| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 625, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 7.936507936507937e-08, | |
| "logits/chosen": -0.2010916769504547, | |
| "logits/rejected": 0.09005054831504822, | |
| "logps/chosen": -540.942626953125, | |
| "logps/rejected": -796.8775634765625, | |
| "loss": 0.2182, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.936507936507937e-07, | |
| "logits/chosen": -0.15247675776481628, | |
| "logits/rejected": -0.14707261323928833, | |
| "logps/chosen": -501.8849792480469, | |
| "logps/rejected": -774.5216064453125, | |
| "loss": 0.2086, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": 3.564198050298728e-05, | |
| "rewards/margins": -2.1308711438905448e-05, | |
| "rewards/rejected": 5.695069921785034e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5873015873015873e-06, | |
| "logits/chosen": -0.1420287936925888, | |
| "logits/rejected": -0.10379602760076523, | |
| "logps/chosen": -489.9579162597656, | |
| "logps/rejected": -771.2081298828125, | |
| "loss": 0.2095, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.0012544477358460426, | |
| "rewards/margins": 0.0017321283230558038, | |
| "rewards/rejected": -0.0029865759424865246, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.380952380952381e-06, | |
| "logits/chosen": -0.16283434629440308, | |
| "logits/rejected": -0.14502206444740295, | |
| "logps/chosen": -532.1380615234375, | |
| "logps/rejected": -800.226806640625, | |
| "loss": 0.1996, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.007352855056524277, | |
| "rewards/margins": 0.009955727495253086, | |
| "rewards/rejected": -0.017308581620454788, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1746031746031746e-06, | |
| "logits/chosen": -0.19371934235095978, | |
| "logits/rejected": -0.15917012095451355, | |
| "logps/chosen": -529.4437255859375, | |
| "logps/rejected": -794.9968872070312, | |
| "loss": 0.1897, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.028293650597333908, | |
| "rewards/margins": 0.02773173525929451, | |
| "rewards/rejected": -0.05602538585662842, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.968253968253968e-06, | |
| "logits/chosen": -0.16935278475284576, | |
| "logits/rejected": -0.12196620553731918, | |
| "logps/chosen": -556.66064453125, | |
| "logps/rejected": -913.4127197265625, | |
| "loss": 0.1623, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.06831763684749603, | |
| "rewards/margins": 0.09133367240428925, | |
| "rewards/rejected": -0.15965132415294647, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.761904761904762e-06, | |
| "logits/chosen": -0.18819646537303925, | |
| "logits/rejected": -0.21151557564735413, | |
| "logps/chosen": -641.8372802734375, | |
| "logps/rejected": -1064.3094482421875, | |
| "loss": 0.1532, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.14167475700378418, | |
| "rewards/margins": 0.16748657822608948, | |
| "rewards/rejected": -0.30916133522987366, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.998086282661188e-06, | |
| "logits/chosen": -0.19726331532001495, | |
| "logits/rejected": -0.2387177050113678, | |
| "logps/chosen": -630.7346801757812, | |
| "logps/rejected": -1035.373046875, | |
| "loss": 0.1511, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.13452503085136414, | |
| "rewards/margins": 0.14926694333553314, | |
| "rewards/rejected": -0.2837919592857361, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.988720025682995e-06, | |
| "logits/chosen": -0.23227183520793915, | |
| "logits/rejected": -0.15486109256744385, | |
| "logps/chosen": -612.7208862304688, | |
| "logps/rejected": -1018.8267822265625, | |
| "loss": 0.142, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.11648458242416382, | |
| "rewards/margins": 0.137832373380661, | |
| "rewards/rejected": -0.25431695580482483, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.9715789537359126e-06, | |
| "logits/chosen": -0.269733726978302, | |
| "logits/rejected": -0.2005140334367752, | |
| "logps/chosen": -666.8603515625, | |
| "logps/rejected": -959.771484375, | |
| "loss": 0.1449, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.1299312561750412, | |
| "rewards/margins": 0.12538622319698334, | |
| "rewards/rejected": -0.25531744956970215, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.946716615897932e-06, | |
| "logits/chosen": -0.21513569355010986, | |
| "logits/rejected": -0.22341570258140564, | |
| "logps/chosen": -636.3090209960938, | |
| "logps/rejected": -982.4715576171875, | |
| "loss": 0.1376, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.1491270661354065, | |
| "rewards/margins": 0.12507781386375427, | |
| "rewards/rejected": -0.27420490980148315, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.9142106826480114e-06, | |
| "logits/chosen": -0.26347213983535767, | |
| "logits/rejected": -0.1899401843547821, | |
| "logps/chosen": -682.718017578125, | |
| "logps/rejected": -1051.3671875, | |
| "loss": 0.1456, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.1585748940706253, | |
| "rewards/margins": 0.13942097127437592, | |
| "rewards/rejected": -0.2979958653450012, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.874162703221823e-06, | |
| "logits/chosen": -0.28930234909057617, | |
| "logits/rejected": -0.21862976253032684, | |
| "logps/chosen": -624.7803344726562, | |
| "logps/rejected": -1044.067626953125, | |
| "loss": 0.1322, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.1492077112197876, | |
| "rewards/margins": 0.15546968579292297, | |
| "rewards/rejected": -0.3046773374080658, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.826697788369752e-06, | |
| "logits/chosen": -0.2196696251630783, | |
| "logits/rejected": -0.22875969111919403, | |
| "logps/chosen": -621.0065307617188, | |
| "logps/rejected": -944.654296875, | |
| "loss": 0.1432, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.15334677696228027, | |
| "rewards/margins": 0.11417926847934723, | |
| "rewards/rejected": -0.2675260007381439, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.7719642195082224e-06, | |
| "logits/chosen": -0.23348715901374817, | |
| "logits/rejected": -0.21704678237438202, | |
| "logps/chosen": -599.64453125, | |
| "logps/rejected": -991.4964599609375, | |
| "loss": 0.1471, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.13741596043109894, | |
| "rewards/margins": 0.13189604878425598, | |
| "rewards/rejected": -0.2693119943141937, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.710132985485355e-06, | |
| "logits/chosen": -0.2200555056333542, | |
| "logits/rejected": -0.20334219932556152, | |
| "logps/chosen": -661.56884765625, | |
| "logps/rejected": -1054.980712890625, | |
| "loss": 0.1376, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.16635623574256897, | |
| "rewards/margins": 0.14565840363502502, | |
| "rewards/rejected": -0.312014639377594, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.641397248408122e-06, | |
| "logits/chosen": -0.20752374827861786, | |
| "logits/rejected": -0.24948814511299133, | |
| "logps/chosen": -719.3741455078125, | |
| "logps/rejected": -1020.0384521484375, | |
| "loss": 0.1338, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.1749049723148346, | |
| "rewards/margins": 0.13863129913806915, | |
| "rewards/rejected": -0.31353622674942017, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5659717401997655e-06, | |
| "logits/chosen": -0.2512062191963196, | |
| "logits/rejected": -0.25224044919013977, | |
| "logps/chosen": -652.2489624023438, | |
| "logps/rejected": -1064.6927490234375, | |
| "loss": 0.1404, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.16943642497062683, | |
| "rewards/margins": 0.14737632870674133, | |
| "rewards/rejected": -0.3168127238750458, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.4840920917726425e-06, | |
| "logits/chosen": -0.26564353704452515, | |
| "logits/rejected": -0.31829017400741577, | |
| "logps/chosen": -643.8062744140625, | |
| "logps/rejected": -1146.453857421875, | |
| "loss": 0.1238, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.1721421331167221, | |
| "rewards/margins": 0.17220233380794525, | |
| "rewards/rejected": -0.34434446692466736, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.396014096912182e-06, | |
| "logits/chosen": -0.30384334921836853, | |
| "logits/rejected": -0.3049886226654053, | |
| "logps/chosen": -700.3992919921875, | |
| "logps/rejected": -1094.860107421875, | |
| "loss": 0.1293, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.190101757645607, | |
| "rewards/margins": 0.16493478417396545, | |
| "rewards/rejected": -0.35503652691841125, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.302012913171584e-06, | |
| "logits/chosen": -0.22822122275829315, | |
| "logits/rejected": -0.21796974539756775, | |
| "logps/chosen": -703.4880981445312, | |
| "logps/rejected": -1098.6363525390625, | |
| "loss": 0.1291, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.18932084739208221, | |
| "rewards/margins": 0.18700966238975525, | |
| "rewards/rejected": -0.37633052468299866, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.202382202273702e-06, | |
| "logits/chosen": -0.2829930782318115, | |
| "logits/rejected": -0.2941269278526306, | |
| "logps/chosen": -712.0001220703125, | |
| "logps/rejected": -1121.9896240234375, | |
| "loss": 0.1212, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.19748489558696747, | |
| "rewards/margins": 0.1557844579219818, | |
| "rewards/rejected": -0.3532693684101105, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.097433212705492e-06, | |
| "logits/chosen": -0.2563570737838745, | |
| "logits/rejected": -0.20130082964897156, | |
| "logps/chosen": -719.8380126953125, | |
| "logps/rejected": -1181.190185546875, | |
| "loss": 0.1349, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.2003975212574005, | |
| "rewards/margins": 0.1832597851753235, | |
| "rewards/rejected": -0.3836573362350464, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.987493807371033e-06, | |
| "logits/chosen": -0.19761434197425842, | |
| "logits/rejected": -0.24285447597503662, | |
| "logps/chosen": -713.5037841796875, | |
| "logps/rejected": -1127.9371337890625, | |
| "loss": 0.1377, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.20124192535877228, | |
| "rewards/margins": 0.16942095756530762, | |
| "rewards/rejected": -0.3706628680229187, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.872907439340758e-06, | |
| "logits/chosen": -0.22070392966270447, | |
| "logits/rejected": -0.2395116537809372, | |
| "logps/chosen": -689.1747436523438, | |
| "logps/rejected": -1246.12646484375, | |
| "loss": 0.1219, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -0.18724720180034637, | |
| "rewards/margins": 0.221848726272583, | |
| "rewards/rejected": -0.4090959429740906, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.75403207889666e-06, | |
| "logits/chosen": -0.2952207028865814, | |
| "logits/rejected": -0.286382257938385, | |
| "logps/chosen": -680.5881958007812, | |
| "logps/rejected": -1075.620849609375, | |
| "loss": 0.1351, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.20417407155036926, | |
| "rewards/margins": 0.16537046432495117, | |
| "rewards/rejected": -0.36954453587532043, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.631239095225417e-06, | |
| "logits/chosen": -0.28091198205947876, | |
| "logits/rejected": -0.25016146898269653, | |
| "logps/chosen": -717.04150390625, | |
| "logps/rejected": -1180.3306884765625, | |
| "loss": 0.1185, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.2006029188632965, | |
| "rewards/margins": 0.1787930279970169, | |
| "rewards/rejected": -0.3793959617614746, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5049120962530608e-06, | |
| "logits/chosen": -0.23318138718605042, | |
| "logits/rejected": -0.1964143067598343, | |
| "logps/chosen": -722.5912475585938, | |
| "logps/rejected": -1105.978759765625, | |
| "loss": 0.1251, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.20104451477527618, | |
| "rewards/margins": 0.16880543529987335, | |
| "rewards/rejected": -0.36984992027282715, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3754457302455464e-06, | |
| "logits/chosen": -0.21553221344947815, | |
| "logits/rejected": -0.2545991837978363, | |
| "logps/chosen": -708.6617431640625, | |
| "logps/rejected": -1200.54248046875, | |
| "loss": 0.1241, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.1849481165409088, | |
| "rewards/margins": 0.22220449149608612, | |
| "rewards/rejected": -0.40715259313583374, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2432444529190714e-06, | |
| "logits/chosen": -0.2878536283969879, | |
| "logits/rejected": -0.240807443857193, | |
| "logps/chosen": -685.4102783203125, | |
| "logps/rejected": -1073.6112060546875, | |
| "loss": 0.1278, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.19928036630153656, | |
| "rewards/margins": 0.167768195271492, | |
| "rewards/rejected": -0.36704859137535095, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1087212639117057e-06, | |
| "logits/chosen": -0.22543080151081085, | |
| "logits/rejected": -0.25053372979164124, | |
| "logps/chosen": -713.0574340820312, | |
| "logps/rejected": -1165.50244140625, | |
| "loss": 0.1199, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.20086824893951416, | |
| "rewards/margins": 0.19784381985664368, | |
| "rewards/rejected": -0.39871203899383545, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9722964165636263e-06, | |
| "logits/chosen": -0.2477823793888092, | |
| "logits/rejected": -0.23944005370140076, | |
| "logps/chosen": -741.0659790039062, | |
| "logps/rejected": -1150.4737548828125, | |
| "loss": 0.1334, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.21726730465888977, | |
| "rewards/margins": 0.18077899515628815, | |
| "rewards/rejected": -0.3980463147163391, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8343961050366275e-06, | |
| "logits/chosen": -0.206426739692688, | |
| "logits/rejected": -0.294622004032135, | |
| "logps/chosen": -698.1879272460938, | |
| "logps/rejected": -1074.371337890625, | |
| "loss": 0.1279, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.1868043690919876, | |
| "rewards/margins": 0.1699267327785492, | |
| "rewards/rejected": -0.3567310869693756, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.695451132874385e-06, | |
| "logits/chosen": -0.2724359631538391, | |
| "logits/rejected": -0.21938621997833252, | |
| "logps/chosen": -708.3114013671875, | |
| "logps/rejected": -1099.7318115234375, | |
| "loss": 0.1196, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.18547296524047852, | |
| "rewards/margins": 0.1704384684562683, | |
| "rewards/rejected": -0.3559114336967468, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5558955671628964e-06, | |
| "logits/chosen": -0.25434714555740356, | |
| "logits/rejected": -0.25090768933296204, | |
| "logps/chosen": -695.5904541015625, | |
| "logps/rejected": -1176.5301513671875, | |
| "loss": 0.131, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.21099016070365906, | |
| "rewards/margins": 0.20053663849830627, | |
| "rewards/rejected": -0.4115268290042877, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4161653824955654e-06, | |
| "logits/chosen": -0.28552037477493286, | |
| "logits/rejected": -0.24306419491767883, | |
| "logps/chosen": -764.7316284179688, | |
| "logps/rejected": -1231.4193115234375, | |
| "loss": 0.1166, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.24171674251556396, | |
| "rewards/margins": 0.21239931881427765, | |
| "rewards/rejected": -0.4541160464286804, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2766970989791697e-06, | |
| "logits/chosen": -0.25456491112709045, | |
| "logits/rejected": -0.21178500354290009, | |
| "logps/chosen": -765.3328247070312, | |
| "logps/rejected": -1261.208984375, | |
| "loss": 0.1103, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.22941815853118896, | |
| "rewards/margins": 0.21526849269866943, | |
| "rewards/rejected": -0.4446867108345032, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1379264185356545e-06, | |
| "logits/chosen": -0.2724260687828064, | |
| "logits/rejected": -0.18538828194141388, | |
| "logps/chosen": -741.7236328125, | |
| "logps/rejected": -1078.375732421875, | |
| "loss": 0.1161, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.21514591574668884, | |
| "rewards/margins": 0.16624660789966583, | |
| "rewards/rejected": -0.38139253854751587, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.000286863759934e-06, | |
| "logits/chosen": -0.27519670128822327, | |
| "logits/rejected": -0.2832408845424652, | |
| "logps/chosen": -682.8367309570312, | |
| "logps/rejected": -1131.5003662109375, | |
| "loss": 0.1198, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.20102819800376892, | |
| "rewards/margins": 0.18376831710338593, | |
| "rewards/rejected": -0.38479650020599365, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8642084235859764e-06, | |
| "logits/chosen": -0.2652140259742737, | |
| "logits/rejected": -0.2889960706233978, | |
| "logps/chosen": -731.1627807617188, | |
| "logps/rejected": -1137.467041015625, | |
| "loss": 0.1208, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.19975660741329193, | |
| "rewards/margins": 0.21678335964679718, | |
| "rewards/rejected": -0.4165399968624115, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7301162099921013e-06, | |
| "logits/chosen": -0.26221686601638794, | |
| "logits/rejected": -0.2641231119632721, | |
| "logps/chosen": -755.5460205078125, | |
| "logps/rejected": -1093.4853515625, | |
| "loss": 0.1228, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.21798157691955566, | |
| "rewards/margins": 0.17120864987373352, | |
| "rewards/rejected": -0.3891902267932892, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5984291299420117e-06, | |
| "logits/chosen": -0.2737795114517212, | |
| "logits/rejected": -0.26781201362609863, | |
| "logps/chosen": -703.7815551757812, | |
| "logps/rejected": -1142.2947998046875, | |
| "loss": 0.1225, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.2116294652223587, | |
| "rewards/margins": 0.18610379099845886, | |
| "rewards/rejected": -0.39773327112197876, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4695585767104092e-06, | |
| "logits/chosen": -0.29544955492019653, | |
| "logits/rejected": -0.2773270308971405, | |
| "logps/chosen": -686.2037353515625, | |
| "logps/rejected": -1173.82958984375, | |
| "loss": 0.1319, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.22286900877952576, | |
| "rewards/margins": 0.183340385556221, | |
| "rewards/rejected": -0.40620937943458557, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3439071446815452e-06, | |
| "logits/chosen": -0.2415129840373993, | |
| "logits/rejected": -0.27339568734169006, | |
| "logps/chosen": -718.8626708984375, | |
| "logps/rejected": -1240.2945556640625, | |
| "loss": 0.1158, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.21683505177497864, | |
| "rewards/margins": 0.2326025515794754, | |
| "rewards/rejected": -0.44943755865097046, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2218673716356919e-06, | |
| "logits/chosen": -0.27473658323287964, | |
| "logits/rejected": -0.2448565512895584, | |
| "logps/chosen": -734.1240844726562, | |
| "logps/rejected": -1157.1583251953125, | |
| "loss": 0.1151, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.22244882583618164, | |
| "rewards/margins": 0.20606684684753418, | |
| "rewards/rejected": -0.4285156726837158, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.103820512452661e-06, | |
| "logits/chosen": -0.24782344698905945, | |
| "logits/rejected": -0.26655644178390503, | |
| "logps/chosen": -712.7034912109375, | |
| "logps/rejected": -1142.8284912109375, | |
| "loss": 0.1273, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.21107585728168488, | |
| "rewards/margins": 0.19358885288238525, | |
| "rewards/rejected": -0.40466469526290894, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.901353480633468e-07, | |
| "logits/chosen": -0.2719441056251526, | |
| "logits/rejected": -0.26369568705558777, | |
| "logps/chosen": -686.2820434570312, | |
| "logps/rejected": -1074.0950927734375, | |
| "loss": 0.1207, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.21289560198783875, | |
| "rewards/margins": 0.17246314883232117, | |
| "rewards/rejected": -0.3853587806224823, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.811670333701544e-07, | |
| "logits/chosen": -0.2672122120857239, | |
| "logits/rejected": -0.2347395420074463, | |
| "logps/chosen": -752.274169921875, | |
| "logps/rejected": -1232.9998779296875, | |
| "loss": 0.1098, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -0.2272864133119583, | |
| "rewards/margins": 0.21149499714374542, | |
| "rewards/rejected": -0.4387814402580261, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.772559877354341e-07, | |
| "logits/chosen": -0.24777153134346008, | |
| "logits/rejected": -0.2399456799030304, | |
| "logps/chosen": -736.2609252929688, | |
| "logps/rejected": -1130.094970703125, | |
| "loss": 0.1168, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.22519740462303162, | |
| "rewards/margins": 0.17224054038524628, | |
| "rewards/rejected": -0.3974379599094391, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.787268315040604e-07, | |
| "logits/chosen": -0.2684328854084015, | |
| "logits/rejected": -0.24628722667694092, | |
| "logps/chosen": -689.9044799804688, | |
| "logps/rejected": -1155.0946044921875, | |
| "loss": 0.1174, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.21575360000133514, | |
| "rewards/margins": 0.1959463357925415, | |
| "rewards/rejected": -0.41169995069503784, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.858873718824829e-07, | |
| "logits/chosen": -0.26447051763534546, | |
| "logits/rejected": -0.26331770420074463, | |
| "logps/chosen": -707.0438842773438, | |
| "logps/rejected": -1152.1182861328125, | |
| "loss": 0.1172, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.20341674983501434, | |
| "rewards/margins": 0.19374600052833557, | |
| "rewards/rejected": -0.3971627652645111, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.990276413423817e-07, | |
| "logits/chosen": -0.23977680504322052, | |
| "logits/rejected": -0.22561779618263245, | |
| "logps/chosen": -729.1976318359375, | |
| "logps/rejected": -1142.5386962890625, | |
| "loss": 0.1182, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.21028919517993927, | |
| "rewards/margins": 0.1970864236354828, | |
| "rewards/rejected": -0.40737563371658325, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.184189915529796e-07, | |
| "logits/chosen": -0.26442182064056396, | |
| "logits/rejected": -0.25705209374427795, | |
| "logps/chosen": -737.6463623046875, | |
| "logps/rejected": -1247.0609130859375, | |
| "loss": 0.1102, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.2177223414182663, | |
| "rewards/margins": 0.20199593901634216, | |
| "rewards/rejected": -0.41971832513809204, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4431324567258176e-07, | |
| "logits/chosen": -0.26106202602386475, | |
| "logits/rejected": -0.2545424997806549, | |
| "logps/chosen": -700.1468505859375, | |
| "logps/rejected": -1189.3382568359375, | |
| "loss": 0.1211, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.20997758209705353, | |
| "rewards/margins": 0.2070033997297287, | |
| "rewards/rejected": -0.4169809818267822, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.769419116476052e-07, | |
| "logits/chosen": -0.2785791754722595, | |
| "logits/rejected": -0.26757797598838806, | |
| "logps/chosen": -753.193115234375, | |
| "logps/rejected": -1155.57763671875, | |
| "loss": 0.1094, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.21952767670154572, | |
| "rewards/margins": 0.1966555118560791, | |
| "rewards/rejected": -0.4161831736564636, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1651545897676512e-07, | |
| "logits/chosen": -0.21319365501403809, | |
| "logits/rejected": -0.20390382409095764, | |
| "logps/chosen": -736.0490112304688, | |
| "logps/rejected": -1217.077880859375, | |
| "loss": 0.1175, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.21741120517253876, | |
| "rewards/margins": 0.21254892647266388, | |
| "rewards/rejected": -0.42996007204055786, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6322266119983222e-07, | |
| "logits/chosen": -0.24104240536689758, | |
| "logits/rejected": -0.2172488421201706, | |
| "logps/chosen": -765.8303833007812, | |
| "logps/rejected": -1205.9599609375, | |
| "loss": 0.11, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.21993395686149597, | |
| "rewards/margins": 0.19853533804416656, | |
| "rewards/rejected": -0.41846928000450134, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1723000616502167e-07, | |
| "logits/chosen": -0.26349014043807983, | |
| "logits/rejected": -0.3146611452102661, | |
| "logps/chosen": -755.0218505859375, | |
| "logps/rejected": -1200.8770751953125, | |
| "loss": 0.1129, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -0.21991780400276184, | |
| "rewards/margins": 0.22101625800132751, | |
| "rewards/rejected": -0.44093409180641174, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.868117591737585e-08, | |
| "logits/chosen": -0.20703573524951935, | |
| "logits/rejected": -0.21116182208061218, | |
| "logps/chosen": -741.433837890625, | |
| "logps/rejected": -1168.3992919921875, | |
| "loss": 0.1149, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.2117561548948288, | |
| "rewards/margins": 0.19863612949848175, | |
| "rewards/rejected": -0.41039222478866577, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.769659783295383e-08, | |
| "logits/chosen": -0.23616275191307068, | |
| "logits/rejected": -0.2213321179151535, | |
| "logps/chosen": -802.316650390625, | |
| "logps/rejected": -1224.220458984375, | |
| "loss": 0.1123, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.2234838753938675, | |
| "rewards/margins": 0.21354734897613525, | |
| "rewards/rejected": -0.43703120946884155, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4373068401120358e-08, | |
| "logits/chosen": -0.25830012559890747, | |
| "logits/rejected": -0.24621865153312683, | |
| "logps/chosen": -802.4405517578125, | |
| "logps/rejected": -1169.389892578125, | |
| "loss": 0.1136, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.23679308593273163, | |
| "rewards/margins": 0.1784859597682953, | |
| "rewards/rejected": -0.4152790606021881, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.78345083022425e-09, | |
| "logits/chosen": -0.24250511825084686, | |
| "logits/rejected": -0.24350687861442566, | |
| "logps/chosen": -742.060546875, | |
| "logps/rejected": -1132.1630859375, | |
| "loss": 0.1141, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.2040189504623413, | |
| "rewards/margins": 0.19322746992111206, | |
| "rewards/rejected": -0.39724642038345337, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.764474213677654e-10, | |
| "logits/chosen": -0.2109394073486328, | |
| "logits/rejected": -0.22563035786151886, | |
| "logps/chosen": -744.6256103515625, | |
| "logps/rejected": -1185.656982421875, | |
| "loss": 0.1153, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.23164710402488708, | |
| "rewards/margins": 0.20466098189353943, | |
| "rewards/rejected": -0.4363080859184265, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 625, | |
| "total_flos": 0.0, | |
| "train_loss": 0.13047290132045747, | |
| "train_runtime": 8392.5195, | |
| "train_samples_per_second": 3.575, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 625, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |