| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984301412872841, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0020931449502878076, | |
| "grad_norm": 9.686688861295458, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -2.7144877910614014, | |
| "logits/rejected": -2.6456987857818604, | |
| "logps/chosen": -307.5989990234375, | |
| "logps/rejected": -353.280517578125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020931449502878074, | |
| "grad_norm": 8.910692539294239, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.464939832687378, | |
| "logits/rejected": -2.4213643074035645, | |
| "logps/chosen": -299.6020812988281, | |
| "logps/rejected": -269.6002502441406, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4027777910232544, | |
| "rewards/chosen": -0.0003792633942794055, | |
| "rewards/margins": -0.000643148785457015, | |
| "rewards/rejected": 0.00026388533296994865, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04186289900575615, | |
| "grad_norm": 7.998647811770768, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.5575523376464844, | |
| "logits/rejected": -2.4738364219665527, | |
| "logps/chosen": -300.9610290527344, | |
| "logps/rejected": -276.1901550292969, | |
| "loss": 0.6924, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.003328660037368536, | |
| "rewards/margins": 0.0013169770827516913, | |
| "rewards/rejected": 0.0020116830710321665, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 8.21834105065241, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.532283306121826, | |
| "logits/rejected": -2.4365248680114746, | |
| "logps/chosen": -283.2300109863281, | |
| "logps/rejected": -240.62158203125, | |
| "loss": 0.6892, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": 0.024711979553103447, | |
| "rewards/margins": 0.00843757577240467, | |
| "rewards/rejected": 0.016274403780698776, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0837257980115123, | |
| "grad_norm": 8.23700971574819, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.4243757724761963, | |
| "logits/rejected": -2.3694825172424316, | |
| "logps/chosen": -255.69534301757812, | |
| "logps/rejected": -253.03646850585938, | |
| "loss": 0.6803, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.051530830562114716, | |
| "rewards/margins": 0.027234002947807312, | |
| "rewards/rejected": 0.024296827614307404, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10465724751439037, | |
| "grad_norm": 8.490471908136655, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.4612836837768555, | |
| "logits/rejected": -2.366600275039673, | |
| "logps/chosen": -266.41839599609375, | |
| "logps/rejected": -266.593017578125, | |
| "loss": 0.6671, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": 0.05303896218538284, | |
| "rewards/margins": 0.05404331535100937, | |
| "rewards/rejected": -0.0010043552611023188, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 11.448244491004864, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.5066006183624268, | |
| "logits/rejected": -2.4322540760040283, | |
| "logps/chosen": -249.62960815429688, | |
| "logps/rejected": -252.5321807861328, | |
| "loss": 0.6449, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.015312853269279003, | |
| "rewards/margins": 0.12173323333263397, | |
| "rewards/rejected": -0.13704606890678406, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14652014652014653, | |
| "grad_norm": 12.222640628288522, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.4458515644073486, | |
| "logits/rejected": -2.4075169563293457, | |
| "logps/chosen": -304.5162048339844, | |
| "logps/rejected": -312.8282775878906, | |
| "loss": 0.6171, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.16042451560497284, | |
| "rewards/margins": 0.20845112204551697, | |
| "rewards/rejected": -0.368875652551651, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1674515960230246, | |
| "grad_norm": 13.355609497034044, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.5683891773223877, | |
| "logits/rejected": -2.4449405670166016, | |
| "logps/chosen": -335.45166015625, | |
| "logps/rejected": -294.0103454589844, | |
| "loss": 0.6038, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.12419579923152924, | |
| "rewards/margins": 0.2938229441642761, | |
| "rewards/rejected": -0.41801875829696655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 16.831516529666843, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.5068366527557373, | |
| "logits/rejected": -2.4493064880371094, | |
| "logps/chosen": -292.50592041015625, | |
| "logps/rejected": -333.88446044921875, | |
| "loss": 0.5754, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.45342904329299927, | |
| "rewards/margins": 0.37515172362327576, | |
| "rewards/rejected": -0.8285807371139526, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20931449502878074, | |
| "grad_norm": 14.56937938894116, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.426504373550415, | |
| "logits/rejected": -2.359724521636963, | |
| "logps/chosen": -338.78021240234375, | |
| "logps/rejected": -367.85589599609375, | |
| "loss": 0.5745, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.47890257835388184, | |
| "rewards/margins": 0.4808070659637451, | |
| "rewards/rejected": -0.959709644317627, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2302459445316588, | |
| "grad_norm": 20.002857020935622, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.346848726272583, | |
| "logits/rejected": -2.258775472640991, | |
| "logps/chosen": -342.98046875, | |
| "logps/rejected": -371.9133605957031, | |
| "loss": 0.5817, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5712723731994629, | |
| "rewards/margins": 0.45406538248062134, | |
| "rewards/rejected": -1.02533757686615, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 20.278919360241712, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.105543613433838, | |
| "logits/rejected": -2.0793662071228027, | |
| "logps/chosen": -311.2692565917969, | |
| "logps/rejected": -360.46136474609375, | |
| "loss": 0.5614, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.4692445397377014, | |
| "rewards/margins": 0.5657081604003906, | |
| "rewards/rejected": -1.0349526405334473, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 15.445489902432293, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.0017788410186768, | |
| "logits/rejected": -1.9253467321395874, | |
| "logps/chosen": -359.1274719238281, | |
| "logps/rejected": -392.7835998535156, | |
| "loss": 0.5493, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.8196575045585632, | |
| "rewards/margins": 0.4658297002315521, | |
| "rewards/rejected": -1.285487174987793, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29304029304029305, | |
| "grad_norm": 23.527632580214693, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.021439790725708, | |
| "logits/rejected": -1.9034312963485718, | |
| "logps/chosen": -354.5030517578125, | |
| "logps/rejected": -370.05926513671875, | |
| "loss": 0.5458, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.44141221046447754, | |
| "rewards/margins": 0.6134065389633179, | |
| "rewards/rejected": -1.0548187494277954, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 20.969211081154093, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -1.808932900428772, | |
| "logits/rejected": -1.6896326541900635, | |
| "logps/chosen": -336.74200439453125, | |
| "logps/rejected": -351.64471435546875, | |
| "loss": 0.5563, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.6119992733001709, | |
| "rewards/margins": 0.46776071190834045, | |
| "rewards/rejected": -1.0797598361968994, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3349031920460492, | |
| "grad_norm": 16.819029625662488, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -1.6512082815170288, | |
| "logits/rejected": -1.4201633930206299, | |
| "logps/chosen": -328.9945373535156, | |
| "logps/rejected": -369.72216796875, | |
| "loss": 0.554, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.5794623494148254, | |
| "rewards/margins": 0.6330465078353882, | |
| "rewards/rejected": -1.2125087976455688, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35583464154892724, | |
| "grad_norm": 21.61579841438679, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -1.3276041746139526, | |
| "logits/rejected": -1.2030179500579834, | |
| "logps/chosen": -326.64642333984375, | |
| "logps/rejected": -355.4222106933594, | |
| "loss": 0.5522, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.7154276967048645, | |
| "rewards/margins": 0.5184686779975891, | |
| "rewards/rejected": -1.2338963747024536, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 23.330504007228544, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -1.4597452878952026, | |
| "logits/rejected": -1.335136890411377, | |
| "logps/chosen": -339.64739990234375, | |
| "logps/rejected": -370.96832275390625, | |
| "loss": 0.512, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7294771075248718, | |
| "rewards/margins": 0.5706949234008789, | |
| "rewards/rejected": -1.3001720905303955, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3976975405546834, | |
| "grad_norm": 20.457726835632624, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -1.406663179397583, | |
| "logits/rejected": -1.3718976974487305, | |
| "logps/chosen": -341.9677429199219, | |
| "logps/rejected": -390.94769287109375, | |
| "loss": 0.5365, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.8832741975784302, | |
| "rewards/margins": 0.5474223494529724, | |
| "rewards/rejected": -1.4306964874267578, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4186289900575615, | |
| "grad_norm": 45.24370174913431, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -1.4117847681045532, | |
| "logits/rejected": -1.3832786083221436, | |
| "logps/chosen": -319.09490966796875, | |
| "logps/rejected": -393.9739074707031, | |
| "loss": 0.5489, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.7460958957672119, | |
| "rewards/margins": 0.5401177406311035, | |
| "rewards/rejected": -1.286213755607605, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 21.103872725510552, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -1.6419851779937744, | |
| "logits/rejected": -1.4086859226226807, | |
| "logps/chosen": -358.5858154296875, | |
| "logps/rejected": -359.4571228027344, | |
| "loss": 0.5454, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.5668836832046509, | |
| "rewards/margins": 0.5663009285926819, | |
| "rewards/rejected": -1.1331846714019775, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4604918890633176, | |
| "grad_norm": 25.972160767293996, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -1.3075270652770996, | |
| "logits/rejected": -1.233934998512268, | |
| "logps/chosen": -309.5500793457031, | |
| "logps/rejected": -373.98223876953125, | |
| "loss": 0.5313, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7272627353668213, | |
| "rewards/margins": 0.6237983703613281, | |
| "rewards/rejected": -1.351061224937439, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48142333856619574, | |
| "grad_norm": 24.174538019841428, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -1.3718620538711548, | |
| "logits/rejected": -1.1529138088226318, | |
| "logps/chosen": -327.77801513671875, | |
| "logps/rejected": -397.93426513671875, | |
| "loss": 0.5333, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.7222844958305359, | |
| "rewards/margins": 0.8920642733573914, | |
| "rewards/rejected": -1.6143487691879272, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023547880690737, | |
| "grad_norm": 23.429710656248453, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -1.2215020656585693, | |
| "logits/rejected": -1.0080658197402954, | |
| "logps/chosen": -363.2150573730469, | |
| "logps/rejected": -406.48297119140625, | |
| "loss": 0.5051, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.8090720176696777, | |
| "rewards/margins": 0.6936684846878052, | |
| "rewards/rejected": -1.5027403831481934, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5232862375719518, | |
| "grad_norm": 21.330001354953637, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -1.1898741722106934, | |
| "logits/rejected": -0.9075394868850708, | |
| "logps/chosen": -360.69354248046875, | |
| "logps/rejected": -408.86773681640625, | |
| "loss": 0.5165, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.8794037103652954, | |
| "rewards/margins": 0.8506172299385071, | |
| "rewards/rejected": -1.7300208806991577, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54421768707483, | |
| "grad_norm": 27.49314240274799, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -1.4158910512924194, | |
| "logits/rejected": -1.3504579067230225, | |
| "logps/chosen": -341.8540344238281, | |
| "logps/rejected": -389.331787109375, | |
| "loss": 0.521, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.8279495239257812, | |
| "rewards/margins": 0.6607829332351685, | |
| "rewards/rejected": -1.4887326955795288, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.565149136577708, | |
| "grad_norm": 23.74859118455523, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -1.4965870380401611, | |
| "logits/rejected": -1.3825056552886963, | |
| "logps/chosen": -359.75787353515625, | |
| "logps/rejected": -420.7306213378906, | |
| "loss": 0.5209, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.7315458655357361, | |
| "rewards/margins": 0.853925347328186, | |
| "rewards/rejected": -1.5854711532592773, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5860805860805861, | |
| "grad_norm": 26.82116388601379, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -1.37690007686615, | |
| "logits/rejected": -1.2291252613067627, | |
| "logps/chosen": -337.71820068359375, | |
| "logps/rejected": -374.30706787109375, | |
| "loss": 0.5319, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.7968527674674988, | |
| "rewards/margins": 0.657835066318512, | |
| "rewards/rejected": -1.4546878337860107, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6070120355834642, | |
| "grad_norm": 22.28753319365226, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -1.51715087890625, | |
| "logits/rejected": -1.3852065801620483, | |
| "logps/chosen": -324.92340087890625, | |
| "logps/rejected": -390.6435241699219, | |
| "loss": 0.5218, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.8132198452949524, | |
| "rewards/margins": 0.717974066734314, | |
| "rewards/rejected": -1.5311939716339111, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6279434850863422, | |
| "grad_norm": 26.95577178592052, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -1.4670779705047607, | |
| "logits/rejected": -1.297649621963501, | |
| "logps/chosen": -403.04913330078125, | |
| "logps/rejected": -428.7928161621094, | |
| "loss": 0.5019, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.8886338472366333, | |
| "rewards/margins": 0.7050989270210266, | |
| "rewards/rejected": -1.5937328338623047, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6488749345892203, | |
| "grad_norm": 24.13195804406926, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -1.4607212543487549, | |
| "logits/rejected": -1.3081916570663452, | |
| "logps/chosen": -383.5493469238281, | |
| "logps/rejected": -418.50146484375, | |
| "loss": 0.5041, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.8673523664474487, | |
| "rewards/margins": 0.8634257316589355, | |
| "rewards/rejected": -1.7307780981063843, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6698063840920984, | |
| "grad_norm": 27.24325859713386, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -1.382027506828308, | |
| "logits/rejected": -1.1992334127426147, | |
| "logps/chosen": -328.2751770019531, | |
| "logps/rejected": -391.3571472167969, | |
| "loss": 0.4938, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.7451919317245483, | |
| "rewards/margins": 0.9388906359672546, | |
| "rewards/rejected": -1.6840826272964478, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6907378335949764, | |
| "grad_norm": 27.216778454048367, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -1.4924046993255615, | |
| "logits/rejected": -1.4578219652175903, | |
| "logps/chosen": -335.9745788574219, | |
| "logps/rejected": -404.7709655761719, | |
| "loss": 0.5135, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7176355123519897, | |
| "rewards/margins": 0.7227510213851929, | |
| "rewards/rejected": -1.4403865337371826, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7116692830978545, | |
| "grad_norm": 28.17952431515267, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -1.3802480697631836, | |
| "logits/rejected": -1.277980089187622, | |
| "logps/chosen": -361.58062744140625, | |
| "logps/rejected": -420.6078186035156, | |
| "loss": 0.507, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.7701286673545837, | |
| "rewards/margins": 0.8021227717399597, | |
| "rewards/rejected": -1.572251558303833, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7326007326007326, | |
| "grad_norm": 22.12939500514488, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -1.3996691703796387, | |
| "logits/rejected": -1.2072045803070068, | |
| "logps/chosen": -330.85455322265625, | |
| "logps/rejected": -388.41876220703125, | |
| "loss": 0.478, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.7518431544303894, | |
| "rewards/margins": 0.8218573331832886, | |
| "rewards/rejected": -1.5737005472183228, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7535321821036107, | |
| "grad_norm": 19.503424794677688, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -1.2794829607009888, | |
| "logits/rejected": -1.063652753829956, | |
| "logps/chosen": -383.0862121582031, | |
| "logps/rejected": -429.30682373046875, | |
| "loss": 0.5281, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.8063914179801941, | |
| "rewards/margins": 0.8900289535522461, | |
| "rewards/rejected": -1.696420431137085, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7744636316064888, | |
| "grad_norm": 26.907332636471214, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -1.2005360126495361, | |
| "logits/rejected": -1.0235365629196167, | |
| "logps/chosen": -340.9770202636719, | |
| "logps/rejected": -393.8902587890625, | |
| "loss": 0.5067, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.7764495015144348, | |
| "rewards/margins": 0.8340295553207397, | |
| "rewards/rejected": -1.6104789972305298, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7953950811093669, | |
| "grad_norm": 29.80515491184858, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -0.9497655034065247, | |
| "logits/rejected": -0.878097414970398, | |
| "logps/chosen": -346.8731994628906, | |
| "logps/rejected": -461.755126953125, | |
| "loss": 0.5097, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0337427854537964, | |
| "rewards/margins": 0.9731820225715637, | |
| "rewards/rejected": -2.006925106048584, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 24.841605469293125, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -1.209467887878418, | |
| "logits/rejected": -1.031672716140747, | |
| "logps/chosen": -429.25628662109375, | |
| "logps/rejected": -471.14501953125, | |
| "loss": 0.4938, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.9466843605041504, | |
| "rewards/margins": 0.7874827980995178, | |
| "rewards/rejected": -1.7341673374176025, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.837257980115123, | |
| "grad_norm": 30.69453014687722, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -1.067712664604187, | |
| "logits/rejected": -0.8794199228286743, | |
| "logps/chosen": -337.4797058105469, | |
| "logps/rejected": -408.54315185546875, | |
| "loss": 0.5135, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.8585729598999023, | |
| "rewards/margins": 0.849725604057312, | |
| "rewards/rejected": -1.7082984447479248, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.858189429618001, | |
| "grad_norm": 59.96090300586583, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -1.267378807067871, | |
| "logits/rejected": -1.0611450672149658, | |
| "logps/chosen": -386.88427734375, | |
| "logps/rejected": -436.865234375, | |
| "loss": 0.5168, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.9016349911689758, | |
| "rewards/margins": 0.8452868461608887, | |
| "rewards/rejected": -1.7469215393066406, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 28.101172543050826, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -1.0743638277053833, | |
| "logits/rejected": -0.9579585790634155, | |
| "logps/chosen": -352.7454528808594, | |
| "logps/rejected": -439.5741271972656, | |
| "loss": 0.5045, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.0042593479156494, | |
| "rewards/margins": 0.8400126695632935, | |
| "rewards/rejected": -1.8442720174789429, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9000523286237572, | |
| "grad_norm": 25.50390969555994, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -1.0834466218948364, | |
| "logits/rejected": -0.8814092874526978, | |
| "logps/chosen": -379.11468505859375, | |
| "logps/rejected": -440.3974609375, | |
| "loss": 0.4848, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.9338687062263489, | |
| "rewards/margins": 0.8329343795776367, | |
| "rewards/rejected": -1.7668030261993408, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9209837781266352, | |
| "grad_norm": 24.719240527877194, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -1.2495243549346924, | |
| "logits/rejected": -1.0566151142120361, | |
| "logps/chosen": -346.22894287109375, | |
| "logps/rejected": -386.73565673828125, | |
| "loss": 0.4991, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.8856858015060425, | |
| "rewards/margins": 0.7673971056938171, | |
| "rewards/rejected": -1.653083086013794, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9419152276295133, | |
| "grad_norm": 27.94864423555306, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -0.9761242866516113, | |
| "logits/rejected": -0.8453343510627747, | |
| "logps/chosen": -365.3951416015625, | |
| "logps/rejected": -443.78924560546875, | |
| "loss": 0.5018, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.9703672528266907, | |
| "rewards/margins": 0.9006422162055969, | |
| "rewards/rejected": -1.8710094690322876, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9628466771323915, | |
| "grad_norm": 44.89769499800862, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -1.1586068868637085, | |
| "logits/rejected": -0.8781611323356628, | |
| "logps/chosen": -361.3426818847656, | |
| "logps/rejected": -399.6344299316406, | |
| "loss": 0.5109, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.9305132031440735, | |
| "rewards/margins": 0.8112867474555969, | |
| "rewards/rejected": -1.7417999505996704, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9837781266352695, | |
| "grad_norm": 22.89147164994492, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -1.140649676322937, | |
| "logits/rejected": -0.9729598164558411, | |
| "logps/chosen": -378.35162353515625, | |
| "logps/rejected": -425.26666259765625, | |
| "loss": 0.4908, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.828622043132782, | |
| "rewards/margins": 0.7690635919570923, | |
| "rewards/rejected": -1.59768545627594, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "eval_logits/chosen": -0.9970575571060181, | |
| "eval_logits/rejected": -0.8678091764450073, | |
| "eval_logps/chosen": -339.41973876953125, | |
| "eval_logps/rejected": -429.93389892578125, | |
| "eval_loss": 0.508111834526062, | |
| "eval_rewards/accuracies": 0.78125, | |
| "eval_rewards/chosen": -0.8088568449020386, | |
| "eval_rewards/margins": 0.9086635112762451, | |
| "eval_rewards/rejected": -1.7175202369689941, | |
| "eval_runtime": 167.8174, | |
| "eval_samples_per_second": 11.918, | |
| "eval_steps_per_second": 0.191, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5463379294117542, | |
| "train_runtime": 13584.2132, | |
| "train_samples_per_second": 4.5, | |
| "train_steps_per_second": 0.035 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |