{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 87414, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.005719907566293729, "grad_norm": 7.0386528968811035, "learning_rate": 4.971400462168532e-05, "loss": 7.2889, "step": 500 }, { "epoch": 0.011439815132587457, "grad_norm": 6.592615604400635, "learning_rate": 4.942800924337063e-05, "loss": 5.389, "step": 1000 }, { "epoch": 0.017159722698881188, "grad_norm": 6.088059425354004, "learning_rate": 4.914201386505594e-05, "loss": 4.8994, "step": 1500 }, { "epoch": 0.022879630265174915, "grad_norm": 6.355687618255615, "learning_rate": 4.885601848674126e-05, "loss": 4.6219, "step": 2000 }, { "epoch": 0.028599537831468645, "grad_norm": 6.1159186363220215, "learning_rate": 4.8570023108426574e-05, "loss": 4.4233, "step": 2500 }, { "epoch": 0.034319445397762376, "grad_norm": 6.636866092681885, "learning_rate": 4.828402773011189e-05, "loss": 4.2802, "step": 3000 }, { "epoch": 0.0400393529640561, "grad_norm": 5.87605619430542, "learning_rate": 4.7998032351797196e-05, "loss": 4.1453, "step": 3500 }, { "epoch": 0.04575926053034983, "grad_norm": 6.0966691970825195, "learning_rate": 4.771203697348251e-05, "loss": 4.0822, "step": 4000 }, { "epoch": 0.05147916809664356, "grad_norm": 5.99679708480835, "learning_rate": 4.7426041595167824e-05, "loss": 3.9941, "step": 4500 }, { "epoch": 0.05719907566293729, "grad_norm": 5.738614082336426, "learning_rate": 4.714004621685314e-05, "loss": 3.9104, "step": 5000 }, { "epoch": 0.06291898322923102, "grad_norm": 6.038717746734619, "learning_rate": 4.6854050838538446e-05, "loss": 3.8528, "step": 5500 }, { "epoch": 0.06863889079552475, "grad_norm": 5.92971134185791, "learning_rate": 4.656805546022377e-05, "loss": 3.78, "step": 6000 }, { "epoch": 0.07435879836181847, "grad_norm": 5.791173458099365, "learning_rate": 4.628206008190908e-05, "loss": 3.7407, "step": 6500 }, { "epoch": 0.0800787059281122, "grad_norm": 6.2113566398620605, "learning_rate": 4.5996064703594396e-05, "loss": 3.6834, "step": 7000 }, { "epoch": 0.08579861349440593, "grad_norm": 5.970207214355469, "learning_rate": 4.57100693252797e-05, "loss": 3.663, "step": 7500 }, { "epoch": 0.09151852106069966, "grad_norm": 5.8883514404296875, "learning_rate": 4.542407394696502e-05, "loss": 3.6161, "step": 8000 }, { "epoch": 0.09723842862699339, "grad_norm": 6.351154804229736, "learning_rate": 4.513807856865033e-05, "loss": 3.6193, "step": 8500 }, { "epoch": 0.10295833619328712, "grad_norm": 5.69071102142334, "learning_rate": 4.4852083190335646e-05, "loss": 3.545, "step": 9000 }, { "epoch": 0.10867824375958085, "grad_norm": 5.988426208496094, "learning_rate": 4.456608781202096e-05, "loss": 3.5088, "step": 9500 }, { "epoch": 0.11439815132587458, "grad_norm": 6.251087665557861, "learning_rate": 4.4280092433706274e-05, "loss": 3.5158, "step": 10000 }, { "epoch": 0.1201180588921683, "grad_norm": 5.840632438659668, "learning_rate": 4.399409705539159e-05, "loss": 3.5045, "step": 10500 }, { "epoch": 0.12583796645846204, "grad_norm": 5.9697723388671875, "learning_rate": 