| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.002721690169595319, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.4021127119941484e-05, | |
| "grad_norm": 33.091712951660156, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "logits/chosen": -3.133824586868286, | |
| "logits/rejected": -3.1356313228607178, | |
| "logps/chosen": -317.8847961425781, | |
| "logps/rejected": -306.8866271972656, | |
| "loss": 0.8544, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.10576057434082031, | |
| "rewards/margins": 0.2207910716533661, | |
| "rewards/rejected": -0.3265516459941864, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 6.804225423988297e-05, | |
| "grad_norm": 38.49541473388672, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "logits/chosen": -3.1433358192443848, | |
| "logits/rejected": -3.1124918460845947, | |
| "logps/chosen": -212.93820190429688, | |
| "logps/rejected": -198.22511291503906, | |
| "loss": 1.02, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.28638574481010437, | |
| "rewards/margins": -0.37813109159469604, | |
| "rewards/rejected": 0.0917452871799469, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00010206338135982445, | |
| "grad_norm": 27.203516006469727, | |
| "learning_rate": 3e-06, | |
| "logits/chosen": -3.2104451656341553, | |
| "logits/rejected": -3.2103281021118164, | |
| "logps/chosen": -235.89254760742188, | |
| "logps/rejected": -206.0758514404297, | |
| "loss": 0.605, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.40017586946487427, | |
| "rewards/margins": 0.24887371063232422, | |
| "rewards/rejected": 0.15130215883255005, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00013608450847976594, | |
| "grad_norm": 26.42736053466797, | |
| "learning_rate": 4.000000000000001e-06, | |
| "logits/chosen": -3.0741634368896484, | |
| "logits/rejected": -2.953117847442627, | |
| "logps/chosen": -428.7364196777344, | |
| "logps/rejected": -192.357666015625, | |
| "loss": 0.4914, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.4332427978515625, | |
| "rewards/margins": 0.5804362893104553, | |
| "rewards/rejected": -0.1471935510635376, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.00017010563559970743, | |
| "grad_norm": 35.623634338378906, | |
| "learning_rate": 5e-06, | |
| "logits/chosen": -2.985285758972168, | |
| "logits/rejected": -2.8854246139526367, | |
| "logps/chosen": -342.38604736328125, | |
| "logps/rejected": -211.78237915039062, | |
| "loss": 0.8273, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": -0.3475755751132965, | |
| "rewards/margins": -0.19371166825294495, | |
| "rewards/rejected": -0.15386392176151276, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0002041267627196489, | |
| "grad_norm": 31.70409393310547, | |
| "learning_rate": 6e-06, | |
| "logits/chosen": -3.074589967727661, | |
| "logits/rejected": -2.85581636428833, | |
| "logps/chosen": -593.6435546875, | |
| "logps/rejected": -259.5284729003906, | |
| "loss": 0.5473, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.2495414763689041, | |
| "rewards/margins": 0.35888367891311646, | |
| "rewards/rejected": -0.10934218764305115, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00023814788983959038, | |
| "grad_norm": 32.41214370727539, | |
| "learning_rate": 7.000000000000001e-06, | |
| "logits/chosen": -2.976675033569336, | |
| "logits/rejected": -2.978133201599121, | |
| "logps/chosen": -218.11688232421875, | |
| "logps/rejected": -231.7698974609375, | |
| "loss": 0.7124, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": -0.08671779930591583, | |
| "rewards/margins": 0.21950224041938782, | |
| "rewards/rejected": -0.30622005462646484, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00027216901695953187, | |
| "grad_norm": 27.637351989746094, | |
| "learning_rate": 8.000000000000001e-06, | |
| "logits/chosen": -2.997600793838501, | |
| "logits/rejected": -3.1078853607177734, | |
| "logps/chosen": -165.71444702148438, | |
| "logps/rejected": -335.5938720703125, | |
| "loss": 0.6085, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.057393088936805725, | |
| "rewards/margins": 0.34109365940093994, | |
| "rewards/rejected": -0.3984867036342621, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00030619014407947336, | |
| "grad_norm": 32.246559143066406, | |
| "learning_rate": 9e-06, | |
| "logits/chosen": -3.0760390758514404, | |
| "logits/rejected": -2.938197135925293, | |
| "logps/chosen": -463.1333312988281, | |
| "logps/rejected": -173.99951171875, | |
| "loss": 0.6398, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.5581258535385132, | |
| "rewards/margins": 0.7653559446334839, | |
| "rewards/rejected": -0.20723000168800354, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.00034021127119941485, | |
| "grad_norm": 39.84626388549805, | |
| "learning_rate": 1e-05, | |
| "logits/chosen": -3.192596673965454, | |
| "logits/rejected": -2.9751501083374023, | |
| "logps/chosen": -473.14532470703125, | |
| "logps/rejected": -110.90155029296875, | |
| "loss": 0.9794, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": -0.3494076132774353, | |
| "rewards/margins": -0.31546899676322937, | |
| "rewards/rejected": -0.03393859416246414, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00037423239831935634, | |
| "grad_norm": 39.444889068603516, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "logits/chosen": -3.096883535385132, | |
| "logits/rejected": -3.043616771697998, | |
| "logps/chosen": -391.7703552246094, | |
| "logps/rejected": -271.8331604003906, | |
| "loss": 0.8029, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6240158081054688, | |
| "rewards/margins": 0.4651091992855072, | |
| "rewards/rejected": 0.15890654921531677, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0004082535254392978, | |
| "grad_norm": 40.78356170654297, | |
| "learning_rate": 1.2e-05, | |
| "logits/chosen": -3.000021457672119, | |
| "logits/rejected": -3.0893115997314453, | |
| "logps/chosen": -255.94607543945312, | |
| "logps/rejected": -405.3093566894531, | |
| "loss": 0.6297, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.33697396516799927, | |
| "rewards/margins": 0.3315456509590149, | |
| "rewards/rejected": 0.005428321659564972, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.00044227465255923927, | |
| "grad_norm": 35.69674301147461, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "logits/chosen": -3.011014699935913, | |
| "logits/rejected": -3.0214414596557617, | |
| "logps/chosen": -223.4480438232422, | |
| "logps/rejected": -261.62469482421875, | |
| "loss": 0.815, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.7416718006134033, | |
| "rewards/margins": 0.19626998901367188, | |
| "rewards/rejected": 0.5454018115997314, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00047629577967918076, | |
| "grad_norm": 32.88405990600586, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "logits/chosen": -3.065549373626709, | |
| "logits/rejected": -3.1094508171081543, | |
| "logps/chosen": -169.08047485351562, | |
| "logps/rejected": -227.85162353515625, | |
| "loss": 0.8115, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 0.37659722566604614, | |
| "rewards/margins": -0.11390496790409088, | |
| "rewards/rejected": 0.4905022084712982, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0005103169067991223, | |
| "grad_norm": 20.918012619018555, | |
| "learning_rate": 1.5e-05, | |
| "logits/chosen": -3.045846462249756, | |
| "logits/rejected": -2.9114990234375, | |
| "logps/chosen": -573.1253662109375, | |
| "logps/rejected": -273.9326477050781, | |
| "loss": 0.3105, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.7755428552627563, | |
| "rewards/margins": 1.082038164138794, | |
| "rewards/rejected": -0.3064952790737152, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0005443380339190637, | |
| "grad_norm": 38.434749603271484, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "logits/chosen": -3.1250810623168945, | |
| "logits/rejected": -2.96527099609375, | |
| "logps/chosen": -396.9795837402344, | |
| "logps/rejected": -174.92918395996094, | |
| "loss": 0.9971, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.2244594693183899, | |
| "rewards/margins": -0.37520748376846313, | |
| "rewards/rejected": 0.5996670126914978, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0005783591610390052, | |
| "grad_norm": 37.71885681152344, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "logits/chosen": -2.970050096511841, | |
| "logits/rejected": -3.0346829891204834, | |
| "logps/chosen": -236.67071533203125, | |
| "logps/rejected": -403.0212097167969, | |
| "loss": 1.0307, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.9259212613105774, | |
| "rewards/margins": -0.23367927968502045, | |
| "rewards/rejected": 1.1596004962921143, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0006123802881589467, | |
| "grad_norm": 25.964290618896484, | |
| "learning_rate": 1.8e-05, | |
| "logits/chosen": -3.0690722465515137, | |
| "logits/rejected": -3.0348947048187256, | |
| "logps/chosen": -312.81317138671875, | |
| "logps/rejected": -259.230712890625, | |
| "loss": 0.6989, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.6348915100097656, | |
| "rewards/margins": 0.6418477892875671, | |
| "rewards/rejected": -0.006956290453672409, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.0006464014152788882, | |
| "grad_norm": 27.076108932495117, | |
| "learning_rate": 1.9e-05, | |
| "logits/chosen": -3.135136842727661, | |
| "logits/rejected": -2.911625862121582, | |
| "logps/chosen": -473.7462158203125, | |
| "logps/rejected": -141.9491729736328, | |
| "loss": 0.6027, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.9424793720245361, | |
| "rewards/margins": 0.9380611181259155, | |
| "rewards/rejected": 0.00441819429397583, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0006804225423988297, | |
| "grad_norm": 51.820159912109375, | |
| "learning_rate": 2e-05, | |
| "logits/chosen": -2.8403027057647705, | |
| "logits/rejected": -3.0664803981781006, | |
| "logps/chosen": -219.79153442382812, | |
| "logps/rejected": -231.70004272460938, | |
| "loss": 1.1022, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.052263081073760986, | |
| "rewards/margins": -0.6957410573959351, | |
| "rewards/rejected": 0.7480041980743408, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0007144436695187712, | |
| "grad_norm": 45.9146614074707, | |
| "learning_rate": 2.1e-05, | |
| "logits/chosen": -2.9764764308929443, | |
| "logits/rejected": -2.9391565322875977, | |
| "logps/chosen": -458.0147705078125, | |
| "logps/rejected": -218.37648010253906, | |
| "loss": 1.0772, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.8049196004867554, | |
| "rewards/margins": -0.448437362909317, | |
| "rewards/rejected": 1.2533570528030396, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.0007484647966387127, | |
| "grad_norm": 22.854082107543945, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "logits/chosen": -3.1029181480407715, | |
| "logits/rejected": -3.044419527053833, | |
| "logps/chosen": -154.8139190673828, | |
| "logps/rejected": -155.980712890625, | |
| "loss": 0.4065, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.45841503143310547, | |
| "rewards/margins": 0.8753025531768799, | |
| "rewards/rejected": -0.41688746213912964, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0007824859237586541, | |
| "grad_norm": 43.94650650024414, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "logits/chosen": -2.9984076023101807, | |
| "logits/rejected": -3.114562511444092, | |
| "logps/chosen": -263.8316650390625, | |
| "logps/rejected": -340.1427307128906, | |
| "loss": 1.0427, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 0.6933631896972656, | |
| "rewards/margins": -0.14637237787246704, | |
| "rewards/rejected": 0.8397356271743774, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0008165070508785956, | |
| "grad_norm": 49.11759948730469, | |
| "learning_rate": 2.4e-05, | |
| "logits/chosen": -3.025609016418457, | |
| "logits/rejected": -3.0236239433288574, | |
| "logps/chosen": -412.2383117675781, | |
| "logps/rejected": -643.138427734375, | |
| "loss": 1.0374, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.1175220012664795, | |
| "rewards/margins": -0.31503716111183167, | |
| "rewards/rejected": 1.4325592517852783, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.000850528177998537, | |
| "grad_norm": 37.57383728027344, | |
| "learning_rate": 2.5e-05, | |
| "logits/chosen": -3.024618625640869, | |
| "logits/rejected": -3.123701572418213, | |
| "logps/chosen": -201.70530700683594, | |
| "logps/rejected": -224.99038696289062, | |
| "loss": 0.7477, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 0.9206753373146057, | |
| "rewards/margins": 0.5947309732437134, | |
| "rewards/rejected": 0.32594436407089233, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0008845493051184785, | |
| "grad_norm": 40.297794342041016, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "logits/chosen": -3.1100215911865234, | |
| "logits/rejected": -3.0195024013519287, | |
| "logps/chosen": -226.2320098876953, | |
| "logps/rejected": -193.0676727294922, | |
| "loss": 0.8151, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.9122087359428406, | |
| "rewards/margins": 0.0203973650932312, | |
| "rewards/rejected": 0.8918113708496094, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.00091857043223842, | |
| "grad_norm": 43.394229888916016, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "logits/chosen": -3.0187630653381348, | |
| "logits/rejected": -3.053678035736084, | |
| "logps/chosen": -153.02883911132812, | |
| "logps/rejected": -260.0540771484375, | |
| "loss": 1.3457, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.7145153284072876, | |
| "rewards/margins": -0.806611180305481, | |
| "rewards/rejected": 1.5211265087127686, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0009525915593583615, | |
| "grad_norm": 28.59680938720703, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "logits/chosen": -3.032409191131592, | |
| "logits/rejected": -3.0502161979675293, | |
| "logps/chosen": -240.1644287109375, | |
| "logps/rejected": -245.81436157226562, | |
| "loss": 0.5501, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.0826761722564697, | |
| "rewards/margins": 0.3433869779109955, | |
| "rewards/rejected": 1.7392891645431519, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.000986612686478303, | |
| "grad_norm": 37.069583892822266, | |
| "learning_rate": 2.9e-05, | |
| "logits/chosen": -3.0516104698181152, | |
| "logits/rejected": -3.144774913787842, | |
| "logps/chosen": -296.03125, | |
| "logps/rejected": -292.9052429199219, | |
| "loss": 1.1087, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.8324771523475647, | |
| "rewards/margins": -0.34529268741607666, | |
| "rewards/rejected": 1.1777698993682861, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0010206338135982445, | |
| "grad_norm": 30.238658905029297, | |
| "learning_rate": 3e-05, | |
| "logits/chosen": -3.064671039581299, | |
| "logits/rejected": -2.8963613510131836, | |
| "logps/chosen": -415.6085510253906, | |
| "logps/rejected": -149.9669952392578, | |
| "loss": 0.4892, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.0249770879745483, | |
| "rewards/margins": 0.7549355030059814, | |
| "rewards/rejected": 0.27004164457321167, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.001054654940718186, | |
| "grad_norm": 40.35402297973633, | |
| "learning_rate": 3.1e-05, | |
| "logits/chosen": -3.1762304306030273, | |
| "logits/rejected": -3.061676025390625, | |
| "logps/chosen": -585.9200439453125, | |
| "logps/rejected": -238.16307067871094, | |
| "loss": 0.587, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.726571559906006, | |
| "rewards/margins": 0.9808880090713501, | |
| "rewards/rejected": 1.7456837892532349, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0010886760678381275, | |
| "grad_norm": 33.78396987915039, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "logits/chosen": -3.05906081199646, | |
| "logits/rejected": -2.9525132179260254, | |
| "logps/chosen": -580.87646484375, | |
| "logps/rejected": -223.65457153320312, | |
| "loss": 0.8173, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.3172895908355713, | |
| "rewards/margins": 0.07103176414966583, | |
| "rewards/rejected": 1.2462577819824219, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.001122697194958069, | |
| "grad_norm": 40.97666549682617, | |
| "learning_rate": 3.3e-05, | |
| "logits/chosen": -3.084476947784424, | |
| "logits/rejected": -3.11270809173584, | |
| "logps/chosen": -285.040283203125, | |
| "logps/rejected": -251.92686462402344, | |
| "loss": 1.0645, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.4826446771621704, | |
| "rewards/margins": -0.13718131184577942, | |
| "rewards/rejected": 1.6198259592056274, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0011567183220780105, | |
| "grad_norm": 33.32075881958008, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "logits/chosen": -3.037482500076294, | |
| "logits/rejected": -3.0169787406921387, | |
| "logps/chosen": -467.7403869628906, | |
| "logps/rejected": -308.62274169921875, | |
| "loss": 0.5226, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.4038407802581787, | |
| "rewards/margins": 0.8624236583709717, | |
| "rewards/rejected": 1.541416883468628, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.001190739449197952, | |
| "grad_norm": 26.35866928100586, | |
| "learning_rate": 3.5e-05, | |
| "logits/chosen": -2.969816207885742, | |
| "logits/rejected": -2.9069106578826904, | |
| "logps/chosen": -312.856689453125, | |
| "logps/rejected": -183.8594512939453, | |
| "loss": 0.4495, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.1757181882858276, | |
| "rewards/margins": 0.803225576877594, | |
| "rewards/rejected": 0.37249261140823364, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0012247605763178934, | |
| "grad_norm": 44.118099212646484, | |
| "learning_rate": 3.6e-05, | |
| "logits/chosen": -3.091865301132202, | |
| "logits/rejected": -3.0135951042175293, | |
| "logps/chosen": -724.1255493164062, | |
| "logps/rejected": -335.2945861816406, | |
| "loss": 0.6341, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.7523369789123535, | |
| "rewards/margins": 0.3912923336029053, | |
| "rewards/rejected": 1.3610445261001587, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.001258781703437835, | |
| "grad_norm": 26.80854606628418, | |
| "learning_rate": 3.7e-05, | |
| "logits/chosen": -3.0334794521331787, | |
| "logits/rejected": -2.9878833293914795, | |
| "logps/chosen": -406.20745849609375, | |
| "logps/rejected": -377.2811584472656, | |
| "loss": 0.5577, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.6232261657714844, | |
| "rewards/margins": 0.9015346765518188, | |
| "rewards/rejected": 1.7216914892196655, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0012928028305577764, | |
| "grad_norm": 31.345367431640625, | |
| "learning_rate": 3.8e-05, | |
| "logits/chosen": -3.0986504554748535, | |
| "logits/rejected": -3.062534809112549, | |
| "logps/chosen": -412.54498291015625, | |
| "logps/rejected": -386.44158935546875, | |
| "loss": 0.5479, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.985120177268982, | |
| "rewards/margins": 0.8609979748725891, | |
| "rewards/rejected": 1.1241222620010376, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.001326823957677718, | |
| "grad_norm": 15.26794719696045, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "logits/chosen": -3.0730984210968018, | |
| "logits/rejected": -3.0951075553894043, | |
| "logps/chosen": -305.12982177734375, | |
| "logps/rejected": -285.45001220703125, | |
| "loss": 0.2547, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.55714750289917, | |
| "rewards/margins": 1.3561341762542725, | |
| "rewards/rejected": 1.2010130882263184, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0013608450847976594, | |
| "grad_norm": 22.881174087524414, | |
| "learning_rate": 4e-05, | |
| "logits/chosen": -3.162843942642212, | |
| "logits/rejected": -3.15628981590271, | |
| "logps/chosen": -277.48626708984375, | |
| "logps/rejected": -219.9507293701172, | |
| "loss": 0.5427, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.6990911960601807, | |
| "rewards/margins": 0.5041837692260742, | |
| "rewards/rejected": 2.1949074268341064, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.001394866211917601, | |
| "grad_norm": 15.179062843322754, | |
| "learning_rate": 4.1e-05, | |
| "logits/chosen": -3.138144016265869, | |
| "logits/rejected": -3.1033153533935547, | |
| "logps/chosen": -234.6917724609375, | |
| "logps/rejected": -141.10888671875, | |
| "loss": 0.2626, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.799344539642334, | |
| "rewards/margins": 1.4650464057922363, | |
| "rewards/rejected": 1.3342978954315186, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0014288873390375424, | |
| "grad_norm": 21.09362030029297, | |
| "learning_rate": 4.2e-05, | |
| "logits/chosen": -3.0385375022888184, | |
| "logits/rejected": -3.083676338195801, | |
| "logps/chosen": -168.33697509765625, | |
| "logps/rejected": -179.30813598632812, | |
| "loss": 0.4002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 2.3925881385803223, | |
| "rewards/margins": 0.8477088809013367, | |
| "rewards/rejected": 1.5448791980743408, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.0014629084661574839, | |
| "grad_norm": 38.255069732666016, | |
| "learning_rate": 4.3e-05, | |
| "logits/chosen": -3.080360174179077, | |
| "logits/rejected": -3.029470443725586, | |
| "logps/chosen": -267.5482177734375, | |
| "logps/rejected": -161.17111206054688, | |
| "loss": 0.8857, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.8575999736785889, | |
| "rewards/margins": -0.24116593599319458, | |
| "rewards/rejected": 2.0987658500671387, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.0014969295932774254, | |
| "grad_norm": 27.97095489501953, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "logits/chosen": -3.1322476863861084, | |
| "logits/rejected": -3.1334149837493896, | |
| "logps/chosen": -94.13690185546875, | |
| "logps/rejected": -78.98738098144531, | |
| "loss": 0.9987, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.6137428283691406, | |
| "rewards/margins": -0.4127828776836395, | |
| "rewards/rejected": 2.0265257358551025, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0015309507203973669, | |
| "grad_norm": 45.25028991699219, | |
| "learning_rate": 4.5e-05, | |
| "logits/chosen": -3.121274709701538, | |
| "logits/rejected": -3.1236798763275146, | |
| "logps/chosen": -516.3858642578125, | |
| "logps/rejected": -212.694580078125, | |
| "loss": 0.9457, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 2.419821262359619, | |
| "rewards/margins": -0.38284340500831604, | |
| "rewards/rejected": 2.8026647567749023, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0015649718475173081, | |
| "grad_norm": 23.58992576599121, | |
| "learning_rate": 4.600000000000001e-05, | |
| "logits/chosen": -3.113922357559204, | |
| "logits/rejected": -3.0506935119628906, | |
| "logps/chosen": -318.288330078125, | |
| "logps/rejected": -124.70735168457031, | |
| "loss": 0.5671, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.407118320465088, | |
| "rewards/margins": 0.30923187732696533, | |
| "rewards/rejected": 2.097886323928833, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.0015989929746372496, | |
| "grad_norm": 36.93040084838867, | |
| "learning_rate": 4.7e-05, | |
| "logits/chosen": -3.079291582107544, | |
| "logits/rejected": -3.1158454418182373, | |
| "logps/chosen": -246.26065063476562, | |
| "logps/rejected": -336.46649169921875, | |
| "loss": 0.9352, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 2.214073657989502, | |
| "rewards/margins": -0.2623659074306488, | |
| "rewards/rejected": 2.4764394760131836, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.0016330141017571911, | |
| "grad_norm": 32.5983772277832, | |
| "learning_rate": 4.8e-05, | |
| "logits/chosen": -3.155709981918335, | |
| "logits/rejected": -3.1829142570495605, | |
| "logps/chosen": -204.55328369140625, | |
| "logps/rejected": -219.52647399902344, | |
| "loss": 0.8805, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.155839681625366, | |
| "rewards/margins": -0.002133697271347046, | |
| "rewards/rejected": 2.157973289489746, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.0016670352288771326, | |
| "grad_norm": 35.626869201660156, | |
| "learning_rate": 4.9e-05, | |
| "logits/chosen": -3.10532546043396, | |
| "logits/rejected": -3.113755464553833, | |
| "logps/chosen": -390.5479736328125, | |
| "logps/rejected": -351.7720031738281, | |
| "loss": 0.7657, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.5670056343078613, | |
| "rewards/margins": -0.04349440336227417, | |
| "rewards/rejected": 2.6105000972747803, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.001701056355997074, | |
| "grad_norm": 37.51042938232422, | |
| "learning_rate": 5e-05, | |
| "logits/chosen": -3.122190237045288, | |
| "logits/rejected": -3.0837321281433105, | |
| "logps/chosen": -251.68600463867188, | |
| "logps/rejected": -242.4346923828125, | |
| "loss": 0.9539, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.0532169342041016, | |
| "rewards/margins": -0.32518482208251953, | |
| "rewards/rejected": 2.378401756286621, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0017350774831170156, | |
| "grad_norm": 43.16423416137695, | |
| "learning_rate": 4.999939076763487e-05, | |
| "logits/chosen": -3.0424001216888428, | |
| "logits/rejected": -3.0246453285217285, | |
| "logps/chosen": -228.8073272705078, | |
| "logps/rejected": -399.57781982421875, | |
| "loss": 1.204, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.1423535346984863, | |
| "rewards/margins": -0.628890335559845, | |
| "rewards/rejected": 2.7712438106536865, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.001769098610236957, | |
| "grad_norm": 32.16966247558594, | |
| "learning_rate": 4.999756310023261e-05, | |
| "logits/chosen": -3.049051284790039, | |
| "logits/rejected": -3.061192750930786, | |
| "logps/chosen": -219.27503967285156, | |
| "logps/rejected": -238.98226928710938, | |
| "loss": 0.9616, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.1868481636047363, | |
| "rewards/margins": -0.2618050277233124, | |
| "rewards/rejected": 2.448652744293213, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0018031197373568986, | |
| "grad_norm": 38.5218505859375, | |
| "learning_rate": 4.999451708687114e-05, | |
| "logits/chosen": -3.088409185409546, | |
| "logits/rejected": -3.0845394134521484, | |
| "logps/chosen": -131.70074462890625, | |
| "logps/rejected": -205.93624877929688, | |
| "loss": 1.0419, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.822286605834961, | |
| "rewards/margins": -0.18211710453033447, | |
| "rewards/rejected": 2.004403829574585, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.00183714086447684, | |
| "grad_norm": 31.46576690673828, | |
| "learning_rate": 4.999025287600886e-05, | |
| "logits/chosen": -3.143968105316162, | |
| "logits/rejected": -3.032649278640747, | |
| "logps/chosen": -403.4815673828125, | |
| "logps/rejected": -195.69281005859375, | |
| "loss": 0.567, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.2466557025909424, | |
| "rewards/margins": 0.6339659094810486, | |
| "rewards/rejected": 1.612689733505249, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0018711619915967816, | |
| "grad_norm": 22.339370727539062, | |
| "learning_rate": 4.99847706754774e-05, | |
| "logits/chosen": -3.1364409923553467, | |
| "logits/rejected": -3.0678646564483643, | |
| "logps/chosen": -272.07330322265625, | |
| "logps/rejected": -94.81082153320312, | |
| "loss": 0.4856, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 1.9992039203643799, | |
| "rewards/margins": 0.5075141191482544, | |
| "rewards/rejected": 1.491689682006836, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.001905183118716723, | |
| "grad_norm": 38.88309860229492, | |
| "learning_rate": 4.997807075247146e-05, | |
| "logits/chosen": -3.0754241943359375, | |
| "logits/rejected": -3.1445722579956055, | |
| "logps/chosen": -161.1916961669922, | |
| "logps/rejected": -230.43589782714844, | |
| "loss": 0.8925, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.070554256439209, | |
| "rewards/margins": -0.027551889419555664, | |
| "rewards/rejected": 2.0981061458587646, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.0019392042458366645, | |
| "grad_norm": 29.006982803344727, | |
| "learning_rate": 4.997015343353585e-05, | |
| "logits/chosen": -3.079509973526001, | |
| "logits/rejected": -3.119426727294922, | |
| "logps/chosen": -236.40664672851562, | |
| "logps/rejected": -259.257568359375, | |
| "loss": 0.5812, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.4261646270751953, | |
| "rewards/margins": 0.7371253371238708, | |
| "rewards/rejected": 1.6890392303466797, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.001973225372956606, | |
| "grad_norm": 18.193532943725586, | |
| "learning_rate": 4.996101910454953e-05, | |
| "logits/chosen": -3.070995569229126, | |
| "logits/rejected": -2.8317549228668213, | |
| "logps/chosen": -730.03759765625, | |
| "logps/rejected": -133.17929077148438, | |
| "loss": 0.3401, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 3.711276054382324, | |
| "rewards/margins": 1.7915010452270508, | |
| "rewards/rejected": 1.9197750091552734, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.0020072465000765477, | |
| "grad_norm": 35.50777053833008, | |
| "learning_rate": 4.995066821070679e-05, | |
| "logits/chosen": -3.093824863433838, | |
| "logits/rejected": -3.083796739578247, | |
| "logps/chosen": -238.11883544921875, | |
| "logps/rejected": -151.69790649414062, | |
| "loss": 0.6251, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.0124969482421875, | |
| "rewards/margins": 0.640554666519165, | |
| "rewards/rejected": 1.371942400932312, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.002041267627196489, | |
| "grad_norm": 40.34638977050781, | |
| "learning_rate": 4.993910125649561e-05, | |
| "logits/chosen": -3.1637063026428223, | |
| "logits/rejected": -3.055285930633545, | |
| "logps/chosen": -365.89434814453125, | |
| "logps/rejected": -187.11392211914062, | |
| "loss": 1.1514, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.089510679244995, | |
| "rewards/margins": -0.3508464992046356, | |
| "rewards/rejected": 2.440357208251953, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0020752887543164307, | |
| "grad_norm": 31.27332878112793, | |
| "learning_rate": 4.992631880567301e-05, | |
| "logits/chosen": -2.9931535720825195, | |
| "logits/rejected": -2.994088649749756, | |
| "logps/chosen": -127.87042236328125, | |
| "logps/rejected": -236.82327270507812, | |
| "loss": 0.8745, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.9324662685394287, | |
| "rewards/margins": -0.040864601731300354, | |
| "rewards/rejected": 1.9733309745788574, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.002109309881436372, | |
| "grad_norm": 24.1131591796875, | |
| "learning_rate": 4.991232148123761e-05, | |
| "logits/chosen": -3.094635009765625, | |
| "logits/rejected": -3.060533046722412, | |
| "logps/chosen": -369.5483093261719, | |
| "logps/rejected": -167.15155029296875, | |
| "loss": 0.46, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.38918399810791, | |
| "rewards/margins": 1.1014902591705322, | |
| "rewards/rejected": 1.2876936197280884, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.0021433310085563133, | |
| "grad_norm": 49.346961975097656, | |
| "learning_rate": 4.989710996539926e-05, | |
| "logits/chosen": -3.1134653091430664, | |
| "logits/rejected": -3.0323257446289062, | |
| "logps/chosen": -273.39361572265625, | |
| "logps/rejected": -157.84652709960938, | |
| "loss": 1.023, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 1.1458964347839355, | |
| "rewards/margins": -0.5081260204315186, | |
| "rewards/rejected": 1.654022455215454, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.002177352135676255, | |
| "grad_norm": 41.145633697509766, | |
| "learning_rate": 4.988068499954578e-05, | |
| "logits/chosen": -2.9834160804748535, | |
| "logits/rejected": -2.9958994388580322, | |
| "logps/chosen": -334.5246887207031, | |
| "logps/rejected": -386.3387145996094, | |
| "loss": 1.5388, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.305959939956665, | |
| "rewards/margins": 0.07403802871704102, | |
| "rewards/rejected": 2.231921911239624, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0022113732627961962, | |
| "grad_norm": 33.07326126098633, | |
| "learning_rate": 4.9863047384206835e-05, | |
| "logits/chosen": -3.0006370544433594, | |
| "logits/rejected": -3.020918846130371, | |
| "logps/chosen": -344.7704162597656, | |
| "logps/rejected": -200.45114135742188, | |
| "loss": 0.5845, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.643819808959961, | |
| "rewards/margins": 0.6079027652740479, | |
| "rewards/rejected": 1.035917043685913, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.002245394389916138, | |
| "grad_norm": 51.295345306396484, | |
| "learning_rate": 4.984419797901491e-05, | |
| "logits/chosen": -3.039350986480713, | |
| "logits/rejected": -3.030729293823242, | |
| "logps/chosen": -307.3952331542969, | |
| "logps/rejected": -238.56224060058594, | |
| "loss": 1.2331, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 0.6686372756958008, | |
| "rewards/margins": -0.6910279989242554, | |
| "rewards/rejected": 1.3596653938293457, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0022794155170360792, | |
| "grad_norm": 27.732166290283203, | |
| "learning_rate": 4.982413770266342e-05, | |
| "logits/chosen": -3.092078685760498, | |
| "logits/rejected": -3.1104204654693604, | |
| "logps/chosen": -206.7743682861328, | |
| "logps/rejected": -112.66790771484375, | |
| "loss": 0.5552, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.273049235343933, | |
| "rewards/margins": 0.48671725392341614, | |
| "rewards/rejected": 0.7863319516181946, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.002313436644156021, | |
| "grad_norm": 46.03068161010742, | |
| "learning_rate": 4.980286753286195e-05, | |
| "logits/chosen": -2.9219119548797607, | |
| "logits/rejected": -2.973864793777466, | |
| "logps/chosen": -455.3492431640625, | |
| "logps/rejected": -422.7337646484375, | |
| "loss": 0.8967, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.788267731666565, | |
| "rewards/margins": -0.24189908802509308, | |
| "rewards/rejected": 2.0301668643951416, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.002347457771275962, | |
| "grad_norm": 24.625539779663086, | |
| "learning_rate": 4.978038850628854e-05, | |
| "logits/chosen": -2.9824914932250977, | |
| "logits/rejected": -2.923306703567505, | |
| "logps/chosen": -320.2785339355469, | |
| "logps/rejected": -199.95407104492188, | |
| "loss": 0.5542, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.4030070304870605, | |
| "rewards/margins": 1.2227776050567627, | |
| "rewards/rejected": 1.180229663848877, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.002381478898395904, | |
| "grad_norm": 52.87644958496094, | |
| "learning_rate": 4.975670171853926e-05, | |
| "logits/chosen": -2.9576680660247803, | |
| "logits/rejected": -3.113696336746216, | |
| "logps/chosen": -226.1188201904297, | |
| "logps/rejected": -472.0656433105469, | |
| "loss": 1.5976, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 1.8890628814697266, | |
| "rewards/margins": -1.2377982139587402, | |
| "rewards/rejected": 3.126861095428467, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.002415500025515845, | |
| "grad_norm": 41.89684295654297, | |
| "learning_rate": 4.9731808324074717e-05, | |
| "logits/chosen": -2.963118553161621, | |
| "logits/rejected": -2.95782732963562, | |
| "logps/chosen": -376.9123229980469, | |
| "logps/rejected": -408.6007080078125, | |
| "loss": 0.8305, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.2748968601226807, | |
| "rewards/margins": 1.4728538990020752, | |
| "rewards/rejected": 0.8020429611206055, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.002449521152635787, | |
| "grad_norm": 33.66620635986328, | |
| "learning_rate": 4.9705709536163824e-05, | |
| "logits/chosen": -2.951019287109375, | |
| "logits/rejected": -2.8525712490081787, | |
| "logps/chosen": -334.75689697265625, | |
| "logps/rejected": -256.43963623046875, | |
| "loss": 0.5852, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 2.2168543338775635, | |
| "rewards/margins": 1.3764129877090454, | |
| "rewards/rejected": 0.8404413461685181, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.002483542279755728, | |
| "grad_norm": 48.537376403808594, | |
| "learning_rate": 4.96784066268247e-05, | |
| "logits/chosen": -2.96649169921875, | |
| "logits/rejected": -2.877800703048706, | |
| "logps/chosen": -342.9018859863281, | |
| "logps/rejected": -166.73826599121094, | |
| "loss": 1.3147, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 1.4513309001922607, | |
| "rewards/margins": -0.9881281852722168, | |
| "rewards/rejected": 2.4394590854644775, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.00251756340687567, | |
| "grad_norm": 35.8758544921875, | |
| "learning_rate": 4.964990092676263e-05, | |
| "logits/chosen": -3.007376194000244, | |
| "logits/rejected": -3.0710248947143555, | |
| "logps/chosen": -322.75604248046875, | |
| "logps/rejected": -288.9136047363281, | |
| "loss": 1.0483, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.328976035118103, | |
| "rewards/margins": -0.057621777057647705, | |
| "rewards/rejected": 1.3865978717803955, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.002551584533995611, | |
| "grad_norm": 52.42377471923828, | |
| "learning_rate": 4.962019382530521e-05, | |
| "logits/chosen": -2.952801465988159, | |
| "logits/rejected": -2.9580321311950684, | |
| "logps/chosen": -272.15576171875, | |
| "logps/rejected": -230.2147674560547, | |
| "loss": 1.4341, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 1.2654911279678345, | |
| "rewards/margins": -0.9303737282752991, | |
| "rewards/rejected": 2.1958649158477783, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.002585605661115553, | |
| "grad_norm": 54.157386779785156, | |
| "learning_rate": 4.9589286770334654e-05, | |
| "logits/chosen": -2.997375965118408, | |
| "logits/rejected": -3.091611385345459, | |
| "logps/chosen": -219.73159790039062, | |
| "logps/rejected": -503.797119140625, | |
| "loss": 1.3568, | |
| "rewards/accuracies": 0.25, | |
| "rewards/chosen": 1.520772933959961, | |
| "rewards/margins": -0.4154503643512726, | |
| "rewards/rejected": 1.9362232685089111, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.002619626788235494, | |
| "grad_norm": 40.647274017333984, | |
| "learning_rate": 4.9557181268217227e-05, | |
| "logits/chosen": -3.0128493309020996, | |
| "logits/rejected": -2.9981260299682617, | |
| "logps/chosen": -319.1977233886719, | |
| "logps/rejected": -250.11123657226562, | |
| "loss": 0.7922, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 1.9633705615997314, | |
| "rewards/margins": -0.027412444353103638, | |
| "rewards/rejected": 1.9907830953598022, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.002653647915355436, | |
| "grad_norm": 46.99341583251953, | |
| "learning_rate": 4.952387888372979e-05, | |
| "logits/chosen": -3.0307843685150146, | |
| "logits/rejected": -3.070122480392456, | |
| "logps/chosen": -134.52757263183594, | |
| "logps/rejected": -225.51242065429688, | |
| "loss": 1.182, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 1.2856706380844116, | |
| "rewards/margins": -0.7930065393447876, | |
| "rewards/rejected": 2.078677177429199, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.002687669042475377, | |
| "grad_norm": 26.92460823059082, | |
| "learning_rate": 4.94893812399836e-05, | |
| "logits/chosen": -2.995112895965576, | |
| "logits/rejected": -2.950528144836426, | |
| "logps/chosen": -560.2005615234375, | |
| "logps/rejected": -383.92523193359375, | |
| "loss": 0.5477, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 1.8620712757110596, | |
| "rewards/margins": 0.3632301390171051, | |
| "rewards/rejected": 1.4988411664962769, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.002721690169595319, | |
| "grad_norm": 18.80535125732422, | |
| "learning_rate": 4.9453690018345144e-05, | |
| "logits/chosen": -3.0887956619262695, | |
| "logits/rejected": -3.013925790786743, | |
| "logps/chosen": -201.7890167236328, | |
| "logps/rejected": -120.17092895507812, | |
| "loss": 0.6049, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 2.207493305206299, | |
| "rewards/margins": 0.5624717473983765, | |
| "rewards/rejected": 1.645021677017212, | |
| "step": 80 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |