File size: 12,487 Bytes
39714c6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 |
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8316008316008316,
"eval_steps": 50,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04158004158004158,
"grad_norm": 57.63412014296559,
"learning_rate": 5e-07,
"logits/chosen": -2.7145180702209473,
"logits/rejected": -2.7382097244262695,
"logps/chosen": -265.24481201171875,
"logps/rejected": -201.1150665283203,
"loss": 0.6915,
"rewards/accuracies": 0.375,
"rewards/chosen": 0.01524554193019867,
"rewards/margins": 0.008573563769459724,
"rewards/rejected": 0.006671978626400232,
"step": 5
},
{
"epoch": 0.08316008316008316,
"grad_norm": 52.91443405505758,
"learning_rate": 1e-06,
"logits/chosen": -2.690777063369751,
"logits/rejected": -2.689495325088501,
"logps/chosen": -259.32598876953125,
"logps/rejected": -236.3219757080078,
"loss": 0.6616,
"rewards/accuracies": 0.6312500238418579,
"rewards/chosen": 0.4881957471370697,
"rewards/margins": 0.10135326534509659,
"rewards/rejected": 0.3868425786495209,
"step": 10
},
{
"epoch": 0.12474012474012475,
"grad_norm": 53.28329333963272,
"learning_rate": 9.994965332706572e-07,
"logits/chosen": -2.5831141471862793,
"logits/rejected": -2.580876350402832,
"logps/chosen": -237.8153839111328,
"logps/rejected": -192.7242889404297,
"loss": 0.6359,
"rewards/accuracies": 0.6875,
"rewards/chosen": 1.1336857080459595,
"rewards/margins": 0.47578057646751404,
"rewards/rejected": 0.6579049825668335,
"step": 15
},
{
"epoch": 0.16632016632016633,
"grad_norm": 59.51771261732688,
"learning_rate": 9.979871469976195e-07,
"logits/chosen": -2.5097970962524414,
"logits/rejected": -2.491774797439575,
"logps/chosen": -241.24349975585938,
"logps/rejected": -206.7909698486328,
"loss": 0.6385,
"rewards/accuracies": 0.6875,
"rewards/chosen": 1.5591249465942383,
"rewards/margins": 0.980699896812439,
"rewards/rejected": 0.5784249305725098,
"step": 20
},
{
"epoch": 0.2079002079002079,
"grad_norm": 43.932979300169734,
"learning_rate": 9.954748808839674e-07,
"logits/chosen": -2.428896427154541,
"logits/rejected": -2.426640510559082,
"logps/chosen": -260.3008117675781,
"logps/rejected": -223.4193878173828,
"loss": 0.6287,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": 1.4255732297897339,
"rewards/margins": 0.9219636917114258,
"rewards/rejected": 0.5036097764968872,
"step": 25
},
{
"epoch": 0.2494802494802495,
"grad_norm": 51.9520224365334,
"learning_rate": 9.919647942993147e-07,
"logits/chosen": -2.270231008529663,
"logits/rejected": -2.2373533248901367,
"logps/chosen": -239.1925811767578,
"logps/rejected": -199.03244018554688,
"loss": 0.6013,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 1.0962295532226562,
"rewards/margins": 0.8090855479240417,
"rewards/rejected": 0.28714415431022644,
"step": 30
},
{
"epoch": 0.2910602910602911,
"grad_norm": 42.07074967218912,
"learning_rate": 9.874639560909118e-07,
"logits/chosen": -2.2837703227996826,
"logits/rejected": -2.2773120403289795,
"logps/chosen": -245.8435516357422,
"logps/rejected": -218.3318328857422,
"loss": 0.5782,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": 0.842710018157959,
"rewards/margins": 0.7606425285339355,
"rewards/rejected": 0.08206750452518463,
"step": 35
},
{
"epoch": 0.33264033264033266,
"grad_norm": 40.52902304303911,
"learning_rate": 9.819814303479267e-07,
"logits/chosen": -2.3677680492401123,
"logits/rejected": -2.3257031440734863,
"logps/chosen": -255.2162322998047,
"logps/rejected": -206.1927490234375,
"loss": 0.5732,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": 0.774177610874176,
"rewards/margins": 0.9143842458724976,
"rewards/rejected": -0.14020657539367676,
"step": 40
},
{
"epoch": 0.37422037422037424,
"grad_norm": 42.34370895122762,
"learning_rate": 9.755282581475767e-07,
"logits/chosen": -2.39888596534729,
"logits/rejected": -2.3836829662323,
"logps/chosen": -237.30508422851562,
"logps/rejected": -216.06588745117188,
"loss": 0.5958,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.7707290649414062,
"rewards/margins": 1.1094779968261719,
"rewards/rejected": -0.3387489914894104,
"step": 45
},
{
"epoch": 0.4158004158004158,
"grad_norm": 46.42422960165206,
"learning_rate": 9.681174353198686e-07,
"logits/chosen": -2.3984861373901367,
"logits/rejected": -2.3900020122528076,
"logps/chosen": -233.5660400390625,
"logps/rejected": -216.6641082763672,
"loss": 0.635,
"rewards/accuracies": 0.7437499761581421,
"rewards/chosen": 0.9606113433837891,
"rewards/margins": 1.124942660331726,
"rewards/rejected": -0.16433146595954895,
"step": 50
},
{
"epoch": 0.4158004158004158,
"eval_logits/chosen": -2.3899078369140625,
"eval_logits/rejected": -2.3907740116119385,
"eval_logps/chosen": -237.2779541015625,
"eval_logps/rejected": -211.79153442382812,
"eval_loss": 0.598716676235199,
"eval_rewards/accuracies": 0.7013888955116272,
"eval_rewards/chosen": 0.8451206684112549,
"eval_rewards/margins": 0.8839507102966309,
"eval_rewards/rejected": -0.038830023258924484,
"eval_runtime": 229.0134,
"eval_samples_per_second": 14.934,
"eval_steps_per_second": 0.236,
"step": 50
},
{
"epoch": 0.4573804573804574,
"grad_norm": 38.39211344385862,
"learning_rate": 9.597638862757253e-07,
"logits/chosen": -2.456921339035034,
"logits/rejected": -2.4394099712371826,
"logps/chosen": -237.6243133544922,
"logps/rejected": -233.6071014404297,
"loss": 0.6009,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": 0.8705752491950989,
"rewards/margins": 1.126349687576294,
"rewards/rejected": -0.2557744085788727,
"step": 55
},
{
"epoch": 0.498960498960499,
"grad_norm": 37.3024427630972,
"learning_rate": 9.504844339512094e-07,
"logits/chosen": -2.521387815475464,
"logits/rejected": -2.4951186180114746,
"logps/chosen": -266.1732177734375,
"logps/rejected": -242.85897827148438,
"loss": 0.554,
"rewards/accuracies": 0.7437499761581421,
"rewards/chosen": 0.9916666150093079,
"rewards/margins": 1.3694941997528076,
"rewards/rejected": -0.3778277039527893,
"step": 60
},
{
"epoch": 0.5405405405405406,
"grad_norm": 34.5258394053866,
"learning_rate": 9.402977659283689e-07,
"logits/chosen": -2.4310925006866455,
"logits/rejected": -2.4042844772338867,
"logps/chosen": -223.9085693359375,
"logps/rejected": -203.2917022705078,
"loss": 0.5364,
"rewards/accuracies": 0.6625000238418579,
"rewards/chosen": 0.5567647814750671,
"rewards/margins": 0.8208805918693542,
"rewards/rejected": -0.26411566138267517,
"step": 65
},
{
"epoch": 0.5821205821205822,
"grad_norm": 37.746071231755344,
"learning_rate": 9.29224396800933e-07,
"logits/chosen": -2.3321871757507324,
"logits/rejected": -2.2954535484313965,
"logps/chosen": -243.26156616210938,
"logps/rejected": -226.8199005126953,
"loss": 0.5423,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.38314875960350037,
"rewards/margins": 1.1013362407684326,
"rewards/rejected": -0.7181875109672546,
"step": 70
},
{
"epoch": 0.6237006237006237,
"grad_norm": 57.07804647920337,
"learning_rate": 9.172866268606513e-07,
"logits/chosen": -2.26948881149292,
"logits/rejected": -2.2403252124786377,
"logps/chosen": -251.2090606689453,
"logps/rejected": -240.20169067382812,
"loss": 0.5844,
"rewards/accuracies": 0.7562500238418579,
"rewards/chosen": 0.9941379427909851,
"rewards/margins": 1.1448538303375244,
"rewards/rejected": -0.1507158726453781,
"step": 75
},
{
"epoch": 0.6652806652806653,
"grad_norm": 36.48822907311496,
"learning_rate": 9.045084971874737e-07,
"logits/chosen": -2.212411403656006,
"logits/rejected": -2.1935043334960938,
"logps/chosen": -272.8021545410156,
"logps/rejected": -241.2564239501953,
"loss": 0.5453,
"rewards/accuracies": 0.78125,
"rewards/chosen": 1.1616096496582031,
"rewards/margins": 1.850441336631775,
"rewards/rejected": -0.6888317465782166,
"step": 80
},
{
"epoch": 0.7068607068607069,
"grad_norm": 38.5389032336389,
"learning_rate": 8.909157412340149e-07,
"logits/chosen": -2.180285930633545,
"logits/rejected": -2.126051902770996,
"logps/chosen": -253.0643310546875,
"logps/rejected": -237.400390625,
"loss": 0.5269,
"rewards/accuracies": 0.75,
"rewards/chosen": 0.9046308398246765,
"rewards/margins": 1.5041005611419678,
"rewards/rejected": -0.599469780921936,
"step": 85
},
{
"epoch": 0.7484407484407485,
"grad_norm": 31.08445554774815,
"learning_rate": 8.765357330018055e-07,
"logits/chosen": -2.146207332611084,
"logits/rejected": -2.1198434829711914,
"logps/chosen": -246.0504608154297,
"logps/rejected": -242.79714965820312,
"loss": 0.5741,
"rewards/accuracies": 0.7749999761581421,
"rewards/chosen": 0.4694552421569824,
"rewards/margins": 1.537427544593811,
"rewards/rejected": -1.0679724216461182,
"step": 90
},
{
"epoch": 0.7900207900207901,
"grad_norm": 32.25150255167185,
"learning_rate": 8.613974319136957e-07,
"logits/chosen": -2.114973306655884,
"logits/rejected": -2.0754525661468506,
"logps/chosen": -249.64132690429688,
"logps/rejected": -233.89352416992188,
"loss": 0.4787,
"rewards/accuracies": 0.7437499761581421,
"rewards/chosen": -0.09324260801076889,
"rewards/margins": 1.446690320968628,
"rewards/rejected": -1.5399329662322998,
"step": 95
},
{
"epoch": 0.8316008316008316,
"grad_norm": 34.22313791470385,
"learning_rate": 8.455313244934324e-07,
"logits/chosen": -2.027057409286499,
"logits/rejected": -1.9819841384887695,
"logps/chosen": -250.86947631835938,
"logps/rejected": -241.2573699951172,
"loss": 0.4933,
"rewards/accuracies": 0.8187500238418579,
"rewards/chosen": 0.10764148086309433,
"rewards/margins": 1.9850435256958008,
"rewards/rejected": -1.8774023056030273,
"step": 100
},
{
"epoch": 0.8316008316008316,
"eval_logits/chosen": -1.9530320167541504,
"eval_logits/rejected": -1.9127507209777832,
"eval_logps/chosen": -247.99232482910156,
"eval_logps/rejected": -229.55447387695312,
"eval_loss": 0.5284922122955322,
"eval_rewards/accuracies": 0.7523148059844971,
"eval_rewards/chosen": -0.22631923854351044,
"eval_rewards/margins": 1.5888041257858276,
"eval_rewards/rejected": -1.815123438835144,
"eval_runtime": 227.9451,
"eval_samples_per_second": 15.004,
"eval_steps_per_second": 0.237,
"step": 100
}
],
"logging_steps": 5,
"max_steps": 360,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1178822762299392.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}
|