4.37081016770769e-05, "loss": 3.4466, "step": 11000 }, { "epoch": 0.13155787402475577, "grad_norm": 6.149275779724121, "learning_rate": 4.342210629876222e-05, "loss": 3.4544, "step": 11500 }, { "epoch": 0.1372777815910495, "grad_norm": 5.605130672454834, "learning_rate": 4.3136110920447525e-05, "loss": 3.4116, "step": 12000 }, { "epoch": 0.1429976891573432, "grad_norm": 5.5137152671813965, "learning_rate": 4.285011554213284e-05, "loss": 3.3919, "step": 12500 }, { "epoch": 0.14871759672363694, "grad_norm": 5.848428726196289, "learning_rate": 4.256412016381815e-05, "loss": 3.3695, "step": 13000 }, { "epoch": 0.15443750428993067, "grad_norm": 5.794093132019043, "learning_rate": 4.227812478550347e-05, "loss": 3.3509, "step": 13500 }, { "epoch": 0.1601574118562244, "grad_norm": 5.879131317138672, "learning_rate": 4.199212940718878e-05, "loss": 3.3551, "step": 14000 }, { "epoch": 0.16587731942251813, "grad_norm": 5.5501179695129395, "learning_rate": 4.1706134028874096e-05, "loss": 3.3327, "step": 14500 }, { "epoch": 0.17159722698881186, "grad_norm": 5.30273962020874, "learning_rate": 4.142013865055941e-05, "loss": 3.3045, "step": 15000 }, { "epoch": 0.1773171345551056, "grad_norm": 6.049214839935303, "learning_rate": 4.1134143272244725e-05, "loss": 3.2868, "step": 15500 }, { "epoch": 0.18303704212139932, "grad_norm": 5.499197483062744, "learning_rate": 4.084814789393003e-05, "loss": 3.2837, "step": 16000 }, { "epoch": 0.18875694968769305, "grad_norm": 5.448641777038574, "learning_rate": 4.0562152515615347e-05, "loss": 3.2751, "step": 16500 }, { "epoch": 0.19447685725398678, "grad_norm": 5.422529697418213, "learning_rate": 4.027615713730066e-05, "loss": 3.2585, "step": 17000 }, { "epoch": 0.2001967648202805, "grad_norm": 5.639166831970215, "learning_rate": 3.9990161758985975e-05, "loss": 3.2587, "step": 17500 }, { "epoch": 0.20591667238657424, "grad_norm": 5.255904674530029, "learning_rate": 3.970416638067129e-05, "loss": 3.2505, "step": 18000 }, { "epoch": 0.21163657995286797, "grad_norm": 5.453869819641113, "learning_rate": 3.9418171002356604e-05, "loss": 3.2382, "step": 18500 }, { "epoch": 0.2173564875191617, "grad_norm": 5.905749797821045, "learning_rate": 3.913217562404192e-05, "loss": 3.2309, "step": 19000 }, { "epoch": 0.22307639508545543, "grad_norm": 5.28553581237793, "learning_rate": 3.884618024572723e-05, "loss": 3.2308, "step": 19500 }, { "epoch": 0.22879630265174916, "grad_norm": 5.1942830085754395, "learning_rate": 3.8560184867412547e-05, "loss": 3.2163, "step": 20000 }, { "epoch": 0.23451621021804286, "grad_norm": 6.12723970413208, "learning_rate": 3.8274189489097854e-05, "loss": 3.2082, "step": 20500 }, { "epoch": 0.2402361177843366, "grad_norm": 5.664548873901367, "learning_rate": 3.798819411078317e-05, "loss": 3.1802, "step": 21000 }, { "epoch": 0.24595602535063033, "grad_norm": 5.903208255767822, "learning_rate": 3.770219873246848e-05, "loss": 3.1946, "step": 21500 }, { "epoch": 0.2516759329169241, "grad_norm": 5.729937553405762, "learning_rate": 3.7416203354153804e-05, "loss": 3.1646, "step": 22000 }, { "epoch": 0.2573958404832178, "grad_norm": 6.068752765655518, "learning_rate": 3.713020797583911e-05, "loss": 3.2006, "step": 22500 }, { "epoch": 0.26311574804951154, "grad_norm": 5.97099494934082, "learning_rate": 3.6844212597524425e-05, "loss": 3.1437, "step": 23000 }, { "epoch": 0.26883565561580525, "grad_norm": 5.777164936065674, "learning_rate": 3.655821721920974e-05, "loss": 3.1407, "step": 23500 }, { "epoch": 0.274555563182099, "grad_norm": 4.982606887817383, "learning_rate": 3.6272221840895054e-05, "loss": 3.1282, "step": 24000 }, { "epoch": 0.2802754707483927, "grad_norm": 5.495816230773926, "learning_rate": 3.598622646258037e-05, "loss": 3.1434, "step": 24500 }, { "epoch": 0.2859953783146864, "grad_norm": 5.898298263549805, "learning_rate": 3.5700231084265676e-05, "loss": 3.1446, "step": 25000 }, { "epoch": 0.29171528588098017, "grad_norm": 5.729229927062988, "learning_rate": 3.541423570595099e-05, "loss": 3.1254, "step": 25500 }, { "epoch": 0.29743519344727387, "grad_norm": 6.3333821296691895, "learning_rate": 3.512824032763631e-05, "loss": 3.1543, "step": 26000 }, { "epoch": 0.30315510101356763, "grad_norm": 6.0027756690979, "learning_rate": 3.4842244949321625e-05, "loss": 3.1379, "step": 26500 }, { "epoch": 0.30887500857986133, "grad_norm": 5.95717716217041, "learning_rate": 3.455624957100693e-05, "loss": 3.1188, "step": 27000 }, { "epoch": 0.3145949161461551, "grad_norm": 6.262216567993164, "learning_rate": 3.427025419269225e-05, "loss": 3.1096, "step": 27500 }, { "epoch": 0.3203148237124488, "grad_norm": 6.436416149139404, "learning_rate": 3.398425881437756e-05, "loss": 3.1294, "step": 28000 }, { "epoch": 0.32603473127874255, "grad_norm": 5.524046421051025, "learning_rate": 3.3698263436062876e-05, "loss": 3.0965, "step": 28500 }, { "epoch": 0.33175463884503625, "grad_norm": 5.237400531768799, "learning_rate": 3.341226805774818e-05, "loss": 3.0826, "step": 29000 }, { "epoch": 0.33747454641133, "grad_norm": 5.551352500915527, "learning_rate": 3.3126272679433504e-05, "loss": 3.0666, "step": 29500 }, { "epoch": 0.3431944539776237, "grad_norm": 5.407064914703369, "learning_rate": 3.284027730111882e-05, "loss": 3.0738, "step": 30000 }, { "epoch": 0.3489143615439175, "grad_norm": 5.428358554840088, "learning_rate": 3.255428192280413e-05, "loss": 3.0628, "step": 30500 }, { "epoch": 0.3546342691102112, "grad_norm": 6.178744792938232, "learning_rate": 3.226828654448944e-05, "loss": 3.0683, "step": 31000 }, { "epoch": 0.36035417667650493, "grad_norm": 5.695249080657959, "learning_rate": 3.1982291166174755e-05, "loss": 3.0732, "step": 31500 }, { "epoch": 0.36607408424279864, "grad_norm": 5.676379203796387, "learning_rate": 3.169629578786007e-05, "loss": 3.0531, "step": 32000 }, { "epoch": 0.37179399180909234, "grad_norm": 5.420720100402832, "learning_rate": 3.141030040954538e-05, "loss": 3.0472, "step": 32500 }, { "epoch": 0.3775138993753861, "grad_norm": 5.6645379066467285, "learning_rate": 3.11243050312307e-05, "loss": 3.0391, "step": 33000 }, { "epoch": 0.3832338069416798, "grad_norm": 6.123884201049805, "learning_rate": 3.083830965291601e-05, "loss": 3.0527, "step": 33500 }, { "epoch": 0.38895371450797356, "grad_norm": 5.331460475921631, "learning_rate": 3.0552314274601326e-05, "loss": 3.049, "step": 34000 }, { "epoch": 0.39467362207426726, "grad_norm": 6.356675624847412, "learning_rate": 3.0266318896286637e-05, "loss": 3.0416, "step": 34500 }, { "epoch": 0.400393529640561, "grad_norm": 6.067440986633301, "learning_rate": 2.9980323517971955e-05, "loss": 3.0303, "step": 35000 }, { "epoch": 0.4061134372068547, "grad_norm": 5.3762030601501465, "learning_rate": 2.9694328139657262e-05, "loss": 3.0363, "step": 35500 }, { "epoch": 0.4118333447731485, "grad_norm": 5.837817192077637, "learning_rate": 2.9408332761342576e-05, "loss": 3.0284, "step": 36000 }, { "epoch": 0.4175532523394422, "grad_norm": 5.484166622161865, "learning_rate": 2.912233738302789e-05, "loss": 2.9946, "step": 36500 }, { "epoch": 0.42327315990573594, "grad_norm": 5.776547908782959, "learning_rate": 2.883634200471321e-05, "loss": 3.037, "step": 37000 }, { "epoch": 0.42899306747202964, "grad_norm": 5.481433868408203, "learning_rate": 2.8550346626398516e-05, "loss": 3.0289, "step": 37500 }, { "epoch": 0.4347129750383234, "grad_norm": 5.784084320068359, "learning_rate": 2.826435124808383e-05, "loss": 2.9928, "step": 38000 }, { "epoch": 0.4404328826046171, "grad_norm": 5.899621486663818, "learning_rate": 2.7978355869769148e-05, "loss": 3.0009, "step": 38500 }, { "epoch": 0.44615279017091086, "grad_norm": 5.488452434539795, "learning_rate": 2.7692360491454462e-05, "loss": 3.0099, "step": 39000 }, { "epoch": 0.45187269773720457, "grad_norm": 5.848759174346924, "learning_rate": 2.740636511313977e-05, "loss": 3.0009, "step": 39500 }, { "epoch": 0.4575926053034983, "grad_norm": 5.612068176269531, "learning_rate": 2.7120369734825084e-05, "loss": 2.9828, "step": 40000 }, { "epoch": 0.463312512869792, "grad_norm": 5.79826021194458, "learning_rate": 2.68343743565104e-05, "loss": 3.0152, "step": 40500 }, { "epoch": 0.46903242043608573, "grad_norm": 5.842123508453369, "learning_rate": 2.6548378978195716e-05, "loss": 2.9719, "step": 41000 }, { "epoch": 0.4747523280023795, "grad_norm": 5.69782018661499, "learning_rate": 2.626238359988103e-05, "loss": 2.975, "step": 41500 }, { "epoch": 0.4804722355686732, "grad_norm": 6.189919948577881, "learning_rate": 2.5976388221566338e-05, "loss": 2.9709, "step": 42000 }, { "epoch": 0.48619214313496695, "grad_norm": 5.761579513549805, "learning_rate": 2.5690392843251655e-05, "loss": 2.9583, "step": 42500 }, { "epoch": 0.49191205070126065, "grad_norm": 6.164900779724121, "learning_rate": 2.540439746493697e-05, "loss": 2.9742, "step": 43000 }, { "epoch": 0.4976319582675544, "grad_norm": 5.3809285163879395, "learning_rate": 2.5118402086622284e-05, "loss": 2.9849, "step": 43500 }, { "epoch": 0.5033518658338482, "grad_norm": 5.787545680999756, "learning_rate": 2.4832406708307595e-05, "loss": 2.963, "step": 44000 }, { "epoch": 0.5090717734001419, "grad_norm": 5.825649261474609, "learning_rate": 2.454641132999291e-05, "loss": 2.9634, "step": 44500 }, { "epoch": 0.5147916809664356, "grad_norm": 5.936666488647461, "learning_rate": 2.4260415951678223e-05, "loss": 2.9678, "step": 45000 }, { "epoch": 0.5205115885327293, "grad_norm": 5.980503082275391, "learning_rate": 2.3974420573363534e-05, "loss": 2.9675, "step": 45500 }, { "epoch": 0.5262314960990231, "grad_norm": 5.755555629730225, "learning_rate": 2.368842519504885e-05, "loss": 2.9597, "step": 46000 }, { "epoch": 0.5319514036653168, "grad_norm": 5.1978936195373535, "learning_rate": 2.3402429816734163e-05, "loss": 2.947, "step": 46500 }, { "epoch": 0.5376713112316105, "grad_norm": 5.265974521636963, "learning_rate": 2.3116434438419477e-05, "loss": 2.9521, "step": 47000 }, { "epoch": 0.5433912187979042, "grad_norm": 6.028165340423584, "learning_rate": 2.2830439060104788e-05, "loss": 2.9579, "step": 47500 }, { "epoch": 0.549111126364198, "grad_norm": 5.533000946044922, "learning_rate": 2.2544443681790102e-05, "loss": 2.9541, "step": 48000 }, { "epoch": 0.5548310339304917, "grad_norm": 5.428481101989746, "learning_rate": 2.225844830347542e-05, "loss": 2.9532, "step": 48500 }, { "epoch": 0.5605509414967854, "grad_norm": 5.905336856842041, "learning_rate": 2.197245292516073e-05, "loss": 2.94, "step": 49000 }, { "epoch": 0.5662708490630791, "grad_norm": 6.032477855682373, "learning_rate": 2.1686457546846045e-05, "loss": 2.9435, "step": 49500 }, { "epoch": 0.5719907566293728, "grad_norm": 5.996410369873047, "learning_rate": 2.1400462168531356e-05, "loss": 2.9358, "step": 50000 }, { "epoch": 0.5777106641956666, "grad_norm": 5.634001731872559, "learning_rate": 2.1114466790216674e-05, "loss": 2.9361, "step": 50500 }, { "epoch": 0.5834305717619603, "grad_norm": 5.509332656860352, "learning_rate": 2.0828471411901985e-05, "loss": 2.9314, "step": 51000 }, { "epoch": 0.589150479328254, "grad_norm": 6.294771194458008, "learning_rate": 2.05424760335873e-05, "loss": 2.9238, "step": 51500 }, { "epoch": 0.5948703868945477, "grad_norm": 5.542776107788086, "learning_rate": 2.025648065527261e-05, "loss": 2.9226, "step": 52000 }, { "epoch": 0.6005902944608416, "grad_norm": 5.870414733886719, "learning_rate": 1.9970485276957927e-05, "loss": 2.9202, "step": 52500 }, { "epoch": 0.6063102020271353, "grad_norm": 6.047051429748535, "learning_rate": 1.9684489898643238e-05, "loss": 2.9285, "step": 53000 }, { "epoch": 0.612030109593429, "grad_norm": 5.594234943389893, "learning_rate": 1.9398494520328553e-05, "loss": 2.9307, "step": 53500 }, { "epoch": 0.6177500171597227, "grad_norm": 5.4298295974731445, "learning_rate": 1.9112499142013863e-05, "loss": 2.9382, "step": 54000 }, { "epoch": 0.6234699247260165, "grad_norm": 6.184563636779785, "learning_rate": 1.882650376369918e-05, "loss": 2.9341, "step": 54500 }, { "epoch": 0.6291898322923102, "grad_norm": 5.776815414428711, "learning_rate": 1.8540508385384492e-05, "loss": 2.9287, "step": 55000 }, { "epoch": 0.6349097398586039, "grad_norm": 5.83139181137085, "learning_rate": 1.8254513007069806e-05, "loss": 2.9007, "step": 55500 }, { "epoch": 0.6406296474248976, "grad_norm": 5.469008922576904, "learning_rate": 1.7968517628755117e-05, "loss": 2.904, "step": 56000 }, { "epoch": 0.6463495549911913, "grad_norm": 6.898833751678467, "learning_rate": 1.7682522250440435e-05, "loss": 2.9105, "step": 56500 }, { "epoch": 0.6520694625574851, "grad_norm": 5.798022747039795, "learning_rate": 1.739652687212575e-05, "loss": 2.8995, "step": 57000 }, { "epoch": 0.6577893701237788, "grad_norm": 5.57025146484375, "learning_rate": 1.711053149381106e-05, "loss": 2.9093, "step": 57500 }, { "epoch": 0.6635092776900725, "grad_norm": 5.779621124267578, "learning_rate": 1.6824536115496374e-05, "loss": 2.9118, "step": 58000 }, { "epoch": 0.6692291852563662, "grad_norm": 5.683529853820801, "learning_rate": 1.653854073718169e-05, "loss": 2.9102, "step": 58500 }, { "epoch": 0.67494909282266, "grad_norm": 6.1109538078308105, "learning_rate": 1.6252545358867003e-05, "loss": 2.9102, "step": 59000 }, { "epoch": 0.6806690003889537, "grad_norm": 5.536868095397949, "learning_rate": 1.5966549980552314e-05, "loss": 2.8966, "step": 59500 }, { "epoch": 0.6863889079552474, "grad_norm": 6.560556888580322, "learning_rate": 1.5680554602237628e-05, "loss": 2.8896, "step": 60000 }, { "epoch": 0.6921088155215411, "grad_norm": 5.969814300537109, "learning_rate": 1.5394559223922942e-05, "loss": 2.8961, "step": 60500 }, { "epoch": 0.697828723087835, "grad_norm": 5.238883018493652, "learning_rate": 1.5108563845608257e-05, "loss": 2.882, "step": 61000 }, { "epoch": 0.7035486306541286, "grad_norm": 5.538156509399414, "learning_rate": 1.4822568467293567e-05, "loss": 2.8955, "step": 61500 }, { "epoch": 0.7092685382204224, "grad_norm": 6.287049770355225, "learning_rate": 1.4536573088978883e-05, "loss": 2.8932, "step": 62000 }, { "epoch": 0.714988445786716, "grad_norm": 6.05150032043457, "learning_rate": 1.4250577710664196e-05, "loss": 2.8868, "step": 62500 }, { "epoch": 0.7207083533530099, "grad_norm": 5.857907295227051, "learning_rate": 1.396458233234951e-05, "loss": 2.8872, "step": 63000 }, { "epoch": 0.7264282609193036, "grad_norm": 5.748091697692871, "learning_rate": 1.3678586954034823e-05, "loss": 2.8964, "step": 63500 }, { "epoch": 0.7321481684855973, "grad_norm": 5.55509614944458, "learning_rate": 1.3392591575720137e-05, "loss": 2.8758, "step": 64000 }, { "epoch": 0.737868076051891, "grad_norm": 6.135568618774414, "learning_rate": 1.3106596197405451e-05, "loss": 2.8685, "step": 64500 }, { "epoch": 0.7435879836181847, "grad_norm": 5.816187381744385, "learning_rate": 1.2820600819090764e-05, "loss": 2.8802, "step": 65000 }, { "epoch": 0.7493078911844785, "grad_norm": 6.309732437133789, "learning_rate": 1.2534605440776078e-05, "loss": 2.8859, "step": 65500 }, { "epoch": 0.7550277987507722, "grad_norm": 5.721366882324219, "learning_rate": 1.2248610062461391e-05, "loss": 2.8702, "step": 66000 }, { "epoch": 0.7607477063170659, "grad_norm": 5.648174285888672, "learning_rate": 1.1962614684146704e-05, "loss": 2.8603, "step": 66500 }, { "epoch": 0.7664676138833596, "grad_norm": 6.4422688484191895, "learning_rate": 1.1676619305832018e-05, "loss": 2.875, "step": 67000 }, { "epoch": 0.7721875214496534, "grad_norm": 5.673862934112549, "learning_rate": 1.1390623927517332e-05, "loss": 2.8743, "step": 67500 }, { "epoch": 0.7779074290159471, "grad_norm": 5.87379789352417, "learning_rate": 1.1104628549202645e-05, "loss": 2.8678, "step": 68000 }, { "epoch": 0.7836273365822408, "grad_norm": 5.968353271484375, "learning_rate": 1.0818633170887959e-05, "loss": 2.8797, "step": 68500 }, { "epoch": 0.7893472441485345, "grad_norm": 5.516451835632324, "learning_rate": 1.0532637792573273e-05, "loss": 2.8716, "step": 69000 }, { "epoch": 0.7950671517148283, "grad_norm": 6.277103900909424, "learning_rate": 1.0246642414258586e-05, "loss": 2.8483, "step": 69500 }, { "epoch": 0.800787059281122, "grad_norm": 5.54793643951416, "learning_rate": 9.9606470359439e-06, "loss": 2.8462, "step": 70000 }, { "epoch": 0.8065069668474157, "grad_norm": 5.989738464355469, "learning_rate": 9.674651657629213e-06, "loss": 2.8672, "step": 70500 }, { "epoch": 0.8122268744137094, "grad_norm": 5.7795844078063965, "learning_rate": 9.388656279314527e-06, "loss": 2.8566, "step": 71000 }, { "epoch": 0.8179467819800033, "grad_norm": 5.732882976531982, "learning_rate": 9.10266090099984e-06, "loss": 2.8497, "step": 71500 }, { "epoch": 0.823666689546297, "grad_norm": 6.427890777587891, "learning_rate": 8.816665522685154e-06, "loss": 2.8656, "step": 72000 }, { "epoch": 0.8293865971125907, "grad_norm": 6.1445746421813965, "learning_rate": 8.530670144370468e-06, "loss": 2.8425, "step": 72500 }, { "epoch": 0.8351065046788844, "grad_norm": 6.342021465301514, "learning_rate": 8.24467476605578e-06, "loss": 2.859, "step": 73000 }, { "epoch": 0.8408264122451781, "grad_norm": 6.117573261260986, "learning_rate": 7.958679387741095e-06, "loss": 2.8631, "step": 73500 }, { "epoch": 0.8465463198114719, "grad_norm": 6.145172595977783, "learning_rate": 7.672684009426408e-06, "loss": 2.8536, "step": 74000 }, { "epoch": 0.8522662273777656, "grad_norm": 5.709812641143799, "learning_rate": 7.386688631111721e-06, "loss": 2.8546, "step": 74500 }, { "epoch": 0.8579861349440593, "grad_norm": 6.244381904602051, "learning_rate": 7.100693252797034e-06, "loss": 2.8277, "step": 75000 }, { "epoch": 0.863706042510353, "grad_norm": 6.229698181152344, "learning_rate": 6.814697874482348e-06, "loss": 2.85, "step": 75500 }, { "epoch": 0.8694259500766468, "grad_norm": 5.480854511260986, "learning_rate": 6.528702496167661e-06, "loss": 2.8576, "step": 76000 }, { "epoch": 0.8751458576429405, "grad_norm": 6.088413715362549, "learning_rate": 6.2427071178529756e-06, "loss": 2.8618, "step": 76500 }, { "epoch": 0.8808657652092342, "grad_norm": 6.159554481506348, "learning_rate": 5.956711739538289e-06, "loss": 2.8319, "step": 77000 }, { "epoch": 0.8865856727755279, "grad_norm": 6.287491798400879, "learning_rate": 5.6707163612236024e-06, "loss": 2.8622, "step": 77500 }, { "epoch": 0.8923055803418217, "grad_norm": 6.237947940826416, "learning_rate": 5.384720982908916e-06, "loss": 2.8539, "step": 78000 }, { "epoch": 0.8980254879081154, "grad_norm": 6.096075057983398, "learning_rate": 5.09872560459423e-06, "loss": 2.8256, "step": 78500 }, { "epoch": 0.9037453954744091, "grad_norm": 6.158285140991211, "learning_rate": 4.812730226279544e-06, "loss": 2.8476, "step": 79000 }, { "epoch": 0.9094653030407028, "grad_norm": 6.239354133605957, "learning_rate": 4.526734847964857e-06, "loss": 2.8538, "step": 79500 }, { "epoch": 0.9151852106069966, "grad_norm": 5.273146629333496, "learning_rate": 4.2407394696501705e-06, "loss": 2.849, "step": 80000 }, { "epoch": 0.9209051181732903, "grad_norm": 6.121274948120117, "learning_rate": 3.954744091335484e-06, "loss": 2.8454, "step": 80500 }, { "epoch": 0.926625025739584, "grad_norm": 6.213692665100098, "learning_rate": 3.6687487130207977e-06, "loss": 2.8431, "step": 81000 }, { "epoch": 0.9323449333058778, "grad_norm": 6.369662761688232, "learning_rate": 3.382753334706111e-06, "loss": 2.8589, "step": 81500 }, { "epoch": 0.9380648408721715, "grad_norm": 5.934210300445557, "learning_rate": 3.096757956391425e-06, "loss": 2.8371, "step": 82000 }, { "epoch": 0.9437847484384653, "grad_norm": 6.598156452178955, "learning_rate": 2.8107625780767385e-06, "loss": 2.8282, "step": 82500 }, { "epoch": 0.949504656004759, "grad_norm": 5.81234073638916, "learning_rate": 2.524767199762052e-06, "loss": 2.8244, "step": 83000 }, { "epoch": 0.9552245635710527, "grad_norm": 6.145198822021484, "learning_rate": 2.2387718214473658e-06, "loss": 2.8294, "step": 83500 }, { "epoch": 0.9609444711373464, "grad_norm": 5.321970462799072, "learning_rate": 1.952776443132679e-06, "loss": 2.8425, "step": 84000 }, { "epoch": 0.9666643787036402, "grad_norm": 5.8561248779296875, "learning_rate": 1.6667810648179926e-06, "loss": 2.8618, "step": 84500 }, { "epoch": 0.9723842862699339, "grad_norm": 6.001429080963135, "learning_rate": 1.3807856865033063e-06, "loss": 2.8456, "step": 85000 }, { "epoch": 0.9781041938362276, "grad_norm": 6.074833393096924, "learning_rate": 1.0947903081886197e-06, "loss": 2.8408, "step": 85500 }, { "epoch": 0.9838241014025213, "grad_norm": 5.579460144042969, "learning_rate": 8.087949298739332e-07, "loss": 2.8427, "step": 86000 }, { "epoch": 0.9895440089688151, "grad_norm": 5.968313217163086, "learning_rate": 5.227995515592468e-07, "loss": 2.8345, "step": 86500 }, { "epoch": 0.9952639165351088, "grad_norm": 5.920719146728516, "learning_rate": 2.3680417324456038e-07, "loss": 2.8337, "step": 87000 }, { "epoch": 1.0, "step": 87414, "total_flos": 4.634223291773338e+16, "train_loss": 3.132497888326312, "train_runtime": 20429.0413, "train_samples_per_second": 34.231, "train_steps_per_second": 4.279 } ], "logging_steps": 500, "max_steps": 87414, "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": 4.634223291773338e+16, "train_batch_size": 8, "trial_name": null, "trial_params": null }