File size: 9,778 Bytes
3cc9f34 | 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 | {
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.9968652037617555,
"eval_steps": 500,
"global_step": 159,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 3.125e-08,
"logits/chosen": -2.0781588554382324,
"logits/rejected": -1.9858250617980957,
"logps/chosen": -453.2818603515625,
"logps/pi_response": -300.6171875,
"logps/ref_response": -300.6171875,
"logps/rejected": -393.94146728515625,
"loss": 0.6931,
"rewards/accuracies": 0.0,
"rewards/chosen": 0.0,
"rewards/margins": 0.0,
"rewards/rejected": 0.0,
"step": 1
},
{
"epoch": 0.06,
"learning_rate": 3.1249999999999997e-07,
"logits/chosen": -2.1030797958374023,
"logits/rejected": -2.0334737300872803,
"logps/chosen": -286.7795715332031,
"logps/pi_response": -192.09445190429688,
"logps/ref_response": -192.19212341308594,
"logps/rejected": -342.5543518066406,
"loss": 0.6687,
"rewards/accuracies": 0.4930555522441864,
"rewards/chosen": -0.049170926213264465,
"rewards/margins": 0.10100988298654556,
"rewards/rejected": -0.15018080174922943,
"step": 10
},
{
"epoch": 0.13,
"learning_rate": 4.990353313429303e-07,
"logits/chosen": -2.0767107009887695,
"logits/rejected": -2.057764768600464,
"logps/chosen": -252.1197967529297,
"logps/pi_response": -182.93057250976562,
"logps/ref_response": -184.64120483398438,
"logps/rejected": -385.4762268066406,
"loss": 0.7043,
"rewards/accuracies": 0.574999988079071,
"rewards/chosen": -0.17670336365699768,
"rewards/margins": 0.7420909404754639,
"rewards/rejected": -0.9187943339347839,
"step": 20
},
{
"epoch": 0.19,
"learning_rate": 4.882681251368548e-07,
"logits/chosen": -2.079202175140381,
"logits/rejected": -2.032349109649658,
"logps/chosen": -291.5123596191406,
"logps/pi_response": -202.2268829345703,
"logps/ref_response": -192.75753784179688,
"logps/rejected": -345.61810302734375,
"loss": 0.7312,
"rewards/accuracies": 0.637499988079071,
"rewards/chosen": -0.801366925239563,
"rewards/margins": 0.6456394195556641,
"rewards/rejected": -1.4470064640045166,
"step": 30
},
{
"epoch": 0.25,
"learning_rate": 4.6604720940421207e-07,
"logits/chosen": -2.147491931915283,
"logits/rejected": -2.0794780254364014,
"logps/chosen": -282.3897705078125,
"logps/pi_response": -198.35958862304688,
"logps/ref_response": -189.3666534423828,
"logps/rejected": -389.67364501953125,
"loss": 0.8956,
"rewards/accuracies": 0.668749988079071,
"rewards/chosen": -0.9823302030563354,
"rewards/margins": 0.8020473718643188,
"rewards/rejected": -1.7843774557113647,
"step": 40
},
{
"epoch": 0.31,
"learning_rate": 4.3344075855595097e-07,
"logits/chosen": -2.226250171661377,
"logits/rejected": -2.1124267578125,
"logps/chosen": -302.74493408203125,
"logps/pi_response": -224.74600219726562,
"logps/ref_response": -208.11865234375,
"logps/rejected": -445.48974609375,
"loss": 0.8241,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": -1.6344465017318726,
"rewards/margins": 1.1299909353256226,
"rewards/rejected": -2.764437198638916,
"step": 50
},
{
"epoch": 0.38,
"learning_rate": 3.920161866827889e-07,
"logits/chosen": -2.119874954223633,
"logits/rejected": -2.081808090209961,
"logps/chosen": -285.70880126953125,
"logps/pi_response": -211.44058227539062,
"logps/ref_response": -197.88754272460938,
"logps/rejected": -408.64111328125,
"loss": 0.7844,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": -1.3133118152618408,
"rewards/margins": 0.963096022605896,
"rewards/rejected": -2.2764081954956055,
"step": 60
},
{
"epoch": 0.44,
"learning_rate": 3.4376480090239047e-07,
"logits/chosen": -2.0851354598999023,
"logits/rejected": -2.0019078254699707,
"logps/chosen": -282.61920166015625,
"logps/pi_response": -194.21969604492188,
"logps/ref_response": -180.25257873535156,
"logps/rejected": -390.8282775878906,
"loss": 0.8222,
"rewards/accuracies": 0.6499999761581421,
"rewards/chosen": -1.2267100811004639,
"rewards/margins": 1.4668184518814087,
"rewards/rejected": -2.693528652191162,
"step": 70
},
{
"epoch": 0.5,
"learning_rate": 2.910060778827554e-07,
