File size: 12,508 Bytes
663dbab |
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": 1.4925373134328357,
"eval_steps": 50,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07462686567164178,
"grad_norm": 56.38858835713434,
"learning_rate": 5e-07,
"logits/chosen": -2.731353282928467,
"logits/rejected": -2.7158660888671875,
"logps/chosen": -256.7725524902344,
"logps/rejected": -205.1663055419922,
"loss": 0.6903,
"rewards/accuracies": 0.35624998807907104,
"rewards/chosen": 0.014972714707255363,
"rewards/margins": 0.006492167711257935,
"rewards/rejected": 0.008480546995997429,
"step": 5
},
{
"epoch": 0.14925373134328357,
"grad_norm": 48.210654865798084,
"learning_rate": 1e-06,
"logits/chosen": -2.6910791397094727,
"logits/rejected": -2.686784267425537,
"logps/chosen": -261.28240966796875,
"logps/rejected": -210.28414916992188,
"loss": 0.6445,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 0.5361045002937317,
"rewards/margins": 0.2673659920692444,
"rewards/rejected": 0.2687385082244873,
"step": 10
},
{
"epoch": 0.22388059701492538,
"grad_norm": 48.18867539560807,
"learning_rate": 9.983100718730718e-07,
"logits/chosen": -2.5146777629852295,
"logits/rejected": -2.5069146156311035,
"logps/chosen": -274.51788330078125,
"logps/rejected": -212.06613159179688,
"loss": 0.6397,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": 1.0824434757232666,
"rewards/margins": 0.4525600075721741,
"rewards/rejected": 0.6298834085464478,
"step": 15
},
{
"epoch": 0.29850746268656714,
"grad_norm": 56.03620889669535,
"learning_rate": 9.932517109205849e-07,
"logits/chosen": -2.346874237060547,
"logits/rejected": -2.3457634449005127,
"logps/chosen": -233.57192993164062,
"logps/rejected": -216.9462890625,
"loss": 0.6508,
"rewards/accuracies": 0.675000011920929,
"rewards/chosen": 1.2721847295761108,
"rewards/margins": 0.8419507741928101,
"rewards/rejected": 0.4302339553833008,
"step": 20
},
{
"epoch": 0.373134328358209,
"grad_norm": 44.42398594582461,
"learning_rate": 9.848591102083375e-07,
"logits/chosen": -2.321877956390381,
"logits/rejected": -2.3033697605133057,
"logps/chosen": -276.81353759765625,
"logps/rejected": -215.10543823242188,
"loss": 0.5836,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": 1.4672718048095703,
"rewards/margins": 0.9575842618942261,
"rewards/rejected": 0.5096874237060547,
"step": 25
},
{
"epoch": 0.44776119402985076,
"grad_norm": 47.04899305167527,
"learning_rate": 9.731890013043367e-07,
"logits/chosen": -2.3766517639160156,
"logits/rejected": -2.375422954559326,
"logps/chosen": -253.60995483398438,
"logps/rejected": -239.47940063476562,
"loss": 0.634,
"rewards/accuracies": 0.6312500238418579,
"rewards/chosen": 1.2893927097320557,
"rewards/margins": 0.6847591996192932,
"rewards/rejected": 0.6046335697174072,
"step": 30
},
{
"epoch": 0.5223880597014925,
"grad_norm": 39.80811236547202,
"learning_rate": 9.583202707897073e-07,
"logits/chosen": -2.4474780559539795,
"logits/rejected": -2.450472354888916,
"logps/chosen": -257.7427673339844,
"logps/rejected": -207.629638671875,
"loss": 0.5649,
"rewards/accuracies": 0.71875,
"rewards/chosen": 1.217116355895996,
"rewards/margins": 1.1533949375152588,
"rewards/rejected": 0.06372135877609253,
"step": 35
},
{
"epoch": 0.5970149253731343,
"grad_norm": 42.3182391334653,
"learning_rate": 9.403534270080829e-07,
"logits/chosen": -2.4868321418762207,
"logits/rejected": -2.4954237937927246,
"logps/chosen": -243.6182403564453,
"logps/rejected": -235.5696563720703,
"loss": 0.5926,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": 1.1232950687408447,
"rewards/margins": 1.261273741722107,
"rewards/rejected": -0.13797876238822937,
"step": 40
},
{
"epoch": 0.6716417910447762,
"grad_norm": 38.544673892587376,
"learning_rate": 9.19409920658098e-07,
"logits/chosen": -2.4237208366394043,
"logits/rejected": -2.400038480758667,
"logps/chosen": -247.0177764892578,
"logps/rejected": -214.0535888671875,
"loss": 0.6227,
"rewards/accuracies": 0.762499988079071,
"rewards/chosen": 1.1583524942398071,
"rewards/margins": 1.3776183128356934,
"rewards/rejected": -0.21926572918891907,
"step": 45
},
{
"epoch": 0.746268656716418,
"grad_norm": 44.95262854721761,
"learning_rate": 8.956313238215823e-07,
"logits/chosen": -2.3861770629882812,
"logits/rejected": -2.361314058303833,
"logps/chosen": -251.87680053710938,
"logps/rejected": -241.48794555664062,
"loss": 0.5946,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": 1.3533507585525513,
"rewards/margins": 1.143169641494751,
"rewards/rejected": 0.2101811170578003,
"step": 50
},
{
"epoch": 0.746268656716418,
"eval_logits/chosen": -2.419147253036499,
"eval_logits/rejected": -2.3880856037139893,
"eval_logps/chosen": -247.12657165527344,
"eval_logps/rejected": -217.0163116455078,
"eval_loss": 0.5850147008895874,
"eval_rewards/accuracies": 0.7333333492279053,
"eval_rewards/chosen": 1.2898719310760498,
"eval_rewards/margins": 1.370118498802185,
"eval_rewards/rejected": -0.0802464708685875,
"eval_runtime": 126.7763,
"eval_samples_per_second": 14.987,
"eval_steps_per_second": 0.237,
"step": 50
},
{
"epoch": 0.8208955223880597,
"grad_norm": 34.637816378109704,
"learning_rate": 8.691783729769873e-07,
"logits/chosen": -2.4367661476135254,
"logits/rejected": -2.4355311393737793,
"logps/chosen": -237.86105346679688,
"logps/rejected": -211.58084106445312,
"loss": 0.555,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": 1.1501390933990479,
"rewards/margins": 1.354534387588501,
"rewards/rejected": -0.20439541339874268,
"step": 55
},
{
"epoch": 0.8955223880597015,
"grad_norm": 39.91245885039146,
"learning_rate": 8.402298824670029e-07,
"logits/chosen": -2.5068681240081787,
"logits/rejected": -2.485455274581909,
"logps/chosen": -256.1775817871094,
"logps/rejected": -231.7941436767578,
"loss": 0.611,
"rewards/accuracies": 0.78125,
"rewards/chosen": 1.03065025806427,
"rewards/margins": 1.9949699640274048,
"rewards/rejected": -0.9643197059631348,
"step": 60
},
{
"epoch": 0.9701492537313433,
"grad_norm": 39.08249314957454,
"learning_rate": 8.089815357650089e-07,
"logits/chosen": -2.5430567264556885,
"logits/rejected": -2.543178081512451,
"logps/chosen": -232.22262573242188,
"logps/rejected": -225.10220336914062,
"loss": 0.5475,
"rewards/accuracies": 0.731249988079071,
"rewards/chosen": 0.8175865411758423,
"rewards/margins": 1.328970193862915,
"rewards/rejected": -0.5113834738731384,
"step": 65
},
{
"epoch": 1.044776119402985,
"grad_norm": 20.186385290734442,
"learning_rate": 7.756445627110522e-07,
"logits/chosen": -2.552123546600342,
"logits/rejected": -2.5313704013824463,
"logps/chosen": -234.0012664794922,
"logps/rejected": -240.6761474609375,
"loss": 0.3553,
"rewards/accuracies": 0.831250011920929,
"rewards/chosen": 1.4629334211349487,
"rewards/margins": 2.36128568649292,
"rewards/rejected": -0.8983524441719055,
"step": 70
},
{
"epoch": 1.1194029850746268,
"grad_norm": 21.595036305155492,
"learning_rate": 7.404443116588547e-07,
"logits/chosen": -2.5418105125427246,
"logits/rejected": -2.5285840034484863,
"logps/chosen": -242.90499877929688,
"logps/rejected": -216.7215576171875,
"loss": 0.2463,
"rewards/accuracies": 0.918749988079071,
"rewards/chosen": 2.360262393951416,
"rewards/margins": 3.110288381576538,
"rewards/rejected": -0.7500261068344116,
"step": 75
},
{
"epoch": 1.1940298507462686,
"grad_norm": 22.4039757471544,
"learning_rate": 7.036187261857288e-07,
"logits/chosen": -2.456780195236206,
"logits/rejected": -2.439521312713623,
"logps/chosen": -233.08560180664062,
"logps/rejected": -244.88961791992188,
"loss": 0.2423,
"rewards/accuracies": 0.90625,
"rewards/chosen": 2.1394240856170654,
"rewards/margins": 3.141387939453125,
"rewards/rejected": -1.0019636154174805,
"step": 80
},
{
"epoch": 1.2686567164179103,
"grad_norm": 25.42866646351196,
"learning_rate": 6.654167366624008e-07,
"logits/chosen": -2.3945281505584717,
"logits/rejected": -2.3779776096343994,
"logps/chosen": -230.88217163085938,
"logps/rejected": -222.7549285888672,
"loss": 0.2769,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 1.6151447296142578,
"rewards/margins": 2.812251091003418,
"rewards/rejected": -1.1971065998077393,
"step": 85
},
{
"epoch": 1.3432835820895521,
"grad_norm": 23.51747857481762,
"learning_rate": 6.260965775552713e-07,
"logits/chosen": -2.361176013946533,
"logits/rejected": -2.3344523906707764,
"logps/chosen": -236.94775390625,
"logps/rejected": -242.43069458007812,
"loss": 0.2743,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 1.946207046508789,
"rewards/margins": 3.319734573364258,
"rewards/rejected": -1.3735275268554688,
"step": 90
},
{
"epoch": 1.417910447761194,
"grad_norm": 25.346384805721517,
"learning_rate": 5.859240418356614e-07,
"logits/chosen": -2.3584346771240234,
"logits/rejected": -2.3219895362854004,
"logps/chosen": -246.46530151367188,
"logps/rejected": -231.7371063232422,
"loss": 0.2576,
"rewards/accuracies": 0.90625,
"rewards/chosen": 2.247493267059326,
"rewards/margins": 3.8193671703338623,
"rewards/rejected": -1.571873426437378,
"step": 95
},
{
"epoch": 1.4925373134328357,
"grad_norm": 17.435014853431674,
"learning_rate": 5.451706842957421e-07,
"logits/chosen": -2.4034814834594727,
"logits/rejected": -2.3993496894836426,
"logps/chosen": -248.6347198486328,
"logps/rejected": -231.65682983398438,
"loss": 0.255,
"rewards/accuracies": 0.9125000238418579,
"rewards/chosen": 2.421250820159912,
"rewards/margins": 3.7864270210266113,
"rewards/rejected": -1.3651763200759888,
"step": 100
},
{
"epoch": 1.4925373134328357,
"eval_logits/chosen": -2.4342453479766846,
"eval_logits/rejected": -2.4099276065826416,
"eval_logps/chosen": -245.49957275390625,
"eval_logps/rejected": -221.9740753173828,
"eval_loss": 0.5856931805610657,
"eval_rewards/accuracies": 0.7958333492279053,
"eval_rewards/chosen": 1.4525729417800903,
"eval_rewards/margins": 2.0285959243774414,
"eval_rewards/rejected": -0.5760230422019958,
"eval_runtime": 126.6219,
"eval_samples_per_second": 15.005,
"eval_steps_per_second": 0.237,
"step": 100
}
],
"logging_steps": 5,
"max_steps": 201,
"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
}
|