 1344 |
-
"step": 440
|
| 1345 |
-
},
|
| 1346 |
-
{
|
| 1347 |
-
"epoch": 0.8882235528942116,
|
| 1348 |
-
"grad_norm": 14.249216111089481,
|
| 1349 |
-
"learning_rate": 3.7726983193843485e-08,
|
| 1350 |
-
"logits/chosen": 0.517819344997406,
|
| 1351 |
-
"logits/rejected": 0.5205559730529785,
|
| 1352 |
-
"logps/chosen": -1.0231516361236572,
|
| 1353 |
-
"logps/rejected": -1.148457646369934,
|
| 1354 |
-
"loss": 0.5199,
|
| 1355 |
-
"rewards/accuracies": 0.699999988079071,
|
| 1356 |
-
"rewards/chosen": -1.4192359447479248,
|
| 1357 |
-
"rewards/margins": 0.889240562915802,
|
| 1358 |
-
"rewards/rejected": -2.308476448059082,
|
| 1359 |
-
"step": 445
|
| 1360 |
-
},
|
| 1361 |
-
{
|
| 1362 |
-
"epoch": 0.8982035928143712,
|
| 1363 |
-
"grad_norm": 17.926055468482602,
|
| 1364 |
-
"learning_rate": 3.135900525405427e-08,
|
| 1365 |
-
"logits/chosen": 0.5483254194259644,
|
| 1366 |
-
"logits/rejected": 0.5569725036621094,
|
| 1367 |
-
"logps/chosen": -1.160649299621582,
|
| 1368 |
-
"logps/rejected": -1.2841194868087769,
|
| 1369 |
-
"loss": 0.5171,
|
| 1370 |
-
"rewards/accuracies": 0.762499988079071,
|
| 1371 |
-
"rewards/chosen": -1.7122529745101929,
|
| 1372 |
-
"rewards/margins": 0.9245160222053528,
|
| 1373 |
-
"rewards/rejected": -2.6367690563201904,
|
| 1374 |
-
"step": 450
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 0.908183632734531,
|
| 1378 |
-
"grad_norm": 15.183540302368447,
|
| 1379 |
-
"learning_rate": 2.5561994177751732e-08,
|
| 1380 |
-
"logits/chosen": 0.6299312114715576,
|
| 1381 |
-
"logits/rejected": 0.6476989984512329,
|
| 1382 |
-
"logps/chosen": -1.0939618349075317,
|
| 1383 |
-
"logps/rejected": -1.187743902206421,
|
| 1384 |
-
"loss": 0.4643,
|
| 1385 |
-
"rewards/accuracies": 0.75,
|
| 1386 |
-
"rewards/chosen": -1.5630865097045898,
|
| 1387 |
-
"rewards/margins": 0.8125492334365845,
|
| 1388 |
-
"rewards/rejected": -2.3756356239318848,
|
| 1389 |
-
"step": 455
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 0.9181636726546906,
|
| 1393 |
-
"grad_norm": 15.769731536967763,
|
| 1394 |
-
"learning_rate": 2.0343012729971243e-08,
|
| 1395 |
-
"logits/chosen": 0.5560911893844604,
|
| 1396 |
-
"logits/rejected": 0.563130259513855,
|
| 1397 |
-
"logps/chosen": -1.1199856996536255,
|
| 1398 |
-
"logps/rejected": -1.2025185823440552,
|
| 1399 |
-
"loss": 0.4948,
|
| 1400 |
-
"rewards/accuracies": 0.7250000238418579,
|
| 1401 |
-
"rewards/chosen": -1.6438310146331787,
|
| 1402 |
-
"rewards/margins": 0.8590097427368164,
|
| 1403 |
-
"rewards/rejected": -2.502840518951416,
|
| 1404 |
-
"step": 460
|
| 1405 |
-
},
|
| 1406 |
-
{
|
| 1407 |
-
"epoch": 0.9281437125748503,
|
| 1408 |
-
"grad_norm": 21.7191545301948,
|
| 1409 |
-
"learning_rate": 1.570841943568446e-08,
|
| 1410 |
-
"logits/chosen": 0.6068538427352905,
|
| 1411 |
-
"logits/rejected": 0.6409121751785278,
|
| 1412 |
-
"logps/chosen": -1.211014747619629,
|
| 1413 |
-
"logps/rejected": -1.3489148616790771,
|
| 1414 |
-
"loss": 0.4823,
|
| 1415 |
-
"rewards/accuracies": 0.8999999761581421,
|
| 1416 |
-
"rewards/chosen": -1.6730806827545166,
|
| 1417 |
-
"rewards/margins": 1.099481463432312,
|
| 1418 |
-
"rewards/rejected": -2.772562026977539,
|
| 1419 |
-
"step": 465
|
| 1420 |
-
},
|
| 1421 |
-
{
|
| 1422 |
-
"epoch": 0.93812375249501,
|
| 1423 |
-
"grad_norm": 23.87949640105234,
|
| 1424 |
-
"learning_rate": 1.166386083291604e-08,
|
| 1425 |
-
"logits/chosen": 0.5642676949501038,
|
| 1426 |
-
"logits/rejected": 0.5937134027481079,
|
| 1427 |
-
"logps/chosen": -1.0865347385406494,
|
| 1428 |
-
"logps/rejected": -1.2261168956756592,
|
| 1429 |
-
"loss": 0.5012,
|
| 1430 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 1431 |
-
"rewards/chosen": -1.5355206727981567,
|
| 1432 |
-
"rewards/margins": 0.9762239456176758,
|
| 1433 |
-
"rewards/rejected": -2.511744260787964,
|
| 1434 |
-
"step": 470
|
| 1435 |
-
},
|
| 1436 |
-
{
|
| 1437 |
-
"epoch": 0.9481037924151696,
|
| 1438 |
-
"grad_norm": 11.821044673852356,
|
| 1439 |
-
"learning_rate": 8.214264593307096e-09,
|
| 1440 |
-
"logits/chosen": 0.5303646922111511,
|
| 1441 |
-
"logits/rejected": 0.550706684589386,
|
| 1442 |
-
"logps/chosen": -1.212477207183838,
|
| 1443 |
-
"logps/rejected": -1.27964186668396,
|
| 1444 |
-
"loss": 0.4881,
|
| 1445 |
-
"rewards/accuracies": 0.762499988079071,
|
| 1446 |
-
"rewards/chosen": -1.8850059509277344,
|
| 1447 |
-
"rewards/margins": 0.7182316780090332,
|
| 1448 |
-
"rewards/rejected": -2.603238105773926,
|
| 1449 |
-
"step": 475
|
| 1450 |
-
},
|
| 1451 |
-
{
|
| 1452 |
-
"epoch": 0.9580838323353293,
|
| 1453 |
-
"grad_norm": 18.686466292842862,
|
| 1454 |
-
"learning_rate": 5.3638335185058335e-09,
|
| 1455 |
-
"logits/chosen": 0.5736523866653442,
|
| 1456 |
-
"logits/rejected": 0.625065267086029,
|
| 1457 |
-
"logps/chosen": -1.2084397077560425,
|
| 1458 |
-
"logps/rejected": -1.3002406358718872,
|
| 1459 |
-
"loss": 0.5266,
|
| 1460 |
-
"rewards/accuracies": 0.8125,
|
| 1461 |
-
"rewards/chosen": -1.7023632526397705,
|
| 1462 |
-
"rewards/margins": 0.9809869527816772,
|
| 1463 |
-
"rewards/rejected": -2.6833503246307373,
|
| 1464 |
-
"step": 480
|
| 1465 |
-
},
|
| 1466 |
-
{
|
| 1467 |
-
"epoch": 0.9680638722554891,
|
| 1468 |
-
"grad_norm": 26.284460492309865,
|
| 1469 |
-
"learning_rate": 3.116040419701815e-09,
|
| 1470 |
-
"logits/chosen": 0.5889401435852051,
|
| 1471 |
-
"logits/rejected": 0.5715131759643555,
|
| 1472 |
-
"logps/chosen": -1.1137585639953613,
|
| 1473 |
-
"logps/rejected": -1.2124450206756592,
|
| 1474 |
-
"loss": 0.5171,
|
| 1475 |
-
"rewards/accuracies": 0.7875000238418579,
|
| 1476 |
-
"rewards/chosen": -1.5402872562408447,
|
| 1477 |
-
"rewards/margins": 0.7413524389266968,
|
| 1478 |
-
"rewards/rejected": -2.281639575958252,
|
| 1479 |
-
"step": 485
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.9780439121756487,
|
| 1483 |
-
"grad_norm": 30.03419204450396,
|
| 1484 |
-
"learning_rate": 1.4736238865398765e-09,
|
| 1485 |
-
"logits/chosen": 0.6326584815979004,
|
| 1486 |
-
"logits/rejected": 0.6649195551872253,
|
| 1487 |
-
"logps/chosen": -1.1648736000061035,
|
| 1488 |
-
"logps/rejected": -1.2586228847503662,
|
| 1489 |
-
"loss": 0.519,
|
| 1490 |
-
"rewards/accuracies": 0.762499988079071,
|
| 1491 |
-
"rewards/chosen": -1.403749942779541,
|
| 1492 |
-
"rewards/margins": 1.0453355312347412,
|
| 1493 |
-
"rewards/rejected": -2.4490854740142822,
|
| 1494 |
-
"step": 490
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 0.9880239520958084,
|
| 1498 |
-
"grad_norm": 33.25313115299262,
|
| 1499 |
-
"learning_rate": 4.3858495057080836e-10,
|
| 1500 |
-
"logits/chosen": 0.5985925197601318,
|
| 1501 |
-
"logits/rejected": 0.5628663301467896,
|
| 1502 |
-
"logps/chosen": -1.2969096899032593,
|
| 1503 |
-
"logps/rejected": -1.4230639934539795,
|
| 1504 |
-
"loss": 0.4764,
|
| 1505 |
-
"rewards/accuracies": 0.8125,
|
| 1506 |
-
"rewards/chosen": -1.819708228111267,
|
| 1507 |
-
"rewards/margins": 1.287508487701416,
|
| 1508 |
-
"rewards/rejected": -3.1072165966033936,
|
| 1509 |
-
"step": 495
|
| 1510 |
-
},
|
| 1511 |
-
{
|
| 1512 |
-
"epoch": 0.998003992015968,
|
| 1513 |
-
"grad_norm": 30.55644687750284,
|
| 1514 |
-
"learning_rate": 1.2184647302626582e-11,
|
| 1515 |
-
"logits/chosen": 0.5701574087142944,
|
| 1516 |
-
"logits/rejected": 0.5943929553031921,
|
| 1517 |
-
"logps/chosen": -1.1318881511688232,
|
| 1518 |
-
"logps/rejected": -1.2142232656478882,
|
| 1519 |
-
"loss": 0.4976,
|
| 1520 |
-
"rewards/accuracies": 0.762499988079071,
|
| 1521 |
-
"rewards/chosen": -1.6143200397491455,
|
| 1522 |
-
"rewards/margins": 0.6937130093574524,
|
| 1523 |
-
"rewards/rejected": -2.3080332279205322,
|
| 1524 |
-
"step": 500
|
| 1525 |
-
},
|
| 1526 |
-
{
|
| 1527 |
-
"epoch": 1.0,
|
| 1528 |
-
"step": 501,
|
| 1529 |
-
"total_flos": 0.0,
|
| 1530 |
-
"train_loss": 0.559710203114146,
|
| 1531 |
-
"train_runtime": 2572.2381,
|
| 1532 |
-
"train_samples_per_second": 24.931,
|
| 1533 |
-
"train_steps_per_second": 0.195
|
| 1534 |
-
}
|
| 1535 |
-
],
|
| 1536 |
-
"logging_steps": 5,
|
| 1537 |
-
"max_steps": 501,
|
| 1538 |
-
"num_input_tokens_seen": 0,
|
| 1539 |
-
"num_train_epochs": 1,
|
| 1540 |
-
"save_steps": 101,
|
| 1541 |
-
"stateful_callbacks": {
|
| 1542 |
-
"TrainerControl": {
|
| 1543 |
-
"args": {
|
| 1544 |
-
"should_epoch_stop": false,
|
| 1545 |
-
"should_evaluate": false,
|
| 1546 |
-
"should_log": false,
|
| 1547 |
-
"should_save": true,
|
| 1548 |
-
"should_training_stop": true
|
| 1549 |
-
},
|
| 1550 |
-
"attributes": {}
|
| 1551 |
-
}
|
| 1552 |
-
},
|
| 1553 |
-
"total_flos": 0.0,
|
| 1554 |
-
"train_batch_size": 4,
|
| 1555 |
-
"trial_name": null,
|
| 1556 |
-
"trial_params": null
|
| 1557 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|