"logits/chosen": -2.161830186843872,
"logits/rejected": -2.095283269882202,
"logps/chosen": -327.2430114746094,
"logps/pi_response": -215.4277801513672,
"logps/ref_response": -201.3350372314453,
"logps/rejected": -383.038818359375,
"loss": 0.6609,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": -1.1827173233032227,
"rewards/margins": 1.14064621925354,
"rewards/rejected": -2.323363780975342,
"step": 80
},
{
"epoch": 0.56,
"learning_rate": 2.3627616503391812e-07,
"logits/chosen": -2.1371817588806152,
"logits/rejected": -2.031243085861206,
"logps/chosen": -288.8699951171875,
"logps/pi_response": -221.05953979492188,
"logps/ref_response": -203.6631317138672,
"logps/rejected": -432.8627014160156,
"loss": 0.6938,
"rewards/accuracies": 0.7124999761581421,
"rewards/chosen": -1.5247129201889038,
"rewards/margins": 1.1340636014938354,
"rewards/rejected": -2.6587765216827393,
"step": 90
},
{
"epoch": 0.63,
"learning_rate": 1.8220596619089573e-07,
"logits/chosen": -2.113327741622925,
"logits/rejected": -2.029888391494751,
"logps/chosen": -297.3309326171875,
"logps/pi_response": -229.23782348632812,
"logps/ref_response": -213.95852661132812,
"logps/rejected": -410.16015625,
"loss": 0.7029,
"rewards/accuracies": 0.71875,
"rewards/chosen": -1.5872786045074463,
"rewards/margins": 0.9572550654411316,
"rewards/rejected": -2.5445337295532227,
"step": 100
},
{
"epoch": 0.69,
"learning_rate": 1.3139467229135998e-07,
"logits/chosen": -2.10136342048645,
"logits/rejected": -2.0246944427490234,
"logps/chosen": -301.06085205078125,
"logps/pi_response": -213.1235809326172,
"logps/ref_response": -197.9862823486328,
"logps/rejected": -417.035400390625,
"loss": 0.6701,
"rewards/accuracies": 0.6812499761581421,
"rewards/chosen": -1.4077179431915283,
"rewards/margins": 1.052346110343933,
"rewards/rejected": -2.460064172744751,
"step": 110
},
{
"epoch": 0.75,
"learning_rate": 8.628481651367875e-08,
"logits/chosen": -2.0771682262420654,
"logits/rejected": -2.0391485691070557,
"logps/chosen": -286.8796691894531,
"logps/pi_response": -202.24591064453125,
"logps/ref_response": -183.84140014648438,
"logps/rejected": -393.9871520996094,
"loss": 0.582,
"rewards/accuracies": 0.7250000238418579,
"rewards/chosen": -1.8573909997940063,
"rewards/margins": 1.027687668800354,
"rewards/rejected": -2.8850789070129395,
"step": 120
},
{
"epoch": 0.82,
"learning_rate": 4.904486005914027e-08,
"logits/chosen": -2.0720248222351074,
"logits/rejected": -2.039855480194092,
"logps/chosen": -284.1001892089844,
"logps/pi_response": -205.81741333007812,
"logps/ref_response": -192.61219787597656,
"logps/rejected": -424.58856201171875,
"loss": 0.6267,
"rewards/accuracies": 0.706250011920929,
"rewards/chosen": -1.078873872756958,
"rewards/margins": 1.561975121498108,
"rewards/rejected": -2.6408488750457764,
"step": 130
},
{
"epoch": 0.88,
"learning_rate": 2.1464952759020856e-08,
"logits/chosen": -2.074803590774536,
"logits/rejected": -1.9720996618270874,
"logps/chosen": -289.3224182128906,
"logps/pi_response": -216.6945343017578,
"logps/ref_response": -199.8957977294922,
"logps/rejected": -415.46209716796875,
"loss": 0.6077,
"rewards/accuracies": 0.71875,
"rewards/chosen": -1.5722277164459229,
"rewards/margins": 1.58167564868927,
"rewards/rejected": -3.1539032459259033,
"step": 140
},
{
"epoch": 0.94,
"learning_rate": 4.8708793644441086e-09,
"logits/chosen": -2.0950675010681152,
"logits/rejected": -2.0208332538604736,
"logps/chosen": -298.1050720214844,
"logps/pi_response": -232.5490264892578,
"logps/ref_response": -216.212158203125,
"logps/rejected": -405.6884460449219,
"loss": 0.6254,
"rewards/accuracies": 0.699999988079071,
"rewards/chosen": -1.4602974653244019,
"rewards/margins": 0.9969442486763,
"rewards/rejected": -2.4572417736053467,
"step": 150
},
{
"epoch": 1.0,
"step": 159,
"total_flos": 0.0,
"train_loss": 0.6964452761524128,
"train_runtime": 4539.1381,
"train_samples_per_second": 4.489,
"train_steps_per_second": 0.035
}
],
"logging_steps": 10,
"max_steps": 159,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}
|