diff --git "a/trainer_state.json" "b/trainer_state.json" --- "a/trainer_state.json" +++ "b/trainer_state.json" @@ -3,3114 +3,14055 @@ "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, - "global_step": 441, + "global_step": 2004, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { - "epoch": 0.006802721088435374, - "grad_norm": 0.7543969951894244, - "learning_rate": 1.4285714285714286e-06, - "loss": 0.9624, + "epoch": 0.0014970059880239522, + "grad_norm": 0.4412629429939014, + "learning_rate": 3.278688524590164e-07, + "loss": 1.0214, "step": 1 }, { - "epoch": 0.013605442176870748, - "grad_norm": 0.44681902466146967, - "learning_rate": 2.8571428571428573e-06, - "loss": 0.964, + "epoch": 0.0029940119760479044, + "grad_norm": 0.44857961209660113, + "learning_rate": 6.557377049180328e-07, + "loss": 1.0321, "step": 2 }, { - "epoch": 0.02040816326530612, - "grad_norm": 0.5151164741690802, - "learning_rate": 4.2857142857142855e-06, - "loss": 1.0721, + "epoch": 0.004491017964071856, + "grad_norm": 0.4453725376646987, + "learning_rate": 9.836065573770493e-07, + "loss": 1.0352, "step": 3 }, { - "epoch": 0.027210884353741496, - "grad_norm": 0.5012473323244383, - "learning_rate": 5.7142857142857145e-06, - "loss": 1.0208, + "epoch": 0.005988023952095809, + "grad_norm": 0.4051365182123174, + "learning_rate": 1.3114754098360657e-06, + "loss": 0.959, "step": 4 }, { - "epoch": 0.034013605442176874, - "grad_norm": 0.47766942036038285, - "learning_rate": 7.1428571428571436e-06, - "loss": 0.987, + "epoch": 0.0074850299401197605, + "grad_norm": 0.3685112349775458, + "learning_rate": 1.6393442622950819e-06, + "loss": 0.9052, "step": 5 }, { - "epoch": 0.04081632653061224, - "grad_norm": 0.4930453957288559, - "learning_rate": 8.571428571428571e-06, - "loss": 1.0436, + "epoch": 0.008982035928143712, + "grad_norm": 0.46048753426928685, + "learning_rate": 1.9672131147540985e-06, + "loss": 0.9808, "step": 6 }, { - "epoch": 0.047619047619047616, - "grad_norm": 0.4570728289003672, - "learning_rate": 1e-05, - "loss": 0.9853, + "epoch": 0.010479041916167664, + "grad_norm": 0.4364944016404722, + "learning_rate": 2.295081967213115e-06, + "loss": 1.0203, "step": 7 }, { - "epoch": 0.05442176870748299, - "grad_norm": 0.39629873030193696, - "learning_rate": 1.1428571428571429e-05, - "loss": 0.8791, + "epoch": 0.011976047904191617, + "grad_norm": 0.4197052350095643, + "learning_rate": 2.6229508196721314e-06, + "loss": 0.936, "step": 8 }, { - "epoch": 0.061224489795918366, - "grad_norm": 0.49589342201146847, - "learning_rate": 1.2857142857142859e-05, - "loss": 1.0681, + "epoch": 0.01347305389221557, + "grad_norm": 0.3935251532018477, + "learning_rate": 2.9508196721311478e-06, + "loss": 0.931, "step": 9 }, { - "epoch": 0.06802721088435375, - "grad_norm": 0.39261176159121086, - "learning_rate": 1.4285714285714287e-05, - "loss": 0.9111, + "epoch": 0.014970059880239521, + "grad_norm": 0.40165315264572143, + "learning_rate": 3.2786885245901638e-06, + "loss": 0.9519, "step": 10 }, { - "epoch": 0.07482993197278912, - "grad_norm": 0.3716281292996566, - "learning_rate": 1.5714285714285715e-05, - "loss": 0.8437, + "epoch": 0.016467065868263474, + "grad_norm": 0.4327103578358591, + "learning_rate": 3.6065573770491806e-06, + "loss": 0.9975, "step": 11 }, { - "epoch": 0.08163265306122448, - "grad_norm": 0.4413102077225305, - "learning_rate": 1.7142857142857142e-05, - "loss": 0.9862, + "epoch": 0.017964071856287425, + "grad_norm": 0.4829902386185703, + "learning_rate": 3.934426229508197e-06, + "loss": 1.0322, "step": 12 }, { - "epoch": 0.08843537414965986, - "grad_norm": 0.42728639625517184, - "learning_rate": 1.8571428571428575e-05, - "loss": 0.9318, + "epoch": 0.019461077844311378, + "grad_norm": 0.43745126656959527, + "learning_rate": 4.2622950819672135e-06, + "loss": 0.9339, "step": 13 }, { - "epoch": 0.09523809523809523, - "grad_norm": 0.3865701816492697, - "learning_rate": 2e-05, - "loss": 0.8794, + "epoch": 0.020958083832335328, + "grad_norm": 0.40770435507966485, + "learning_rate": 4.59016393442623e-06, + "loss": 0.9558, "step": 14 }, { - "epoch": 0.10204081632653061, - "grad_norm": 0.40870933803528964, - "learning_rate": 1.9999729347501484e-05, - "loss": 0.9287, + "epoch": 0.02245508982035928, + "grad_norm": 0.3880526820488479, + "learning_rate": 4.918032786885246e-06, + "loss": 0.9233, "step": 15 }, { - "epoch": 0.10884353741496598, - "grad_norm": 0.3488567698718625, - "learning_rate": 1.9998917404656488e-05, - "loss": 0.802, + "epoch": 0.023952095808383235, + "grad_norm": 0.41942806957859513, + "learning_rate": 5.245901639344263e-06, + "loss": 0.97, "step": 16 }, { - "epoch": 0.11564625850340136, - "grad_norm": 0.32774310053561123, - "learning_rate": 1.9997564215415886e-05, - "loss": 0.7316, + "epoch": 0.025449101796407185, + "grad_norm": 0.38973129771254217, + "learning_rate": 5.573770491803278e-06, + "loss": 0.8961, "step": 17 }, { - "epoch": 0.12244897959183673, - "grad_norm": 0.3962817834034302, - "learning_rate": 1.9995669853028485e-05, - "loss": 0.8542, + "epoch": 0.02694610778443114, + "grad_norm": 0.4152019760637627, + "learning_rate": 5.9016393442622956e-06, + "loss": 0.9692, "step": 18 }, { - "epoch": 0.1292517006802721, - "grad_norm": 0.3339991775697425, - "learning_rate": 1.9993234420037072e-05, - "loss": 0.7316, + "epoch": 0.02844311377245509, + "grad_norm": 0.46572881246006315, + "learning_rate": 6.229508196721312e-06, + "loss": 0.9336, "step": 19 }, { - "epoch": 0.1360544217687075, - "grad_norm": 0.3613596947800494, - "learning_rate": 1.999025804827285e-05, - "loss": 0.7685, + "epoch": 0.029940119760479042, + "grad_norm": 0.3928467539453223, + "learning_rate": 6.5573770491803276e-06, + "loss": 0.9064, "step": 20 }, { - "epoch": 0.14285714285714285, - "grad_norm": 0.3427806922992633, - "learning_rate": 1.9986740898848306e-05, - "loss": 0.7005, + "epoch": 0.03143712574850299, + "grad_norm": 0.3955023428706724, + "learning_rate": 6.885245901639345e-06, + "loss": 0.8755, "step": 21 }, { - "epoch": 0.14965986394557823, - "grad_norm": 0.34368082813141, - "learning_rate": 1.99826831621485e-05, - "loss": 0.6939, + "epoch": 0.03293413173652695, + "grad_norm": 0.3655402984227431, + "learning_rate": 7.213114754098361e-06, + "loss": 0.8226, "step": 22 }, { - "epoch": 0.1564625850340136, - "grad_norm": 0.3300141228352275, - "learning_rate": 1.997808505782075e-05, - "loss": 0.6756, + "epoch": 0.0344311377245509, + "grad_norm": 0.38433031729962397, + "learning_rate": 7.540983606557377e-06, + "loss": 0.8473, "step": 23 }, { - "epoch": 0.16326530612244897, - "grad_norm": 0.3440934696087154, - "learning_rate": 1.9972946834762732e-05, - "loss": 0.6643, + "epoch": 0.03592814371257485, + "grad_norm": 0.3573419875384817, + "learning_rate": 7.868852459016394e-06, + "loss": 0.8282, "step": 24 }, { - "epoch": 0.17006802721088435, - "grad_norm": 0.35675869839992724, - "learning_rate": 1.9967268771109037e-05, - "loss": 0.6623, + "epoch": 0.0374251497005988, + "grad_norm": 0.4093172699149451, + "learning_rate": 8.19672131147541e-06, + "loss": 0.8879, "step": 25 }, { - "epoch": 0.17687074829931973, - "grad_norm": 0.34399594353909607, - "learning_rate": 1.996105117421608e-05, - "loss": 0.6182, + "epoch": 0.038922155688622756, + "grad_norm": 0.3820938437320782, + "learning_rate": 8.524590163934427e-06, + "loss": 0.864, "step": 26 }, { - "epoch": 0.1836734693877551, - "grad_norm": 0.3634094292173927, - "learning_rate": 1.9954294380645497e-05, - "loss": 0.6286, + "epoch": 0.040419161676646706, + "grad_norm": 0.39326220628777386, + "learning_rate": 8.852459016393443e-06, + "loss": 0.8862, "step": 27 }, { - "epoch": 0.19047619047619047, - "grad_norm": 0.37888057387883767, - "learning_rate": 1.9946998756145894e-05, - "loss": 0.6367, + "epoch": 0.041916167664670656, + "grad_norm": 0.36807756007538617, + "learning_rate": 9.18032786885246e-06, + "loss": 0.8257, "step": 28 }, { - "epoch": 0.19727891156462585, - "grad_norm": 0.36856282647607674, - "learning_rate": 1.9939164695633067e-05, - "loss": 0.5763, + "epoch": 0.04341317365269461, + "grad_norm": 0.35588261728997433, + "learning_rate": 9.508196721311476e-06, + "loss": 0.8128, "step": 29 }, { - "epoch": 0.20408163265306123, - "grad_norm": 0.33791502181439126, - "learning_rate": 1.9930792623168638e-05, - "loss": 0.5111, + "epoch": 0.04491017964071856, + "grad_norm": 0.3715077903083305, + "learning_rate": 9.836065573770493e-06, + "loss": 0.8099, "step": 30 }, { - "epoch": 0.2108843537414966, - "grad_norm": 0.3507039073715319, - "learning_rate": 1.992188299193706e-05, - "loss": 0.5191, + "epoch": 0.04640718562874251, + "grad_norm": 0.3596909054144244, + "learning_rate": 1.0163934426229509e-05, + "loss": 0.7715, "step": 31 }, { - "epoch": 0.21768707482993196, - "grad_norm": 0.3530833992530626, - "learning_rate": 1.9912436284221134e-05, - "loss": 0.4889, + "epoch": 0.04790419161676647, + "grad_norm": 0.384702217147705, + "learning_rate": 1.0491803278688525e-05, + "loss": 0.833, "step": 32 }, { - "epoch": 0.22448979591836735, - "grad_norm": 0.3592183944530985, - "learning_rate": 1.9902453011375865e-05, - "loss": 0.4832, + "epoch": 0.04940119760479042, + "grad_norm": 0.36512927940401047, + "learning_rate": 1.0819672131147544e-05, + "loss": 0.7903, "step": 33 }, { - "epoch": 0.23129251700680273, - "grad_norm": 0.34918368964463997, - "learning_rate": 1.98919337138008e-05, - "loss": 0.444, + "epoch": 0.05089820359281437, + "grad_norm": 0.39841041594176824, + "learning_rate": 1.1147540983606557e-05, + "loss": 0.7641, "step": 34 }, { - "epoch": 0.23809523809523808, - "grad_norm": 0.38954703063221796, - "learning_rate": 1.9880878960910772e-05, - "loss": 0.4722, + "epoch": 0.05239520958083832, + "grad_norm": 0.41324753544403525, + "learning_rate": 1.1475409836065575e-05, + "loss": 0.7774, "step": 35 }, { - "epoch": 0.24489795918367346, - "grad_norm": 0.37889828104003, - "learning_rate": 1.9869289351105087e-05, - "loss": 0.4284, + "epoch": 0.05389221556886228, + "grad_norm": 0.3805767081935722, + "learning_rate": 1.1803278688524591e-05, + "loss": 0.7746, "step": 36 }, { - "epoch": 0.25170068027210885, - "grad_norm": 0.3879345796013965, - "learning_rate": 1.9857165511735105e-05, - "loss": 0.4192, + "epoch": 0.05538922155688623, + "grad_norm": 0.3640692670562054, + "learning_rate": 1.2131147540983608e-05, + "loss": 0.741, "step": 37 }, { - "epoch": 0.2585034013605442, - "grad_norm": 0.36130568820722925, - "learning_rate": 1.9844508099070313e-05, - "loss": 0.3716, + "epoch": 0.05688622754491018, + "grad_norm": 0.3727384829726358, + "learning_rate": 1.2459016393442624e-05, + "loss": 0.7203, "step": 38 }, { - "epoch": 0.2653061224489796, - "grad_norm": 0.36791653912252376, - "learning_rate": 1.9831317798262787e-05, - "loss": 0.3653, + "epoch": 0.058383233532934134, + "grad_norm": 0.3695211453830474, + "learning_rate": 1.2786885245901642e-05, + "loss": 0.6895, "step": 39 }, { - "epoch": 0.272108843537415, - "grad_norm": 0.3761185645158447, - "learning_rate": 1.98175953233101e-05, - "loss": 0.3465, + "epoch": 0.059880239520958084, + "grad_norm": 0.41877844793582925, + "learning_rate": 1.3114754098360655e-05, + "loss": 0.6759, "step": 40 }, { - "epoch": 0.2789115646258503, - "grad_norm": 0.4350230270095709, - "learning_rate": 1.980334141701667e-05, - "loss": 0.348, + "epoch": 0.061377245508982034, + "grad_norm": 0.33931923326947677, + "learning_rate": 1.3442622950819673e-05, + "loss": 0.6196, "step": 41 }, { - "epoch": 0.2857142857142857, - "grad_norm": 0.32213563899302, - "learning_rate": 1.978855685095358e-05, - "loss": 0.2679, + "epoch": 0.06287425149700598, + "grad_norm": 0.3684580530246632, + "learning_rate": 1.377049180327869e-05, + "loss": 0.6655, "step": 42 }, { - "epoch": 0.2925170068027211, - "grad_norm": 0.3768678079788929, - "learning_rate": 1.977324242541677e-05, - "loss": 0.2842, + "epoch": 0.06437125748502993, + "grad_norm": 0.32458089242640203, + "learning_rate": 1.4098360655737706e-05, + "loss": 0.5865, "step": 43 }, { - "epoch": 0.29931972789115646, - "grad_norm": 0.3675295235354873, - "learning_rate": 1.9757398969383752e-05, - "loss": 0.25, + "epoch": 0.0658682634730539, + "grad_norm": 0.3895546564318695, + "learning_rate": 1.4426229508196722e-05, + "loss": 0.6705, "step": 44 }, { - "epoch": 0.30612244897959184, - "grad_norm": 0.3702046286490698, - "learning_rate": 1.974102734046872e-05, - "loss": 0.2398, + "epoch": 0.06736526946107785, + "grad_norm": 0.3483102383245026, + "learning_rate": 1.4754098360655739e-05, + "loss": 0.5904, "step": 45 }, { - "epoch": 0.3129251700680272, - "grad_norm": 0.39377885956971614, - "learning_rate": 1.9724128424876117e-05, - "loss": 0.2288, + "epoch": 0.0688622754491018, + "grad_norm": 0.3789944257657541, + "learning_rate": 1.5081967213114754e-05, + "loss": 0.6053, "step": 46 }, { - "epoch": 0.3197278911564626, - "grad_norm": 0.35607191659617016, - "learning_rate": 1.9706703137352695e-05, - "loss": 0.1986, + "epoch": 0.07035928143712575, + "grad_norm": 0.37752386838956564, + "learning_rate": 1.5409836065573772e-05, + "loss": 0.5844, "step": 47 }, { - "epoch": 0.32653061224489793, - "grad_norm": 0.37797721697894693, - "learning_rate": 1.968875242113798e-05, - "loss": 0.1921, + "epoch": 0.0718562874251497, + "grad_norm": 0.3630635575321055, + "learning_rate": 1.5737704918032788e-05, + "loss": 0.5291, "step": 48 }, { - "epoch": 0.3333333333333333, - "grad_norm": 0.4458500482446309, - "learning_rate": 1.9670277247913205e-05, - "loss": 0.1808, + "epoch": 0.07335329341317365, + "grad_norm": 0.3914577758366373, + "learning_rate": 1.6065573770491805e-05, + "loss": 0.5326, "step": 49 }, { - "epoch": 0.3401360544217687, - "grad_norm": 0.4904856763671931, - "learning_rate": 1.965127861774873e-05, - "loss": 0.178, + "epoch": 0.0748502994011976, + "grad_norm": 0.40325362044689916, + "learning_rate": 1.639344262295082e-05, + "loss": 0.5288, "step": 50 }, { - "epoch": 0.3469387755102041, - "grad_norm": 0.42047437091507667, - "learning_rate": 1.96317575590499e-05, - "loss": 0.1556, + "epoch": 0.07634730538922156, + "grad_norm": 0.4137060583375249, + "learning_rate": 1.6721311475409837e-05, + "loss": 0.4758, "step": 51 }, { - "epoch": 0.35374149659863946, - "grad_norm": 0.3659087858039602, - "learning_rate": 1.9611715128501378e-05, - "loss": 0.1316, + "epoch": 0.07784431137724551, + "grad_norm": 0.37431549278431586, + "learning_rate": 1.7049180327868854e-05, + "loss": 0.4133, "step": 52 }, { - "epoch": 0.36054421768707484, - "grad_norm": 0.41650852039639386, - "learning_rate": 1.9591152411009942e-05, - "loss": 0.1366, + "epoch": 0.07934131736526946, + "grad_norm": 0.4434060283126579, + "learning_rate": 1.737704918032787e-05, + "loss": 0.4636, "step": 53 }, { - "epoch": 0.3673469387755102, - "grad_norm": 0.42327758680290756, - "learning_rate": 1.9570070519645767e-05, - "loss": 0.1239, + "epoch": 0.08083832335329341, + "grad_norm": 0.44913599309601093, + "learning_rate": 1.7704918032786887e-05, + "loss": 0.4453, "step": 54 }, { - "epoch": 0.3741496598639456, - "grad_norm": 0.4263037625862704, - "learning_rate": 1.9548470595582166e-05, - "loss": 0.1226, + "epoch": 0.08233532934131736, + "grad_norm": 0.4208028028608017, + "learning_rate": 1.8032786885245903e-05, + "loss": 0.3823, "step": 55 }, { - "epoch": 0.38095238095238093, - "grad_norm": 0.4203561676384866, - "learning_rate": 1.9526353808033827e-05, - "loss": 0.1203, + "epoch": 0.08383233532934131, + "grad_norm": 0.4340233256708269, + "learning_rate": 1.836065573770492e-05, + "loss": 0.3575, "step": 56 }, { - "epoch": 0.3877551020408163, - "grad_norm": 0.3960865903458932, - "learning_rate": 1.9503721354193507e-05, - "loss": 0.1121, + "epoch": 0.08532934131736528, + "grad_norm": 0.4231169164061207, + "learning_rate": 1.8688524590163936e-05, + "loss": 0.3157, "step": 57 }, { - "epoch": 0.3945578231292517, - "grad_norm": 0.3811191794773781, - "learning_rate": 1.948057445916724e-05, - "loss": 0.1095, + "epoch": 0.08682634730538923, + "grad_norm": 0.4174421584158068, + "learning_rate": 1.9016393442622952e-05, + "loss": 0.2966, "step": 58 }, { - "epoch": 0.4013605442176871, - "grad_norm": 0.34427762247965304, - "learning_rate": 1.9456914375908026e-05, - "loss": 0.1019, + "epoch": 0.08832335329341318, + "grad_norm": 0.4611062696725442, + "learning_rate": 1.934426229508197e-05, + "loss": 0.2824, "step": 59 }, { - "epoch": 0.40816326530612246, - "grad_norm": 0.3250633041475826, - "learning_rate": 1.9432742385147988e-05, - "loss": 0.0961, + "epoch": 0.08982035928143713, + "grad_norm": 0.5834947622453618, + "learning_rate": 1.9672131147540985e-05, + "loss": 0.2742, "step": 60 }, { - "epoch": 0.41496598639455784, - "grad_norm": 0.3391653472941772, - "learning_rate": 1.9408059795329073e-05, - "loss": 0.0907, + "epoch": 0.09131736526946108, + "grad_norm": 0.4356716457731799, + "learning_rate": 2e-05, + "loss": 0.232, "step": 61 }, { - "epoch": 0.4217687074829932, - "grad_norm": 0.3071859632068275, - "learning_rate": 1.9382867942532195e-05, - "loss": 0.0887, + "epoch": 0.09281437125748503, + "grad_norm": 0.5040703847152936, + "learning_rate": 1.9999986928541335e-05, + "loss": 0.215, "step": 62 }, { - "epoch": 0.42857142857142855, - "grad_norm": 0.28368480944915603, - "learning_rate": 1.9357168190404937e-05, - "loss": 0.0844, + "epoch": 0.09431137724550898, + "grad_norm": 0.6246157723635114, + "learning_rate": 1.9999947714199512e-05, + "loss": 0.2105, "step": 63 }, { - "epoch": 0.43537414965986393, - "grad_norm": 0.37481710014760405, - "learning_rate": 1.9330961930087724e-05, - "loss": 0.0891, + "epoch": 0.09580838323353294, + "grad_norm": 0.620783787770869, + "learning_rate": 1.999988235707705e-05, + "loss": 0.1869, "step": 64 }, { - "epoch": 0.4421768707482993, - "grad_norm": 0.3263398981365655, - "learning_rate": 1.9304250580138524e-05, - "loss": 0.0866, + "epoch": 0.09730538922155689, + "grad_norm": 0.4844829922546262, + "learning_rate": 1.9999790857344808e-05, + "loss": 0.1693, "step": 65 }, { - "epoch": 0.4489795918367347, - "grad_norm": 0.2880232601443931, - "learning_rate": 1.9277035586456056e-05, - "loss": 0.0851, + "epoch": 0.09880239520958084, + "grad_norm": 0.39690065232341765, + "learning_rate": 1.9999673215241996e-05, + "loss": 0.1559, "step": 66 }, { - "epoch": 0.4557823129251701, - "grad_norm": 0.38682915757490255, - "learning_rate": 1.9249318422201524e-05, - "loss": 0.0813, + "epoch": 0.10029940119760479, + "grad_norm": 0.39690065232341765, + "learning_rate": 1.9999673215241996e-05, + "loss": 0.1393, "step": 67 }, { - "epoch": 0.46258503401360546, - "grad_norm": 0.20937614024944473, - "learning_rate": 1.9221100587718884e-05, - "loss": 0.0782, + "epoch": 0.10179640718562874, + "grad_norm": 0.3744306931664112, + "learning_rate": 1.9999529431076165e-05, + "loss": 0.1444, "step": 68 }, { - "epoch": 0.46938775510204084, - "grad_norm": 0.24997247747455145, - "learning_rate": 1.919238361045362e-05, - "loss": 0.0764, + "epoch": 0.10329341317365269, + "grad_norm": 0.3555775828949075, + "learning_rate": 1.9999359505223208e-05, + "loss": 0.131, "step": 69 }, { - "epoch": 0.47619047619047616, - "grad_norm": 0.23920563707712375, - "learning_rate": 1.916316904487005e-05, - "loss": 0.0783, + "epoch": 0.10479041916167664, + "grad_norm": 0.446887895690596, + "learning_rate": 1.999916343812736e-05, + "loss": 0.1295, "step": 70 }, { - "epoch": 0.48299319727891155, - "grad_norm": 0.23894151439100778, - "learning_rate": 1.9133458472367216e-05, - "loss": 0.0761, + "epoch": 0.1062874251497006, + "grad_norm": 0.4182793503525473, + "learning_rate": 1.99989412303012e-05, + "loss": 0.1129, "step": 71 }, { - "epoch": 0.4897959183673469, - "grad_norm": 0.24006873105227072, - "learning_rate": 1.9103253501193256e-05, - "loss": 0.076, + "epoch": 0.10778443113772455, + "grad_norm": 0.3512235169414461, + "learning_rate": 1.999869288232564e-05, + "loss": 0.1115, "step": 72 }, { - "epoch": 0.4965986394557823, - "grad_norm": 0.2551200414134792, - "learning_rate": 1.9072555766358346e-05, - "loss": 0.0766, + "epoch": 0.1092814371257485, + "grad_norm": 0.39358701307726796, + "learning_rate": 1.9998418394849935e-05, + "loss": 0.1059, "step": 73 }, { - "epoch": 0.5034013605442177, - "grad_norm": 0.22157625780163037, - "learning_rate": 1.904136692954622e-05, - "loss": 0.0736, + "epoch": 0.11077844311377245, + "grad_norm": 0.37079524723615265, + "learning_rate": 1.999811776859168e-05, + "loss": 0.1063, "step": 74 }, { - "epoch": 0.5102040816326531, - "grad_norm": 0.22027442259887026, - "learning_rate": 1.900968867902419e-05, - "loss": 0.0714, + "epoch": 0.1122754491017964, + "grad_norm": 0.35175920574697694, + "learning_rate": 1.9997791004336794e-05, + "loss": 0.0978, "step": 75 }, { - "epoch": 0.5170068027210885, - "grad_norm": 0.18204627775070914, - "learning_rate": 1.89775227295518e-05, - "loss": 0.0714, + "epoch": 0.11377245508982035, + "grad_norm": 0.44421104255896743, + "learning_rate": 1.9997438102939538e-05, + "loss": 0.0972, "step": 76 }, { - "epoch": 0.5238095238095238, - "grad_norm": 0.2620995135551117, - "learning_rate": 1.8944870822287957e-05, - "loss": 0.0729, + "epoch": 0.11526946107784432, + "grad_norm": 0.2935792679601046, + "learning_rate": 1.99970590653225e-05, + "loss": 0.0865, "step": 77 }, { - "epoch": 0.5306122448979592, - "grad_norm": 0.29878189909698066, - "learning_rate": 1.891173472469672e-05, - "loss": 0.0744, + "epoch": 0.11676646706586827, + "grad_norm": 0.32149928982255116, + "learning_rate": 1.999665389247659e-05, + "loss": 0.0848, "step": 78 }, { - "epoch": 0.5374149659863946, - "grad_norm": 0.20917807504184172, - "learning_rate": 1.8878116230451615e-05, - "loss": 0.0709, + "epoch": 0.11826347305389222, + "grad_norm": 0.3346624453038523, + "learning_rate": 1.999622258546105e-05, + "loss": 0.0841, "step": 79 }, { - "epoch": 0.54421768707483, - "grad_norm": 0.2280652332605586, - "learning_rate": 1.884401715933853e-05, - "loss": 0.0691, + "epoch": 0.11976047904191617, + "grad_norm": 0.2307198733032352, + "learning_rate": 1.9995765145403444e-05, + "loss": 0.078, "step": 80 }, { - "epoch": 0.5510204081632653, - "grad_norm": 0.16319230454307865, - "learning_rate": 1.8809439357157226e-05, - "loss": 0.0664, + "epoch": 0.12125748502994012, + "grad_norm": 0.2501179187728697, + "learning_rate": 1.9995281573499652e-05, + "loss": 0.0789, "step": 81 }, { - "epoch": 0.5578231292517006, - "grad_norm": 0.17045276025315298, - "learning_rate": 1.8774384695621407e-05, - "loss": 0.0672, + "epoch": 0.12275449101796407, + "grad_norm": 0.3027539443814983, + "learning_rate": 1.9994771871013874e-05, + "loss": 0.0773, "step": 82 }, { - "epoch": 0.564625850340136, - "grad_norm": 0.18557421713410696, - "learning_rate": 1.8738855072257428e-05, - "loss": 0.0684, + "epoch": 0.12425149700598802, + "grad_norm": 0.4231966349076508, + "learning_rate": 1.999423603927862e-05, + "loss": 0.0789, "step": 83 }, { - "epoch": 0.5714285714285714, - "grad_norm": 0.1857928022764888, - "learning_rate": 1.8702852410301556e-05, - "loss": 0.0713, + "epoch": 0.12574850299401197, + "grad_norm": 0.2635202834499579, + "learning_rate": 1.9993674079694708e-05, + "loss": 0.0675, "step": 84 }, { - "epoch": 0.5782312925170068, - "grad_norm": 0.19053752905827265, - "learning_rate": 1.8666378658595863e-05, - "loss": 0.067, + "epoch": 0.12724550898203593, + "grad_norm": 0.43234380467645195, + "learning_rate": 1.999308599373127e-05, + "loss": 0.0787, "step": 85 }, { - "epoch": 0.5850340136054422, - "grad_norm": 0.26375035657754126, - "learning_rate": 1.8629435791482765e-05, - "loss": 0.0643, + "epoch": 0.12874251497005987, + "grad_norm": 0.1825788301606033, + "learning_rate": 1.999247178292573e-05, + "loss": 0.0721, "step": 86 }, { - "epoch": 0.5918367346938775, - "grad_norm": 0.20416646049963377, - "learning_rate": 1.8592025808698116e-05, - "loss": 0.0664, + "epoch": 0.13023952095808383, + "grad_norm": 0.2078262527726928, + "learning_rate": 1.9991831448883815e-05, + "loss": 0.072, "step": 87 }, { - "epoch": 0.5986394557823129, - "grad_norm": 0.22529978535979228, - "learning_rate": 1.8554150735262975e-05, - "loss": 0.0666, + "epoch": 0.1317365269461078, + "grad_norm": 0.20140626332699332, + "learning_rate": 1.9991164993279543e-05, + "loss": 0.0692, "step": 88 }, { - "epoch": 0.6054421768707483, - "grad_norm": 0.2343565249290847, - "learning_rate": 1.8515812621373998e-05, - "loss": 0.0671, + "epoch": 0.13323353293413173, + "grad_norm": 0.27122788318338725, + "learning_rate": 1.999047241785523e-05, + "loss": 0.0728, "step": 89 }, { - "epoch": 0.6122448979591837, - "grad_norm": 0.23897319920733306, - "learning_rate": 1.8477013542292446e-05, - "loss": 0.066, + "epoch": 0.1347305389221557, + "grad_norm": 0.2420008393498157, + "learning_rate": 1.998975372442146e-05, + "loss": 0.074, "step": 90 }, { - "epoch": 0.6190476190476191, - "grad_norm": 0.24710547661980992, - "learning_rate": 1.8437755598231857e-05, - "loss": 0.0657, + "epoch": 0.13622754491017963, + "grad_norm": 0.29824238890853016, + "learning_rate": 1.9989008914857115e-05, + "loss": 0.0676, "step": 91 }, { - "epoch": 0.6258503401360545, - "grad_norm": 0.25014419464910725, - "learning_rate": 1.8398040914244363e-05, - "loss": 0.0684, + "epoch": 0.1377245508982036, + "grad_norm": 0.2912749464147097, + "learning_rate": 1.9988237991109342e-05, + "loss": 0.0683, "step": 92 }, { - "epoch": 0.6326530612244898, - "grad_norm": 0.17847236067983924, - "learning_rate": 1.8357871640105648e-05, - "loss": 0.0631, + "epoch": 0.13922155688622753, + "grad_norm": 0.22500666240199066, + "learning_rate": 1.9987440955193563e-05, + "loss": 0.0695, "step": 93 }, { - "epoch": 0.6394557823129252, - "grad_norm": 0.24262416375321524, - "learning_rate": 1.8317249950198598e-05, - "loss": 0.0657, + "epoch": 0.1407185628742515, + "grad_norm": 0.4333093969766628, + "learning_rate": 1.998661780919346e-05, + "loss": 0.0709, "step": 94 }, { - "epoch": 0.6462585034013606, - "grad_norm": 0.19587079364109738, - "learning_rate": 1.8276178043395588e-05, - "loss": 0.0661, + "epoch": 0.14221556886227546, + "grad_norm": 0.1895065089384485, + "learning_rate": 1.9985768555260976e-05, + "loss": 0.0665, "step": 95 }, { - "epoch": 0.6530612244897959, - "grad_norm": 0.2511062200421889, - "learning_rate": 1.8234658142939454e-05, - "loss": 0.0626, + "epoch": 0.1437125748502994, + "grad_norm": 0.27388821802849495, + "learning_rate": 1.9984893195616313e-05, + "loss": 0.0668, "step": 96 }, { - "epoch": 0.6598639455782312, - "grad_norm": 0.19039392887739032, - "learning_rate": 1.8192692496323158e-05, - "loss": 0.0645, + "epoch": 0.14520958083832336, + "grad_norm": 0.2133616131184573, + "learning_rate": 1.9983991732547912e-05, + "loss": 0.0644, "step": 97 }, { - "epoch": 0.6666666666666666, - "grad_norm": 0.1999976259553887, - "learning_rate": 1.8150283375168112e-05, - "loss": 0.0657, + "epoch": 0.1467065868263473, + "grad_norm": 0.23231083558556673, + "learning_rate": 1.998306416841246e-05, + "loss": 0.0659, "step": 98 }, { - "epoch": 0.673469387755102, - "grad_norm": 0.20368044763239615, - "learning_rate": 1.8107433075101254e-05, - "loss": 0.0665, + "epoch": 0.14820359281437126, + "grad_norm": 0.3238616061638123, + "learning_rate": 1.998211050563488e-05, + "loss": 0.0676, "step": 99 }, { - "epoch": 0.6802721088435374, - "grad_norm": 0.22553964175771693, - "learning_rate": 1.8064143915630723e-05, - "loss": 0.0661, + "epoch": 0.1497005988023952, + "grad_norm": 0.19406634289412697, + "learning_rate": 1.998113074670833e-05, + "loss": 0.0651, "step": 100 }, { - "epoch": 0.6870748299319728, - "grad_norm": 0.28951843622254164, - "learning_rate": 1.8020418240020362e-05, - "loss": 0.069, + "epoch": 0.15119760479041916, + "grad_norm": 0.17418625172312732, + "learning_rate": 1.9980124894194184e-05, + "loss": 0.0641, "step": 101 }, { - "epoch": 0.6938775510204082, - "grad_norm": 0.2057412466161773, - "learning_rate": 1.7976258415162836e-05, - "loss": 0.0618, + "epoch": 0.15269461077844312, + "grad_norm": 0.21916068973423491, + "learning_rate": 1.9979092950722033e-05, + "loss": 0.0624, "step": 102 }, { - "epoch": 0.7006802721088435, - "grad_norm": 0.1836029567698062, - "learning_rate": 1.7931666831451536e-05, - "loss": 0.064, + "epoch": 0.15419161676646706, + "grad_norm": 0.23896804696125767, + "learning_rate": 1.9978034918989674e-05, + "loss": 0.0644, "step": 103 }, { - "epoch": 0.7074829931972789, - "grad_norm": 0.219074746713116, - "learning_rate": 1.7886645902651166e-05, - "loss": 0.0629, + "epoch": 0.15568862275449102, + "grad_norm": 0.19653741457543372, + "learning_rate": 1.9976950801763114e-05, + "loss": 0.0638, "step": 104 }, { - "epoch": 0.7142857142857143, - "grad_norm": 0.1826642649944216, - "learning_rate": 1.7841198065767107e-05, - "loss": 0.06, + "epoch": 0.15718562874251496, + "grad_norm": 0.33352908376943624, + "learning_rate": 1.9975840601876553e-05, + "loss": 0.063, "step": 105 }, { - "epoch": 0.7210884353741497, - "grad_norm": 0.14348426307016002, - "learning_rate": 1.779532578091347e-05, - "loss": 0.0598, + "epoch": 0.15868263473053892, + "grad_norm": 0.19174789139029536, + "learning_rate": 1.9974704322232375e-05, + "loss": 0.0619, "step": 106 }, { - "epoch": 0.7278911564625851, - "grad_norm": 0.1892487521305465, - "learning_rate": 1.7749031531179962e-05, - "loss": 0.0608, + "epoch": 0.1601796407185629, + "grad_norm": 0.31382911722883733, + "learning_rate": 1.997354196580115e-05, + "loss": 0.063, "step": 107 }, { - "epoch": 0.7346938775510204, - "grad_norm": 0.1852039687458771, - "learning_rate": 1.7702317822497457e-05, - "loss": 0.0612, + "epoch": 0.16167664670658682, + "grad_norm": 0.19646342349922372, + "learning_rate": 1.9972353535621613e-05, + "loss": 0.0614, "step": 108 }, { - "epoch": 0.7414965986394558, - "grad_norm": 0.20838595163721002, - "learning_rate": 1.7655187183502344e-05, - "loss": 0.0636, + "epoch": 0.1631736526946108, + "grad_norm": 0.36904190804708314, + "learning_rate": 1.9971139034800666e-05, + "loss": 0.0626, "step": 109 }, { - "epoch": 0.7482993197278912, - "grad_norm": 0.1981664363252372, - "learning_rate": 1.7607642165399665e-05, - "loss": 0.0586, + "epoch": 0.16467065868263472, + "grad_norm": 0.2720984468904841, + "learning_rate": 1.9969898466513375e-05, + "loss": 0.0624, "step": 110 }, { - "epoch": 0.7551020408163265, - "grad_norm": 0.2820295964012395, - "learning_rate": 1.755968534182501e-05, - "loss": 0.0627, + "epoch": 0.1661676646706587, + "grad_norm": 0.20744167748661432, + "learning_rate": 1.9968631834002947e-05, + "loss": 0.0595, "step": 111 }, { - "epoch": 0.7619047619047619, - "grad_norm": 0.16519458200620668, - "learning_rate": 1.7511319308705198e-05, - "loss": 0.059, + "epoch": 0.16766467065868262, + "grad_norm": 0.494627859147702, + "learning_rate": 1.9967339140580718e-05, + "loss": 0.0629, "step": 112 }, { - "epoch": 0.7687074829931972, - "grad_norm": 0.26268272830133443, - "learning_rate": 1.746254668411778e-05, - "loss": 0.0649, + "epoch": 0.1691616766467066, + "grad_norm": 0.21122512164165028, + "learning_rate": 1.9966020389626173e-05, + "loss": 0.0611, "step": 113 }, { - "epoch": 0.7755102040816326, - "grad_norm": 0.19201979033727093, - "learning_rate": 1.7413370108149288e-05, - "loss": 0.0639, + "epoch": 0.17065868263473055, + "grad_norm": 0.22452536057932843, + "learning_rate": 1.9964675584586918e-05, + "loss": 0.0613, "step": 114 }, { - "epoch": 0.782312925170068, - "grad_norm": 0.23222721849071837, - "learning_rate": 1.7363792242752354e-05, - "loss": 0.0611, + "epoch": 0.1721556886227545, + "grad_norm": 0.1937038289344653, + "learning_rate": 1.996330472897866e-05, + "loss": 0.0568, "step": 115 }, { - "epoch": 0.7891156462585034, - "grad_norm": 0.19754689775387121, - "learning_rate": 1.731381577160161e-05, - "loss": 0.0611, + "epoch": 0.17365269461077845, + "grad_norm": 0.21370437758625413, + "learning_rate": 1.996190782638521e-05, + "loss": 0.0623, "step": 116 }, { - "epoch": 0.7959183673469388, - "grad_norm": 0.26191489705008636, - "learning_rate": 1.726344339994841e-05, - "loss": 0.0628, + "epoch": 0.1751497005988024, + "grad_norm": 0.2026742633701451, + "learning_rate": 1.9960484880458486e-05, + "loss": 0.0603, "step": 117 }, { - "epoch": 0.8027210884353742, - "grad_norm": 0.22002163098655456, - "learning_rate": 1.7212677854474402e-05, - "loss": 0.0597, + "epoch": 0.17664670658682635, + "grad_norm": 0.22896351028877954, + "learning_rate": 1.995903589491848e-05, + "loss": 0.0599, "step": 118 }, { - "epoch": 0.8095238095238095, - "grad_norm": 0.21168658257559084, - "learning_rate": 1.7161521883143936e-05, + "epoch": 0.1781437125748503, + "grad_norm": 0.3143604037843644, + "learning_rate": 1.995756087355327e-05, "loss": 0.06, "step": 119 }, { - "epoch": 0.8163265306122449, - "grad_norm": 0.2222174632266417, - "learning_rate": 1.7109978255055295e-05, - "loss": 0.0603, + "epoch": 0.17964071856287425, + "grad_norm": 0.24857151161776553, + "learning_rate": 1.9956059820218982e-05, + "loss": 0.0598, "step": 120 }, { - "epoch": 0.8231292517006803, - "grad_norm": 0.2520406444086263, - "learning_rate": 1.705804976029083e-05, - "loss": 0.0618, + "epoch": 0.18113772455089822, + "grad_norm": 0.23471075757390752, + "learning_rate": 1.9954532738839814e-05, + "loss": 0.0583, "step": 121 }, { - "epoch": 0.8299319727891157, - "grad_norm": 0.1873330783063175, - "learning_rate": 1.7005739209765906e-05, - "loss": 0.0599, + "epoch": 0.18263473053892215, + "grad_norm": 0.1933238932543813, + "learning_rate": 1.9952979633408002e-05, + "loss": 0.0572, "step": 122 }, { - "epoch": 0.8367346938775511, - "grad_norm": 0.24145186470793584, - "learning_rate": 1.6953049435076768e-05, - "loss": 0.0607, + "epoch": 0.18413173652694612, + "grad_norm": 0.21990365595016542, + "learning_rate": 1.9951400507983812e-05, + "loss": 0.0585, "step": 123 }, { - "epoch": 0.8435374149659864, - "grad_norm": 0.24291351924884128, - "learning_rate": 1.6899983288347248e-05, - "loss": 0.0588, + "epoch": 0.18562874251497005, + "grad_norm": 0.19297438870815392, + "learning_rate": 1.9949795366695545e-05, + "loss": 0.0577, "step": 124 }, { - "epoch": 0.8503401360544217, - "grad_norm": 0.19252112685578338, - "learning_rate": 1.6846543642074382e-05, - "loss": 0.0605, + "epoch": 0.18712574850299402, + "grad_norm": 0.2355462451136706, + "learning_rate": 1.9948164213739502e-05, + "loss": 0.0558, "step": 125 }, { - "epoch": 0.8571428571428571, - "grad_norm": 0.2211891440249708, - "learning_rate": 1.679273338897293e-05, - "loss": 0.0571, + "epoch": 0.18862275449101795, + "grad_norm": 0.19921818148554826, + "learning_rate": 1.9946507053379998e-05, + "loss": 0.061, "step": 126 }, { - "epoch": 0.8639455782312925, - "grad_norm": 0.22450038327268926, - "learning_rate": 1.6738555441818785e-05, - "loss": 0.0589, + "epoch": 0.19011976047904192, + "grad_norm": 0.24443384088619968, + "learning_rate": 1.994482388994933e-05, + "loss": 0.0605, "step": 127 }, { - "epoch": 0.8707482993197279, - "grad_norm": 0.21870971416073584, - "learning_rate": 1.668401273329129e-05, - "loss": 0.0596, + "epoch": 0.19161676646706588, + "grad_norm": 0.17312333530517943, + "learning_rate": 1.9943114727847782e-05, + "loss": 0.0604, "step": 128 }, { - "epoch": 0.8775510204081632, - "grad_norm": 0.2290461205947771, - "learning_rate": 1.6629108215814523e-05, - "loss": 0.0572, + "epoch": 0.19311377245508982, + "grad_norm": 0.22275107608465397, + "learning_rate": 1.9941379571543597e-05, + "loss": 0.0559, "step": 129 }, { - "epoch": 0.8843537414965986, - "grad_norm": 0.218824327466051, - "learning_rate": 1.6573844861397444e-05, - "loss": 0.059, + "epoch": 0.19461077844311378, + "grad_norm": 0.21284771446183928, + "learning_rate": 1.9939618425572985e-05, + "loss": 0.0577, "step": 130 }, { - "epoch": 0.891156462585034, - "grad_norm": 0.25616970456692517, - "learning_rate": 1.6518225661473045e-05, - "loss": 0.0607, + "epoch": 0.19610778443113772, + "grad_norm": 0.23636170247374946, + "learning_rate": 1.9937831294540094e-05, + "loss": 0.0576, "step": 131 }, { - "epoch": 0.8979591836734694, - "grad_norm": 0.22446082647506313, - "learning_rate": 1.6462253626736413e-05, - "loss": 0.0602, + "epoch": 0.19760479041916168, + "grad_norm": 0.21444998693247203, + "learning_rate": 1.9936018183117003e-05, + "loss": 0.0575, "step": 132 }, { - "epoch": 0.9047619047619048, - "grad_norm": 0.19346153297308857, - "learning_rate": 1.6405931786981753e-05, - "loss": 0.0598, + "epoch": 0.19910179640718562, + "grad_norm": 0.2216207633715879, + "learning_rate": 1.993417909604372e-05, + "loss": 0.0549, "step": 133 }, { - "epoch": 0.9115646258503401, - "grad_norm": 0.19485485871091884, - "learning_rate": 1.63492631909384e-05, - "loss": 0.0583, + "epoch": 0.20059880239520958, + "grad_norm": 0.2520990703952128, + "learning_rate": 1.993231403812815e-05, + "loss": 0.0588, "step": 134 }, { - "epoch": 0.9183673469387755, - "grad_norm": 0.2164924347710122, - "learning_rate": 1.629225090610577e-05, - "loss": 0.0591, + "epoch": 0.20209580838323354, + "grad_norm": 0.22625505105344712, + "learning_rate": 1.99304230142461e-05, + "loss": 0.0545, "step": 135 }, { - "epoch": 0.9251700680272109, - "grad_norm": 0.19853198517283607, - "learning_rate": 1.6234898018587336e-05, - "loss": 0.0578, + "epoch": 0.20359281437125748, + "grad_norm": 0.17746500333059273, + "learning_rate": 1.9928506029341256e-05, + "loss": 0.0588, "step": 136 }, { - "epoch": 0.9319727891156463, - "grad_norm": 0.22323813166846368, - "learning_rate": 1.6177207632923558e-05, - "loss": 0.061, + "epoch": 0.20508982035928144, + "grad_norm": 0.17244614526395785, + "learning_rate": 1.9926563088425177e-05, + "loss": 0.0561, "step": 137 }, { - "epoch": 0.9387755102040817, - "grad_norm": 0.2832642613760529, - "learning_rate": 1.6119182871923834e-05, - "loss": 0.0576, + "epoch": 0.20658682634730538, + "grad_norm": 0.17124109763728518, + "learning_rate": 1.992459419657728e-05, + "loss": 0.0562, "step": 138 }, { - "epoch": 0.9455782312925171, - "grad_norm": 0.1775474704870787, - "learning_rate": 1.606082687649748e-05, - "loss": 0.0575, + "epoch": 0.20808383233532934, + "grad_norm": 0.1984353051118203, + "learning_rate": 1.9922599358944823e-05, + "loss": 0.0563, "step": 139 }, { - "epoch": 0.9523809523809523, - "grad_norm": 0.1974886171121568, - "learning_rate": 1.6002142805483686e-05, - "loss": 0.057, + "epoch": 0.20958083832335328, + "grad_norm": 0.2614654611155798, + "learning_rate": 1.9920578580742892e-05, + "loss": 0.0556, "step": 140 }, { - "epoch": 0.9591836734693877, - "grad_norm": 0.23607667812519084, - "learning_rate": 1.5943133835480536e-05, - "loss": 0.0563, + "epoch": 0.21107784431137724, + "grad_norm": 0.22618621587850632, + "learning_rate": 1.9918531867254386e-05, + "loss": 0.056, "step": 141 }, { - "epoch": 0.9659863945578231, - "grad_norm": 0.2558776300805798, - "learning_rate": 1.588380316067307e-05, - "loss": 0.0581, + "epoch": 0.2125748502994012, + "grad_norm": 0.1686791845721939, + "learning_rate": 1.9916459223830018e-05, + "loss": 0.0549, "step": 142 }, { - "epoch": 0.9727891156462585, - "grad_norm": 0.2571443627011699, - "learning_rate": 1.582415399266036e-05, - "loss": 0.0628, + "epoch": 0.21407185628742514, + "grad_norm": 0.27469186421828573, + "learning_rate": 1.991436065588828e-05, + "loss": 0.0531, "step": 143 }, { - "epoch": 0.9795918367346939, - "grad_norm": 0.3450295396522308, - "learning_rate": 1.5764189560281677e-05, - "loss": 0.0582, + "epoch": 0.2155688622754491, + "grad_norm": 0.28877488507125365, + "learning_rate": 1.9912236168915443e-05, + "loss": 0.0562, "step": 144 }, { - "epoch": 0.9863945578231292, - "grad_norm": 0.26588782094294083, - "learning_rate": 1.5703913109441715e-05, - "loss": 0.0583, + "epoch": 0.21706586826347304, + "grad_norm": 0.1726289274307483, + "learning_rate": 1.991008576846553e-05, + "loss": 0.0538, "step": 145 }, { - "epoch": 0.9931972789115646, - "grad_norm": 0.24694959346022663, - "learning_rate": 1.564332790293487e-05, - "loss": 0.055, + "epoch": 0.218562874251497, + "grad_norm": 0.1912420614588707, + "learning_rate": 1.990790946016032e-05, + "loss": 0.0528, "step": 146 }, { - "epoch": 1.0, - "grad_norm": 0.27312319281208475, - "learning_rate": 1.5582437220268648e-05, - "loss": 0.0626, + "epoch": 0.22005988023952097, + "grad_norm": 0.23379848312481374, + "learning_rate": 1.9905707249689318e-05, + "loss": 0.052, "step": 147 }, { - "epoch": 1.0068027210884354, - "grad_norm": 0.30750819635972554, - "learning_rate": 1.5521244357486132e-05, - "loss": 0.0593, + "epoch": 0.2215568862275449, + "grad_norm": 0.24847391624574525, + "learning_rate": 1.990347914280974e-05, + "loss": 0.0502, "step": 148 }, { - "epoch": 1.0136054421768708, - "grad_norm": 0.24494203240771864, - "learning_rate": 1.5459752626987563e-05, - "loss": 0.054, + "epoch": 0.22305389221556887, + "grad_norm": 0.3441419819000447, + "learning_rate": 1.990122514534651e-05, + "loss": 0.0575, "step": 149 }, { - "epoch": 1.0204081632653061, - "grad_norm": 0.22210056662122235, - "learning_rate": 1.5397965357351035e-05, - "loss": 0.0558, + "epoch": 0.2245508982035928, + "grad_norm": 0.3955212700237536, + "learning_rate": 1.989894526319224e-05, + "loss": 0.0542, "step": 150 }, { - "epoch": 1.0272108843537415, - "grad_norm": 0.23782219601214885, - "learning_rate": 1.5335885893152335e-05, - "loss": 0.0564, + "epoch": 0.22604790419161677, + "grad_norm": 0.2380289767466844, + "learning_rate": 1.98966395023072e-05, + "loss": 0.0537, "step": 151 }, { - "epoch": 1.034013605442177, - "grad_norm": 0.21427038780601312, - "learning_rate": 1.5273517594783878e-05, - "loss": 0.0523, + "epoch": 0.2275449101796407, + "grad_norm": 0.26238101314438406, + "learning_rate": 1.989430786871932e-05, + "loss": 0.054, "step": 152 }, { - "epoch": 1.0408163265306123, - "grad_norm": 0.32614094149577777, - "learning_rate": 1.521086383827282e-05, - "loss": 0.0568, + "epoch": 0.22904191616766467, + "grad_norm": 0.31325198919582375, + "learning_rate": 1.989195036852418e-05, + "loss": 0.0563, "step": 153 }, { - "epoch": 1.0476190476190477, - "grad_norm": 0.1902175321443073, - "learning_rate": 1.5147928015098309e-05, - "loss": 0.0545, + "epoch": 0.23053892215568864, + "grad_norm": 0.27559129084895034, + "learning_rate": 1.9889567007884965e-05, + "loss": 0.0527, "step": 154 }, { - "epoch": 1.054421768707483, - "grad_norm": 0.2087024945007221, - "learning_rate": 1.5084713532007906e-05, - "loss": 0.0541, + "epoch": 0.23203592814371257, + "grad_norm": 0.2036748289588086, + "learning_rate": 1.988715779303248e-05, + "loss": 0.0523, "step": 155 }, { - "epoch": 1.0612244897959184, - "grad_norm": 0.30608404804040057, - "learning_rate": 1.5021223810833165e-05, - "loss": 0.0556, + "epoch": 0.23353293413173654, + "grad_norm": 0.21158749273476493, + "learning_rate": 1.988472273026511e-05, + "loss": 0.0517, "step": 156 }, { - "epoch": 1.0680272108843538, - "grad_norm": 0.27839206067610106, - "learning_rate": 1.4957462288304421e-05, - "loss": 0.0534, + "epoch": 0.23502994011976047, + "grad_norm": 0.3777409399476317, + "learning_rate": 1.9882261825948828e-05, + "loss": 0.0545, "step": 157 }, { - "epoch": 1.0748299319727892, - "grad_norm": 0.2213087101191798, - "learning_rate": 1.489343241586475e-05, - "loss": 0.0547, + "epoch": 0.23652694610778444, + "grad_norm": 0.16198901337349092, + "learning_rate": 1.9879775086517145e-05, + "loss": 0.049, "step": 158 }, { - "epoch": 1.0816326530612246, - "grad_norm": 0.20871107281249643, - "learning_rate": 1.4829137659483144e-05, - "loss": 0.0543, + "epoch": 0.23802395209580837, + "grad_norm": 0.28600911159833825, + "learning_rate": 1.9877262518471133e-05, + "loss": 0.0511, "step": 159 }, { - "epoch": 1.08843537414966, - "grad_norm": 0.23030377388618803, - "learning_rate": 1.4764581499466895e-05, - "loss": 0.0556, + "epoch": 0.23952095808383234, + "grad_norm": 0.21326314303753005, + "learning_rate": 1.9874724128379373e-05, + "loss": 0.0511, "step": 160 }, { - "epoch": 1.0952380952380953, - "grad_norm": 0.1932290087273829, - "learning_rate": 1.4699767430273202e-05, - "loss": 0.0524, + "epoch": 0.2410179640718563, + "grad_norm": 0.23896462883526792, + "learning_rate": 1.9872159922877955e-05, + "loss": 0.052, "step": 161 }, { - "epoch": 1.1020408163265305, - "grad_norm": 0.40409724353966364, - "learning_rate": 1.4634698960320018e-05, - "loss": 0.0527, + "epoch": 0.24251497005988024, + "grad_norm": 0.24778993752258044, + "learning_rate": 1.9869569908670466e-05, + "loss": 0.0503, "step": 162 }, { - "epoch": 1.1088435374149659, - "grad_norm": 0.22092199900707887, - "learning_rate": 1.4569379611796137e-05, - "loss": 0.0531, + "epoch": 0.2440119760479042, + "grad_norm": 0.2678509446705851, + "learning_rate": 1.9866954092527954e-05, + "loss": 0.0509, "step": 163 }, { - "epoch": 1.1156462585034013, - "grad_norm": 0.23715254386252377, - "learning_rate": 1.4503812920470535e-05, - "loss": 0.052, + "epoch": 0.24550898203592814, + "grad_norm": 0.4976820467779877, + "learning_rate": 1.9864312481288934e-05, + "loss": 0.0452, "step": 164 }, { - "epoch": 1.1224489795918366, - "grad_norm": 0.25222406455269336, - "learning_rate": 1.443800243550098e-05, - "loss": 0.0507, + "epoch": 0.2470059880239521, + "grad_norm": 0.21451893299135652, + "learning_rate": 1.9861645081859334e-05, + "loss": 0.0503, "step": 165 }, { - "epoch": 1.129251700680272, - "grad_norm": 0.20150273343577169, - "learning_rate": 1.4371951719241906e-05, - "loss": 0.0572, + "epoch": 0.24850299401197604, + "grad_norm": 0.18172472245421414, + "learning_rate": 1.9858951901212525e-05, + "loss": 0.0467, "step": 166 }, { - "epoch": 1.1360544217687074, - "grad_norm": 0.2981137590723175, - "learning_rate": 1.4305664347051586e-05, - "loss": 0.0557, + "epoch": 0.25, + "grad_norm": 0.1831777829268529, + "learning_rate": 1.9856232946389264e-05, + "loss": 0.046, "step": 167 }, { - "epoch": 1.1428571428571428, - "grad_norm": 0.36363358730765705, - "learning_rate": 1.423914390709861e-05, - "loss": 0.0499, + "epoch": 0.25149700598802394, + "grad_norm": 0.24695897142692572, + "learning_rate": 1.985348822449769e-05, + "loss": 0.0449, "step": 168 }, { - "epoch": 1.1496598639455782, - "grad_norm": 0.3489948294285827, - "learning_rate": 1.4172394000167625e-05, - "loss": 0.0559, + "epoch": 0.25299401197604793, + "grad_norm": 0.24320724433114305, + "learning_rate": 1.985071774271331e-05, + "loss": 0.0514, "step": 169 }, { - "epoch": 1.1564625850340136, - "grad_norm": 0.25108294665044295, - "learning_rate": 1.4105418239464452e-05, - "loss": 0.0516, + "epoch": 0.25449101796407186, + "grad_norm": 0.29316721920718963, + "learning_rate": 1.9847921508278977e-05, + "loss": 0.0455, "step": 170 }, { - "epoch": 1.163265306122449, - "grad_norm": 0.31647443965414784, - "learning_rate": 1.4038220250420487e-05, - "loss": 0.0535, + "epoch": 0.2559880239520958, + "grad_norm": 0.24473542108780105, + "learning_rate": 1.984509952850485e-05, + "loss": 0.0477, "step": 171 }, { - "epoch": 1.1700680272108843, - "grad_norm": 0.3026605870439961, - "learning_rate": 1.3970803670496453e-05, - "loss": 0.0501, + "epoch": 0.25748502994011974, + "grad_norm": 0.29195204962744054, + "learning_rate": 1.9842251810768415e-05, + "loss": 0.0454, "step": 172 }, { - "epoch": 1.1768707482993197, - "grad_norm": 0.3310780887181829, - "learning_rate": 1.390317214898551e-05, - "loss": 0.0541, + "epoch": 0.25898203592814373, + "grad_norm": 0.26054012611614774, + "learning_rate": 1.983937836251444e-05, + "loss": 0.0475, "step": 173 }, { - "epoch": 1.183673469387755, - "grad_norm": 0.23883640286772895, - "learning_rate": 1.3835329346815716e-05, - "loss": 0.052, + "epoch": 0.26047904191616766, + "grad_norm": 0.2683808403595366, + "learning_rate": 1.983647919125495e-05, + "loss": 0.0469, "step": 174 }, { - "epoch": 1.1904761904761905, - "grad_norm": 0.3668380546284328, - "learning_rate": 1.3767278936351853e-05, - "loss": 0.0521, + "epoch": 0.2619760479041916, + "grad_norm": 0.2798959888989543, + "learning_rate": 1.983355430456923e-05, + "loss": 0.0457, "step": 175 }, { - "epoch": 1.1972789115646258, - "grad_norm": 0.3165490998015608, - "learning_rate": 1.3699024601196641e-05, - "loss": 0.05, + "epoch": 0.2634730538922156, + "grad_norm": 0.3230989454804457, + "learning_rate": 1.983060371010378e-05, + "loss": 0.0481, "step": 176 }, { - "epoch": 1.2040816326530612, - "grad_norm": 0.28757144157751063, - "learning_rate": 1.3630570035991352e-05, - "loss": 0.0523, + "epoch": 0.26497005988023953, + "grad_norm": 0.28580432616804236, + "learning_rate": 1.9827627415572328e-05, + "loss": 0.0428, "step": 177 }, { - "epoch": 1.2108843537414966, - "grad_norm": 0.24922250771310617, - "learning_rate": 1.3561918946215807e-05, - "loss": 0.0476, + "epoch": 0.26646706586826346, + "grad_norm": 0.3713352257679464, + "learning_rate": 1.982462542875576e-05, + "loss": 0.0467, "step": 178 }, { - "epoch": 1.217687074829932, - "grad_norm": 0.3109432126479953, - "learning_rate": 1.34930750479878e-05, - "loss": 0.0528, + "epoch": 0.2679640718562874, + "grad_norm": 0.2792642088474226, + "learning_rate": 1.982159775750216e-05, + "loss": 0.0468, "step": 179 }, { - "epoch": 1.2244897959183674, - "grad_norm": 0.3191250919868236, - "learning_rate": 1.3424042067861944e-05, - "loss": 0.0488, + "epoch": 0.2694610778443114, + "grad_norm": 0.26225012059799724, + "learning_rate": 1.9818544409726734e-05, + "loss": 0.0481, "step": 180 }, { - "epoch": 1.2312925170068028, - "grad_norm": 0.33504846480374884, - "learning_rate": 1.335482374262795e-05, - "loss": 0.0485, + "epoch": 0.27095808383233533, + "grad_norm": 0.320954008862723, + "learning_rate": 1.981546539341183e-05, + "loss": 0.0444, "step": 181 }, { - "epoch": 1.2380952380952381, - "grad_norm": 0.41636172795109355, - "learning_rate": 1.3285423819108349e-05, - "loss": 0.0481, + "epoch": 0.27245508982035926, + "grad_norm": 0.3027600817067956, + "learning_rate": 1.9812360716606886e-05, + "loss": 0.0461, "step": 182 }, { - "epoch": 1.2448979591836735, - "grad_norm": 0.43892808658376165, - "learning_rate": 1.3215846053955683e-05, - "loss": 0.0486, + "epoch": 0.27395209580838326, + "grad_norm": 0.28110581468430873, + "learning_rate": 1.9809230387428444e-05, + "loss": 0.0445, "step": 183 }, { - "epoch": 1.251700680272109, - "grad_norm": 0.32837726008368157, - "learning_rate": 1.3146094213449148e-05, - "loss": 0.0451, + "epoch": 0.2754491017964072, + "grad_norm": 0.30945413021599966, + "learning_rate": 1.980607441406009e-05, + "loss": 0.0441, "step": 184 }, { - "epoch": 1.2585034013605443, - "grad_norm": 0.24228613984412564, - "learning_rate": 1.3076172073290726e-05, - "loss": 0.046, + "epoch": 0.27694610778443113, + "grad_norm": 0.29814499120640103, + "learning_rate": 1.9802892804752462e-05, + "loss": 0.0434, "step": 185 }, { - "epoch": 1.2653061224489797, - "grad_norm": 0.42623267395045844, - "learning_rate": 1.3006083418400799e-05, - "loss": 0.0495, + "epoch": 0.27844311377245506, + "grad_norm": 0.34230417502487365, + "learning_rate": 1.9799685567823215e-05, + "loss": 0.0444, "step": 186 }, { - "epoch": 1.272108843537415, - "grad_norm": 0.3067088420671033, - "learning_rate": 1.2935832042713288e-05, - "loss": 0.0506, + "epoch": 0.27994011976047906, + "grad_norm": 0.2518591743527993, + "learning_rate": 1.9796452711657002e-05, + "loss": 0.0397, "step": 187 }, { - "epoch": 1.2789115646258504, - "grad_norm": 0.3652294320037095, - "learning_rate": 1.2865421748970257e-05, - "loss": 0.0493, + "epoch": 0.281437125748503, + "grad_norm": 0.3292942776587782, + "learning_rate": 1.9793194244705453e-05, + "loss": 0.0452, "step": 188 }, { - "epoch": 1.2857142857142856, - "grad_norm": 0.4171203977029359, - "learning_rate": 1.2794856348516095e-05, - "loss": 0.0458, + "epoch": 0.28293413173652693, + "grad_norm": 0.2582695780716865, + "learning_rate": 1.9789910175487147e-05, + "loss": 0.0418, "step": 189 }, { - "epoch": 1.2925170068027212, - "grad_norm": 0.2701855372697879, - "learning_rate": 1.2724139661091188e-05, - "loss": 0.0432, + "epoch": 0.2844311377245509, + "grad_norm": 0.2730893776861321, + "learning_rate": 1.97866005125876e-05, + "loss": 0.0402, "step": 190 }, { - "epoch": 1.2993197278911564, - "grad_norm": 0.3259307296746982, - "learning_rate": 1.2653275514625165e-05, - "loss": 0.0463, + "epoch": 0.28592814371257486, + "grad_norm": 0.3111167647085528, + "learning_rate": 1.9783265264659244e-05, + "loss": 0.0395, "step": 191 }, { - "epoch": 1.306122448979592, - "grad_norm": 0.2550766456669367, - "learning_rate": 1.2582267745029685e-05, - "loss": 0.0447, + "epoch": 0.2874251497005988, + "grad_norm": 0.39076695341386136, + "learning_rate": 1.977990444042138e-05, + "loss": 0.0514, "step": 192 }, { - "epoch": 1.3129251700680271, - "grad_norm": 0.4220940884592931, - "learning_rate": 1.2511120195990797e-05, - "loss": 0.048, + "epoch": 0.28892215568862273, + "grad_norm": 0.43768952581558807, + "learning_rate": 1.9776518048660188e-05, + "loss": 0.046, "step": 193 }, { - "epoch": 1.3197278911564627, - "grad_norm": 0.36984656348421135, - "learning_rate": 1.2439836718760887e-05, - "loss": 0.0411, + "epoch": 0.2904191616766467, + "grad_norm": 0.3165460062789067, + "learning_rate": 1.9773106098228685e-05, + "loss": 0.0449, "step": 194 }, { - "epoch": 1.3265306122448979, - "grad_norm": 0.36136346549964, - "learning_rate": 1.2368421171950193e-05, - "loss": 0.046, + "epoch": 0.29191616766467066, + "grad_norm": 0.28739375717770865, + "learning_rate": 1.97696685980467e-05, + "loss": 0.0434, "step": 195 }, { - "epoch": 1.3333333333333333, - "grad_norm": 0.3788271776627097, - "learning_rate": 1.2296877421317958e-05, - "loss": 0.0437, + "epoch": 0.2934131736526946, + "grad_norm": 0.32294252936754736, + "learning_rate": 1.976620555710087e-05, + "loss": 0.0447, "step": 196 }, { - "epoch": 1.3401360544217686, - "grad_norm": 0.28520654057079897, - "learning_rate": 1.2225209339563144e-05, - "loss": 0.0418, + "epoch": 0.2949101796407186, + "grad_norm": 0.3345999711440401, + "learning_rate": 1.9762716984444585e-05, + "loss": 0.0456, "step": 197 }, { - "epoch": 1.346938775510204, - "grad_norm": 0.49694987958709724, - "learning_rate": 1.215342080611484e-05, - "loss": 0.0448, + "epoch": 0.2964071856287425, + "grad_norm": 0.31106180218416374, + "learning_rate": 1.9759202889197994e-05, + "loss": 0.0389, "step": 198 }, { - "epoch": 1.3537414965986394, - "grad_norm": 0.442980712849271, - "learning_rate": 1.2081515706922226e-05, - "loss": 0.0477, + "epoch": 0.29790419161676646, + "grad_norm": 0.25582092602450185, + "learning_rate": 1.975566328054797e-05, + "loss": 0.0433, "step": 199 }, { - "epoch": 1.3605442176870748, - "grad_norm": 0.30908251467526043, - "learning_rate": 1.2009497934244257e-05, - "loss": 0.0486, + "epoch": 0.2994011976047904, + "grad_norm": 0.2037709453766463, + "learning_rate": 1.9752098167748083e-05, + "loss": 0.0348, "step": 200 }, { - "epoch": 1.3673469387755102, - "grad_norm": 0.31690703603059645, - "learning_rate": 1.1937371386438954e-05, - "loss": 0.0434, + "epoch": 0.3008982035928144, + "grad_norm": 0.34634898932379976, + "learning_rate": 1.9748507560118573e-05, + "loss": 0.0398, "step": 201 }, { - "epoch": 1.3741496598639455, - "grad_norm": 0.3622937591780268, - "learning_rate": 1.186513996775239e-05, - "loss": 0.0442, + "epoch": 0.3023952095808383, + "grad_norm": 0.38910405891750727, + "learning_rate": 1.974489146704634e-05, + "loss": 0.0393, "step": 202 }, { - "epoch": 1.380952380952381, - "grad_norm": 0.30716164525104495, - "learning_rate": 1.1792807588107358e-05, - "loss": 0.0436, + "epoch": 0.30389221556886226, + "grad_norm": 0.33295863522793434, + "learning_rate": 1.9741249897984907e-05, + "loss": 0.0384, "step": 203 }, { - "epoch": 1.3877551020408163, - "grad_norm": 0.3321484566845795, - "learning_rate": 1.1720378162891709e-05, - "loss": 0.0464, + "epoch": 0.30538922155688625, + "grad_norm": 0.21725882145501477, + "learning_rate": 1.9737582862454398e-05, + "loss": 0.034, "step": 204 }, { - "epoch": 1.3945578231292517, - "grad_norm": 0.3019164729518717, - "learning_rate": 1.1647855612746423e-05, - "loss": 0.0421, + "epoch": 0.3068862275449102, + "grad_norm": 0.3222703155967036, + "learning_rate": 1.9733890370041505e-05, + "loss": 0.0379, "step": 205 }, { - "epoch": 1.401360544217687, - "grad_norm": 0.28208298182130315, - "learning_rate": 1.1575243863353383e-05, - "loss": 0.0384, + "epoch": 0.3083832335329341, + "grad_norm": 0.36543573520717576, + "learning_rate": 1.9730172430399494e-05, + "loss": 0.0368, "step": 206 }, { - "epoch": 1.4081632653061225, - "grad_norm": 0.27187917628455394, - "learning_rate": 1.150254684522286e-05, - "loss": 0.0395, + "epoch": 0.30988023952095806, + "grad_norm": 0.2972736130414979, + "learning_rate": 1.972642905324813e-05, + "loss": 0.0387, "step": 207 }, { - "epoch": 1.4149659863945578, - "grad_norm": 0.29967306578610386, - "learning_rate": 1.142976849348078e-05, - "loss": 0.0408, + "epoch": 0.31137724550898205, + "grad_norm": 0.27528054613918207, + "learning_rate": 1.9722660248373707e-05, + "loss": 0.0337, "step": 208 }, { - "epoch": 1.4217687074829932, - "grad_norm": 0.22436230612440464, - "learning_rate": 1.1356912747655687e-05, - "loss": 0.0339, + "epoch": 0.312874251497006, + "grad_norm": 0.2514931870827675, + "learning_rate": 1.971886602562897e-05, + "loss": 0.0359, "step": 209 }, { - "epoch": 1.4285714285714286, - "grad_norm": 0.42439154027724657, - "learning_rate": 1.1283983551465512e-05, - "loss": 0.0476, + "epoch": 0.3143712574850299, + "grad_norm": 0.33119244111260354, + "learning_rate": 1.9715046394933126e-05, + "loss": 0.0377, "step": 210 }, { - "epoch": 1.435374149659864, - "grad_norm": 0.31700756503083577, - "learning_rate": 1.1210984852604084e-05, - "loss": 0.0358, + "epoch": 0.3158682634730539, + "grad_norm": 0.2644246425467484, + "learning_rate": 1.971120136627181e-05, + "loss": 0.0391, "step": 211 }, { - "epoch": 1.4421768707482994, - "grad_norm": 0.3378079110420009, - "learning_rate": 1.1137920602527448e-05, - "loss": 0.0379, + "epoch": 0.31736526946107785, + "grad_norm": 0.23488800200566334, + "learning_rate": 1.970733094969704e-05, + "loss": 0.0305, "step": 212 }, { - "epoch": 1.4489795918367347, - "grad_norm": 0.24349092505484088, - "learning_rate": 1.1064794756239978e-05, - "loss": 0.0363, + "epoch": 0.3188622754491018, + "grad_norm": 0.3336820637038372, + "learning_rate": 1.970343515532722e-05, + "loss": 0.0347, "step": 213 }, { - "epoch": 1.4557823129251701, - "grad_norm": 0.3400711419864756, - "learning_rate": 1.099161127208027e-05, - "loss": 0.0377, + "epoch": 0.3203592814371258, + "grad_norm": 0.32537350598910286, + "learning_rate": 1.969951399334709e-05, + "loss": 0.0383, "step": 214 }, { - "epoch": 1.4625850340136055, - "grad_norm": 0.5100291242139185, - "learning_rate": 1.0918374111506893e-05, - "loss": 0.0382, + "epoch": 0.3218562874251497, + "grad_norm": 0.4477662129282192, + "learning_rate": 1.9695567474007713e-05, + "loss": 0.0375, "step": 215 }, { - "epoch": 1.469387755102041, - "grad_norm": 0.28802764191587965, - "learning_rate": 1.0845087238883945e-05, - "loss": 0.032, + "epoch": 0.32335329341317365, + "grad_norm": 0.39632766587050555, + "learning_rate": 1.9691595607626443e-05, + "loss": 0.0306, "step": 216 }, { - "epoch": 1.4761904761904763, - "grad_norm": 0.2607923120434538, - "learning_rate": 1.0771754621266466e-05, - "loss": 0.0365, + "epoch": 0.3248502994011976, + "grad_norm": 0.31164368475238186, + "learning_rate": 1.9687598404586895e-05, + "loss": 0.0317, "step": 217 }, { - "epoch": 1.4829931972789114, - "grad_norm": 0.2640244555130478, - "learning_rate": 1.0698380228185685e-05, - "loss": 0.0433, + "epoch": 0.3263473053892216, + "grad_norm": 0.29356080957866204, + "learning_rate": 1.9683575875338922e-05, + "loss": 0.0319, "step": 218 }, { - "epoch": 1.489795918367347, - "grad_norm": 0.2849877896554485, - "learning_rate": 1.0624968031434174e-05, - "loss": 0.0426, + "epoch": 0.3278443113772455, + "grad_norm": 0.3170039835076496, + "learning_rate": 1.9679528030398596e-05, + "loss": 0.0329, "step": 219 }, { - "epoch": 1.4965986394557822, - "grad_norm": 0.34138049373172324, - "learning_rate": 1.0551522004850821e-05, - "loss": 0.042, + "epoch": 0.32934131736526945, + "grad_norm": 0.30665504238301144, + "learning_rate": 1.9675454880348158e-05, + "loss": 0.0279, "step": 220 }, { - "epoch": 1.5034013605442178, - "grad_norm": 0.4028334423985101, - "learning_rate": 1.0478046124105746e-05, - "loss": 0.032, + "epoch": 0.33083832335329344, + "grad_norm": 0.25789142501667217, + "learning_rate": 1.967135643583601e-05, + "loss": 0.029, "step": 221 }, { - "epoch": 1.510204081632653, - "grad_norm": 0.29976572412890073, - "learning_rate": 1.0404544366485094e-05, - "loss": 0.0318, + "epoch": 0.3323353293413174, + "grad_norm": 0.29710874117629504, + "learning_rate": 1.966723270757669e-05, + "loss": 0.0339, "step": 222 }, { - "epoch": 1.5170068027210886, - "grad_norm": 0.3575056341024037, - "learning_rate": 1.033102071067573e-05, - "loss": 0.0328, + "epoch": 0.3338323353293413, + "grad_norm": 0.3797724903024257, + "learning_rate": 1.966308370635082e-05, + "loss": 0.0317, "step": 223 }, { - "epoch": 1.5238095238095237, - "grad_norm": 0.42003758608591496, - "learning_rate": 1.0257479136549889e-05, - "loss": 0.0412, + "epoch": 0.33532934131736525, + "grad_norm": 0.3234195979276272, + "learning_rate": 1.96589094430051e-05, + "loss": 0.0282, "step": 224 }, { - "epoch": 1.5306122448979593, - "grad_norm": 0.2646979747322718, - "learning_rate": 1.0183923624949721e-05, - "loss": 0.0359, + "epoch": 0.33682634730538924, + "grad_norm": 0.3322138017326269, + "learning_rate": 1.965470992845227e-05, + "loss": 0.0289, "step": 225 }, { - "epoch": 1.5374149659863945, - "grad_norm": 0.2822553257788332, - "learning_rate": 1.0110358157471825e-05, - "loss": 0.0306, + "epoch": 0.3383233532934132, + "grad_norm": 0.24818051146540246, + "learning_rate": 1.9650485173671092e-05, + "loss": 0.0221, "step": 226 }, { - "epoch": 1.54421768707483, - "grad_norm": 0.46122155273972015, - "learning_rate": 1.0036786716251721e-05, - "loss": 0.0457, + "epoch": 0.3398203592814371, + "grad_norm": 0.42659472820256933, + "learning_rate": 1.96462351897063e-05, + "loss": 0.031, "step": 227 }, { - "epoch": 1.5510204081632653, - "grad_norm": 0.44028200523172295, - "learning_rate": 9.963213283748282e-06, - "loss": 0.0348, + "epoch": 0.3413173652694611, + "grad_norm": 0.4005980652683074, + "learning_rate": 1.9641959987668603e-05, + "loss": 0.0365, "step": 228 }, { - "epoch": 1.5578231292517006, - "grad_norm": 0.33390274335640047, - "learning_rate": 9.889641842528179e-06, - "loss": 0.0385, + "epoch": 0.34281437125748504, + "grad_norm": 0.4685928568147018, + "learning_rate": 1.9637659578734615e-05, + "loss": 0.0327, "step": 229 }, { - "epoch": 1.564625850340136, - "grad_norm": 0.35548363557576734, - "learning_rate": 9.816076375050284e-06, - "loss": 0.0316, + "epoch": 0.344311377245509, + "grad_norm": 0.3863947245851545, + "learning_rate": 1.9633333974146867e-05, + "loss": 0.0267, "step": 230 }, { - "epoch": 1.5714285714285714, - "grad_norm": 0.2988165557824134, - "learning_rate": 9.742520863450116e-06, - "loss": 0.0349, + "epoch": 0.3458083832335329, + "grad_norm": 0.5573187059327918, + "learning_rate": 1.9628983185213747e-05, + "loss": 0.0279, "step": 231 }, { - "epoch": 1.5782312925170068, - "grad_norm": 0.31453672481100403, - "learning_rate": 9.668979289324274e-06, - "loss": 0.0318, + "epoch": 0.3473053892215569, + "grad_norm": 0.3063425475105623, + "learning_rate": 1.962460722330949e-05, + "loss": 0.0254, "step": 232 }, { - "epoch": 1.5850340136054422, - "grad_norm": 0.2910996810127783, - "learning_rate": 9.595455633514908e-06, - "loss": 0.0254, + "epoch": 0.34880239520958084, + "grad_norm": 0.25300385871670333, + "learning_rate": 1.9620206099874135e-05, + "loss": 0.0238, "step": 233 }, { - "epoch": 1.5918367346938775, - "grad_norm": 0.27024499311466754, - "learning_rate": 9.521953875894256e-06, - "loss": 0.0317, + "epoch": 0.3502994011976048, + "grad_norm": 0.33330643618442385, + "learning_rate": 1.96157798264135e-05, + "loss": 0.023, "step": 234 }, { - "epoch": 1.598639455782313, - "grad_norm": 0.3192737285714962, - "learning_rate": 9.448477995149182e-06, - "loss": 0.0346, + "epoch": 0.35179640718562877, + "grad_norm": 0.3374148649272243, + "learning_rate": 1.9611328414499164e-05, + "loss": 0.0245, "step": 235 }, { - "epoch": 1.6054421768707483, - "grad_norm": 0.31462197089436433, - "learning_rate": 9.37503196856583e-06, - "loss": 0.0347, + "epoch": 0.3532934131736527, + "grad_norm": 0.38153658162458814, + "learning_rate": 1.9606851875768404e-05, + "loss": 0.0227, "step": 236 }, { - "epoch": 1.6122448979591837, - "grad_norm": 0.4780207494130082, - "learning_rate": 9.301619771814317e-06, - "loss": 0.0344, + "epoch": 0.35479041916167664, + "grad_norm": 0.31854741219630106, + "learning_rate": 1.9602350221924208e-05, + "loss": 0.0194, "step": 237 }, { - "epoch": 1.619047619047619, - "grad_norm": 0.3603665494734086, - "learning_rate": 9.228245378733537e-06, - "loss": 0.0284, + "epoch": 0.3562874251497006, + "grad_norm": 0.3245371702157616, + "learning_rate": 1.9597823464735204e-05, + "loss": 0.0231, "step": 238 }, { - "epoch": 1.6258503401360545, - "grad_norm": 0.3481977799989387, - "learning_rate": 9.154912761116056e-06, - "loss": 0.0319, + "epoch": 0.35778443113772457, + "grad_norm": 0.3260937932412784, + "learning_rate": 1.9593271616035666e-05, + "loss": 0.0201, "step": 239 }, { - "epoch": 1.6326530612244898, - "grad_norm": 0.31718948127093954, - "learning_rate": 9.081625888493107e-06, - "loss": 0.0347, + "epoch": 0.3592814371257485, + "grad_norm": 0.313463122278002, + "learning_rate": 1.9588694687725448e-05, + "loss": 0.0203, "step": 240 }, { - "epoch": 1.6394557823129252, - "grad_norm": 0.39602394734020263, - "learning_rate": 9.00838872791973e-06, - "loss": 0.0389, + "epoch": 0.36077844311377244, + "grad_norm": 0.4452480165482146, + "learning_rate": 1.9584092691769978e-05, + "loss": 0.0213, "step": 241 }, { - "epoch": 1.6462585034013606, - "grad_norm": 0.2811227178639146, - "learning_rate": 8.935205243760022e-06, - "loss": 0.0301, + "epoch": 0.36227544910179643, + "grad_norm": 0.38506762755559276, + "learning_rate": 1.9579465640200213e-05, + "loss": 0.0227, "step": 242 }, { - "epoch": 1.6530612244897958, - "grad_norm": 0.2738235813523902, - "learning_rate": 8.862079397472552e-06, - "loss": 0.0295, + "epoch": 0.36377245508982037, + "grad_norm": 0.39214871815864943, + "learning_rate": 1.957481354511262e-05, + "loss": 0.0199, "step": 243 }, { - "epoch": 1.6598639455782314, - "grad_norm": 0.325061391409321, - "learning_rate": 8.78901514739592e-06, - "loss": 0.029, + "epoch": 0.3652694610778443, + "grad_norm": 0.3539339228414202, + "learning_rate": 1.957013641866913e-05, + "loss": 0.0225, "step": 244 }, { - "epoch": 1.6666666666666665, - "grad_norm": 0.3027200222791084, - "learning_rate": 8.71601644853449e-06, - "loss": 0.0338, + "epoch": 0.36676646706586824, + "grad_norm": 0.3539330765036243, + "learning_rate": 1.9565434273097114e-05, + "loss": 0.0207, "step": 245 }, { - "epoch": 1.6734693877551021, - "grad_norm": 0.3891063981452845, - "learning_rate": 8.643087252344313e-06, - "loss": 0.0299, + "epoch": 0.36826347305389223, + "grad_norm": 0.3753514827965324, + "learning_rate": 1.9560707120689354e-05, + "loss": 0.021, "step": 246 }, { - "epoch": 1.6802721088435373, - "grad_norm": 0.31302794618115964, - "learning_rate": 8.57023150651922e-06, - "loss": 0.03, + "epoch": 0.36976047904191617, + "grad_norm": 0.40608828993861595, + "learning_rate": 1.9555954973804013e-05, + "loss": 0.0181, "step": 247 }, { - "epoch": 1.6870748299319729, - "grad_norm": 0.3449817518564228, - "learning_rate": 8.49745315477714e-06, - "loss": 0.0317, + "epoch": 0.3712574850299401, + "grad_norm": 0.3303053398457487, + "learning_rate": 1.9551177844864577e-05, + "loss": 0.0208, "step": 248 }, { - "epoch": 1.693877551020408, - "grad_norm": 0.29810889691448594, - "learning_rate": 8.424756136646624e-06, - "loss": 0.0291, + "epoch": 0.3727544910179641, + "grad_norm": 0.41949758179222085, + "learning_rate": 1.954637574635986e-05, + "loss": 0.0212, "step": 249 }, { - "epoch": 1.7006802721088436, - "grad_norm": 0.3798859940115675, - "learning_rate": 8.352144387253582e-06, - "loss": 0.0322, + "epoch": 0.37425149700598803, + "grad_norm": 0.2764763651977154, + "learning_rate": 1.954154869084395e-05, + "loss": 0.0159, "step": 250 }, { - "epoch": 1.7074829931972788, - "grad_norm": 0.21879122540865556, - "learning_rate": 8.279621837108295e-06, - "loss": 0.0317, + "epoch": 0.37574850299401197, + "grad_norm": 0.39212670974147856, + "learning_rate": 1.9536696690936177e-05, + "loss": 0.0189, "step": 251 }, { - "epoch": 1.7142857142857144, - "grad_norm": 0.35646635601921695, - "learning_rate": 8.207192411892645e-06, - "loss": 0.0266, + "epoch": 0.3772455089820359, + "grad_norm": 0.32025921970118243, + "learning_rate": 1.9531819759321083e-05, + "loss": 0.0169, "step": 252 }, { - "epoch": 1.7210884353741496, - "grad_norm": 0.32902401726950553, - "learning_rate": 8.134860032247613e-06, - "loss": 0.0281, + "epoch": 0.3787425149700599, + "grad_norm": 0.3429001255459339, + "learning_rate": 1.9526917908748395e-05, + "loss": 0.0232, "step": 253 }, { - "epoch": 1.7278911564625852, - "grad_norm": 0.2954321296424936, - "learning_rate": 8.062628613561051e-06, - "loss": 0.0288, + "epoch": 0.38023952095808383, + "grad_norm": 0.37665262959840423, + "learning_rate": 1.9521991152032974e-05, + "loss": 0.023, "step": 254 }, { - "epoch": 1.7346938775510203, - "grad_norm": 0.33576585838310874, - "learning_rate": 7.990502065755748e-06, - "loss": 0.0283, + "epoch": 0.38173652694610777, + "grad_norm": 0.2649408332614438, + "learning_rate": 1.9517039502054805e-05, + "loss": 0.0153, "step": 255 }, { - "epoch": 1.741496598639456, - "grad_norm": 0.372785587382324, - "learning_rate": 7.918484293077777e-06, - "loss": 0.0325, + "epoch": 0.38323353293413176, + "grad_norm": 0.2772926318508054, + "learning_rate": 1.9512062971758937e-05, + "loss": 0.0177, "step": 256 }, { - "epoch": 1.748299319727891, - "grad_norm": 0.2699962005993758, - "learning_rate": 7.846579193885165e-06, - "loss": 0.0245, + "epoch": 0.3847305389221557, + "grad_norm": 0.333717122655441, + "learning_rate": 1.950706157415548e-05, + "loss": 0.0175, "step": 257 }, { - "epoch": 1.7551020408163265, - "grad_norm": 0.4069704653536772, - "learning_rate": 7.774790660436857e-06, - "loss": 0.0331, + "epoch": 0.38622754491017963, + "grad_norm": 0.2881853890457447, + "learning_rate": 1.9502035322319546e-05, + "loss": 0.0197, "step": 258 }, { - "epoch": 1.7619047619047619, - "grad_norm": 0.2543285657041178, - "learning_rate": 7.703122578682047e-06, - "loss": 0.0195, + "epoch": 0.38772455089820357, + "grad_norm": 0.3011392842260328, + "learning_rate": 1.949698422939122e-05, + "loss": 0.0194, "step": 259 }, { - "epoch": 1.7687074829931972, - "grad_norm": 0.49851302483614535, - "learning_rate": 7.631578828049809e-06, - "loss": 0.0228, + "epoch": 0.38922155688622756, + "grad_norm": 0.25182071202500195, + "learning_rate": 1.9491908308575537e-05, + "loss": 0.0168, "step": 260 }, { - "epoch": 1.7755102040816326, - "grad_norm": 0.34904705107956824, - "learning_rate": 7.560163281239116e-06, - "loss": 0.0248, + "epoch": 0.3907185628742515, + "grad_norm": 0.33730858021035043, + "learning_rate": 1.948680757314243e-05, + "loss": 0.0132, "step": 261 }, { - "epoch": 1.782312925170068, - "grad_norm": 0.3079169045046351, - "learning_rate": 7.488879804009206e-06, - "loss": 0.029, + "epoch": 0.39221556886227543, + "grad_norm": 0.3785903917651171, + "learning_rate": 1.9481682036426707e-05, + "loss": 0.0151, "step": 262 }, { - "epoch": 1.7891156462585034, - "grad_norm": 0.32573423282858954, - "learning_rate": 7.4177322549703165e-06, - "loss": 0.0246, + "epoch": 0.3937125748502994, + "grad_norm": 0.21775764169004, + "learning_rate": 1.9476531711828027e-05, + "loss": 0.0133, "step": 263 }, { - "epoch": 1.7959183673469388, - "grad_norm": 0.3112809895061056, - "learning_rate": 7.346724485374837e-06, - "loss": 0.0299, + "epoch": 0.39520958083832336, + "grad_norm": 0.22229766801022924, + "learning_rate": 1.9471356612810832e-05, + "loss": 0.0103, "step": 264 }, { - "epoch": 1.8027210884353742, - "grad_norm": 0.2874275900338384, - "learning_rate": 7.275860338908815e-06, - "loss": 0.0299, + "epoch": 0.3967065868263473, + "grad_norm": 0.3032194913024956, + "learning_rate": 1.9466156752904344e-05, + "loss": 0.0131, "step": 265 }, { - "epoch": 1.8095238095238095, - "grad_norm": 0.37722251427133746, - "learning_rate": 7.2051436514839064e-06, - "loss": 0.0295, + "epoch": 0.39820359281437123, + "grad_norm": 0.3924806246674395, + "learning_rate": 1.9460932145702514e-05, + "loss": 0.0164, "step": 266 }, { - "epoch": 1.816326530612245, - "grad_norm": 0.4495052816684312, - "learning_rate": 7.134578251029745e-06, - "loss": 0.0287, + "epoch": 0.3997005988023952, + "grad_norm": 0.2958650726177419, + "learning_rate": 1.9455682804863986e-05, + "loss": 0.0132, "step": 267 }, { - "epoch": 1.8231292517006803, - "grad_norm": 0.4383184010849996, - "learning_rate": 7.064167957286714e-06, - "loss": 0.0276, + "epoch": 0.40119760479041916, + "grad_norm": 0.32232453177932663, + "learning_rate": 1.9450408744112072e-05, + "loss": 0.016, "step": 268 }, { - "epoch": 1.8299319727891157, - "grad_norm": 0.34023669062637585, - "learning_rate": 6.993916581599203e-06, - "loss": 0.0271, + "epoch": 0.4026946107784431, + "grad_norm": 0.3492194562524053, + "learning_rate": 1.9445109977234704e-05, + "loss": 0.0145, "step": 269 }, { - "epoch": 1.836734693877551, - "grad_norm": 0.4710800088561054, - "learning_rate": 6.923827926709277e-06, - "loss": 0.0409, + "epoch": 0.4041916167664671, + "grad_norm": 0.374197844485198, + "learning_rate": 1.9439786518084404e-05, + "loss": 0.0135, "step": 270 }, { - "epoch": 1.8435374149659864, - "grad_norm": 0.5275268904537102, - "learning_rate": 6.853905786550855e-06, - "loss": 0.0296, + "epoch": 0.405688622754491, + "grad_norm": 0.3268312149590773, + "learning_rate": 1.9434438380578248e-05, + "loss": 0.0164, "step": 271 }, { - "epoch": 1.8503401360544216, - "grad_norm": 0.33912114616353456, - "learning_rate": 6.784153946044321e-06, - "loss": 0.0292, + "epoch": 0.40718562874251496, + "grad_norm": 0.22923876520135972, + "learning_rate": 1.9429065578697833e-05, + "loss": 0.0121, "step": 272 }, { - "epoch": 1.8571428571428572, - "grad_norm": 0.5301308829618651, - "learning_rate": 6.714576180891653e-06, - "loss": 0.0279, + "epoch": 0.4086826347305389, + "grad_norm": 0.29090023648655766, + "learning_rate": 1.942366812648922e-05, + "loss": 0.0103, "step": 273 }, { - "epoch": 1.8639455782312924, - "grad_norm": 0.36662643442740994, - "learning_rate": 6.645176257372054e-06, - "loss": 0.0279, + "epoch": 0.4101796407185629, + "grad_norm": 0.3117702117434219, + "learning_rate": 1.941824603806293e-05, + "loss": 0.0154, "step": 274 }, { - "epoch": 1.870748299319728, - "grad_norm": 0.6307178798575221, - "learning_rate": 6.5759579321380576e-06, - "loss": 0.0255, + "epoch": 0.4116766467065868, + "grad_norm": 0.38775196821617675, + "learning_rate": 1.9412799327593884e-05, + "loss": 0.0106, "step": 275 }, { - "epoch": 1.8775510204081631, - "grad_norm": 0.3391198098709979, - "learning_rate": 6.5069249520122026e-06, - "loss": 0.0299, + "epoch": 0.41317365269461076, + "grad_norm": 0.2310944482235676, + "learning_rate": 1.9407328009321367e-05, + "loss": 0.0091, "step": 276 }, { - "epoch": 1.8843537414965987, - "grad_norm": 0.4817806870006288, - "learning_rate": 6.438081053784197e-06, - "loss": 0.0293, + "epoch": 0.41467065868263475, + "grad_norm": 0.23882380072426646, + "learning_rate": 1.940183209754901e-05, + "loss": 0.0096, "step": 277 }, { - "epoch": 1.891156462585034, - "grad_norm": 0.4572796789157729, - "learning_rate": 6.36942996400865e-06, - "loss": 0.0309, + "epoch": 0.4161676646706587, + "grad_norm": 0.47743508487331154, + "learning_rate": 1.939631160664472e-05, + "loss": 0.0138, "step": 278 }, { - "epoch": 1.8979591836734695, - "grad_norm": 0.3943436654831781, - "learning_rate": 6.300975398803362e-06, - "loss": 0.0274, + "epoch": 0.4176646706586826, + "grad_norm": 0.24267162900266767, + "learning_rate": 1.939076655104068e-05, + "loss": 0.0127, "step": 279 }, { - "epoch": 1.9047619047619047, - "grad_norm": 0.39513158230874584, - "learning_rate": 6.232721063648148e-06, - "loss": 0.0232, + "epoch": 0.41916167664670656, + "grad_norm": 0.25724021933042784, + "learning_rate": 1.9385196945233276e-05, + "loss": 0.008, "step": 280 }, { - "epoch": 1.9115646258503403, - "grad_norm": 0.3405190746199494, - "learning_rate": 6.1646706531842845e-06, - "loss": 0.0249, + "epoch": 0.42065868263473055, + "grad_norm": 0.2863894617393607, + "learning_rate": 1.9379602803783088e-05, + "loss": 0.0118, "step": 281 }, { - "epoch": 1.9183673469387754, - "grad_norm": 0.3938020668564346, - "learning_rate": 6.09682785101449e-06, - "loss": 0.0277, + "epoch": 0.4221556886227545, + "grad_norm": 0.24028996046153225, + "learning_rate": 1.9373984141314827e-05, + "loss": 0.0123, "step": 282 }, { - "epoch": 1.925170068027211, - "grad_norm": 0.3802867002071383, - "learning_rate": 6.029196329503548e-06, - "loss": 0.0264, + "epoch": 0.4236526946107784, + "grad_norm": 0.24877798633512982, + "learning_rate": 1.936834097251732e-05, + "loss": 0.01, "step": 283 }, { - "epoch": 1.9319727891156462, - "grad_norm": 0.28045602092848787, - "learning_rate": 5.961779749579516e-06, - "loss": 0.0209, + "epoch": 0.4251497005988024, + "grad_norm": 0.24480091385137137, + "learning_rate": 1.936267331214346e-05, + "loss": 0.0088, "step": 284 }, { - "epoch": 1.9387755102040818, - "grad_norm": 0.33135168744952337, - "learning_rate": 5.8945817605355495e-06, - "loss": 0.0192, + "epoch": 0.42664670658682635, + "grad_norm": 0.378474089158859, + "learning_rate": 1.9356981175010155e-05, + "loss": 0.0138, "step": 285 }, { - "epoch": 1.945578231292517, - "grad_norm": 0.310525514277901, - "learning_rate": 5.827605999832375e-06, - "loss": 0.0179, + "epoch": 0.4281437125748503, + "grad_norm": 0.2731112425501055, + "learning_rate": 1.9351264575998322e-05, + "loss": 0.0111, "step": 286 }, { - "epoch": 1.9523809523809523, - "grad_norm": 0.4333347605749311, - "learning_rate": 5.760856092901394e-06, - "loss": 0.0299, + "epoch": 0.4296407185628742, + "grad_norm": 0.45385449828176233, + "learning_rate": 1.934552353005281e-05, + "loss": 0.0144, "step": 287 }, { - "epoch": 1.9591836734693877, - "grad_norm": 0.2600623965752405, - "learning_rate": 5.694335652948415e-06, - "loss": 0.0222, + "epoch": 0.4311377245508982, + "grad_norm": 0.2426064120408043, + "learning_rate": 1.9339758052182396e-05, + "loss": 0.0113, "step": 288 }, { - "epoch": 1.965986394557823, - "grad_norm": 0.3654991906697444, - "learning_rate": 5.628048280758096e-06, - "loss": 0.0299, + "epoch": 0.43263473053892215, + "grad_norm": 0.2840729140180023, + "learning_rate": 1.933396815745972e-05, + "loss": 0.0088, "step": 289 }, { - "epoch": 1.9727891156462585, - "grad_norm": 0.2836245750881348, - "learning_rate": 5.561997564499024e-06, - "loss": 0.0266, + "epoch": 0.4341317365269461, + "grad_norm": 0.4447034848523609, + "learning_rate": 1.932815386102125e-05, + "loss": 0.0104, "step": 290 }, { - "epoch": 1.9795918367346939, - "grad_norm": 0.3772557920038287, - "learning_rate": 5.4961870795294644e-06, - "loss": 0.028, + "epoch": 0.4356287425149701, + "grad_norm": 0.27318626797446804, + "learning_rate": 1.9322315178067255e-05, + "loss": 0.0082, "step": 291 }, { - "epoch": 1.9863945578231292, - "grad_norm": 0.2390418620516205, - "learning_rate": 5.430620388203866e-06, - "loss": 0.0181, + "epoch": 0.437125748502994, + "grad_norm": 0.28620936956656057, + "learning_rate": 1.931645212386176e-05, + "loss": 0.0113, "step": 292 }, { - "epoch": 1.9931972789115646, - "grad_norm": 0.26917774514431425, - "learning_rate": 5.365301039679985e-06, - "loss": 0.0252, + "epoch": 0.43862275449101795, + "grad_norm": 0.26386101205038476, + "learning_rate": 1.9310564713732495e-05, + "loss": 0.0082, "step": 293 }, { - "epoch": 2.0, - "grad_norm": 0.3027370695631011, - "learning_rate": 5.300232569726805e-06, - "loss": 0.0187, + "epoch": 0.44011976047904194, + "grad_norm": 0.40806356163691765, + "learning_rate": 1.9304652963070868e-05, + "loss": 0.0107, "step": 294 }, { - "epoch": 2.006802721088435, - "grad_norm": 0.5374555027441916, - "learning_rate": 5.2354185005331095e-06, - "loss": 0.0282, + "epoch": 0.4416167664670659, + "grad_norm": 0.19248450148831384, + "learning_rate": 1.929871688733192e-05, + "loss": 0.0091, "step": 295 }, { - "epoch": 2.0136054421768708, - "grad_norm": 0.23930806738660212, - "learning_rate": 5.170862340516858e-06, - "loss": 0.0149, + "epoch": 0.4431137724550898, + "grad_norm": 0.28001071490300894, + "learning_rate": 1.929275650203429e-05, + "loss": 0.0086, "step": 296 }, { - "epoch": 2.020408163265306, - "grad_norm": 0.32072539854622056, - "learning_rate": 5.106567584135251e-06, - "loss": 0.0255, + "epoch": 0.44461077844311375, + "grad_norm": 0.19281433588980926, + "learning_rate": 1.9286771822760155e-05, + "loss": 0.0075, "step": 297 }, { - "epoch": 2.0272108843537415, - "grad_norm": 0.3765856363755405, - "learning_rate": 5.042537711695584e-06, - "loss": 0.0265, + "epoch": 0.44610778443113774, + "grad_norm": 0.30951373486991235, + "learning_rate": 1.9280762865155217e-05, + "loss": 0.0126, "step": 298 }, { - "epoch": 2.0340136054421767, - "grad_norm": 0.375918284328471, - "learning_rate": 4.97877618916684e-06, - "loss": 0.0296, + "epoch": 0.4476047904191617, + "grad_norm": 0.24670484913137125, + "learning_rate": 1.9274729644928646e-05, + "loss": 0.0079, "step": 299 }, { - "epoch": 2.0408163265306123, - "grad_norm": 0.28544218268890903, - "learning_rate": 4.915286467992098e-06, - "loss": 0.0266, + "epoch": 0.4491017964071856, + "grad_norm": 0.21597250129732382, + "learning_rate": 1.9268672177853037e-05, + "loss": 0.0082, "step": 300 }, { - "epoch": 2.0476190476190474, - "grad_norm": 0.29041194755799743, - "learning_rate": 4.852071984901696e-06, - "loss": 0.0243, + "epoch": 0.4505988023952096, + "grad_norm": 0.25565404173278816, + "learning_rate": 1.926259047976438e-05, + "loss": 0.0078, "step": 301 }, { - "epoch": 2.054421768707483, - "grad_norm": 0.289138208611435, - "learning_rate": 4.789136161727184e-06, - "loss": 0.0193, + "epoch": 0.45209580838323354, + "grad_norm": 0.25018467551033313, + "learning_rate": 1.9256484566562002e-05, + "loss": 0.0082, "step": 302 }, { - "epoch": 2.061224489795918, - "grad_norm": 0.39699377159750365, - "learning_rate": 4.7264824052161255e-06, - "loss": 0.0213, + "epoch": 0.4535928143712575, + "grad_norm": 0.24949067973854458, + "learning_rate": 1.9250354454208545e-05, + "loss": 0.0084, "step": 303 }, { - "epoch": 2.068027210884354, - "grad_norm": 0.2981351983066942, - "learning_rate": 4.664114106847667e-06, - "loss": 0.0157, + "epoch": 0.4550898203592814, + "grad_norm": 0.31313857717111415, + "learning_rate": 1.924420015872991e-05, + "loss": 0.0102, "step": 304 }, { - "epoch": 2.074829931972789, - "grad_norm": 0.41346371414139993, - "learning_rate": 4.602034642648968e-06, - "loss": 0.0222, + "epoch": 0.4565868263473054, + "grad_norm": 0.20485306938289613, + "learning_rate": 1.923802169621522e-05, + "loss": 0.0075, "step": 305 }, { - "epoch": 2.0816326530612246, - "grad_norm": 0.46281644559656343, - "learning_rate": 4.5402473730124395e-06, - "loss": 0.0228, + "epoch": 0.45808383233532934, + "grad_norm": 0.20454714142094618, + "learning_rate": 1.9231819082816786e-05, + "loss": 0.0076, "step": 306 }, { - "epoch": 2.0884353741496597, - "grad_norm": 0.36453555154639505, - "learning_rate": 4.478755642513868e-06, - "loss": 0.0214, + "epoch": 0.4595808383233533, + "grad_norm": 0.24763142577873085, + "learning_rate": 1.9225592334750036e-05, + "loss": 0.0081, "step": 307 }, { - "epoch": 2.0952380952380953, - "grad_norm": 0.3747842217307271, - "learning_rate": 4.417562779731355e-06, - "loss": 0.0237, + "epoch": 0.46107784431137727, + "grad_norm": 0.27799312574579305, + "learning_rate": 1.9219341468293518e-05, + "loss": 0.0079, "step": 308 }, { - "epoch": 2.1020408163265305, - "grad_norm": 0.25516596155835825, - "learning_rate": 4.356672097065134e-06, - "loss": 0.0227, + "epoch": 0.4625748502994012, + "grad_norm": 0.16117009627843906, + "learning_rate": 1.921306649978881e-05, + "loss": 0.006, "step": 309 }, { - "epoch": 2.108843537414966, - "grad_norm": 0.3978996621254664, - "learning_rate": 4.2960868905582895e-06, - "loss": 0.0251, + "epoch": 0.46407185628742514, + "grad_norm": 0.2451738774041042, + "learning_rate": 1.920676744564052e-05, + "loss": 0.0103, "step": 310 }, { - "epoch": 2.1156462585034013, - "grad_norm": 0.37774170305452065, - "learning_rate": 4.235810439718327e-06, - "loss": 0.0185, + "epoch": 0.4655688622754491, + "grad_norm": 0.23167970747659175, + "learning_rate": 1.9200444322316206e-05, + "loss": 0.0069, "step": 311 }, { - "epoch": 2.122448979591837, - "grad_norm": 0.5276653229249155, - "learning_rate": 4.175846007339644e-06, - "loss": 0.0175, + "epoch": 0.46706586826347307, + "grad_norm": 0.19444027025835037, + "learning_rate": 1.919409714634636e-05, + "loss": 0.0064, "step": 312 }, { - "epoch": 2.129251700680272, - "grad_norm": 0.45335169134199393, - "learning_rate": 4.1161968393269324e-06, - "loss": 0.0176, + "epoch": 0.468562874251497, + "grad_norm": 0.259521752410209, + "learning_rate": 1.9187725934324352e-05, + "loss": 0.0087, "step": 313 }, { - "epoch": 2.1360544217687076, - "grad_norm": 0.3501639808861222, - "learning_rate": 4.0568661645194656e-06, - "loss": 0.0229, + "epoch": 0.47005988023952094, + "grad_norm": 0.1870803982415873, + "learning_rate": 1.918133070290639e-05, + "loss": 0.0099, "step": 314 }, { - "epoch": 2.142857142857143, - "grad_norm": 0.36922699495113326, - "learning_rate": 3.997857194516319e-06, - "loss": 0.0247, + "epoch": 0.47155688622754494, + "grad_norm": 0.2782883408394447, + "learning_rate": 1.917491146881147e-05, + "loss": 0.0079, "step": 315 }, { - "epoch": 2.1496598639455784, - "grad_norm": 0.3844861347028257, - "learning_rate": 3.939173123502523e-06, - "loss": 0.0199, + "epoch": 0.47305389221556887, + "grad_norm": 0.30227323802831124, + "learning_rate": 1.916846824882135e-05, + "loss": 0.0068, "step": 316 }, { - "epoch": 2.1564625850340136, - "grad_norm": 0.645854225510273, - "learning_rate": 3.8808171280761665e-06, - "loss": 0.0208, + "epoch": 0.4745508982035928, + "grad_norm": 0.17450053993957124, + "learning_rate": 1.916200105978048e-05, + "loss": 0.0055, "step": 317 }, { - "epoch": 2.163265306122449, - "grad_norm": 0.3101470346726489, - "learning_rate": 3.822792367076446e-06, - "loss": 0.0168, + "epoch": 0.47604790419161674, + "grad_norm": 0.22986677982746617, + "learning_rate": 1.9155509918595977e-05, + "loss": 0.0094, "step": 318 }, { - "epoch": 2.1700680272108843, - "grad_norm": 0.4327622148598966, - "learning_rate": 3.7651019814126656e-06, - "loss": 0.0194, + "epoch": 0.47754491017964074, + "grad_norm": 0.2210009165017658, + "learning_rate": 1.9148994842237585e-05, + "loss": 0.0089, "step": 319 }, { - "epoch": 2.17687074829932, - "grad_norm": 0.3994172850300765, - "learning_rate": 3.7077490938942307e-06, - "loss": 0.0189, + "epoch": 0.47904191616766467, + "grad_norm": 0.3369866058031818, + "learning_rate": 1.9142455847737613e-05, + "loss": 0.0102, "step": 320 }, { - "epoch": 2.183673469387755, - "grad_norm": 0.23341486544176343, - "learning_rate": 3.6507368090616014e-06, - "loss": 0.0154, + "epoch": 0.4805389221556886, + "grad_norm": 0.31529088422929136, + "learning_rate": 1.9135892952190894e-05, + "loss": 0.0067, "step": 321 }, { - "epoch": 2.1904761904761907, - "grad_norm": 0.2800782727579594, - "learning_rate": 3.594068213018249e-06, - "loss": 0.0271, + "epoch": 0.4820359281437126, + "grad_norm": 0.34819262640952137, + "learning_rate": 1.912930617275476e-05, + "loss": 0.0083, "step": 322 }, { - "epoch": 2.197278911564626, - "grad_norm": 0.41245009605814564, - "learning_rate": 3.53774637326359e-06, - "loss": 0.0211, + "epoch": 0.48353293413173654, + "grad_norm": 0.2823658083150663, + "learning_rate": 1.9122695526648968e-05, + "loss": 0.0074, "step": 323 }, { - "epoch": 2.204081632653061, - "grad_norm": 0.44102425000184575, - "learning_rate": 3.481774338526954e-06, - "loss": 0.0203, + "epoch": 0.48502994011976047, + "grad_norm": 0.22578397023227845, + "learning_rate": 1.9116061031155683e-05, + "loss": 0.0054, "step": 324 }, { - "epoch": 2.2108843537414966, - "grad_norm": 0.3612731375372648, - "learning_rate": 3.426155138602558e-06, - "loss": 0.0109, + "epoch": 0.4865269461077844, + "grad_norm": 0.1798278839714749, + "learning_rate": 1.9109402703619396e-05, + "loss": 0.0069, "step": 325 }, { - "epoch": 2.2176870748299318, - "grad_norm": 0.41003267633028256, - "learning_rate": 3.3708917841854782e-06, - "loss": 0.0254, + "epoch": 0.4880239520958084, + "grad_norm": 0.3173459554977556, + "learning_rate": 1.9102720561446933e-05, + "loss": 0.0097, "step": 326 }, { - "epoch": 2.2244897959183674, - "grad_norm": 0.44409978093074687, - "learning_rate": 3.3159872667087077e-06, - "loss": 0.0219, + "epoch": 0.48952095808383234, + "grad_norm": 0.4520342362723284, + "learning_rate": 1.9096014622107363e-05, + "loss": 0.0064, "step": 327 }, { - "epoch": 2.2312925170068025, - "grad_norm": 0.4310743734288755, - "learning_rate": 3.2614445581812183e-06, - "loss": 0.0201, + "epoch": 0.49101796407185627, + "grad_norm": 0.2950917880335438, + "learning_rate": 1.9089284903131956e-05, + "loss": 0.0075, "step": 328 }, { - "epoch": 2.238095238095238, - "grad_norm": 0.3035091737111279, - "learning_rate": 3.207266611027069e-06, - "loss": 0.024, + "epoch": 0.49251497005988026, + "grad_norm": 0.27125581637418156, + "learning_rate": 1.908253142211417e-05, + "loss": 0.0055, "step": 329 }, { - "epoch": 2.2448979591836733, - "grad_norm": 0.3671044820236742, - "learning_rate": 3.1534563579256172e-06, - "loss": 0.024, + "epoch": 0.4940119760479042, + "grad_norm": 0.4292444115677685, + "learning_rate": 1.9075754196709574e-05, + "loss": 0.0089, "step": 330 }, { - "epoch": 2.251700680272109, - "grad_norm": 0.3619582161667144, - "learning_rate": 3.1000167116527525e-06, - "loss": 0.0203, + "epoch": 0.49550898203592814, + "grad_norm": 0.2737159333850911, + "learning_rate": 1.9068953244635805e-05, + "loss": 0.0061, "step": 331 }, { - "epoch": 2.258503401360544, - "grad_norm": 0.450706470571488, - "learning_rate": 3.0469505649232333e-06, - "loss": 0.029, + "epoch": 0.49700598802395207, + "grad_norm": 0.2813618023609554, + "learning_rate": 1.9062128583672543e-05, + "loss": 0.0055, "step": 332 }, { - "epoch": 2.2653061224489797, - "grad_norm": 0.3953038026535905, - "learning_rate": 2.9942607902340946e-06, - "loss": 0.021, + "epoch": 0.49850299401197606, + "grad_norm": 0.21811824525996454, + "learning_rate": 1.9055280231661443e-05, + "loss": 0.0076, "step": 333 }, { - "epoch": 2.272108843537415, - "grad_norm": 0.45052845721729323, - "learning_rate": 2.9419502397091715e-06, - "loss": 0.0239, + "epoch": 0.5, + "grad_norm": 0.3024569403044615, + "learning_rate": 1.904840820650609e-05, + "loss": 0.0076, "step": 334 }, { - "epoch": 2.2789115646258504, - "grad_norm": 0.40042405402317627, - "learning_rate": 2.8900217449447077e-06, - "loss": 0.0162, + "epoch": 0.5014970059880239, + "grad_norm": 0.3747520923531163, + "learning_rate": 1.9041512526171968e-05, + "loss": 0.0073, "step": 335 }, { - "epoch": 2.2857142857142856, - "grad_norm": 0.26311574813539457, - "learning_rate": 2.8384781168560693e-06, - "loss": 0.022, + "epoch": 0.5029940119760479, + "grad_norm": 0.32327667748094846, + "learning_rate": 1.9034593208686397e-05, + "loss": 0.0086, "step": 336 }, { - "epoch": 2.292517006802721, - "grad_norm": 0.3897033453197049, - "learning_rate": 2.7873221455256006e-06, - "loss": 0.0207, + "epoch": 0.5044910179640718, + "grad_norm": 0.2599402605515639, + "learning_rate": 1.9027650272138483e-05, + "loss": 0.008, "step": 337 }, { - "epoch": 2.2993197278911564, - "grad_norm": 0.3677858687841903, - "learning_rate": 2.736556600051593e-06, - "loss": 0.023, + "epoch": 0.5059880239520959, + "grad_norm": 0.309419497369784, + "learning_rate": 1.9020683734679098e-05, + "loss": 0.0076, "step": 338 }, { - "epoch": 2.306122448979592, - "grad_norm": 0.3577516109284015, - "learning_rate": 2.6861842283983953e-06, - "loss": 0.0219, + "epoch": 0.5074850299401198, + "grad_norm": 0.3980869606666559, + "learning_rate": 1.9013693614520795e-05, + "loss": 0.011, "step": 339 }, { - "epoch": 2.312925170068027, - "grad_norm": 0.24621923670001875, - "learning_rate": 2.6362077572476495e-06, - "loss": 0.0166, + "epoch": 0.5089820359281437, + "grad_norm": 0.17974783303905734, + "learning_rate": 1.9006679929937796e-05, + "loss": 0.0064, "step": 340 }, { - "epoch": 2.3197278911564627, - "grad_norm": 0.26774100034818, - "learning_rate": 2.586629891850716e-06, - "loss": 0.0149, + "epoch": 0.5104790419161677, + "grad_norm": 0.33752895891985435, + "learning_rate": 1.899964269926591e-05, + "loss": 0.0058, "step": 341 }, { - "epoch": 2.326530612244898, - "grad_norm": 0.3498362662110048, - "learning_rate": 2.5374533158822225e-06, - "loss": 0.0203, + "epoch": 0.5119760479041916, + "grad_norm": 0.2851367706634093, + "learning_rate": 1.8992581940902517e-05, + "loss": 0.0059, "step": 342 }, { - "epoch": 2.3333333333333335, - "grad_norm": 0.4379991319756946, - "learning_rate": 2.4886806912948034e-06, - "loss": 0.0234, + "epoch": 0.5134730538922155, + "grad_norm": 0.19085718115700934, + "learning_rate": 1.8985497673306495e-05, + "loss": 0.0042, "step": 343 }, { - "epoch": 2.3401360544217686, - "grad_norm": 0.4484515975309493, - "learning_rate": 2.4403146581749925e-06, - "loss": 0.023, + "epoch": 0.5149700598802395, + "grad_norm": 0.2450459731205004, + "learning_rate": 1.897838991499819e-05, + "loss": 0.0084, "step": 344 }, { - "epoch": 2.3469387755102042, - "grad_norm": 0.24634210193643918, - "learning_rate": 2.392357834600336e-06, - "loss": 0.0151, + "epoch": 0.5164670658682635, + "grad_norm": 0.2159471296015911, + "learning_rate": 1.8971258684559354e-05, + "loss": 0.0062, "step": 345 }, { - "epoch": 2.3537414965986394, - "grad_norm": 0.29145191787124425, - "learning_rate": 2.3448128164976593e-06, - "loss": 0.0212, + "epoch": 0.5179640718562875, + "grad_norm": 0.2582075201511502, + "learning_rate": 1.8964104000633104e-05, + "loss": 0.0072, "step": 346 }, { - "epoch": 2.360544217687075, - "grad_norm": 0.40253909206686106, - "learning_rate": 2.297682177502546e-06, - "loss": 0.025, + "epoch": 0.5194610778443114, + "grad_norm": 0.2826246708970663, + "learning_rate": 1.895692588192387e-05, + "loss": 0.0075, "step": 347 }, { - "epoch": 2.36734693877551, - "grad_norm": 0.3130871435447182, - "learning_rate": 2.2509684688200385e-06, - "loss": 0.0189, + "epoch": 0.5209580838323353, + "grad_norm": 0.2612365012660709, + "learning_rate": 1.894972434719735e-05, + "loss": 0.0046, "step": 348 }, { - "epoch": 2.3741496598639458, - "grad_norm": 0.2359349890142643, - "learning_rate": 2.204674219086531e-06, - "loss": 0.0164, + "epoch": 0.5224550898203593, + "grad_norm": 0.15056193683838454, + "learning_rate": 1.8942499415280458e-05, + "loss": 0.0042, "step": 349 }, { - "epoch": 2.380952380952381, - "grad_norm": 0.3417163765611956, - "learning_rate": 2.158801934232897e-06, - "loss": 0.023, + "epoch": 0.5239520958083832, + "grad_norm": 0.22875131238932167, + "learning_rate": 1.8935251105061268e-05, + "loss": 0.0046, "step": 350 }, { - "epoch": 2.387755102040816, - "grad_norm": 0.2911079105732259, - "learning_rate": 2.113354097348834e-06, - "loss": 0.0222, + "epoch": 0.5254491017964071, + "grad_norm": 0.2534927935258981, + "learning_rate": 1.892797943548898e-05, + "loss": 0.0084, "step": 351 }, { - "epoch": 2.3945578231292517, - "grad_norm": 0.3515229678204925, - "learning_rate": 2.0683331685484655e-06, - "loss": 0.0226, + "epoch": 0.5269461077844312, + "grad_norm": 0.21952770072925085, + "learning_rate": 1.8920684425573865e-05, + "loss": 0.0035, "step": 352 }, { - "epoch": 2.4013605442176873, - "grad_norm": 0.2769437555973887, - "learning_rate": 2.0237415848371666e-06, - "loss": 0.0131, + "epoch": 0.5284431137724551, + "grad_norm": 0.22335232414184136, + "learning_rate": 1.8913366094387203e-05, + "loss": 0.0055, "step": 353 }, { - "epoch": 2.4081632653061225, - "grad_norm": 0.33922796737817396, - "learning_rate": 1.979581759979642e-06, - "loss": 0.0241, + "epoch": 0.5299401197604791, + "grad_norm": 0.18343084280328395, + "learning_rate": 1.8906024461061247e-05, + "loss": 0.0064, "step": 354 }, { - "epoch": 2.4149659863945576, - "grad_norm": 0.31493116786046815, - "learning_rate": 1.9358560843692787e-06, - "loss": 0.0172, + "epoch": 0.531437125748503, + "grad_norm": 0.21313883463880326, + "learning_rate": 1.8898659544789166e-05, + "loss": 0.0052, "step": 355 }, { - "epoch": 2.421768707482993, - "grad_norm": 0.3073302581281779, - "learning_rate": 1.892566924898751e-06, - "loss": 0.0216, + "epoch": 0.5329341317365269, + "grad_norm": 0.17900066861920666, + "learning_rate": 1.8891271364825002e-05, + "loss": 0.0054, "step": 356 }, { - "epoch": 2.4285714285714284, - "grad_norm": 0.40858128963320417, - "learning_rate": 1.8497166248318876e-06, - "loss": 0.0225, + "epoch": 0.5344311377245509, + "grad_norm": 0.3186411908486247, + "learning_rate": 1.8883859940483614e-05, + "loss": 0.0085, "step": 357 }, { - "epoch": 2.435374149659864, - "grad_norm": 0.38355656039601815, - "learning_rate": 1.807307503676846e-06, - "loss": 0.0187, + "epoch": 0.5359281437125748, + "grad_norm": 0.23748527458193902, + "learning_rate": 1.8876425291140626e-05, + "loss": 0.0064, "step": 358 }, { - "epoch": 2.442176870748299, - "grad_norm": 0.31714048371803, - "learning_rate": 1.7653418570605474e-06, - "loss": 0.0181, + "epoch": 0.5374251497005988, + "grad_norm": 0.18189619860086126, + "learning_rate": 1.8868967436232378e-05, + "loss": 0.0061, "step": 359 }, { - "epoch": 2.4489795918367347, - "grad_norm": 0.25496567244495894, - "learning_rate": 1.7238219566044145e-06, - "loss": 0.0249, + "epoch": 0.5389221556886228, + "grad_norm": 0.3081977476306483, + "learning_rate": 1.886148639525588e-05, + "loss": 0.0049, "step": 360 }, { - "epoch": 2.45578231292517, - "grad_norm": 0.2810148889000208, - "learning_rate": 1.6827500498014026e-06, - "loss": 0.0195, + "epoch": 0.5404191616766467, + "grad_norm": 0.3589434090609591, + "learning_rate": 1.885398218776876e-05, + "loss": 0.0052, "step": 361 }, { - "epoch": 2.4625850340136055, - "grad_norm": 0.34079615823562126, - "learning_rate": 1.6421283598943526e-06, - "loss": 0.0242, + "epoch": 0.5419161676646707, + "grad_norm": 0.2719093022438886, + "learning_rate": 1.8846454833389202e-05, + "loss": 0.0063, "step": 362 }, { - "epoch": 2.4693877551020407, - "grad_norm": 0.27849042979199606, - "learning_rate": 1.601959085755641e-06, - "loss": 0.0215, + "epoch": 0.5434131736526946, + "grad_norm": 0.2134250055694977, + "learning_rate": 1.8838904351795905e-05, + "loss": 0.0047, "step": 363 }, { - "epoch": 2.4761904761904763, - "grad_norm": 0.31547844011000803, - "learning_rate": 1.5622444017681438e-06, - "loss": 0.0209, + "epoch": 0.5449101796407185, + "grad_norm": 0.36939383043202, + "learning_rate": 1.8831330762728028e-05, + "loss": 0.0061, "step": 364 }, { - "epoch": 2.4829931972789114, - "grad_norm": 0.3280004034375483, - "learning_rate": 1.5229864577075548e-06, - "loss": 0.0206, + "epoch": 0.5464071856287425, + "grad_norm": 0.3092372268044756, + "learning_rate": 1.882373408598515e-05, + "loss": 0.0055, "step": 365 }, { - "epoch": 2.489795918367347, - "grad_norm": 0.36138147578834934, - "learning_rate": 1.4841873786260019e-06, - "loss": 0.0206, + "epoch": 0.5479041916167665, + "grad_norm": 0.22805534768018146, + "learning_rate": 1.8816114341427196e-05, + "loss": 0.0065, "step": 366 }, { - "epoch": 2.496598639455782, - "grad_norm": 0.34871288825979213, - "learning_rate": 1.445849264737026e-06, - "loss": 0.0197, + "epoch": 0.5494011976047904, + "grad_norm": 0.20186378709158326, + "learning_rate": 1.88084715489744e-05, + "loss": 0.0055, "step": 367 }, { - "epoch": 2.503401360544218, - "grad_norm": 0.2929843679149074, - "learning_rate": 1.4079741913018863e-06, - "loss": 0.0191, + "epoch": 0.5508982035928144, + "grad_norm": 0.2447957214059236, + "learning_rate": 1.8800805728607257e-05, + "loss": 0.0059, "step": 368 }, { - "epoch": 2.510204081632653, - "grad_norm": 0.3729046126118177, - "learning_rate": 1.3705642085172367e-06, - "loss": 0.017, + "epoch": 0.5523952095808383, + "grad_norm": 0.24624052829362808, + "learning_rate": 1.879311690036645e-05, + "loss": 0.005, "step": 369 }, { - "epoch": 2.5170068027210886, - "grad_norm": 0.23294615156735132, - "learning_rate": 1.3336213414041387e-06, - "loss": 0.0129, + "epoch": 0.5538922155688623, + "grad_norm": 0.4598990071533753, + "learning_rate": 1.8785405084352825e-05, + "loss": 0.0119, "step": 370 }, { - "epoch": 2.5238095238095237, - "grad_norm": 0.28575171918313647, - "learning_rate": 1.2971475896984475e-06, - "loss": 0.02, + "epoch": 0.5553892215568862, + "grad_norm": 0.2254413046985076, + "learning_rate": 1.8777670300727314e-05, + "loss": 0.0047, "step": 371 }, { - "epoch": 2.5306122448979593, - "grad_norm": 0.24125146893347207, - "learning_rate": 1.2611449277425715e-06, - "loss": 0.0143, + "epoch": 0.5568862275449101, + "grad_norm": 0.2817060159533667, + "learning_rate": 1.8769912569710902e-05, + "loss": 0.0054, "step": 372 }, { - "epoch": 2.5374149659863945, - "grad_norm": 0.32756177057759744, - "learning_rate": 1.2256153043785911e-06, - "loss": 0.0162, + "epoch": 0.5583832335329342, + "grad_norm": 0.23401568986125731, + "learning_rate": 1.876213191158456e-05, + "loss": 0.0066, "step": 373 }, { - "epoch": 2.54421768707483, - "grad_norm": 0.3138440305938392, - "learning_rate": 1.1905606428427775e-06, - "loss": 0.0193, + "epoch": 0.5598802395209581, + "grad_norm": 0.31712264048465694, + "learning_rate": 1.8754328346689194e-05, + "loss": 0.0051, "step": 374 }, { - "epoch": 2.5510204081632653, - "grad_norm": 0.3278031624256675, - "learning_rate": 1.1559828406614716e-06, - "loss": 0.0236, + "epoch": 0.561377245508982, + "grad_norm": 0.1853362893894113, + "learning_rate": 1.8746501895425603e-05, + "loss": 0.0031, "step": 375 }, { - "epoch": 2.557823129251701, - "grad_norm": 0.3496069896696055, - "learning_rate": 1.1218837695483853e-06, - "loss": 0.017, + "epoch": 0.562874251497006, + "grad_norm": 0.1512730540107703, + "learning_rate": 1.8738652578254414e-05, + "loss": 0.0031, "step": 376 }, { - "epoch": 2.564625850340136, - "grad_norm": 0.35806348572299274, - "learning_rate": 1.0882652753032797e-06, - "loss": 0.0183, + "epoch": 0.5643712574850299, + "grad_norm": 0.251381383971643, + "learning_rate": 1.8730780415696034e-05, + "loss": 0.0061, "step": 377 }, { - "epoch": 2.571428571428571, - "grad_norm": 0.3416680123893772, - "learning_rate": 1.0551291777120465e-06, - "loss": 0.0159, + "epoch": 0.5658682634730539, + "grad_norm": 0.2090374749619751, + "learning_rate": 1.872288542833059e-05, + "loss": 0.0032, "step": 378 }, { - "epoch": 2.578231292517007, - "grad_norm": 0.2848539912430437, - "learning_rate": 1.0224772704482033e-06, - "loss": 0.021, + "epoch": 0.5673652694610778, + "grad_norm": 0.29710148489922494, + "learning_rate": 1.871496763679788e-05, + "loss": 0.0061, "step": 379 }, { - "epoch": 2.5850340136054424, - "grad_norm": 0.34478335165414964, - "learning_rate": 9.903113209758098e-07, - "loss": 0.02, + "epoch": 0.5688622754491018, + "grad_norm": 0.20476911169490797, + "learning_rate": 1.8707027061797328e-05, + "loss": 0.0041, "step": 380 }, { - "epoch": 2.5918367346938775, - "grad_norm": 0.30259868082341945, - "learning_rate": 9.58633070453785e-07, - "loss": 0.0224, + "epoch": 0.5703592814371258, + "grad_norm": 0.15467140235695107, + "learning_rate": 1.8699063724087905e-05, + "loss": 0.0039, "step": 381 }, { - "epoch": 2.5986394557823127, - "grad_norm": 0.29749976015408053, - "learning_rate": 9.274442336416567e-07, - "loss": 0.0226, + "epoch": 0.5718562874251497, + "grad_norm": 0.35593514914651286, + "learning_rate": 1.8691077644488103e-05, + "loss": 0.0051, "step": 382 }, { - "epoch": 2.6054421768707483, - "grad_norm": 0.3195969770154066, - "learning_rate": 8.967464988067476e-07, - "loss": 0.0168, + "epoch": 0.5733532934131736, + "grad_norm": 0.3031925999503722, + "learning_rate": 1.8683068843875866e-05, + "loss": 0.0054, "step": 383 }, { - "epoch": 2.612244897959184, - "grad_norm": 0.2600960483615313, - "learning_rate": 8.665415276327871e-07, - "loss": 0.0137, + "epoch": 0.5748502994011976, + "grad_norm": 0.392987606385426, + "learning_rate": 1.867503734318853e-05, + "loss": 0.0068, "step": 384 }, { - "epoch": 2.619047619047619, - "grad_norm": 0.26403678914149437, - "learning_rate": 8.368309551299536e-07, - "loss": 0.0178, + "epoch": 0.5763473053892215, + "grad_norm": 0.1615633068715301, + "learning_rate": 1.8666983163422786e-05, + "loss": 0.004, "step": 385 }, { - "epoch": 2.6258503401360542, - "grad_norm": 0.2748498822235369, - "learning_rate": 8.076163895463862e-07, - "loss": 0.0144, + "epoch": 0.5778443113772455, + "grad_norm": 0.2623310704504511, + "learning_rate": 1.8658906325634605e-05, + "loss": 0.0065, "step": 386 }, { - "epoch": 2.63265306122449, - "grad_norm": 0.3305010052406547, - "learning_rate": 7.788994122811178e-07, - "loss": 0.0219, + "epoch": 0.5793413173652695, + "grad_norm": 0.41136291194943037, + "learning_rate": 1.86508068509392e-05, + "loss": 0.0065, "step": 387 }, { - "epoch": 2.6394557823129254, - "grad_norm": 0.246882891291933, - "learning_rate": 7.506815777984788e-07, - "loss": 0.0158, + "epoch": 0.5808383233532934, + "grad_norm": 0.32922655904832243, + "learning_rate": 1.864268476051096e-05, + "loss": 0.0083, "step": 388 }, { - "epoch": 2.6462585034013606, - "grad_norm": 0.25439177703144955, - "learning_rate": 7.229644135439473e-07, - "loss": 0.0148, + "epoch": 0.5823353293413174, + "grad_norm": 0.15328440896027834, + "learning_rate": 1.86345400755834e-05, + "loss": 0.0034, "step": 389 }, { - "epoch": 2.6530612244897958, - "grad_norm": 0.36357656004842953, - "learning_rate": 6.957494198614778e-07, - "loss": 0.0238, + "epoch": 0.5838323353293413, + "grad_norm": 0.23062078223568724, + "learning_rate": 1.8626372817449107e-05, + "loss": 0.0035, "step": 390 }, { - "epoch": 2.6598639455782314, - "grad_norm": 0.2872101510976723, - "learning_rate": 6.690380699122767e-07, - "loss": 0.0204, + "epoch": 0.5853293413173652, + "grad_norm": 0.4137111601133577, + "learning_rate": 1.8618183007459664e-05, + "loss": 0.0057, "step": 391 }, { - "epoch": 2.6666666666666665, - "grad_norm": 0.24331315521025246, - "learning_rate": 6.428318095950648e-07, - "loss": 0.0155, + "epoch": 0.5868263473053892, + "grad_norm": 0.27055771834059683, + "learning_rate": 1.8609970667025633e-05, + "loss": 0.0046, "step": 392 }, { - "epoch": 2.673469387755102, - "grad_norm": 0.25145814555745893, - "learning_rate": 6.171320574678064e-07, - "loss": 0.0175, + "epoch": 0.5883233532934131, + "grad_norm": 0.1400482751380909, + "learning_rate": 1.8601735817616466e-05, + "loss": 0.0033, "step": 393 }, { - "epoch": 2.6802721088435373, - "grad_norm": 0.3359588828649143, - "learning_rate": 5.919402046709288e-07, - "loss": 0.02, + "epoch": 0.5898203592814372, + "grad_norm": 0.2825851174593088, + "learning_rate": 1.859347848076046e-05, + "loss": 0.0062, "step": 394 }, { - "epoch": 2.687074829931973, - "grad_norm": 0.29084282580528553, - "learning_rate": 5.672576148520136e-07, - "loss": 0.0183, + "epoch": 0.5913173652694611, + "grad_norm": 0.3397946721846426, + "learning_rate": 1.85851986780447e-05, + "loss": 0.006, "step": 395 }, { - "epoch": 2.693877551020408, - "grad_norm": 0.344918984609033, - "learning_rate": 5.430856240919779e-07, - "loss": 0.0235, + "epoch": 0.592814371257485, + "grad_norm": 0.27215859756652094, + "learning_rate": 1.8576896431115017e-05, + "loss": 0.0043, "step": 396 }, { - "epoch": 2.7006802721088436, - "grad_norm": 0.2711450817646753, - "learning_rate": 5.19425540832762e-07, - "loss": 0.016, + "epoch": 0.594311377245509, + "grad_norm": 0.15206642506994694, + "learning_rate": 1.8568571761675893e-05, + "loss": 0.0035, "step": 397 }, { - "epoch": 2.707482993197279, - "grad_norm": 0.40375159495643054, - "learning_rate": 4.962786458064972e-07, - "loss": 0.0228, + "epoch": 0.5958083832335329, + "grad_norm": 0.177981847566587, + "learning_rate": 1.8560224691490452e-05, + "loss": 0.0051, "step": 398 }, { - "epoch": 2.7142857142857144, - "grad_norm": 0.24410300959704656, - "learning_rate": 4.73646191966175e-07, - "loss": 0.0169, + "epoch": 0.5973053892215568, + "grad_norm": 0.18243440630634988, + "learning_rate": 1.8551855242380366e-05, + "loss": 0.0028, "step": 399 }, { - "epoch": 2.7210884353741496, - "grad_norm": 0.3592873823963376, - "learning_rate": 4.515294044178331e-07, - "loss": 0.0192, + "epoch": 0.5988023952095808, + "grad_norm": 0.22781016153338413, + "learning_rate": 1.854346343622582e-05, + "loss": 0.0061, "step": 400 }, { - "epoch": 2.727891156462585, - "grad_norm": 0.47656899007676473, - "learning_rate": 4.299294803542331e-07, - "loss": 0.0165, + "epoch": 0.6002994011976048, + "grad_norm": 0.14798262053026728, + "learning_rate": 1.8535049294965436e-05, + "loss": 0.0028, "step": 401 }, { - "epoch": 2.7346938775510203, - "grad_norm": 0.3541427597027374, - "learning_rate": 4.0884758899006007e-07, - "loss": 0.0185, + "epoch": 0.6017964071856288, + "grad_norm": 0.12876490456064732, + "learning_rate": 1.8526612840596238e-05, + "loss": 0.0029, "step": 402 }, { - "epoch": 2.741496598639456, - "grad_norm": 0.4056943661965022, - "learning_rate": 3.882848714986243e-07, - "loss": 0.0194, + "epoch": 0.6032934131736527, + "grad_norm": 0.19321532804478075, + "learning_rate": 1.8518154095173583e-05, + "loss": 0.0032, "step": 403 }, { - "epoch": 2.748299319727891, - "grad_norm": 0.32241558571045703, - "learning_rate": 3.6824244095010064e-07, - "loss": 0.0248, + "epoch": 0.6047904191616766, + "grad_norm": 0.12070559140749956, + "learning_rate": 1.8509673080811096e-05, + "loss": 0.0033, "step": 404 }, { - "epoch": 2.7551020408163263, - "grad_norm": 0.30581795904421494, - "learning_rate": 3.4872138225127137e-07, - "loss": 0.016, + "epoch": 0.6062874251497006, + "grad_norm": 0.18453370264259583, + "learning_rate": 1.8501169819680624e-05, + "loss": 0.005, "step": 405 }, { - "epoch": 2.761904761904762, - "grad_norm": 0.34578165087246426, - "learning_rate": 3.2972275208679625e-07, - "loss": 0.0213, + "epoch": 0.6077844311377245, + "grad_norm": 0.1954819685260244, + "learning_rate": 1.849264433401217e-05, + "loss": 0.0036, "step": 406 }, { - "epoch": 2.7687074829931975, - "grad_norm": 0.3136844616082194, - "learning_rate": 3.112475788620217e-07, - "loss": 0.0186, + "epoch": 0.6092814371257484, + "grad_norm": 0.15334768361544945, + "learning_rate": 1.8484096646093842e-05, + "loss": 0.0035, "step": 407 }, { - "epoch": 2.7755102040816326, - "grad_norm": 0.3330644974135503, - "learning_rate": 2.932968626473065e-07, - "loss": 0.0214, + "epoch": 0.6107784431137725, + "grad_norm": 0.15359417727888636, + "learning_rate": 1.847552677827179e-05, + "loss": 0.0047, "step": 408 }, { - "epoch": 2.782312925170068, - "grad_norm": 0.2734645865898147, - "learning_rate": 2.758715751238872e-07, - "loss": 0.0167, + "epoch": 0.6122754491017964, + "grad_norm": 0.15142354869259234, + "learning_rate": 1.846693475295015e-05, + "loss": 0.0025, "step": 409 }, { - "epoch": 2.7891156462585034, - "grad_norm": 0.3235976546346094, - "learning_rate": 2.589726595312858e-07, - "loss": 0.0205, + "epoch": 0.6137724550898204, + "grad_norm": 0.1305889137402961, + "learning_rate": 1.8458320592590976e-05, + "loss": 0.0039, "step": 410 }, { - "epoch": 2.795918367346939, - "grad_norm": 0.3446098604740126, - "learning_rate": 2.426010306162485e-07, - "loss": 0.0169, + "epoch": 0.6152694610778443, + "grad_norm": 0.24052279035392543, + "learning_rate": 1.8449684319714202e-05, + "loss": 0.0037, "step": 411 }, { - "epoch": 2.802721088435374, - "grad_norm": 0.3270040316033607, - "learning_rate": 2.2675757458323066e-07, - "loss": 0.0184, + "epoch": 0.6167664670658682, + "grad_norm": 0.21510134495795813, + "learning_rate": 1.8441025956897566e-05, + "loss": 0.0047, "step": 412 }, { - "epoch": 2.8095238095238093, - "grad_norm": 0.2879016695364431, - "learning_rate": 2.1144314904642194e-07, - "loss": 0.0177, + "epoch": 0.6182634730538922, + "grad_norm": 0.1378513852811569, + "learning_rate": 1.8432345526776555e-05, + "loss": 0.0037, "step": 413 }, { - "epoch": 2.816326530612245, - "grad_norm": 0.33267463209181264, - "learning_rate": 1.9665858298333006e-07, - "loss": 0.0203, + "epoch": 0.6197604790419161, + "grad_norm": 0.12876211162444984, + "learning_rate": 1.842364305204434e-05, + "loss": 0.0047, "step": 414 }, { - "epoch": 2.8231292517006805, - "grad_norm": 0.23836799115278284, - "learning_rate": 1.824046766899046e-07, - "loss": 0.0144, + "epoch": 0.6212574850299402, + "grad_norm": 0.1870068324397345, + "learning_rate": 1.841491855545173e-05, + "loss": 0.0056, "step": 415 }, { - "epoch": 2.8299319727891157, - "grad_norm": 0.35726854776874595, - "learning_rate": 1.6868220173721472e-07, - "loss": 0.0239, + "epoch": 0.6227544910179641, + "grad_norm": 0.15043391887906607, + "learning_rate": 1.840617205980711e-05, + "loss": 0.0034, "step": 416 }, { - "epoch": 2.836734693877551, - "grad_norm": 0.3457061306540936, - "learning_rate": 1.5549190092968736e-07, - "loss": 0.0229, + "epoch": 0.624251497005988, + "grad_norm": 0.1603990561673152, + "learning_rate": 1.8397403587976366e-05, + "loss": 0.0038, "step": 417 }, { - "epoch": 2.8435374149659864, - "grad_norm": 0.30955214004843584, - "learning_rate": 1.4283448826489798e-07, - "loss": 0.0211, + "epoch": 0.625748502994012, + "grad_norm": 0.1380674767451333, + "learning_rate": 1.838861316288284e-05, + "loss": 0.0021, "step": 418 }, { - "epoch": 2.8503401360544216, - "grad_norm": 0.3263601366659666, - "learning_rate": 1.3071064889491723e-07, - "loss": 0.02, + "epoch": 0.6272455089820359, + "grad_norm": 0.18002621685496464, + "learning_rate": 1.8379800807507276e-05, + "loss": 0.0041, "step": 419 }, { - "epoch": 2.857142857142857, - "grad_norm": 0.32256000869883056, - "learning_rate": 1.1912103908922945e-07, - "loss": 0.0196, + "epoch": 0.6287425149700598, + "grad_norm": 0.16156445172842446, + "learning_rate": 1.8370966544887734e-05, + "loss": 0.0026, "step": 420 }, { - "epoch": 2.8639455782312924, - "grad_norm": 0.28683991197300734, - "learning_rate": 1.0806628619920322e-07, - "loss": 0.0163, + "epoch": 0.6302395209580839, + "grad_norm": 0.13680763335054835, + "learning_rate": 1.836211039811956e-05, + "loss": 0.0029, "step": 421 }, { - "epoch": 2.870748299319728, - "grad_norm": 0.2903311126798398, - "learning_rate": 9.754698862413758e-08, - "loss": 0.0202, + "epoch": 0.6317365269461078, + "grad_norm": 0.20502692603268471, + "learning_rate": 1.8353232390355293e-05, + "loss": 0.0045, "step": 422 }, { - "epoch": 2.877551020408163, - "grad_norm": 0.2239071167217727, - "learning_rate": 8.756371577886891e-08, - "loss": 0.0134, + "epoch": 0.6332335329341318, + "grad_norm": 0.23325599547346149, + "learning_rate": 1.8344332544804647e-05, + "loss": 0.0067, "step": 423 }, { - "epoch": 2.8843537414965987, - "grad_norm": 0.29808971100347104, - "learning_rate": 7.81170080629412e-08, - "loss": 0.0204, + "epoch": 0.6347305389221557, + "grad_norm": 0.29861952724882246, + "learning_rate": 1.8335410884734412e-05, + "loss": 0.006, "step": 424 }, { - "epoch": 2.891156462585034, - "grad_norm": 0.2789565484019186, - "learning_rate": 6.920737683136614e-08, - "loss": 0.0187, + "epoch": 0.6362275449101796, + "grad_norm": 0.1405737685776957, + "learning_rate": 1.832646743346841e-05, + "loss": 0.0025, "step": 425 }, { - "epoch": 2.8979591836734695, - "grad_norm": 0.3158830313532607, - "learning_rate": 6.083530436693408e-08, - "loss": 0.0213, + "epoch": 0.6377245508982036, + "grad_norm": 0.15314449241818123, + "learning_rate": 1.831750221438743e-05, + "loss": 0.0025, "step": 426 }, { - "epoch": 2.9047619047619047, - "grad_norm": 0.2652235833848647, - "learning_rate": 5.300124385410943e-08, - "loss": 0.0184, + "epoch": 0.6392215568862275, + "grad_norm": 0.20564899127887978, + "learning_rate": 1.8308515250929174e-05, + "loss": 0.005, "step": 427 }, { - "epoch": 2.9115646258503403, - "grad_norm": 0.303708914505213, - "learning_rate": 4.570561935450468e-08, - "loss": 0.0228, + "epoch": 0.6407185628742516, + "grad_norm": 0.2250426655272749, + "learning_rate": 1.8299506566588178e-05, + "loss": 0.0063, "step": 428 }, { - "epoch": 2.9183673469387754, - "grad_norm": 0.2552716618292082, - "learning_rate": 3.894882578391879e-08, - "loss": 0.0168, + "epoch": 0.6422155688622755, + "grad_norm": 0.20159344511006713, + "learning_rate": 1.829047618491578e-05, + "loss": 0.0041, "step": 429 }, { - "epoch": 2.925170068027211, - "grad_norm": 0.3437542248362107, - "learning_rate": 3.273122889096536e-08, - "loss": 0.018, + "epoch": 0.6437125748502994, + "grad_norm": 0.30367059362305904, + "learning_rate": 1.8281424129520028e-05, + "loss": 0.0045, "step": 430 }, { - "epoch": 2.931972789115646, - "grad_norm": 0.41181823304903403, - "learning_rate": 2.705316523726853e-08, - "loss": 0.0175, + "epoch": 0.6452095808383234, + "grad_norm": 0.17853724353758313, + "learning_rate": 1.8272350424065637e-05, + "loss": 0.0037, "step": 431 }, { - "epoch": 2.938775510204082, - "grad_norm": 0.31165983112820045, - "learning_rate": 2.1914942179253052e-08, - "loss": 0.02, + "epoch": 0.6467065868263473, + "grad_norm": 0.13452139718446834, + "learning_rate": 1.8263255092273918e-05, + "loss": 0.0031, "step": 432 }, { - "epoch": 2.945578231292517, - "grad_norm": 0.37990351221219176, - "learning_rate": 1.7316837851499845e-08, - "loss": 0.0208, + "epoch": 0.6482035928143712, + "grad_norm": 0.1485870363064914, + "learning_rate": 1.8254138157922726e-05, + "loss": 0.0036, "step": 433 }, { - "epoch": 2.9523809523809526, - "grad_norm": 0.42236543353263456, - "learning_rate": 1.325910115169471e-08, - "loss": 0.0211, + "epoch": 0.6497005988023952, + "grad_norm": 0.1369157060039127, + "learning_rate": 1.8244999644846386e-05, + "loss": 0.0049, "step": 434 }, { - "epoch": 2.9591836734693877, - "grad_norm": 0.3667740733661891, - "learning_rate": 9.74195172715242e-09, - "loss": 0.0185, + "epoch": 0.6511976047904192, + "grad_norm": 0.1422170125322691, + "learning_rate": 1.823583957693563e-05, + "loss": 0.0025, "step": 435 }, { - "epoch": 2.965986394557823, - "grad_norm": 0.2491767102688122, - "learning_rate": 6.7655799629284815e-09, - "loss": 0.0216, + "epoch": 0.6526946107784432, + "grad_norm": 0.18173894064740984, + "learning_rate": 1.8226657978137556e-05, + "loss": 0.0027, "step": 436 }, { - "epoch": 2.9727891156462585, - "grad_norm": 0.27827297191105266, - "learning_rate": 4.330146971515126e-09, - "loss": 0.0135, + "epoch": 0.6541916167664671, + "grad_norm": 0.11382010411529647, + "learning_rate": 1.8217454872455537e-05, + "loss": 0.0029, "step": 437 }, { - "epoch": 2.979591836734694, - "grad_norm": 0.2479372965359603, - "learning_rate": 2.435784584114975e-09, - "loss": 0.0197, + "epoch": 0.655688622754491, + "grad_norm": 0.24972459202509598, + "learning_rate": 1.820823028394918e-05, + "loss": 0.0039, "step": 438 }, { - "epoch": 2.9863945578231292, - "grad_norm": 0.3342210953903805, - "learning_rate": 1.0825953435122938e-09, - "loss": 0.0224, + "epoch": 0.657185628742515, + "grad_norm": 0.19783767385058063, + "learning_rate": 1.8198984236734246e-05, + "loss": 0.0043, "step": 439 }, { - "epoch": 2.9931972789115644, - "grad_norm": 0.3251029484247515, - "learning_rate": 2.706524985174319e-10, - "loss": 0.0255, + "epoch": 0.6586826347305389, + "grad_norm": 0.2542153997631583, + "learning_rate": 1.8189716754982604e-05, + "loss": 0.0053, "step": 440 }, { - "epoch": 3.0, - "grad_norm": 0.30791966633936196, - "learning_rate": 0.0, - "loss": 0.0177, + "epoch": 0.6601796407185628, + "grad_norm": 0.16722569392416708, + "learning_rate": 1.8180427862922157e-05, + "loss": 0.004, "step": 441 }, + { + "epoch": 0.6616766467065869, + "grad_norm": 0.18714906390575894, + "learning_rate": 1.817111758483677e-05, + "loss": 0.0049, + "step": 442 + }, + { + "epoch": 0.6631736526946108, + "grad_norm": 0.14257993994735135, + "learning_rate": 1.8161785945066235e-05, + "loss": 0.0026, + "step": 443 + }, + { + "epoch": 0.6646706586826348, + "grad_norm": 0.16935381936187344, + "learning_rate": 1.8152432968006176e-05, + "loss": 0.0036, + "step": 444 + }, + { + "epoch": 0.6661676646706587, + "grad_norm": 0.1342503188843719, + "learning_rate": 1.8143058678108002e-05, + "loss": 0.0028, + "step": 445 + }, + { + "epoch": 0.6676646706586826, + "grad_norm": 0.25046636983085335, + "learning_rate": 1.8133663099878846e-05, + "loss": 0.0068, + "step": 446 + }, + { + "epoch": 0.6691616766467066, + "grad_norm": 0.19961420082825826, + "learning_rate": 1.812424625788149e-05, + "loss": 0.0035, + "step": 447 + }, + { + "epoch": 0.6706586826347305, + "grad_norm": 0.17568029731674176, + "learning_rate": 1.8114808176734302e-05, + "loss": 0.003, + "step": 448 + }, + { + "epoch": 0.6721556886227545, + "grad_norm": 0.17477245221194124, + "learning_rate": 1.8105348881111182e-05, + "loss": 0.0047, + "step": 449 + }, + { + "epoch": 0.6736526946107785, + "grad_norm": 0.22428133343136628, + "learning_rate": 1.809586839574149e-05, + "loss": 0.0027, + "step": 450 + }, + { + "epoch": 0.6751497005988024, + "grad_norm": 0.19664866060996913, + "learning_rate": 1.808636674540998e-05, + "loss": 0.0034, + "step": 451 + }, + { + "epoch": 0.6766467065868264, + "grad_norm": 0.14083033586251215, + "learning_rate": 1.8076843954956736e-05, + "loss": 0.004, + "step": 452 + }, + { + "epoch": 0.6781437125748503, + "grad_norm": 0.1647022452825819, + "learning_rate": 1.8067300049277113e-05, + "loss": 0.0023, + "step": 453 + }, + { + "epoch": 0.6796407185628742, + "grad_norm": 0.32202530277073854, + "learning_rate": 1.805773505332166e-05, + "loss": 0.0066, + "step": 454 + }, + { + "epoch": 0.6811377245508982, + "grad_norm": 0.14979165173614356, + "learning_rate": 1.8048148992096073e-05, + "loss": 0.0032, + "step": 455 + }, + { + "epoch": 0.6826347305389222, + "grad_norm": 0.1418060842276349, + "learning_rate": 1.8038541890661106e-05, + "loss": 0.0036, + "step": 456 + }, + { + "epoch": 0.6841317365269461, + "grad_norm": 0.16706209980708875, + "learning_rate": 1.802891377413253e-05, + "loss": 0.0019, + "step": 457 + }, + { + "epoch": 0.6856287425149701, + "grad_norm": 0.18905559651264367, + "learning_rate": 1.8019264667681053e-05, + "loss": 0.0023, + "step": 458 + }, + { + "epoch": 0.687125748502994, + "grad_norm": 0.23786995943332953, + "learning_rate": 1.800959459653224e-05, + "loss": 0.0041, + "step": 459 + }, + { + "epoch": 0.688622754491018, + "grad_norm": 0.11727770023510513, + "learning_rate": 1.79999035859665e-05, + "loss": 0.0018, + "step": 460 + }, + { + "epoch": 0.6901197604790419, + "grad_norm": 0.190583040451331, + "learning_rate": 1.7990191661318942e-05, + "loss": 0.0044, + "step": 461 + }, + { + "epoch": 0.6916167664670658, + "grad_norm": 0.19155482816471409, + "learning_rate": 1.798045884797938e-05, + "loss": 0.0034, + "step": 462 + }, + { + "epoch": 0.6931137724550899, + "grad_norm": 0.2432660367460031, + "learning_rate": 1.797070517139223e-05, + "loss": 0.0064, + "step": 463 + }, + { + "epoch": 0.6946107784431138, + "grad_norm": 0.16246882216240055, + "learning_rate": 1.796093065705644e-05, + "loss": 0.0033, + "step": 464 + }, + { + "epoch": 0.6961077844311377, + "grad_norm": 0.10829584717187421, + "learning_rate": 1.7951135330525447e-05, + "loss": 0.0021, + "step": 465 + }, + { + "epoch": 0.6976047904191617, + "grad_norm": 0.13169748817522234, + "learning_rate": 1.794131921740709e-05, + "loss": 0.0034, + "step": 466 + }, + { + "epoch": 0.6991017964071856, + "grad_norm": 0.22261122260930077, + "learning_rate": 1.7931482343363552e-05, + "loss": 0.0032, + "step": 467 + }, + { + "epoch": 0.7005988023952096, + "grad_norm": 0.21429988838318564, + "learning_rate": 1.7921624734111292e-05, + "loss": 0.0064, + "step": 468 + }, + { + "epoch": 0.7020958083832335, + "grad_norm": 0.15085978142323003, + "learning_rate": 1.7911746415420982e-05, + "loss": 0.0019, + "step": 469 + }, + { + "epoch": 0.7035928143712575, + "grad_norm": 0.09480266222976796, + "learning_rate": 1.7901847413117424e-05, + "loss": 0.0018, + "step": 470 + }, + { + "epoch": 0.7050898203592815, + "grad_norm": 0.1840037147393193, + "learning_rate": 1.7891927753079495e-05, + "loss": 0.0033, + "step": 471 + }, + { + "epoch": 0.7065868263473054, + "grad_norm": 0.12251157236145369, + "learning_rate": 1.7881987461240086e-05, + "loss": 0.0019, + "step": 472 + }, + { + "epoch": 0.7080838323353293, + "grad_norm": 0.3770740041300004, + "learning_rate": 1.7872026563586015e-05, + "loss": 0.0061, + "step": 473 + }, + { + "epoch": 0.7095808383233533, + "grad_norm": 0.14686337723150117, + "learning_rate": 1.7862045086157975e-05, + "loss": 0.0036, + "step": 474 + }, + { + "epoch": 0.7110778443113772, + "grad_norm": 0.12202140713465083, + "learning_rate": 1.7852043055050466e-05, + "loss": 0.0028, + "step": 475 + }, + { + "epoch": 0.7125748502994012, + "grad_norm": 0.20749262333762236, + "learning_rate": 1.7842020496411707e-05, + "loss": 0.0021, + "step": 476 + }, + { + "epoch": 0.7140718562874252, + "grad_norm": 0.243735318182604, + "learning_rate": 1.7831977436443594e-05, + "loss": 0.004, + "step": 477 + }, + { + "epoch": 0.7155688622754491, + "grad_norm": 0.2123776492606604, + "learning_rate": 1.7821913901401616e-05, + "loss": 0.0033, + "step": 478 + }, + { + "epoch": 0.7170658682634731, + "grad_norm": 0.0860593828548492, + "learning_rate": 1.781182991759479e-05, + "loss": 0.0012, + "step": 479 + }, + { + "epoch": 0.718562874251497, + "grad_norm": 0.17480422685210906, + "learning_rate": 1.780172551138559e-05, + "loss": 0.0037, + "step": 480 + }, + { + "epoch": 0.7200598802395209, + "grad_norm": 0.2672043506973695, + "learning_rate": 1.779160070918988e-05, + "loss": 0.0046, + "step": 481 + }, + { + "epoch": 0.7215568862275449, + "grad_norm": 0.30149827478860836, + "learning_rate": 1.7781455537476848e-05, + "loss": 0.0052, + "step": 482 + }, + { + "epoch": 0.7230538922155688, + "grad_norm": 0.23829191956670007, + "learning_rate": 1.7771290022768933e-05, + "loss": 0.0032, + "step": 483 + }, + { + "epoch": 0.7245508982035929, + "grad_norm": 0.19471894040554305, + "learning_rate": 1.776110419164176e-05, + "loss": 0.0025, + "step": 484 + }, + { + "epoch": 0.7260479041916168, + "grad_norm": 0.3281317127559268, + "learning_rate": 1.7750898070724057e-05, + "loss": 0.0053, + "step": 485 + }, + { + "epoch": 0.7275449101796407, + "grad_norm": 0.22137467518837572, + "learning_rate": 1.7740671686697605e-05, + "loss": 0.005, + "step": 486 + }, + { + "epoch": 0.7290419161676647, + "grad_norm": 0.16708307413649082, + "learning_rate": 1.7730425066297158e-05, + "loss": 0.0037, + "step": 487 + }, + { + "epoch": 0.7305389221556886, + "grad_norm": 0.0707938538877929, + "learning_rate": 1.772015823631036e-05, + "loss": 0.0023, + "step": 488 + }, + { + "epoch": 0.7320359281437125, + "grad_norm": 0.1183727230604871, + "learning_rate": 1.7709871223577713e-05, + "loss": 0.0024, + "step": 489 + }, + { + "epoch": 0.7335329341317365, + "grad_norm": 0.20204690451609672, + "learning_rate": 1.769956405499246e-05, + "loss": 0.0045, + "step": 490 + }, + { + "epoch": 0.7350299401197605, + "grad_norm": 0.22427020009198373, + "learning_rate": 1.768923675750055e-05, + "loss": 0.0037, + "step": 491 + }, + { + "epoch": 0.7365269461077845, + "grad_norm": 0.09360542375246439, + "learning_rate": 1.767888935810055e-05, + "loss": 0.0026, + "step": 492 + }, + { + "epoch": 0.7380239520958084, + "grad_norm": 0.18339744192805107, + "learning_rate": 1.7668521883843586e-05, + "loss": 0.003, + "step": 493 + }, + { + "epoch": 0.7395209580838323, + "grad_norm": 0.1079849963238365, + "learning_rate": 1.7658134361833253e-05, + "loss": 0.0019, + "step": 494 + }, + { + "epoch": 0.7410179640718563, + "grad_norm": 0.1345391487313826, + "learning_rate": 1.764772681922557e-05, + "loss": 0.0027, + "step": 495 + }, + { + "epoch": 0.7425149700598802, + "grad_norm": 0.14603341686152715, + "learning_rate": 1.7637299283228887e-05, + "loss": 0.0032, + "step": 496 + }, + { + "epoch": 0.7440119760479041, + "grad_norm": 0.16829676525727566, + "learning_rate": 1.762685178110382e-05, + "loss": 0.0029, + "step": 497 + }, + { + "epoch": 0.7455089820359282, + "grad_norm": 0.2128675734460732, + "learning_rate": 1.7616384340163196e-05, + "loss": 0.0019, + "step": 498 + }, + { + "epoch": 0.7470059880239521, + "grad_norm": 0.12654769972937424, + "learning_rate": 1.7605896987771957e-05, + "loss": 0.0026, + "step": 499 + }, + { + "epoch": 0.7485029940119761, + "grad_norm": 0.05725603798182813, + "learning_rate": 1.75953897513471e-05, + "loss": 0.0011, + "step": 500 + }, + { + "epoch": 0.75, + "grad_norm": 0.12305721333189583, + "learning_rate": 1.7584862658357603e-05, + "loss": 0.002, + "step": 501 + }, + { + "epoch": 0.7514970059880239, + "grad_norm": 0.1998661971442016, + "learning_rate": 1.757431573632437e-05, + "loss": 0.0028, + "step": 502 + }, + { + "epoch": 0.7529940119760479, + "grad_norm": 0.21334825352182804, + "learning_rate": 1.7563749012820115e-05, + "loss": 0.0042, + "step": 503 + }, + { + "epoch": 0.7544910179640718, + "grad_norm": 0.08550116259152069, + "learning_rate": 1.7553162515469346e-05, + "loss": 0.0015, + "step": 504 + }, + { + "epoch": 0.7559880239520959, + "grad_norm": 0.3173087453070982, + "learning_rate": 1.754255627194826e-05, + "loss": 0.0027, + "step": 505 + }, + { + "epoch": 0.7574850299401198, + "grad_norm": 0.12840517943608828, + "learning_rate": 1.7531930309984663e-05, + "loss": 0.0021, + "step": 506 + }, + { + "epoch": 0.7589820359281437, + "grad_norm": 0.22380401998833405, + "learning_rate": 1.752128465735792e-05, + "loss": 0.0042, + "step": 507 + }, + { + "epoch": 0.7604790419161677, + "grad_norm": 0.17755574559866727, + "learning_rate": 1.751061934189888e-05, + "loss": 0.0043, + "step": 508 + }, + { + "epoch": 0.7619760479041916, + "grad_norm": 0.13841142360132738, + "learning_rate": 1.749993439148978e-05, + "loss": 0.0027, + "step": 509 + }, + { + "epoch": 0.7634730538922155, + "grad_norm": 0.19949827808136236, + "learning_rate": 1.7489229834064207e-05, + "loss": 0.0015, + "step": 510 + }, + { + "epoch": 0.7649700598802395, + "grad_norm": 0.08462524659982372, + "learning_rate": 1.747850569760699e-05, + "loss": 0.002, + "step": 511 + }, + { + "epoch": 0.7664670658682635, + "grad_norm": 0.185019733009154, + "learning_rate": 1.7467762010154148e-05, + "loss": 0.0028, + "step": 512 + }, + { + "epoch": 0.7679640718562875, + "grad_norm": 0.14537973618978656, + "learning_rate": 1.7456998799792822e-05, + "loss": 0.0045, + "step": 513 + }, + { + "epoch": 0.7694610778443114, + "grad_norm": 0.0919302359343689, + "learning_rate": 1.7446216094661183e-05, + "loss": 0.0017, + "step": 514 + }, + { + "epoch": 0.7709580838323353, + "grad_norm": 0.09806191993657888, + "learning_rate": 1.7435413922948358e-05, + "loss": 0.0015, + "step": 515 + }, + { + "epoch": 0.7724550898203593, + "grad_norm": 0.17581073699332905, + "learning_rate": 1.742459231289439e-05, + "loss": 0.0036, + "step": 516 + }, + { + "epoch": 0.7739520958083832, + "grad_norm": 0.16231931806189956, + "learning_rate": 1.7413751292790113e-05, + "loss": 0.0035, + "step": 517 + }, + { + "epoch": 0.7754491017964071, + "grad_norm": 0.13525922667777576, + "learning_rate": 1.740289089097712e-05, + "loss": 0.0031, + "step": 518 + }, + { + "epoch": 0.7769461077844312, + "grad_norm": 0.20823952255636147, + "learning_rate": 1.739201113584767e-05, + "loss": 0.0023, + "step": 519 + }, + { + "epoch": 0.7784431137724551, + "grad_norm": 0.16208283117260136, + "learning_rate": 1.738111205584462e-05, + "loss": 0.0014, + "step": 520 + }, + { + "epoch": 0.7799401197604791, + "grad_norm": 0.07133110277997594, + "learning_rate": 1.737019367946134e-05, + "loss": 0.0012, + "step": 521 + }, + { + "epoch": 0.781437125748503, + "grad_norm": 0.08862054013594506, + "learning_rate": 1.735925603524165e-05, + "loss": 0.0015, + "step": 522 + }, + { + "epoch": 0.7829341317365269, + "grad_norm": 0.18494335464634515, + "learning_rate": 1.7348299151779748e-05, + "loss": 0.0023, + "step": 523 + }, + { + "epoch": 0.7844311377245509, + "grad_norm": 0.2671256300047652, + "learning_rate": 1.7337323057720122e-05, + "loss": 0.0047, + "step": 524 + }, + { + "epoch": 0.7859281437125748, + "grad_norm": 0.2617413925072033, + "learning_rate": 1.7326327781757482e-05, + "loss": 0.0038, + "step": 525 + }, + { + "epoch": 0.7874251497005988, + "grad_norm": 0.125912865428105, + "learning_rate": 1.731531335263669e-05, + "loss": 0.0028, + "step": 526 + }, + { + "epoch": 0.7889221556886228, + "grad_norm": 0.13267509240686437, + "learning_rate": 1.7304279799152677e-05, + "loss": 0.0033, + "step": 527 + }, + { + "epoch": 0.7904191616766467, + "grad_norm": 0.14976117950694603, + "learning_rate": 1.7293227150150366e-05, + "loss": 0.0029, + "step": 528 + }, + { + "epoch": 0.7919161676646707, + "grad_norm": 0.09070339743848918, + "learning_rate": 1.728215543452461e-05, + "loss": 0.0016, + "step": 529 + }, + { + "epoch": 0.7934131736526946, + "grad_norm": 0.14490201715247109, + "learning_rate": 1.7271064681220105e-05, + "loss": 0.0016, + "step": 530 + }, + { + "epoch": 0.7949101796407185, + "grad_norm": 0.22041533329824736, + "learning_rate": 1.725995491923131e-05, + "loss": 0.0048, + "step": 531 + }, + { + "epoch": 0.7964071856287425, + "grad_norm": 0.18152422489317313, + "learning_rate": 1.724882617760239e-05, + "loss": 0.0024, + "step": 532 + }, + { + "epoch": 0.7979041916167665, + "grad_norm": 0.13200938457839367, + "learning_rate": 1.7237678485427117e-05, + "loss": 0.0025, + "step": 533 + }, + { + "epoch": 0.7994011976047904, + "grad_norm": 0.2569963553188449, + "learning_rate": 1.7226511871848816e-05, + "loss": 0.003, + "step": 534 + }, + { + "epoch": 0.8008982035928144, + "grad_norm": 0.11384609771555894, + "learning_rate": 1.7215326366060266e-05, + "loss": 0.0024, + "step": 535 + }, + { + "epoch": 0.8023952095808383, + "grad_norm": 0.07638650411484138, + "learning_rate": 1.720412199730365e-05, + "loss": 0.0011, + "step": 536 + }, + { + "epoch": 0.8038922155688623, + "grad_norm": 0.12630645830506568, + "learning_rate": 1.719289879487045e-05, + "loss": 0.0017, + "step": 537 + }, + { + "epoch": 0.8053892215568862, + "grad_norm": 0.17122405410573677, + "learning_rate": 1.7181656788101394e-05, + "loss": 0.0018, + "step": 538 + }, + { + "epoch": 0.8068862275449101, + "grad_norm": 0.11978605848911922, + "learning_rate": 1.7170396006386364e-05, + "loss": 0.0022, + "step": 539 + }, + { + "epoch": 0.8083832335329342, + "grad_norm": 0.22805307525076485, + "learning_rate": 1.7159116479164336e-05, + "loss": 0.0017, + "step": 540 + }, + { + "epoch": 0.8098802395209581, + "grad_norm": 0.042607090167363824, + "learning_rate": 1.7147818235923282e-05, + "loss": 0.001, + "step": 541 + }, + { + "epoch": 0.811377245508982, + "grad_norm": 0.3303493277968501, + "learning_rate": 1.71365013062001e-05, + "loss": 0.003, + "step": 542 + }, + { + "epoch": 0.812874251497006, + "grad_norm": 0.20139977963643052, + "learning_rate": 1.712516571958055e-05, + "loss": 0.0051, + "step": 543 + }, + { + "epoch": 0.8143712574850299, + "grad_norm": 0.18852567500825385, + "learning_rate": 1.711381150569917e-05, + "loss": 0.0023, + "step": 544 + }, + { + "epoch": 0.8158682634730539, + "grad_norm": 0.1533989779344055, + "learning_rate": 1.7102438694239172e-05, + "loss": 0.0015, + "step": 545 + }, + { + "epoch": 0.8173652694610778, + "grad_norm": 0.18945005903791085, + "learning_rate": 1.7091047314932418e-05, + "loss": 0.0052, + "step": 546 + }, + { + "epoch": 0.8188622754491018, + "grad_norm": 0.10940559482924138, + "learning_rate": 1.7079637397559284e-05, + "loss": 0.0036, + "step": 547 + }, + { + "epoch": 0.8203592814371258, + "grad_norm": 0.1733647440192747, + "learning_rate": 1.706820897194863e-05, + "loss": 0.0026, + "step": 548 + }, + { + "epoch": 0.8218562874251497, + "grad_norm": 0.2284794974338924, + "learning_rate": 1.7056762067977696e-05, + "loss": 0.0026, + "step": 549 + }, + { + "epoch": 0.8233532934131736, + "grad_norm": 0.14329072634628595, + "learning_rate": 1.7045296715572024e-05, + "loss": 0.0024, + "step": 550 + }, + { + "epoch": 0.8248502994011976, + "grad_norm": 0.31093763718850237, + "learning_rate": 1.703381294470539e-05, + "loss": 0.0033, + "step": 551 + }, + { + "epoch": 0.8263473053892215, + "grad_norm": 0.20477098480741565, + "learning_rate": 1.7022310785399722e-05, + "loss": 0.0028, + "step": 552 + }, + { + "epoch": 0.8278443113772455, + "grad_norm": 0.10382900594838523, + "learning_rate": 1.7010790267725023e-05, + "loss": 0.0017, + "step": 553 + }, + { + "epoch": 0.8293413173652695, + "grad_norm": 0.10870430271363642, + "learning_rate": 1.6999251421799285e-05, + "loss": 0.0014, + "step": 554 + }, + { + "epoch": 0.8308383233532934, + "grad_norm": 0.2938613690271062, + "learning_rate": 1.698769427778842e-05, + "loss": 0.0043, + "step": 555 + }, + { + "epoch": 0.8323353293413174, + "grad_norm": 0.16919545121928964, + "learning_rate": 1.6976118865906167e-05, + "loss": 0.0024, + "step": 556 + }, + { + "epoch": 0.8338323353293413, + "grad_norm": 0.19047981855746474, + "learning_rate": 1.6964525216414033e-05, + "loss": 0.0028, + "step": 557 + }, + { + "epoch": 0.8353293413173652, + "grad_norm": 0.15373361819794654, + "learning_rate": 1.6952913359621204e-05, + "loss": 0.0028, + "step": 558 + }, + { + "epoch": 0.8368263473053892, + "grad_norm": 0.13994232783724012, + "learning_rate": 1.6941283325884458e-05, + "loss": 0.0019, + "step": 559 + }, + { + "epoch": 0.8383233532934131, + "grad_norm": 0.20016507933450822, + "learning_rate": 1.6929635145608098e-05, + "loss": 0.0031, + "step": 560 + }, + { + "epoch": 0.8398203592814372, + "grad_norm": 0.11790210879503302, + "learning_rate": 1.691796884924386e-05, + "loss": 0.0017, + "step": 561 + }, + { + "epoch": 0.8413173652694611, + "grad_norm": 0.060596148104514636, + "learning_rate": 1.6906284467290852e-05, + "loss": 0.0009, + "step": 562 + }, + { + "epoch": 0.842814371257485, + "grad_norm": 0.11618444614587378, + "learning_rate": 1.6894582030295458e-05, + "loss": 0.0018, + "step": 563 + }, + { + "epoch": 0.844311377245509, + "grad_norm": 0.16389923156347438, + "learning_rate": 1.6882861568851258e-05, + "loss": 0.0025, + "step": 564 + }, + { + "epoch": 0.8458083832335329, + "grad_norm": 0.2998453874019442, + "learning_rate": 1.6871123113598954e-05, + "loss": 0.0029, + "step": 565 + }, + { + "epoch": 0.8473053892215568, + "grad_norm": 0.2986815290378479, + "learning_rate": 1.6859366695226303e-05, + "loss": 0.003, + "step": 566 + }, + { + "epoch": 0.8488023952095808, + "grad_norm": 0.21672636750976904, + "learning_rate": 1.6847592344468e-05, + "loss": 0.0035, + "step": 567 + }, + { + "epoch": 0.8502994011976048, + "grad_norm": 0.10471468615303521, + "learning_rate": 1.6835800092105647e-05, + "loss": 0.0015, + "step": 568 + }, + { + "epoch": 0.8517964071856288, + "grad_norm": 0.12014153904526044, + "learning_rate": 1.682398996896762e-05, + "loss": 0.0016, + "step": 569 + }, + { + "epoch": 0.8532934131736527, + "grad_norm": 0.24386062518444432, + "learning_rate": 1.681216200592903e-05, + "loss": 0.0029, + "step": 570 + }, + { + "epoch": 0.8547904191616766, + "grad_norm": 0.4891409878621194, + "learning_rate": 1.6800316233911627e-05, + "loss": 0.003, + "step": 571 + }, + { + "epoch": 0.8562874251497006, + "grad_norm": 0.11158025115959051, + "learning_rate": 1.678845268388371e-05, + "loss": 0.0013, + "step": 572 + }, + { + "epoch": 0.8577844311377245, + "grad_norm": 0.16176832999905424, + "learning_rate": 1.677657138686006e-05, + "loss": 0.0029, + "step": 573 + }, + { + "epoch": 0.8592814371257484, + "grad_norm": 0.2862714242004369, + "learning_rate": 1.6764672373901857e-05, + "loss": 0.0044, + "step": 574 + }, + { + "epoch": 0.8607784431137725, + "grad_norm": 0.23371137768377162, + "learning_rate": 1.6752755676116586e-05, + "loss": 0.0029, + "step": 575 + }, + { + "epoch": 0.8622754491017964, + "grad_norm": 0.16513171198212837, + "learning_rate": 1.6740821324657983e-05, + "loss": 0.0028, + "step": 576 + }, + { + "epoch": 0.8637724550898204, + "grad_norm": 0.0989955200636401, + "learning_rate": 1.672886935072591e-05, + "loss": 0.001, + "step": 577 + }, + { + "epoch": 0.8652694610778443, + "grad_norm": 0.3256050916149061, + "learning_rate": 1.6716899785566324e-05, + "loss": 0.0041, + "step": 578 + }, + { + "epoch": 0.8667664670658682, + "grad_norm": 0.2090094852879972, + "learning_rate": 1.670491266047116e-05, + "loss": 0.0032, + "step": 579 + }, + { + "epoch": 0.8682634730538922, + "grad_norm": 0.15452654886105557, + "learning_rate": 1.669290800677825e-05, + "loss": 0.0024, + "step": 580 + }, + { + "epoch": 0.8697604790419161, + "grad_norm": 0.12984740923736884, + "learning_rate": 1.668088585587127e-05, + "loss": 0.0021, + "step": 581 + }, + { + "epoch": 0.8712574850299402, + "grad_norm": 0.20135825659725679, + "learning_rate": 1.666884623917963e-05, + "loss": 0.0025, + "step": 582 + }, + { + "epoch": 0.8727544910179641, + "grad_norm": 0.2843598279050385, + "learning_rate": 1.6656789188178394e-05, + "loss": 0.0048, + "step": 583 + }, + { + "epoch": 0.874251497005988, + "grad_norm": 0.18200615608698034, + "learning_rate": 1.664471473438822e-05, + "loss": 0.0032, + "step": 584 + }, + { + "epoch": 0.875748502994012, + "grad_norm": 0.10309687291550676, + "learning_rate": 1.6632622909375244e-05, + "loss": 0.0014, + "step": 585 + }, + { + "epoch": 0.8772455089820359, + "grad_norm": 0.06697443098407523, + "learning_rate": 1.662051374475103e-05, + "loss": 0.0008, + "step": 586 + }, + { + "epoch": 0.8787425149700598, + "grad_norm": 0.20351165773382962, + "learning_rate": 1.660838727217246e-05, + "loss": 0.0025, + "step": 587 + }, + { + "epoch": 0.8802395209580839, + "grad_norm": 0.19444272969584192, + "learning_rate": 1.6596243523341678e-05, + "loss": 0.0048, + "step": 588 + }, + { + "epoch": 0.8817365269461078, + "grad_norm": 0.1718287018179276, + "learning_rate": 1.658408253000598e-05, + "loss": 0.0036, + "step": 589 + }, + { + "epoch": 0.8832335329341318, + "grad_norm": 0.1473170594956288, + "learning_rate": 1.657190432395776e-05, + "loss": 0.0033, + "step": 590 + }, + { + "epoch": 0.8847305389221557, + "grad_norm": 0.05320979757128209, + "learning_rate": 1.6559708937034394e-05, + "loss": 0.0009, + "step": 591 + }, + { + "epoch": 0.8862275449101796, + "grad_norm": 0.1948054351678213, + "learning_rate": 1.654749640111818e-05, + "loss": 0.0027, + "step": 592 + }, + { + "epoch": 0.8877245508982036, + "grad_norm": 0.1342630451535342, + "learning_rate": 1.6535266748136255e-05, + "loss": 0.0021, + "step": 593 + }, + { + "epoch": 0.8892215568862275, + "grad_norm": 0.10784490714170987, + "learning_rate": 1.6523020010060497e-05, + "loss": 0.0027, + "step": 594 + }, + { + "epoch": 0.8907185628742516, + "grad_norm": 0.1546876958105241, + "learning_rate": 1.6510756218907453e-05, + "loss": 0.0022, + "step": 595 + }, + { + "epoch": 0.8922155688622755, + "grad_norm": 0.16370527891824135, + "learning_rate": 1.6498475406738248e-05, + "loss": 0.0023, + "step": 596 + }, + { + "epoch": 0.8937125748502994, + "grad_norm": 0.09977139023857336, + "learning_rate": 1.648617760565851e-05, + "loss": 0.0009, + "step": 597 + }, + { + "epoch": 0.8952095808383234, + "grad_norm": 0.14481161212272883, + "learning_rate": 1.647386284781828e-05, + "loss": 0.0026, + "step": 598 + }, + { + "epoch": 0.8967065868263473, + "grad_norm": 0.2579988975748116, + "learning_rate": 1.6461531165411924e-05, + "loss": 0.0039, + "step": 599 + }, + { + "epoch": 0.8982035928143712, + "grad_norm": 0.11325757290508497, + "learning_rate": 1.644918259067806e-05, + "loss": 0.0015, + "step": 600 + }, + { + "epoch": 0.8997005988023952, + "grad_norm": 0.10693055025728564, + "learning_rate": 1.643681715589946e-05, + "loss": 0.0018, + "step": 601 + }, + { + "epoch": 0.9011976047904192, + "grad_norm": 0.10387304790724435, + "learning_rate": 1.6424434893402988e-05, + "loss": 0.0016, + "step": 602 + }, + { + "epoch": 0.9026946107784432, + "grad_norm": 0.08953073156215048, + "learning_rate": 1.6412035835559483e-05, + "loss": 0.0008, + "step": 603 + }, + { + "epoch": 0.9041916167664671, + "grad_norm": 0.1882854229415982, + "learning_rate": 1.6399620014783702e-05, + "loss": 0.0035, + "step": 604 + }, + { + "epoch": 0.905688622754491, + "grad_norm": 0.05376462878864106, + "learning_rate": 1.6387187463534222e-05, + "loss": 0.0007, + "step": 605 + }, + { + "epoch": 0.907185628742515, + "grad_norm": 0.2652074145829586, + "learning_rate": 1.6374738214313354e-05, + "loss": 0.0035, + "step": 606 + }, + { + "epoch": 0.9086826347305389, + "grad_norm": 0.11598887341928915, + "learning_rate": 1.6362272299667078e-05, + "loss": 0.001, + "step": 607 + }, + { + "epoch": 0.9101796407185628, + "grad_norm": 0.0870013457319502, + "learning_rate": 1.6349789752184922e-05, + "loss": 0.0013, + "step": 608 + }, + { + "epoch": 0.9116766467065869, + "grad_norm": 0.17343366028225063, + "learning_rate": 1.6337290604499913e-05, + "loss": 0.0025, + "step": 609 + }, + { + "epoch": 0.9131736526946108, + "grad_norm": 0.0859523102698752, + "learning_rate": 1.6324774889288462e-05, + "loss": 0.0018, + "step": 610 + }, + { + "epoch": 0.9146706586826348, + "grad_norm": 0.07712475151325601, + "learning_rate": 1.6312242639270308e-05, + "loss": 0.0011, + "step": 611 + }, + { + "epoch": 0.9161676646706587, + "grad_norm": 0.07162451662462921, + "learning_rate": 1.6299693887208404e-05, + "loss": 0.0031, + "step": 612 + }, + { + "epoch": 0.9176646706586826, + "grad_norm": 0.1607194705076756, + "learning_rate": 1.628712866590885e-05, + "loss": 0.0021, + "step": 613 + }, + { + "epoch": 0.9191616766467066, + "grad_norm": 0.08490453147422353, + "learning_rate": 1.6274547008220804e-05, + "loss": 0.0015, + "step": 614 + }, + { + "epoch": 0.9206586826347305, + "grad_norm": 0.1508869109445389, + "learning_rate": 1.626194894703638e-05, + "loss": 0.0024, + "step": 615 + }, + { + "epoch": 0.9221556886227545, + "grad_norm": 0.08485546738561354, + "learning_rate": 1.6249334515290593e-05, + "loss": 0.0013, + "step": 616 + }, + { + "epoch": 0.9236526946107785, + "grad_norm": 0.16110417372420824, + "learning_rate": 1.6236703745961244e-05, + "loss": 0.0022, + "step": 617 + }, + { + "epoch": 0.9251497005988024, + "grad_norm": 0.06607135428636862, + "learning_rate": 1.622405667206885e-05, + "loss": 0.0007, + "step": 618 + }, + { + "epoch": 0.9266467065868264, + "grad_norm": 0.2938746182376294, + "learning_rate": 1.6211393326676558e-05, + "loss": 0.0031, + "step": 619 + }, + { + "epoch": 0.9281437125748503, + "grad_norm": 0.14606146095893197, + "learning_rate": 1.6198713742890036e-05, + "loss": 0.0029, + "step": 620 + }, + { + "epoch": 0.9296407185628742, + "grad_norm": 0.05652377223402481, + "learning_rate": 1.6186017953857423e-05, + "loss": 0.0009, + "step": 621 + }, + { + "epoch": 0.9311377245508982, + "grad_norm": 0.18019462860068167, + "learning_rate": 1.617330599276921e-05, + "loss": 0.0043, + "step": 622 + }, + { + "epoch": 0.9326347305389222, + "grad_norm": 0.0804731431661998, + "learning_rate": 1.6160577892858178e-05, + "loss": 0.0012, + "step": 623 + }, + { + "epoch": 0.9341317365269461, + "grad_norm": 0.23044713115642781, + "learning_rate": 1.6147833687399287e-05, + "loss": 0.0026, + "step": 624 + }, + { + "epoch": 0.9356287425149701, + "grad_norm": 0.15040761212205267, + "learning_rate": 1.6135073409709613e-05, + "loss": 0.0033, + "step": 625 + }, + { + "epoch": 0.937125748502994, + "grad_norm": 0.17617733364441931, + "learning_rate": 1.6122297093148242e-05, + "loss": 0.0021, + "step": 626 + }, + { + "epoch": 0.938622754491018, + "grad_norm": 0.2182223824425347, + "learning_rate": 1.6109504771116193e-05, + "loss": 0.0014, + "step": 627 + }, + { + "epoch": 0.9401197604790419, + "grad_norm": 0.20104637613306378, + "learning_rate": 1.6096696477056325e-05, + "loss": 0.0025, + "step": 628 + }, + { + "epoch": 0.9416167664670658, + "grad_norm": 0.10319298502135164, + "learning_rate": 1.608387224445326e-05, + "loss": 0.0016, + "step": 629 + }, + { + "epoch": 0.9431137724550899, + "grad_norm": 0.1744369089861333, + "learning_rate": 1.607103210683328e-05, + "loss": 0.0019, + "step": 630 + }, + { + "epoch": 0.9446107784431138, + "grad_norm": 0.1344166064567352, + "learning_rate": 1.6058176097764254e-05, + "loss": 0.0013, + "step": 631 + }, + { + "epoch": 0.9461077844311377, + "grad_norm": 0.08853174252976158, + "learning_rate": 1.6045304250855534e-05, + "loss": 0.0013, + "step": 632 + }, + { + "epoch": 0.9476047904191617, + "grad_norm": 0.08846484292192285, + "learning_rate": 1.6032416599757894e-05, + "loss": 0.002, + "step": 633 + }, + { + "epoch": 0.9491017964071856, + "grad_norm": 0.12884476368073972, + "learning_rate": 1.6019513178163403e-05, + "loss": 0.0011, + "step": 634 + }, + { + "epoch": 0.9505988023952096, + "grad_norm": 0.15795219235072977, + "learning_rate": 1.6006594019805376e-05, + "loss": 0.0011, + "step": 635 + }, + { + "epoch": 0.9520958083832335, + "grad_norm": 0.13187039248116258, + "learning_rate": 1.5993659158458255e-05, + "loss": 0.0024, + "step": 636 + }, + { + "epoch": 0.9535928143712575, + "grad_norm": 0.15925859946439425, + "learning_rate": 1.5980708627937548e-05, + "loss": 0.0025, + "step": 637 + }, + { + "epoch": 0.9550898203592815, + "grad_norm": 0.13331313394073413, + "learning_rate": 1.5967742462099717e-05, + "loss": 0.0017, + "step": 638 + }, + { + "epoch": 0.9565868263473054, + "grad_norm": 0.08975182166363489, + "learning_rate": 1.59547606948421e-05, + "loss": 0.0014, + "step": 639 + }, + { + "epoch": 0.9580838323353293, + "grad_norm": 0.3449291274347138, + "learning_rate": 1.594176336010283e-05, + "loss": 0.0024, + "step": 640 + }, + { + "epoch": 0.9595808383233533, + "grad_norm": 0.06409320220899797, + "learning_rate": 1.592875049186073e-05, + "loss": 0.0011, + "step": 641 + }, + { + "epoch": 0.9610778443113772, + "grad_norm": 0.17833684537580516, + "learning_rate": 1.5915722124135227e-05, + "loss": 0.002, + "step": 642 + }, + { + "epoch": 0.9625748502994012, + "grad_norm": 0.15331420541889684, + "learning_rate": 1.590267829098628e-05, + "loss": 0.002, + "step": 643 + }, + { + "epoch": 0.9640718562874252, + "grad_norm": 0.5899661957634779, + "learning_rate": 1.5889619026514272e-05, + "loss": 0.0043, + "step": 644 + }, + { + "epoch": 0.9655688622754491, + "grad_norm": 0.1657525859558746, + "learning_rate": 1.5876544364859937e-05, + "loss": 0.0022, + "step": 645 + }, + { + "epoch": 0.9670658682634731, + "grad_norm": 0.18379072480540792, + "learning_rate": 1.5863454340204247e-05, + "loss": 0.003, + "step": 646 + }, + { + "epoch": 0.968562874251497, + "grad_norm": 0.12209910222944766, + "learning_rate": 1.585034898676835e-05, + "loss": 0.0015, + "step": 647 + }, + { + "epoch": 0.9700598802395209, + "grad_norm": 0.10607555085781809, + "learning_rate": 1.583722833881346e-05, + "loss": 0.0019, + "step": 648 + }, + { + "epoch": 0.9715568862275449, + "grad_norm": 0.26474190068427933, + "learning_rate": 1.5824092430640784e-05, + "loss": 0.0038, + "step": 649 + }, + { + "epoch": 0.9730538922155688, + "grad_norm": 0.4156799624786208, + "learning_rate": 1.5810941296591406e-05, + "loss": 0.0028, + "step": 650 + }, + { + "epoch": 0.9745508982035929, + "grad_norm": 0.353881703591171, + "learning_rate": 1.5797774971046243e-05, + "loss": 0.0041, + "step": 651 + }, + { + "epoch": 0.9760479041916168, + "grad_norm": 0.47331385633169554, + "learning_rate": 1.5784593488425898e-05, + "loss": 0.0036, + "step": 652 + }, + { + "epoch": 0.9775449101796407, + "grad_norm": 0.23996897207608175, + "learning_rate": 1.577139688319062e-05, + "loss": 0.0042, + "step": 653 + }, + { + "epoch": 0.9790419161676647, + "grad_norm": 0.14040120360537384, + "learning_rate": 1.5758185189840182e-05, + "loss": 0.0015, + "step": 654 + }, + { + "epoch": 0.9805389221556886, + "grad_norm": 0.16782002665806564, + "learning_rate": 1.5744958442913804e-05, + "loss": 0.0033, + "step": 655 + }, + { + "epoch": 0.9820359281437125, + "grad_norm": 0.3548917708393882, + "learning_rate": 1.5731716676990057e-05, + "loss": 0.0037, + "step": 656 + }, + { + "epoch": 0.9835329341317365, + "grad_norm": 0.19405634340063596, + "learning_rate": 1.571845992668679e-05, + "loss": 0.0024, + "step": 657 + }, + { + "epoch": 0.9850299401197605, + "grad_norm": 0.23475543702616705, + "learning_rate": 1.5705188226661013e-05, + "loss": 0.0045, + "step": 658 + }, + { + "epoch": 0.9865269461077845, + "grad_norm": 0.36667176478027685, + "learning_rate": 1.569190161160882e-05, + "loss": 0.0039, + "step": 659 + }, + { + "epoch": 0.9880239520958084, + "grad_norm": 0.1896831017088073, + "learning_rate": 1.5678600116265293e-05, + "loss": 0.0027, + "step": 660 + }, + { + "epoch": 0.9895209580838323, + "grad_norm": 0.10818436139444451, + "learning_rate": 1.566528377540443e-05, + "loss": 0.0021, + "step": 661 + }, + { + "epoch": 0.9910179640718563, + "grad_norm": 0.1635417574332313, + "learning_rate": 1.5651952623839028e-05, + "loss": 0.0009, + "step": 662 + }, + { + "epoch": 0.9925149700598802, + "grad_norm": 0.2787481879240596, + "learning_rate": 1.5638606696420602e-05, + "loss": 0.0034, + "step": 663 + }, + { + "epoch": 0.9940119760479041, + "grad_norm": 0.12776549267504195, + "learning_rate": 1.5625246028039308e-05, + "loss": 0.0013, + "step": 664 + }, + { + "epoch": 0.9955089820359282, + "grad_norm": 0.19106271050333715, + "learning_rate": 1.5611870653623826e-05, + "loss": 0.0026, + "step": 665 + }, + { + "epoch": 0.9970059880239521, + "grad_norm": 0.1120405670371837, + "learning_rate": 1.5598480608141286e-05, + "loss": 0.0013, + "step": 666 + }, + { + "epoch": 0.9985029940119761, + "grad_norm": 0.06889850891061895, + "learning_rate": 1.5585075926597174e-05, + "loss": 0.0008, + "step": 667 + }, + { + "epoch": 1.0, + "grad_norm": 0.1249403668339043, + "learning_rate": 1.557165664403524e-05, + "loss": 0.0011, + "step": 668 + }, + { + "epoch": 1.001497005988024, + "grad_norm": 0.22792093858397705, + "learning_rate": 1.5558222795537403e-05, + "loss": 0.0021, + "step": 669 + }, + { + "epoch": 1.0029940119760479, + "grad_norm": 0.13959837895980517, + "learning_rate": 1.554477441622366e-05, + "loss": 0.0019, + "step": 670 + }, + { + "epoch": 1.0044910179640718, + "grad_norm": 0.11152024170430974, + "learning_rate": 1.5531311541251995e-05, + "loss": 0.0015, + "step": 671 + }, + { + "epoch": 1.0059880239520957, + "grad_norm": 0.16100791789200442, + "learning_rate": 1.55178342058183e-05, + "loss": 0.0039, + "step": 672 + }, + { + "epoch": 1.0074850299401197, + "grad_norm": 0.16419979563070153, + "learning_rate": 1.5504342445156253e-05, + "loss": 0.0018, + "step": 673 + }, + { + "epoch": 1.0089820359281436, + "grad_norm": 0.18649044296735573, + "learning_rate": 1.549083629453726e-05, + "loss": 0.0029, + "step": 674 + }, + { + "epoch": 1.0104790419161676, + "grad_norm": 0.12322267561572196, + "learning_rate": 1.547731578927033e-05, + "loss": 0.0017, + "step": 675 + }, + { + "epoch": 1.0119760479041917, + "grad_norm": 0.1847824809505218, + "learning_rate": 1.5463780964702017e-05, + "loss": 0.002, + "step": 676 + }, + { + "epoch": 1.0134730538922156, + "grad_norm": 0.09718374857400291, + "learning_rate": 1.5450231856216294e-05, + "loss": 0.0014, + "step": 677 + }, + { + "epoch": 1.0149700598802396, + "grad_norm": 0.11714727810772602, + "learning_rate": 1.5436668499234487e-05, + "loss": 0.0015, + "step": 678 + }, + { + "epoch": 1.0164670658682635, + "grad_norm": 0.17180981362774483, + "learning_rate": 1.5423090929215164e-05, + "loss": 0.0016, + "step": 679 + }, + { + "epoch": 1.0179640718562875, + "grad_norm": 0.1229563131274387, + "learning_rate": 1.5409499181654062e-05, + "loss": 0.0022, + "step": 680 + }, + { + "epoch": 1.0194610778443114, + "grad_norm": 0.08902729595877537, + "learning_rate": 1.5395893292083967e-05, + "loss": 0.0012, + "step": 681 + }, + { + "epoch": 1.0209580838323353, + "grad_norm": 0.08491224818504717, + "learning_rate": 1.5382273296074647e-05, + "loss": 0.0014, + "step": 682 + }, + { + "epoch": 1.0224550898203593, + "grad_norm": 0.12463262583052379, + "learning_rate": 1.5368639229232744e-05, + "loss": 0.0012, + "step": 683 + }, + { + "epoch": 1.0239520958083832, + "grad_norm": 0.08714987099310947, + "learning_rate": 1.5354991127201682e-05, + "loss": 0.0009, + "step": 684 + }, + { + "epoch": 1.0254491017964071, + "grad_norm": 0.10996908003888108, + "learning_rate": 1.5341329025661593e-05, + "loss": 0.0016, + "step": 685 + }, + { + "epoch": 1.026946107784431, + "grad_norm": 0.06688339180384366, + "learning_rate": 1.5327652960329182e-05, + "loss": 0.0011, + "step": 686 + }, + { + "epoch": 1.028443113772455, + "grad_norm": 0.11389018736693612, + "learning_rate": 1.5313962966957683e-05, + "loss": 0.0013, + "step": 687 + }, + { + "epoch": 1.029940119760479, + "grad_norm": 0.1487541170115952, + "learning_rate": 1.5300259081336733e-05, + "loss": 0.0029, + "step": 688 + }, + { + "epoch": 1.031437125748503, + "grad_norm": 0.18232592006619516, + "learning_rate": 1.5286541339292277e-05, + "loss": 0.0019, + "step": 689 + }, + { + "epoch": 1.032934131736527, + "grad_norm": 0.09073197180745511, + "learning_rate": 1.5272809776686507e-05, + "loss": 0.001, + "step": 690 + }, + { + "epoch": 1.034431137724551, + "grad_norm": 0.11198750989897426, + "learning_rate": 1.5259064429417724e-05, + "loss": 0.0011, + "step": 691 + }, + { + "epoch": 1.035928143712575, + "grad_norm": 0.13524279713865056, + "learning_rate": 1.5245305333420284e-05, + "loss": 0.0027, + "step": 692 + }, + { + "epoch": 1.0374251497005988, + "grad_norm": 0.13473254834146128, + "learning_rate": 1.5231532524664472e-05, + "loss": 0.0013, + "step": 693 + }, + { + "epoch": 1.0389221556886228, + "grad_norm": 0.09705487229884018, + "learning_rate": 1.5217746039156426e-05, + "loss": 0.0014, + "step": 694 + }, + { + "epoch": 1.0404191616766467, + "grad_norm": 0.12743903961819494, + "learning_rate": 1.5203945912938047e-05, + "loss": 0.0031, + "step": 695 + }, + { + "epoch": 1.0419161676646707, + "grad_norm": 0.15772232058272861, + "learning_rate": 1.5190132182086887e-05, + "loss": 0.0035, + "step": 696 + }, + { + "epoch": 1.0434131736526946, + "grad_norm": 0.11967873813625308, + "learning_rate": 1.5176304882716074e-05, + "loss": 0.0021, + "step": 697 + }, + { + "epoch": 1.0449101796407185, + "grad_norm": 0.061224633547482255, + "learning_rate": 1.5162464050974195e-05, + "loss": 0.0006, + "step": 698 + }, + { + "epoch": 1.0464071856287425, + "grad_norm": 0.12437711893494492, + "learning_rate": 1.5148609723045227e-05, + "loss": 0.0012, + "step": 699 + }, + { + "epoch": 1.0479041916167664, + "grad_norm": 0.1745613764437353, + "learning_rate": 1.513474193514842e-05, + "loss": 0.0017, + "step": 700 + }, + { + "epoch": 1.0494011976047903, + "grad_norm": 0.16175781608452677, + "learning_rate": 1.5120860723538224e-05, + "loss": 0.0012, + "step": 701 + }, + { + "epoch": 1.0508982035928143, + "grad_norm": 0.10221564506113642, + "learning_rate": 1.5106966124504172e-05, + "loss": 0.002, + "step": 702 + }, + { + "epoch": 1.0523952095808382, + "grad_norm": 0.062271705757527857, + "learning_rate": 1.5093058174370798e-05, + "loss": 0.0008, + "step": 703 + }, + { + "epoch": 1.0538922155688624, + "grad_norm": 0.10864619286621946, + "learning_rate": 1.5079136909497542e-05, + "loss": 0.0009, + "step": 704 + }, + { + "epoch": 1.0553892215568863, + "grad_norm": 0.19647706361842368, + "learning_rate": 1.5065202366278657e-05, + "loss": 0.0014, + "step": 705 + }, + { + "epoch": 1.0568862275449102, + "grad_norm": 0.07888279226925543, + "learning_rate": 1.5051254581143095e-05, + "loss": 0.0008, + "step": 706 + }, + { + "epoch": 1.0583832335329342, + "grad_norm": 0.12025352331354569, + "learning_rate": 1.5037293590554442e-05, + "loss": 0.0012, + "step": 707 + }, + { + "epoch": 1.0598802395209581, + "grad_norm": 0.08615367646336641, + "learning_rate": 1.5023319431010798e-05, + "loss": 0.0013, + "step": 708 + }, + { + "epoch": 1.061377245508982, + "grad_norm": 0.14909244953515943, + "learning_rate": 1.5009332139044696e-05, + "loss": 0.0013, + "step": 709 + }, + { + "epoch": 1.062874251497006, + "grad_norm": 0.10899940946016365, + "learning_rate": 1.4995331751222992e-05, + "loss": 0.0014, + "step": 710 + }, + { + "epoch": 1.06437125748503, + "grad_norm": 0.13011579272063212, + "learning_rate": 1.4981318304146786e-05, + "loss": 0.0026, + "step": 711 + }, + { + "epoch": 1.0658682634730539, + "grad_norm": 0.06507001266187337, + "learning_rate": 1.4967291834451317e-05, + "loss": 0.0008, + "step": 712 + }, + { + "epoch": 1.0673652694610778, + "grad_norm": 0.1452953575075588, + "learning_rate": 1.4953252378805872e-05, + "loss": 0.0027, + "step": 713 + }, + { + "epoch": 1.0688622754491017, + "grad_norm": 0.05183547815818978, + "learning_rate": 1.493919997391368e-05, + "loss": 0.0007, + "step": 714 + }, + { + "epoch": 1.0703592814371257, + "grad_norm": 0.07714623310644146, + "learning_rate": 1.492513465651183e-05, + "loss": 0.0008, + "step": 715 + }, + { + "epoch": 1.0718562874251496, + "grad_norm": 0.1273003039382832, + "learning_rate": 1.4911056463371161e-05, + "loss": 0.0018, + "step": 716 + }, + { + "epoch": 1.0733532934131738, + "grad_norm": 0.14647086320390126, + "learning_rate": 1.4896965431296182e-05, + "loss": 0.002, + "step": 717 + }, + { + "epoch": 1.0748502994011977, + "grad_norm": 0.11031963684292324, + "learning_rate": 1.4882861597124954e-05, + "loss": 0.0017, + "step": 718 + }, + { + "epoch": 1.0763473053892216, + "grad_norm": 0.13321582362567036, + "learning_rate": 1.4868744997729024e-05, + "loss": 0.0015, + "step": 719 + }, + { + "epoch": 1.0778443113772456, + "grad_norm": 0.12391394850850491, + "learning_rate": 1.4854615670013295e-05, + "loss": 0.0014, + "step": 720 + }, + { + "epoch": 1.0793413173652695, + "grad_norm": 0.16741971891552657, + "learning_rate": 1.4840473650915952e-05, + "loss": 0.0014, + "step": 721 + }, + { + "epoch": 1.0808383233532934, + "grad_norm": 0.06084190134638075, + "learning_rate": 1.4826318977408358e-05, + "loss": 0.0008, + "step": 722 + }, + { + "epoch": 1.0823353293413174, + "grad_norm": 0.09523046312621482, + "learning_rate": 1.4812151686494962e-05, + "loss": 0.0011, + "step": 723 + }, + { + "epoch": 1.0838323353293413, + "grad_norm": 0.2782155057266579, + "learning_rate": 1.4797971815213194e-05, + "loss": 0.0027, + "step": 724 + }, + { + "epoch": 1.0853293413173652, + "grad_norm": 0.07866338116148361, + "learning_rate": 1.4783779400633372e-05, + "loss": 0.0019, + "step": 725 + }, + { + "epoch": 1.0868263473053892, + "grad_norm": 0.08163897393875569, + "learning_rate": 1.4769574479858612e-05, + "loss": 0.0013, + "step": 726 + }, + { + "epoch": 1.0883233532934131, + "grad_norm": 0.06549200205338107, + "learning_rate": 1.475535709002472e-05, + "loss": 0.0008, + "step": 727 + }, + { + "epoch": 1.089820359281437, + "grad_norm": 0.10112474427638028, + "learning_rate": 1.4741127268300097e-05, + "loss": 0.0015, + "step": 728 + }, + { + "epoch": 1.091317365269461, + "grad_norm": 0.03922519037579488, + "learning_rate": 1.4726885051885654e-05, + "loss": 0.0005, + "step": 729 + }, + { + "epoch": 1.092814371257485, + "grad_norm": 0.12932978632302858, + "learning_rate": 1.4712630478014696e-05, + "loss": 0.0013, + "step": 730 + }, + { + "epoch": 1.0943113772455089, + "grad_norm": 0.14927429413650972, + "learning_rate": 1.4698363583952839e-05, + "loss": 0.0019, + "step": 731 + }, + { + "epoch": 1.095808383233533, + "grad_norm": 0.08123699864404504, + "learning_rate": 1.4684084406997903e-05, + "loss": 0.0011, + "step": 732 + }, + { + "epoch": 1.097305389221557, + "grad_norm": 0.05959853755308904, + "learning_rate": 1.4669792984479828e-05, + "loss": 0.0012, + "step": 733 + }, + { + "epoch": 1.098802395209581, + "grad_norm": 0.11908929599820205, + "learning_rate": 1.4655489353760555e-05, + "loss": 0.0014, + "step": 734 + }, + { + "epoch": 1.1002994011976048, + "grad_norm": 0.09093165505220649, + "learning_rate": 1.4641173552233954e-05, + "loss": 0.0008, + "step": 735 + }, + { + "epoch": 1.1017964071856288, + "grad_norm": 0.11618048676071026, + "learning_rate": 1.4626845617325703e-05, + "loss": 0.0013, + "step": 736 + }, + { + "epoch": 1.1032934131736527, + "grad_norm": 0.06281753166476825, + "learning_rate": 1.4612505586493204e-05, + "loss": 0.0007, + "step": 737 + }, + { + "epoch": 1.1047904191616766, + "grad_norm": 0.2023556555593652, + "learning_rate": 1.4598153497225482e-05, + "loss": 0.002, + "step": 738 + }, + { + "epoch": 1.1062874251497006, + "grad_norm": 0.11177065874880013, + "learning_rate": 1.4583789387043083e-05, + "loss": 0.0013, + "step": 739 + }, + { + "epoch": 1.1077844311377245, + "grad_norm": 0.14741071238122402, + "learning_rate": 1.4569413293497986e-05, + "loss": 0.0015, + "step": 740 + }, + { + "epoch": 1.1092814371257484, + "grad_norm": 0.24658469283044443, + "learning_rate": 1.455502525417349e-05, + "loss": 0.0012, + "step": 741 + }, + { + "epoch": 1.1107784431137724, + "grad_norm": 0.13681115792701468, + "learning_rate": 1.454062530668413e-05, + "loss": 0.0017, + "step": 742 + }, + { + "epoch": 1.1122754491017963, + "grad_norm": 0.0834969421467918, + "learning_rate": 1.4526213488675564e-05, + "loss": 0.001, + "step": 743 + }, + { + "epoch": 1.1137724550898203, + "grad_norm": 0.1940908997890325, + "learning_rate": 1.4511789837824495e-05, + "loss": 0.0015, + "step": 744 + }, + { + "epoch": 1.1152694610778444, + "grad_norm": 0.10979405041458333, + "learning_rate": 1.4497354391838551e-05, + "loss": 0.0011, + "step": 745 + }, + { + "epoch": 1.1167664670658684, + "grad_norm": 0.11938289413679191, + "learning_rate": 1.4482907188456198e-05, + "loss": 0.0012, + "step": 746 + }, + { + "epoch": 1.1182634730538923, + "grad_norm": 0.1358936092557254, + "learning_rate": 1.4468448265446646e-05, + "loss": 0.0013, + "step": 747 + }, + { + "epoch": 1.1197604790419162, + "grad_norm": 0.1354545576585189, + "learning_rate": 1.4453977660609733e-05, + "loss": 0.0014, + "step": 748 + }, + { + "epoch": 1.1212574850299402, + "grad_norm": 0.08442533262908321, + "learning_rate": 1.4439495411775844e-05, + "loss": 0.001, + "step": 749 + }, + { + "epoch": 1.122754491017964, + "grad_norm": 0.07602658081845277, + "learning_rate": 1.4425001556805796e-05, + "loss": 0.0008, + "step": 750 + }, + { + "epoch": 1.124251497005988, + "grad_norm": 0.13026513890902275, + "learning_rate": 1.4410496133590763e-05, + "loss": 0.0012, + "step": 751 + }, + { + "epoch": 1.125748502994012, + "grad_norm": 0.08108549950525631, + "learning_rate": 1.439597918005215e-05, + "loss": 0.0006, + "step": 752 + }, + { + "epoch": 1.127245508982036, + "grad_norm": 0.2152317708482709, + "learning_rate": 1.4381450734141505e-05, + "loss": 0.002, + "step": 753 + }, + { + "epoch": 1.1287425149700598, + "grad_norm": 0.08773076511235232, + "learning_rate": 1.4366910833840431e-05, + "loss": 0.0021, + "step": 754 + }, + { + "epoch": 1.1302395209580838, + "grad_norm": 0.17027037331891098, + "learning_rate": 1.4352359517160462e-05, + "loss": 0.0013, + "step": 755 + }, + { + "epoch": 1.1317365269461077, + "grad_norm": 0.08918657681225792, + "learning_rate": 1.4337796822142987e-05, + "loss": 0.0015, + "step": 756 + }, + { + "epoch": 1.1332335329341316, + "grad_norm": 0.1305105215164226, + "learning_rate": 1.4323222786859141e-05, + "loss": 0.0011, + "step": 757 + }, + { + "epoch": 1.1347305389221556, + "grad_norm": 0.12593434714214163, + "learning_rate": 1.4308637449409705e-05, + "loss": 0.0011, + "step": 758 + }, + { + "epoch": 1.1362275449101795, + "grad_norm": 0.3028857079800561, + "learning_rate": 1.4294040847925004e-05, + "loss": 0.0035, + "step": 759 + }, + { + "epoch": 1.1377245508982037, + "grad_norm": 0.24334283105322427, + "learning_rate": 1.4279433020564813e-05, + "loss": 0.002, + "step": 760 + }, + { + "epoch": 1.1392215568862276, + "grad_norm": 0.18593254416098137, + "learning_rate": 1.4264814005518251e-05, + "loss": 0.0033, + "step": 761 + }, + { + "epoch": 1.1407185628742516, + "grad_norm": 0.158884487983083, + "learning_rate": 1.4250183841003695e-05, + "loss": 0.0013, + "step": 762 + }, + { + "epoch": 1.1422155688622755, + "grad_norm": 0.14632297769339114, + "learning_rate": 1.4235542565268658e-05, + "loss": 0.0015, + "step": 763 + }, + { + "epoch": 1.1437125748502994, + "grad_norm": 0.15199709628186464, + "learning_rate": 1.4220890216589708e-05, + "loss": 0.0011, + "step": 764 + }, + { + "epoch": 1.1452095808383234, + "grad_norm": 0.07701639398769143, + "learning_rate": 1.4206226833272359e-05, + "loss": 0.0011, + "step": 765 + }, + { + "epoch": 1.1467065868263473, + "grad_norm": 0.06624545307291502, + "learning_rate": 1.4191552453650972e-05, + "loss": 0.0007, + "step": 766 + }, + { + "epoch": 1.1482035928143712, + "grad_norm": 0.18898462608830638, + "learning_rate": 1.4176867116088655e-05, + "loss": 0.0017, + "step": 767 + }, + { + "epoch": 1.1497005988023952, + "grad_norm": 0.11325039624754406, + "learning_rate": 1.4162170858977167e-05, + "loss": 0.0013, + "step": 768 + }, + { + "epoch": 1.151197604790419, + "grad_norm": 0.1260621674366222, + "learning_rate": 1.4147463720736813e-05, + "loss": 0.0023, + "step": 769 + }, + { + "epoch": 1.152694610778443, + "grad_norm": 0.10527624531422942, + "learning_rate": 1.4132745739816341e-05, + "loss": 0.0008, + "step": 770 + }, + { + "epoch": 1.154191616766467, + "grad_norm": 0.22598741769102923, + "learning_rate": 1.4118016954692845e-05, + "loss": 0.0022, + "step": 771 + }, + { + "epoch": 1.1556886227544911, + "grad_norm": 0.11216311604503693, + "learning_rate": 1.4103277403871667e-05, + "loss": 0.0021, + "step": 772 + }, + { + "epoch": 1.157185628742515, + "grad_norm": 0.12672041749180385, + "learning_rate": 1.4088527125886294e-05, + "loss": 0.0021, + "step": 773 + }, + { + "epoch": 1.158682634730539, + "grad_norm": 0.05731251869782339, + "learning_rate": 1.4073766159298257e-05, + "loss": 0.0006, + "step": 774 + }, + { + "epoch": 1.160179640718563, + "grad_norm": 0.12190886299689159, + "learning_rate": 1.4058994542697027e-05, + "loss": 0.0013, + "step": 775 + }, + { + "epoch": 1.1616766467065869, + "grad_norm": 0.16571654034581215, + "learning_rate": 1.4044212314699919e-05, + "loss": 0.0021, + "step": 776 + }, + { + "epoch": 1.1631736526946108, + "grad_norm": 0.12852042640903608, + "learning_rate": 1.402941951395199e-05, + "loss": 0.0017, + "step": 777 + }, + { + "epoch": 1.1646706586826348, + "grad_norm": 0.04996364604118305, + "learning_rate": 1.4014616179125935e-05, + "loss": 0.0004, + "step": 778 + }, + { + "epoch": 1.1661676646706587, + "grad_norm": 0.22739207612882284, + "learning_rate": 1.3999802348921993e-05, + "loss": 0.0036, + "step": 779 + }, + { + "epoch": 1.1676646706586826, + "grad_norm": 0.10329869929004334, + "learning_rate": 1.3984978062067836e-05, + "loss": 0.0008, + "step": 780 + }, + { + "epoch": 1.1691616766467066, + "grad_norm": 0.16774030611607113, + "learning_rate": 1.3970143357318475e-05, + "loss": 0.0025, + "step": 781 + }, + { + "epoch": 1.1706586826347305, + "grad_norm": 0.10032247067801155, + "learning_rate": 1.3955298273456157e-05, + "loss": 0.0008, + "step": 782 + }, + { + "epoch": 1.1721556886227544, + "grad_norm": 0.12720820030778054, + "learning_rate": 1.3940442849290259e-05, + "loss": 0.0028, + "step": 783 + }, + { + "epoch": 1.1736526946107784, + "grad_norm": 0.07501276893929008, + "learning_rate": 1.3925577123657196e-05, + "loss": 0.0013, + "step": 784 + }, + { + "epoch": 1.1751497005988023, + "grad_norm": 0.13032678151078575, + "learning_rate": 1.3910701135420312e-05, + "loss": 0.0014, + "step": 785 + }, + { + "epoch": 1.1766467065868262, + "grad_norm": 0.1189117861891301, + "learning_rate": 1.3895814923469781e-05, + "loss": 0.003, + "step": 786 + }, + { + "epoch": 1.1781437125748502, + "grad_norm": 0.04852723768916087, + "learning_rate": 1.3880918526722497e-05, + "loss": 0.0006, + "step": 787 + }, + { + "epoch": 1.1796407185628743, + "grad_norm": 0.14044921780427475, + "learning_rate": 1.3866011984121997e-05, + "loss": 0.0016, + "step": 788 + }, + { + "epoch": 1.1811377245508983, + "grad_norm": 0.10503085747024221, + "learning_rate": 1.3851095334638323e-05, + "loss": 0.0013, + "step": 789 + }, + { + "epoch": 1.1826347305389222, + "grad_norm": 0.07687725843746662, + "learning_rate": 1.3836168617267956e-05, + "loss": 0.0007, + "step": 790 + }, + { + "epoch": 1.1841317365269461, + "grad_norm": 0.13288540054060996, + "learning_rate": 1.3821231871033684e-05, + "loss": 0.0019, + "step": 791 + }, + { + "epoch": 1.18562874251497, + "grad_norm": 0.10408705042608069, + "learning_rate": 1.3806285134984525e-05, + "loss": 0.0015, + "step": 792 + }, + { + "epoch": 1.187125748502994, + "grad_norm": 0.09545112906104483, + "learning_rate": 1.37913284481956e-05, + "loss": 0.0013, + "step": 793 + }, + { + "epoch": 1.188622754491018, + "grad_norm": 0.19426107849435, + "learning_rate": 1.3776361849768056e-05, + "loss": 0.002, + "step": 794 + }, + { + "epoch": 1.1901197604790419, + "grad_norm": 0.1508486355195202, + "learning_rate": 1.3761385378828946e-05, + "loss": 0.0016, + "step": 795 + }, + { + "epoch": 1.1916167664670658, + "grad_norm": 0.11380042520519763, + "learning_rate": 1.3746399074531139e-05, + "loss": 0.0012, + "step": 796 + }, + { + "epoch": 1.1931137724550898, + "grad_norm": 0.06272949618809354, + "learning_rate": 1.3731402976053203e-05, + "loss": 0.0009, + "step": 797 + }, + { + "epoch": 1.1946107784431137, + "grad_norm": 0.04908194402100617, + "learning_rate": 1.3716397122599311e-05, + "loss": 0.0004, + "step": 798 + }, + { + "epoch": 1.1961077844311376, + "grad_norm": 0.04835379010726109, + "learning_rate": 1.3701381553399147e-05, + "loss": 0.0005, + "step": 799 + }, + { + "epoch": 1.1976047904191618, + "grad_norm": 0.07282438701791324, + "learning_rate": 1.3686356307707784e-05, + "loss": 0.0011, + "step": 800 + }, + { + "epoch": 1.1991017964071857, + "grad_norm": 0.07359224976031348, + "learning_rate": 1.3671321424805602e-05, + "loss": 0.0008, + "step": 801 + }, + { + "epoch": 1.2005988023952097, + "grad_norm": 0.0728588455334413, + "learning_rate": 1.365627694399817e-05, + "loss": 0.0011, + "step": 802 + }, + { + "epoch": 1.2020958083832336, + "grad_norm": 0.08572484612904316, + "learning_rate": 1.3641222904616149e-05, + "loss": 0.0019, + "step": 803 + }, + { + "epoch": 1.2035928143712575, + "grad_norm": 0.10340010986651034, + "learning_rate": 1.3626159346015193e-05, + "loss": 0.0009, + "step": 804 + }, + { + "epoch": 1.2050898203592815, + "grad_norm": 0.07243035348418118, + "learning_rate": 1.3611086307575831e-05, + "loss": 0.0008, + "step": 805 + }, + { + "epoch": 1.2065868263473054, + "grad_norm": 0.1334209046458507, + "learning_rate": 1.3596003828703393e-05, + "loss": 0.0019, + "step": 806 + }, + { + "epoch": 1.2080838323353293, + "grad_norm": 0.0570772989713065, + "learning_rate": 1.3580911948827868e-05, + "loss": 0.001, + "step": 807 + }, + { + "epoch": 1.2095808383233533, + "grad_norm": 0.1204536054401414, + "learning_rate": 1.3565810707403844e-05, + "loss": 0.0009, + "step": 808 + }, + { + "epoch": 1.2110778443113772, + "grad_norm": 0.1867883260775183, + "learning_rate": 1.3550700143910365e-05, + "loss": 0.0025, + "step": 809 + }, + { + "epoch": 1.2125748502994012, + "grad_norm": 0.07727972942708328, + "learning_rate": 1.3535580297850852e-05, + "loss": 0.0007, + "step": 810 + }, + { + "epoch": 1.214071856287425, + "grad_norm": 0.09466637885042935, + "learning_rate": 1.3520451208752992e-05, + "loss": 0.0005, + "step": 811 + }, + { + "epoch": 1.215568862275449, + "grad_norm": 0.1580780788542899, + "learning_rate": 1.3505312916168644e-05, + "loss": 0.001, + "step": 812 + }, + { + "epoch": 1.217065868263473, + "grad_norm": 0.10526784028411676, + "learning_rate": 1.3490165459673716e-05, + "loss": 0.001, + "step": 813 + }, + { + "epoch": 1.218562874251497, + "grad_norm": 0.12653380982480567, + "learning_rate": 1.3475008878868078e-05, + "loss": 0.0013, + "step": 814 + }, + { + "epoch": 1.220059880239521, + "grad_norm": 0.069894431976694, + "learning_rate": 1.345984321337546e-05, + "loss": 0.0014, + "step": 815 + }, + { + "epoch": 1.221556886227545, + "grad_norm": 0.11138606082666604, + "learning_rate": 1.344466850284333e-05, + "loss": 0.001, + "step": 816 + }, + { + "epoch": 1.223053892215569, + "grad_norm": 0.09549692513609892, + "learning_rate": 1.342948478694281e-05, + "loss": 0.0012, + "step": 817 + }, + { + "epoch": 1.2245508982035929, + "grad_norm": 0.08071969012117705, + "learning_rate": 1.3414292105368562e-05, + "loss": 0.0011, + "step": 818 + }, + { + "epoch": 1.2260479041916168, + "grad_norm": 0.014969590068211934, + "learning_rate": 1.339909049783869e-05, + "loss": 0.0002, + "step": 819 + }, + { + "epoch": 1.2275449101796407, + "grad_norm": 0.07241700618747784, + "learning_rate": 1.338388000409463e-05, + "loss": 0.0005, + "step": 820 + }, + { + "epoch": 1.2290419161676647, + "grad_norm": 0.11075869601766639, + "learning_rate": 1.3368660663901045e-05, + "loss": 0.0023, + "step": 821 + }, + { + "epoch": 1.2305389221556886, + "grad_norm": 0.17060039325631357, + "learning_rate": 1.3353432517045739e-05, + "loss": 0.0016, + "step": 822 + }, + { + "epoch": 1.2320359281437125, + "grad_norm": 0.1022635862904923, + "learning_rate": 1.3338195603339525e-05, + "loss": 0.0011, + "step": 823 + }, + { + "epoch": 1.2335329341317365, + "grad_norm": 0.09158836544730396, + "learning_rate": 1.3322949962616142e-05, + "loss": 0.0006, + "step": 824 + }, + { + "epoch": 1.2350299401197604, + "grad_norm": 0.048832084216327266, + "learning_rate": 1.3307695634732144e-05, + "loss": 0.0005, + "step": 825 + }, + { + "epoch": 1.2365269461077844, + "grad_norm": 0.08830232598087197, + "learning_rate": 1.3292432659566791e-05, + "loss": 0.0007, + "step": 826 + }, + { + "epoch": 1.2380239520958083, + "grad_norm": 0.052016398871730235, + "learning_rate": 1.3277161077021957e-05, + "loss": 0.0005, + "step": 827 + }, + { + "epoch": 1.2395209580838324, + "grad_norm": 0.12716710033758089, + "learning_rate": 1.3261880927022008e-05, + "loss": 0.001, + "step": 828 + }, + { + "epoch": 1.2410179640718564, + "grad_norm": 0.08239866964993203, + "learning_rate": 1.3246592249513717e-05, + "loss": 0.0008, + "step": 829 + }, + { + "epoch": 1.2425149700598803, + "grad_norm": 0.07577545667183459, + "learning_rate": 1.3231295084466149e-05, + "loss": 0.0008, + "step": 830 + }, + { + "epoch": 1.2440119760479043, + "grad_norm": 0.08810579822674741, + "learning_rate": 1.3215989471870556e-05, + "loss": 0.0009, + "step": 831 + }, + { + "epoch": 1.2455089820359282, + "grad_norm": 0.04066723916071258, + "learning_rate": 1.3200675451740273e-05, + "loss": 0.0006, + "step": 832 + }, + { + "epoch": 1.2470059880239521, + "grad_norm": 0.14239069700954057, + "learning_rate": 1.3185353064110613e-05, + "loss": 0.0022, + "step": 833 + }, + { + "epoch": 1.248502994011976, + "grad_norm": 0.09145370992887267, + "learning_rate": 1.3170022349038774e-05, + "loss": 0.0011, + "step": 834 + }, + { + "epoch": 1.25, + "grad_norm": 0.09278110626604044, + "learning_rate": 1.3154683346603712e-05, + "loss": 0.001, + "step": 835 + }, + { + "epoch": 1.251497005988024, + "grad_norm": 0.04046285571674997, + "learning_rate": 1.3139336096906056e-05, + "loss": 0.0003, + "step": 836 + }, + { + "epoch": 1.2529940119760479, + "grad_norm": 0.1350937599331085, + "learning_rate": 1.3123980640067999e-05, + "loss": 0.0018, + "step": 837 + }, + { + "epoch": 1.2544910179640718, + "grad_norm": 0.05547862227838062, + "learning_rate": 1.3108617016233178e-05, + "loss": 0.0005, + "step": 838 + }, + { + "epoch": 1.2559880239520957, + "grad_norm": 0.13832547458088876, + "learning_rate": 1.3093245265566588e-05, + "loss": 0.0021, + "step": 839 + }, + { + "epoch": 1.2574850299401197, + "grad_norm": 0.3205689356680666, + "learning_rate": 1.3077865428254474e-05, + "loss": 0.0017, + "step": 840 + }, + { + "epoch": 1.2589820359281436, + "grad_norm": 0.17333145605990935, + "learning_rate": 1.3062477544504218e-05, + "loss": 0.0028, + "step": 841 + }, + { + "epoch": 1.2604790419161676, + "grad_norm": 0.10905825753636433, + "learning_rate": 1.304708165454423e-05, + "loss": 0.001, + "step": 842 + }, + { + "epoch": 1.2619760479041915, + "grad_norm": 0.06230478352428612, + "learning_rate": 1.3031677798623865e-05, + "loss": 0.0005, + "step": 843 + }, + { + "epoch": 1.2634730538922156, + "grad_norm": 0.15076033232514022, + "learning_rate": 1.3016266017013295e-05, + "loss": 0.0015, + "step": 844 + }, + { + "epoch": 1.2649700598802396, + "grad_norm": 0.10330789403132279, + "learning_rate": 1.300084635000341e-05, + "loss": 0.0013, + "step": 845 + }, + { + "epoch": 1.2664670658682635, + "grad_norm": 0.0967360202578523, + "learning_rate": 1.2985418837905719e-05, + "loss": 0.0009, + "step": 846 + }, + { + "epoch": 1.2679640718562875, + "grad_norm": 0.11018485595492865, + "learning_rate": 1.2969983521052243e-05, + "loss": 0.0015, + "step": 847 + }, + { + "epoch": 1.2694610778443114, + "grad_norm": 0.2026144990686223, + "learning_rate": 1.29545404397954e-05, + "loss": 0.0016, + "step": 848 + }, + { + "epoch": 1.2709580838323353, + "grad_norm": 0.11825794065351451, + "learning_rate": 1.2939089634507908e-05, + "loss": 0.0009, + "step": 849 + }, + { + "epoch": 1.2724550898203593, + "grad_norm": 0.1706730641796369, + "learning_rate": 1.2923631145582684e-05, + "loss": 0.0026, + "step": 850 + }, + { + "epoch": 1.2739520958083832, + "grad_norm": 0.057699428000987554, + "learning_rate": 1.2908165013432723e-05, + "loss": 0.0008, + "step": 851 + }, + { + "epoch": 1.2754491017964071, + "grad_norm": 0.142393242455963, + "learning_rate": 1.2892691278491012e-05, + "loss": 0.0017, + "step": 852 + }, + { + "epoch": 1.276946107784431, + "grad_norm": 0.055031594469559006, + "learning_rate": 1.2877209981210405e-05, + "loss": 0.0004, + "step": 853 + }, + { + "epoch": 1.278443113772455, + "grad_norm": 0.0849339018863119, + "learning_rate": 1.2861721162063532e-05, + "loss": 0.0004, + "step": 854 + }, + { + "epoch": 1.2799401197604792, + "grad_norm": 0.07827653949622694, + "learning_rate": 1.2846224861542677e-05, + "loss": 0.0007, + "step": 855 + }, + { + "epoch": 1.281437125748503, + "grad_norm": 0.024359597454629765, + "learning_rate": 1.2830721120159697e-05, + "loss": 0.0003, + "step": 856 + }, + { + "epoch": 1.282934131736527, + "grad_norm": 0.21035448274091229, + "learning_rate": 1.2815209978445898e-05, + "loss": 0.0011, + "step": 857 + }, + { + "epoch": 1.284431137724551, + "grad_norm": 0.07397518183430371, + "learning_rate": 1.2799691476951926e-05, + "loss": 0.0007, + "step": 858 + }, + { + "epoch": 1.285928143712575, + "grad_norm": 0.13982310084632887, + "learning_rate": 1.2784165656247668e-05, + "loss": 0.001, + "step": 859 + }, + { + "epoch": 1.2874251497005988, + "grad_norm": 0.029670153139229417, + "learning_rate": 1.2768632556922151e-05, + "loss": 0.0004, + "step": 860 + }, + { + "epoch": 1.2889221556886228, + "grad_norm": 0.08714604279500297, + "learning_rate": 1.275309221958343e-05, + "loss": 0.0012, + "step": 861 + }, + { + "epoch": 1.2904191616766467, + "grad_norm": 0.02766702455992605, + "learning_rate": 1.2737544684858476e-05, + "loss": 0.0003, + "step": 862 + }, + { + "epoch": 1.2919161676646707, + "grad_norm": 0.15764398213987413, + "learning_rate": 1.2721989993393088e-05, + "loss": 0.0021, + "step": 863 + }, + { + "epoch": 1.2934131736526946, + "grad_norm": 0.3620928680238313, + "learning_rate": 1.2706428185851759e-05, + "loss": 0.0031, + "step": 864 + }, + { + "epoch": 1.2949101796407185, + "grad_norm": 0.15033902468169197, + "learning_rate": 1.26908593029176e-05, + "loss": 0.0017, + "step": 865 + }, + { + "epoch": 1.2964071856287425, + "grad_norm": 0.08093325547150344, + "learning_rate": 1.2675283385292212e-05, + "loss": 0.0006, + "step": 866 + }, + { + "epoch": 1.2979041916167664, + "grad_norm": 0.19391865613837045, + "learning_rate": 1.2659700473695581e-05, + "loss": 0.0026, + "step": 867 + }, + { + "epoch": 1.2994011976047903, + "grad_norm": 0.13771485536341505, + "learning_rate": 1.2644110608865994e-05, + "loss": 0.002, + "step": 868 + }, + { + "epoch": 1.3008982035928143, + "grad_norm": 0.082036361858331, + "learning_rate": 1.2628513831559897e-05, + "loss": 0.0012, + "step": 869 + }, + { + "epoch": 1.3023952095808382, + "grad_norm": 0.13557425572078727, + "learning_rate": 1.2612910182551822e-05, + "loss": 0.0009, + "step": 870 + }, + { + "epoch": 1.3038922155688621, + "grad_norm": 0.1248956204266484, + "learning_rate": 1.259729970263426e-05, + "loss": 0.0012, + "step": 871 + }, + { + "epoch": 1.3053892215568863, + "grad_norm": 0.15323443985781152, + "learning_rate": 1.2581682432617554e-05, + "loss": 0.0021, + "step": 872 + }, + { + "epoch": 1.3068862275449102, + "grad_norm": 0.16495520023315455, + "learning_rate": 1.2566058413329806e-05, + "loss": 0.0011, + "step": 873 + }, + { + "epoch": 1.3083832335329342, + "grad_norm": 0.12612540137072323, + "learning_rate": 1.2550427685616767e-05, + "loss": 0.0019, + "step": 874 + }, + { + "epoch": 1.3098802395209581, + "grad_norm": 0.07140252385386611, + "learning_rate": 1.253479029034171e-05, + "loss": 0.0009, + "step": 875 + }, + { + "epoch": 1.311377245508982, + "grad_norm": 0.07997999603836409, + "learning_rate": 1.2519146268385352e-05, + "loss": 0.0012, + "step": 876 + }, + { + "epoch": 1.312874251497006, + "grad_norm": 0.05031120608143949, + "learning_rate": 1.250349566064573e-05, + "loss": 0.0007, + "step": 877 + }, + { + "epoch": 1.31437125748503, + "grad_norm": 0.10409482982213064, + "learning_rate": 1.2487838508038097e-05, + "loss": 0.0018, + "step": 878 + }, + { + "epoch": 1.3158682634730539, + "grad_norm": 0.06629243450722555, + "learning_rate": 1.2472174851494823e-05, + "loss": 0.0005, + "step": 879 + }, + { + "epoch": 1.3173652694610778, + "grad_norm": 0.11971936347458219, + "learning_rate": 1.2456504731965269e-05, + "loss": 0.0017, + "step": 880 + }, + { + "epoch": 1.3188622754491017, + "grad_norm": 0.20319873175607275, + "learning_rate": 1.2440828190415706e-05, + "loss": 0.0018, + "step": 881 + }, + { + "epoch": 1.3203592814371259, + "grad_norm": 0.04682707557262463, + "learning_rate": 1.2425145267829179e-05, + "loss": 0.0005, + "step": 882 + }, + { + "epoch": 1.3218562874251498, + "grad_norm": 0.3848522460027514, + "learning_rate": 1.2409456005205427e-05, + "loss": 0.0017, + "step": 883 + }, + { + "epoch": 1.3233532934131738, + "grad_norm": 0.20512103175801702, + "learning_rate": 1.239376044356076e-05, + "loss": 0.0024, + "step": 884 + }, + { + "epoch": 1.3248502994011977, + "grad_norm": 0.08312153889808901, + "learning_rate": 1.2378058623927955e-05, + "loss": 0.001, + "step": 885 + }, + { + "epoch": 1.3263473053892216, + "grad_norm": 0.05507532413988543, + "learning_rate": 1.236235058735615e-05, + "loss": 0.0004, + "step": 886 + }, + { + "epoch": 1.3278443113772456, + "grad_norm": 0.08688774454813779, + "learning_rate": 1.2346636374910735e-05, + "loss": 0.0008, + "step": 887 + }, + { + "epoch": 1.3293413173652695, + "grad_norm": 0.18863114337383052, + "learning_rate": 1.233091602767324e-05, + "loss": 0.0014, + "step": 888 + }, + { + "epoch": 1.3308383233532934, + "grad_norm": 0.02191143112881276, + "learning_rate": 1.2315189586741246e-05, + "loss": 0.0002, + "step": 889 + }, + { + "epoch": 1.3323353293413174, + "grad_norm": 0.08898221164791097, + "learning_rate": 1.2299457093228253e-05, + "loss": 0.0006, + "step": 890 + }, + { + "epoch": 1.3338323353293413, + "grad_norm": 0.038820165742322596, + "learning_rate": 1.2283718588263594e-05, + "loss": 0.0004, + "step": 891 + }, + { + "epoch": 1.3353293413173652, + "grad_norm": 0.07647935504218839, + "learning_rate": 1.2267974112992307e-05, + "loss": 0.0011, + "step": 892 + }, + { + "epoch": 1.3368263473053892, + "grad_norm": 0.13982650167833818, + "learning_rate": 1.2252223708575047e-05, + "loss": 0.002, + "step": 893 + }, + { + "epoch": 1.3383233532934131, + "grad_norm": 0.1338195205882971, + "learning_rate": 1.2236467416187965e-05, + "loss": 0.0025, + "step": 894 + }, + { + "epoch": 1.339820359281437, + "grad_norm": 0.1307841640751261, + "learning_rate": 1.2220705277022602e-05, + "loss": 0.0016, + "step": 895 + }, + { + "epoch": 1.341317365269461, + "grad_norm": 0.16651571491950992, + "learning_rate": 1.2204937332285796e-05, + "loss": 0.0024, + "step": 896 + }, + { + "epoch": 1.342814371257485, + "grad_norm": 0.054613620669397775, + "learning_rate": 1.2189163623199547e-05, + "loss": 0.0006, + "step": 897 + }, + { + "epoch": 1.3443113772455089, + "grad_norm": 0.6200338634281016, + "learning_rate": 1.2173384191000936e-05, + "loss": 0.0027, + "step": 898 + }, + { + "epoch": 1.3458083832335328, + "grad_norm": 0.10417162543361413, + "learning_rate": 1.2157599076942004e-05, + "loss": 0.0006, + "step": 899 + }, + { + "epoch": 1.347305389221557, + "grad_norm": 0.28679445691083094, + "learning_rate": 1.214180832228964e-05, + "loss": 0.0018, + "step": 900 + }, + { + "epoch": 1.348802395209581, + "grad_norm": 0.13667869860487905, + "learning_rate": 1.2126011968325486e-05, + "loss": 0.001, + "step": 901 + }, + { + "epoch": 1.3502994011976048, + "grad_norm": 0.14678301118335677, + "learning_rate": 1.211021005634582e-05, + "loss": 0.0014, + "step": 902 + }, + { + "epoch": 1.3517964071856288, + "grad_norm": 0.03643518165194748, + "learning_rate": 1.2094402627661447e-05, + "loss": 0.0004, + "step": 903 + }, + { + "epoch": 1.3532934131736527, + "grad_norm": 0.21051811393884823, + "learning_rate": 1.2078589723597599e-05, + "loss": 0.0019, + "step": 904 + }, + { + "epoch": 1.3547904191616766, + "grad_norm": 0.4465499366437971, + "learning_rate": 1.2062771385493818e-05, + "loss": 0.0019, + "step": 905 + }, + { + "epoch": 1.3562874251497006, + "grad_norm": 0.3280121687211394, + "learning_rate": 1.204694765470386e-05, + "loss": 0.0024, + "step": 906 + }, + { + "epoch": 1.3577844311377245, + "grad_norm": 0.19136289912588866, + "learning_rate": 1.2031118572595568e-05, + "loss": 0.0012, + "step": 907 + }, + { + "epoch": 1.3592814371257484, + "grad_norm": 0.3123777529319237, + "learning_rate": 1.2015284180550786e-05, + "loss": 0.0028, + "step": 908 + }, + { + "epoch": 1.3607784431137724, + "grad_norm": 0.14613074780473378, + "learning_rate": 1.1999444519965228e-05, + "loss": 0.0006, + "step": 909 + }, + { + "epoch": 1.3622754491017965, + "grad_norm": 0.08510506250722642, + "learning_rate": 1.1983599632248393e-05, + "loss": 0.0011, + "step": 910 + }, + { + "epoch": 1.3637724550898205, + "grad_norm": 0.24109722460040886, + "learning_rate": 1.1967749558823436e-05, + "loss": 0.0017, + "step": 911 + }, + { + "epoch": 1.3652694610778444, + "grad_norm": 0.06997668553987517, + "learning_rate": 1.1951894341127075e-05, + "loss": 0.0008, + "step": 912 + }, + { + "epoch": 1.3667664670658684, + "grad_norm": 0.08339177784324386, + "learning_rate": 1.1936034020609476e-05, + "loss": 0.0013, + "step": 913 + }, + { + "epoch": 1.3682634730538923, + "grad_norm": 0.37011844961784346, + "learning_rate": 1.192016863873414e-05, + "loss": 0.0031, + "step": 914 + }, + { + "epoch": 1.3697604790419162, + "grad_norm": 0.19448555543500273, + "learning_rate": 1.1904298236977808e-05, + "loss": 0.0006, + "step": 915 + }, + { + "epoch": 1.3712574850299402, + "grad_norm": 0.37638195330236135, + "learning_rate": 1.1888422856830335e-05, + "loss": 0.002, + "step": 916 + }, + { + "epoch": 1.372754491017964, + "grad_norm": 0.04845927126998666, + "learning_rate": 1.18725425397946e-05, + "loss": 0.0006, + "step": 917 + }, + { + "epoch": 1.374251497005988, + "grad_norm": 0.050269398324577504, + "learning_rate": 1.1856657327386381e-05, + "loss": 0.0003, + "step": 918 + }, + { + "epoch": 1.375748502994012, + "grad_norm": 0.16322026432621767, + "learning_rate": 1.1840767261134262e-05, + "loss": 0.0017, + "step": 919 + }, + { + "epoch": 1.377245508982036, + "grad_norm": 0.1990865675756091, + "learning_rate": 1.1824872382579511e-05, + "loss": 0.004, + "step": 920 + }, + { + "epoch": 1.3787425149700598, + "grad_norm": 0.07465156559192355, + "learning_rate": 1.1808972733275977e-05, + "loss": 0.0008, + "step": 921 + }, + { + "epoch": 1.3802395209580838, + "grad_norm": 0.09208435898453236, + "learning_rate": 1.1793068354789978e-05, + "loss": 0.0013, + "step": 922 + }, + { + "epoch": 1.3817365269461077, + "grad_norm": 0.12422665350230562, + "learning_rate": 1.1777159288700203e-05, + "loss": 0.0013, + "step": 923 + }, + { + "epoch": 1.3832335329341316, + "grad_norm": 0.2436082341006343, + "learning_rate": 1.176124557659759e-05, + "loss": 0.0017, + "step": 924 + }, + { + "epoch": 1.3847305389221556, + "grad_norm": 0.12781022014989502, + "learning_rate": 1.174532726008523e-05, + "loss": 0.0018, + "step": 925 + }, + { + "epoch": 1.3862275449101795, + "grad_norm": 0.16812228483761696, + "learning_rate": 1.1729404380778241e-05, + "loss": 0.0026, + "step": 926 + }, + { + "epoch": 1.3877245508982035, + "grad_norm": 0.0920216290898727, + "learning_rate": 1.1713476980303678e-05, + "loss": 0.0012, + "step": 927 + }, + { + "epoch": 1.3892215568862276, + "grad_norm": 0.11482491332593553, + "learning_rate": 1.1697545100300408e-05, + "loss": 0.002, + "step": 928 + }, + { + "epoch": 1.3907185628742516, + "grad_norm": 0.10002074873744907, + "learning_rate": 1.1681608782419014e-05, + "loss": 0.001, + "step": 929 + }, + { + "epoch": 1.3922155688622755, + "grad_norm": 0.08746209991474958, + "learning_rate": 1.1665668068321684e-05, + "loss": 0.0009, + "step": 930 + }, + { + "epoch": 1.3937125748502994, + "grad_norm": 0.0814270207122144, + "learning_rate": 1.1649722999682094e-05, + "loss": 0.0012, + "step": 931 + }, + { + "epoch": 1.3952095808383234, + "grad_norm": 0.09416853678436857, + "learning_rate": 1.1633773618185302e-05, + "loss": 0.001, + "step": 932 + }, + { + "epoch": 1.3967065868263473, + "grad_norm": 0.14964696805599081, + "learning_rate": 1.161781996552765e-05, + "loss": 0.0012, + "step": 933 + }, + { + "epoch": 1.3982035928143712, + "grad_norm": 0.09483767589434133, + "learning_rate": 1.1601862083416632e-05, + "loss": 0.0013, + "step": 934 + }, + { + "epoch": 1.3997005988023952, + "grad_norm": 0.09788376172150742, + "learning_rate": 1.1585900013570815e-05, + "loss": 0.0015, + "step": 935 + }, + { + "epoch": 1.401197604790419, + "grad_norm": 0.10836219152191484, + "learning_rate": 1.1569933797719703e-05, + "loss": 0.0016, + "step": 936 + }, + { + "epoch": 1.402694610778443, + "grad_norm": 0.23674230633425863, + "learning_rate": 1.155396347760364e-05, + "loss": 0.0016, + "step": 937 + }, + { + "epoch": 1.4041916167664672, + "grad_norm": 0.08196393324885805, + "learning_rate": 1.1537989094973705e-05, + "loss": 0.0014, + "step": 938 + }, + { + "epoch": 1.4056886227544911, + "grad_norm": 0.05782826404362468, + "learning_rate": 1.1522010691591595e-05, + "loss": 0.0007, + "step": 939 + }, + { + "epoch": 1.407185628742515, + "grad_norm": 0.14623254375235795, + "learning_rate": 1.1506028309229513e-05, + "loss": 0.0019, + "step": 940 + }, + { + "epoch": 1.408682634730539, + "grad_norm": 0.10434558213727799, + "learning_rate": 1.1490041989670076e-05, + "loss": 0.0012, + "step": 941 + }, + { + "epoch": 1.410179640718563, + "grad_norm": 0.15631392324819193, + "learning_rate": 1.1474051774706181e-05, + "loss": 0.0013, + "step": 942 + }, + { + "epoch": 1.4116766467065869, + "grad_norm": 0.13437189533675623, + "learning_rate": 1.145805770614092e-05, + "loss": 0.0022, + "step": 943 + }, + { + "epoch": 1.4131736526946108, + "grad_norm": 0.038973494041355694, + "learning_rate": 1.1442059825787448e-05, + "loss": 0.0004, + "step": 944 + }, + { + "epoch": 1.4146706586826348, + "grad_norm": 0.07141700721890615, + "learning_rate": 1.1426058175468895e-05, + "loss": 0.0011, + "step": 945 + }, + { + "epoch": 1.4161676646706587, + "grad_norm": 0.21493859310812677, + "learning_rate": 1.1410052797018245e-05, + "loss": 0.0017, + "step": 946 + }, + { + "epoch": 1.4176646706586826, + "grad_norm": 0.1282522648959389, + "learning_rate": 1.1394043732278225e-05, + "loss": 0.001, + "step": 947 + }, + { + "epoch": 1.4191616766467066, + "grad_norm": 0.07019102461290384, + "learning_rate": 1.1378031023101199e-05, + "loss": 0.0008, + "step": 948 + }, + { + "epoch": 1.4206586826347305, + "grad_norm": 0.05464505705186005, + "learning_rate": 1.1362014711349063e-05, + "loss": 0.0004, + "step": 949 + }, + { + "epoch": 1.4221556886227544, + "grad_norm": 0.07090800295191836, + "learning_rate": 1.1345994838893125e-05, + "loss": 0.0004, + "step": 950 + }, + { + "epoch": 1.4236526946107784, + "grad_norm": 0.1475910596450026, + "learning_rate": 1.1329971447614008e-05, + "loss": 0.0014, + "step": 951 + }, + { + "epoch": 1.4251497005988023, + "grad_norm": 0.11628540378666755, + "learning_rate": 1.1313944579401528e-05, + "loss": 0.0008, + "step": 952 + }, + { + "epoch": 1.4266467065868262, + "grad_norm": 0.11918182166269133, + "learning_rate": 1.12979142761546e-05, + "loss": 0.0013, + "step": 953 + }, + { + "epoch": 1.4281437125748502, + "grad_norm": 0.06010179192377916, + "learning_rate": 1.1281880579781108e-05, + "loss": 0.0008, + "step": 954 + }, + { + "epoch": 1.4296407185628741, + "grad_norm": 0.07923386193741429, + "learning_rate": 1.1265843532197813e-05, + "loss": 0.001, + "step": 955 + }, + { + "epoch": 1.4311377245508983, + "grad_norm": 0.09159193558417823, + "learning_rate": 1.1249803175330237e-05, + "loss": 0.0015, + "step": 956 + }, + { + "epoch": 1.4326347305389222, + "grad_norm": 0.07308876635837779, + "learning_rate": 1.1233759551112553e-05, + "loss": 0.0004, + "step": 957 + }, + { + "epoch": 1.4341317365269461, + "grad_norm": 0.0779800871854366, + "learning_rate": 1.1217712701487471e-05, + "loss": 0.0008, + "step": 958 + }, + { + "epoch": 1.43562874251497, + "grad_norm": 0.06632363235084476, + "learning_rate": 1.120166266840614e-05, + "loss": 0.0005, + "step": 959 + }, + { + "epoch": 1.437125748502994, + "grad_norm": 0.05842006960591898, + "learning_rate": 1.1185609493828033e-05, + "loss": 0.0007, + "step": 960 + }, + { + "epoch": 1.438622754491018, + "grad_norm": 0.19344221257223246, + "learning_rate": 1.1169553219720828e-05, + "loss": 0.0031, + "step": 961 + }, + { + "epoch": 1.4401197604790419, + "grad_norm": 0.1981812238766314, + "learning_rate": 1.1153493888060304e-05, + "loss": 0.0013, + "step": 962 + }, + { + "epoch": 1.4416167664670658, + "grad_norm": 0.055964112494847354, + "learning_rate": 1.1137431540830248e-05, + "loss": 0.0006, + "step": 963 + }, + { + "epoch": 1.4431137724550898, + "grad_norm": 0.09671781357514952, + "learning_rate": 1.1121366220022317e-05, + "loss": 0.0012, + "step": 964 + }, + { + "epoch": 1.4446107784431137, + "grad_norm": 0.23802060163109717, + "learning_rate": 1.1105297967635947e-05, + "loss": 0.002, + "step": 965 + }, + { + "epoch": 1.4461077844311379, + "grad_norm": 0.17887498237379332, + "learning_rate": 1.1089226825678239e-05, + "loss": 0.0014, + "step": 966 + }, + { + "epoch": 1.4476047904191618, + "grad_norm": 0.041592588915209894, + "learning_rate": 1.1073152836163842e-05, + "loss": 0.0004, + "step": 967 + }, + { + "epoch": 1.4491017964071857, + "grad_norm": 0.1005400146875568, + "learning_rate": 1.105707604111486e-05, + "loss": 0.0006, + "step": 968 + }, + { + "epoch": 1.4505988023952097, + "grad_norm": 0.0974962168976635, + "learning_rate": 1.1040996482560724e-05, + "loss": 0.0014, + "step": 969 + }, + { + "epoch": 1.4520958083832336, + "grad_norm": 0.0935782185529421, + "learning_rate": 1.1024914202538088e-05, + "loss": 0.0008, + "step": 970 + }, + { + "epoch": 1.4535928143712575, + "grad_norm": 0.07730096837501058, + "learning_rate": 1.1008829243090724e-05, + "loss": 0.0006, + "step": 971 + }, + { + "epoch": 1.4550898203592815, + "grad_norm": 0.12823405242175784, + "learning_rate": 1.099274164626941e-05, + "loss": 0.0009, + "step": 972 + }, + { + "epoch": 1.4565868263473054, + "grad_norm": 0.07157978147756917, + "learning_rate": 1.0976651454131814e-05, + "loss": 0.001, + "step": 973 + }, + { + "epoch": 1.4580838323353293, + "grad_norm": 0.07310373401364316, + "learning_rate": 1.0960558708742401e-05, + "loss": 0.0005, + "step": 974 + }, + { + "epoch": 1.4595808383233533, + "grad_norm": 0.11186858330092693, + "learning_rate": 1.0944463452172297e-05, + "loss": 0.001, + "step": 975 + }, + { + "epoch": 1.4610778443113772, + "grad_norm": 0.1613611934605809, + "learning_rate": 1.0928365726499194e-05, + "loss": 0.0027, + "step": 976 + }, + { + "epoch": 1.4625748502994012, + "grad_norm": 0.1622550489094645, + "learning_rate": 1.0912265573807249e-05, + "loss": 0.0008, + "step": 977 + }, + { + "epoch": 1.464071856287425, + "grad_norm": 0.15392255395991697, + "learning_rate": 1.0896163036186953e-05, + "loss": 0.0014, + "step": 978 + }, + { + "epoch": 1.465568862275449, + "grad_norm": 0.15486888428359488, + "learning_rate": 1.088005815573504e-05, + "loss": 0.0013, + "step": 979 + }, + { + "epoch": 1.467065868263473, + "grad_norm": 0.16549731334948978, + "learning_rate": 1.0863950974554372e-05, + "loss": 0.0011, + "step": 980 + }, + { + "epoch": 1.468562874251497, + "grad_norm": 0.06451393525137326, + "learning_rate": 1.084784153475381e-05, + "loss": 0.0004, + "step": 981 + }, + { + "epoch": 1.4700598802395208, + "grad_norm": 0.08287888893875361, + "learning_rate": 1.0831729878448131e-05, + "loss": 0.0006, + "step": 982 + }, + { + "epoch": 1.471556886227545, + "grad_norm": 0.14774856291997404, + "learning_rate": 1.0815616047757909e-05, + "loss": 0.0015, + "step": 983 + }, + { + "epoch": 1.473053892215569, + "grad_norm": 0.08417242989544835, + "learning_rate": 1.0799500084809393e-05, + "loss": 0.0009, + "step": 984 + }, + { + "epoch": 1.4745508982035929, + "grad_norm": 0.05265592604561263, + "learning_rate": 1.0783382031734418e-05, + "loss": 0.0008, + "step": 985 + }, + { + "epoch": 1.4760479041916168, + "grad_norm": 0.11539265697300215, + "learning_rate": 1.0767261930670272e-05, + "loss": 0.0015, + "step": 986 + }, + { + "epoch": 1.4775449101796407, + "grad_norm": 0.029457601423187277, + "learning_rate": 1.0751139823759603e-05, + "loss": 0.0002, + "step": 987 + }, + { + "epoch": 1.4790419161676647, + "grad_norm": 0.10750596108363268, + "learning_rate": 1.0735015753150305e-05, + "loss": 0.0015, + "step": 988 + }, + { + "epoch": 1.4805389221556886, + "grad_norm": 0.23998882396666726, + "learning_rate": 1.0718889760995396e-05, + "loss": 0.0016, + "step": 989 + }, + { + "epoch": 1.4820359281437125, + "grad_norm": 0.1009612926927026, + "learning_rate": 1.070276188945293e-05, + "loss": 0.0014, + "step": 990 + }, + { + "epoch": 1.4835329341317365, + "grad_norm": 0.12007396345906915, + "learning_rate": 1.0686632180685868e-05, + "loss": 0.0015, + "step": 991 + }, + { + "epoch": 1.4850299401197604, + "grad_norm": 0.1882263167381516, + "learning_rate": 1.067050067686197e-05, + "loss": 0.0013, + "step": 992 + }, + { + "epoch": 1.4865269461077844, + "grad_norm": 0.16842368805723756, + "learning_rate": 1.0654367420153696e-05, + "loss": 0.0017, + "step": 993 + }, + { + "epoch": 1.4880239520958085, + "grad_norm": 0.13102324447775987, + "learning_rate": 1.0638232452738084e-05, + "loss": 0.0009, + "step": 994 + }, + { + "epoch": 1.4895209580838324, + "grad_norm": 0.14140439476011551, + "learning_rate": 1.0622095816796647e-05, + "loss": 0.0014, + "step": 995 + }, + { + "epoch": 1.4910179640718564, + "grad_norm": 0.048609314581425776, + "learning_rate": 1.0605957554515261e-05, + "loss": 0.0005, + "step": 996 + }, + { + "epoch": 1.4925149700598803, + "grad_norm": 0.09685439693147369, + "learning_rate": 1.058981770808405e-05, + "loss": 0.0008, + "step": 997 + }, + { + "epoch": 1.4940119760479043, + "grad_norm": 0.10476662342644127, + "learning_rate": 1.057367631969728e-05, + "loss": 0.0009, + "step": 998 + }, + { + "epoch": 1.4955089820359282, + "grad_norm": 0.07965591618267445, + "learning_rate": 1.0557533431553251e-05, + "loss": 0.0006, + "step": 999 + }, + { + "epoch": 1.4970059880239521, + "grad_norm": 0.14132503178186984, + "learning_rate": 1.0541389085854177e-05, + "loss": 0.0014, + "step": 1000 + }, + { + "epoch": 1.498502994011976, + "grad_norm": 0.06381511478435144, + "learning_rate": 1.0525243324806091e-05, + "loss": 0.0003, + "step": 1001 + }, + { + "epoch": 1.5, + "grad_norm": 0.19969657755752093, + "learning_rate": 1.0509096190618729e-05, + "loss": 0.0012, + "step": 1002 + }, + { + "epoch": 1.501497005988024, + "grad_norm": 0.16903934501226875, + "learning_rate": 1.0492947725505403e-05, + "loss": 0.0011, + "step": 1003 + }, + { + "epoch": 1.5029940119760479, + "grad_norm": 0.16313780662072883, + "learning_rate": 1.047679797168291e-05, + "loss": 0.0014, + "step": 1004 + }, + { + "epoch": 1.5044910179640718, + "grad_norm": 0.08804790182748146, + "learning_rate": 1.0460646971371424e-05, + "loss": 0.0008, + "step": 1005 + }, + { + "epoch": 1.5059880239520957, + "grad_norm": 0.1368065864574523, + "learning_rate": 1.0444494766794368e-05, + "loss": 0.0014, + "step": 1006 + }, + { + "epoch": 1.5074850299401197, + "grad_norm": 0.1814958223187234, + "learning_rate": 1.0428341400178317e-05, + "loss": 0.0018, + "step": 1007 + }, + { + "epoch": 1.5089820359281436, + "grad_norm": 0.10617237101998808, + "learning_rate": 1.0412186913752884e-05, + "loss": 0.001, + "step": 1008 + }, + { + "epoch": 1.5104790419161676, + "grad_norm": 0.12416915017989684, + "learning_rate": 1.0396031349750612e-05, + "loss": 0.0014, + "step": 1009 + }, + { + "epoch": 1.5119760479041915, + "grad_norm": 0.16700967749894446, + "learning_rate": 1.0379874750406858e-05, + "loss": 0.0018, + "step": 1010 + }, + { + "epoch": 1.5134730538922154, + "grad_norm": 0.07833916487876827, + "learning_rate": 1.0363717157959681e-05, + "loss": 0.0007, + "step": 1011 + }, + { + "epoch": 1.5149700598802394, + "grad_norm": 0.1046908594504945, + "learning_rate": 1.0347558614649747e-05, + "loss": 0.0007, + "step": 1012 + }, + { + "epoch": 1.5164670658682635, + "grad_norm": 0.12447015495058295, + "learning_rate": 1.0331399162720197e-05, + "loss": 0.0018, + "step": 1013 + }, + { + "epoch": 1.5179640718562875, + "grad_norm": 0.12177043291644524, + "learning_rate": 1.0315238844416557e-05, + "loss": 0.0013, + "step": 1014 + }, + { + "epoch": 1.5194610778443114, + "grad_norm": 0.07862091445416813, + "learning_rate": 1.0299077701986613e-05, + "loss": 0.0007, + "step": 1015 + }, + { + "epoch": 1.5209580838323353, + "grad_norm": 0.15981593793118748, + "learning_rate": 1.0282915777680307e-05, + "loss": 0.0012, + "step": 1016 + }, + { + "epoch": 1.5224550898203593, + "grad_norm": 0.09956716437231744, + "learning_rate": 1.0266753113749618e-05, + "loss": 0.001, + "step": 1017 + }, + { + "epoch": 1.5239520958083832, + "grad_norm": 0.12054415585534563, + "learning_rate": 1.0250589752448472e-05, + "loss": 0.0008, + "step": 1018 + }, + { + "epoch": 1.5254491017964071, + "grad_norm": 0.047284868067243004, + "learning_rate": 1.0234425736032607e-05, + "loss": 0.0004, + "step": 1019 + }, + { + "epoch": 1.5269461077844313, + "grad_norm": 0.15543960260097514, + "learning_rate": 1.0218261106759475e-05, + "loss": 0.0018, + "step": 1020 + }, + { + "epoch": 1.5284431137724552, + "grad_norm": 0.09490923813175474, + "learning_rate": 1.020209590688814e-05, + "loss": 0.0007, + "step": 1021 + }, + { + "epoch": 1.5299401197604792, + "grad_norm": 0.06536745934922997, + "learning_rate": 1.0185930178679146e-05, + "loss": 0.0006, + "step": 1022 + }, + { + "epoch": 1.531437125748503, + "grad_norm": 0.1301533353797184, + "learning_rate": 1.0169763964394421e-05, + "loss": 0.002, + "step": 1023 + }, + { + "epoch": 1.532934131736527, + "grad_norm": 0.1502603255655029, + "learning_rate": 1.0153597306297169e-05, + "loss": 0.0011, + "step": 1024 + }, + { + "epoch": 1.534431137724551, + "grad_norm": 0.04144441435798231, + "learning_rate": 1.013743024665175e-05, + "loss": 0.0003, + "step": 1025 + }, + { + "epoch": 1.535928143712575, + "grad_norm": 0.15809185401608605, + "learning_rate": 1.0121262827723573e-05, + "loss": 0.0023, + "step": 1026 + }, + { + "epoch": 1.5374251497005988, + "grad_norm": 0.17137867242074536, + "learning_rate": 1.0105095091778985e-05, + "loss": 0.0012, + "step": 1027 + }, + { + "epoch": 1.5389221556886228, + "grad_norm": 0.10676172512123629, + "learning_rate": 1.008892708108517e-05, + "loss": 0.0009, + "step": 1028 + }, + { + "epoch": 1.5404191616766467, + "grad_norm": 0.10415442540373282, + "learning_rate": 1.007275883791002e-05, + "loss": 0.0014, + "step": 1029 + }, + { + "epoch": 1.5419161676646707, + "grad_norm": 0.048845460680296805, + "learning_rate": 1.0056590404522042e-05, + "loss": 0.0006, + "step": 1030 + }, + { + "epoch": 1.5434131736526946, + "grad_norm": 0.09153026786032244, + "learning_rate": 1.0040421823190239e-05, + "loss": 0.0005, + "step": 1031 + }, + { + "epoch": 1.5449101796407185, + "grad_norm": 0.0979750349520669, + "learning_rate": 1.0024253136183997e-05, + "loss": 0.0008, + "step": 1032 + }, + { + "epoch": 1.5464071856287425, + "grad_norm": 0.12603703868213395, + "learning_rate": 1.000808438577298e-05, + "loss": 0.0007, + "step": 1033 + }, + { + "epoch": 1.5479041916167664, + "grad_norm": 0.09260218797278057, + "learning_rate": 9.991915614227022e-06, + "loss": 0.0006, + "step": 1034 + }, + { + "epoch": 1.5494011976047903, + "grad_norm": 0.150535783399099, + "learning_rate": 9.975746863816006e-06, + "loss": 0.0008, + "step": 1035 + }, + { + "epoch": 1.5508982035928143, + "grad_norm": 0.0866141689732322, + "learning_rate": 9.959578176809764e-06, + "loss": 0.0014, + "step": 1036 + }, + { + "epoch": 1.5523952095808382, + "grad_norm": 0.14207853844047355, + "learning_rate": 9.943409595477958e-06, + "loss": 0.0014, + "step": 1037 + }, + { + "epoch": 1.5538922155688621, + "grad_norm": 0.08207842243713191, + "learning_rate": 9.927241162089984e-06, + "loss": 0.0008, + "step": 1038 + }, + { + "epoch": 1.555389221556886, + "grad_norm": 0.14029627239900275, + "learning_rate": 9.911072918914834e-06, + "loss": 0.0028, + "step": 1039 + }, + { + "epoch": 1.55688622754491, + "grad_norm": 0.09422957915613028, + "learning_rate": 9.894904908221018e-06, + "loss": 0.0006, + "step": 1040 + }, + { + "epoch": 1.5583832335329342, + "grad_norm": 0.26480681530852446, + "learning_rate": 9.87873717227643e-06, + "loss": 0.002, + "step": 1041 + }, + { + "epoch": 1.5598802395209581, + "grad_norm": 0.18537320811580985, + "learning_rate": 9.862569753348255e-06, + "loss": 0.001, + "step": 1042 + }, + { + "epoch": 1.561377245508982, + "grad_norm": 0.12242399214233682, + "learning_rate": 9.846402693702833e-06, + "loss": 0.0009, + "step": 1043 + }, + { + "epoch": 1.562874251497006, + "grad_norm": 0.10685927064129254, + "learning_rate": 9.830236035605578e-06, + "loss": 0.001, + "step": 1044 + }, + { + "epoch": 1.56437125748503, + "grad_norm": 0.13222209724765782, + "learning_rate": 9.814069821320856e-06, + "loss": 0.0023, + "step": 1045 + }, + { + "epoch": 1.5658682634730539, + "grad_norm": 0.0968515135585454, + "learning_rate": 9.79790409311186e-06, + "loss": 0.0008, + "step": 1046 + }, + { + "epoch": 1.5673652694610778, + "grad_norm": 0.05675857658075679, + "learning_rate": 9.781738893240526e-06, + "loss": 0.0005, + "step": 1047 + }, + { + "epoch": 1.568862275449102, + "grad_norm": 0.10548064797653442, + "learning_rate": 9.765574263967397e-06, + "loss": 0.001, + "step": 1048 + }, + { + "epoch": 1.5703592814371259, + "grad_norm": 0.14583652639833644, + "learning_rate": 9.749410247551534e-06, + "loss": 0.0022, + "step": 1049 + }, + { + "epoch": 1.5718562874251498, + "grad_norm": 0.10085574392360613, + "learning_rate": 9.733246886250384e-06, + "loss": 0.0008, + "step": 1050 + }, + { + "epoch": 1.5733532934131738, + "grad_norm": 0.16062888967169808, + "learning_rate": 9.717084222319698e-06, + "loss": 0.0024, + "step": 1051 + }, + { + "epoch": 1.5748502994011977, + "grad_norm": 0.06334101246324102, + "learning_rate": 9.700922298013388e-06, + "loss": 0.0006, + "step": 1052 + }, + { + "epoch": 1.5763473053892216, + "grad_norm": 0.10858659881531557, + "learning_rate": 9.684761155583443e-06, + "loss": 0.001, + "step": 1053 + }, + { + "epoch": 1.5778443113772456, + "grad_norm": 0.05720430794355147, + "learning_rate": 9.668600837279808e-06, + "loss": 0.0003, + "step": 1054 + }, + { + "epoch": 1.5793413173652695, + "grad_norm": 0.18596258294319445, + "learning_rate": 9.652441385350258e-06, + "loss": 0.0008, + "step": 1055 + }, + { + "epoch": 1.5808383233532934, + "grad_norm": 0.02161457070852507, + "learning_rate": 9.636282842040324e-06, + "loss": 0.0002, + "step": 1056 + }, + { + "epoch": 1.5823353293413174, + "grad_norm": 0.12665814776438672, + "learning_rate": 9.620125249593145e-06, + "loss": 0.002, + "step": 1057 + }, + { + "epoch": 1.5838323353293413, + "grad_norm": 0.1479035747876992, + "learning_rate": 9.603968650249388e-06, + "loss": 0.0014, + "step": 1058 + }, + { + "epoch": 1.5853293413173652, + "grad_norm": 0.08875005354112726, + "learning_rate": 9.587813086247118e-06, + "loss": 0.0008, + "step": 1059 + }, + { + "epoch": 1.5868263473053892, + "grad_norm": 0.034598142239245726, + "learning_rate": 9.571658599821684e-06, + "loss": 0.0002, + "step": 1060 + }, + { + "epoch": 1.5883233532934131, + "grad_norm": 0.1382510341790496, + "learning_rate": 9.555505233205635e-06, + "loss": 0.0014, + "step": 1061 + }, + { + "epoch": 1.589820359281437, + "grad_norm": 0.13116321301263414, + "learning_rate": 9.539353028628577e-06, + "loss": 0.001, + "step": 1062 + }, + { + "epoch": 1.591317365269461, + "grad_norm": 0.05547568359029333, + "learning_rate": 9.523202028317095e-06, + "loss": 0.0007, + "step": 1063 + }, + { + "epoch": 1.592814371257485, + "grad_norm": 0.031905668491041365, + "learning_rate": 9.5070522744946e-06, + "loss": 0.0003, + "step": 1064 + }, + { + "epoch": 1.5943113772455089, + "grad_norm": 0.008727577881278311, + "learning_rate": 9.490903809381273e-06, + "loss": 0.0001, + "step": 1065 + }, + { + "epoch": 1.5958083832335328, + "grad_norm": 0.0712191739355828, + "learning_rate": 9.47475667519391e-06, + "loss": 0.0015, + "step": 1066 + }, + { + "epoch": 1.5973053892215567, + "grad_norm": 0.03749421756809963, + "learning_rate": 9.458610914145826e-06, + "loss": 0.001, + "step": 1067 + }, + { + "epoch": 1.5988023952095807, + "grad_norm": 0.02379835401791554, + "learning_rate": 9.442466568446755e-06, + "loss": 0.0003, + "step": 1068 + }, + { + "epoch": 1.6002994011976048, + "grad_norm": 0.12118141203303913, + "learning_rate": 9.426323680302722e-06, + "loss": 0.0013, + "step": 1069 + }, + { + "epoch": 1.6017964071856288, + "grad_norm": 0.12388549366182541, + "learning_rate": 9.410182291915956e-06, + "loss": 0.0009, + "step": 1070 + }, + { + "epoch": 1.6032934131736527, + "grad_norm": 0.117253435172126, + "learning_rate": 9.39404244548474e-06, + "loss": 0.0013, + "step": 1071 + }, + { + "epoch": 1.6047904191616766, + "grad_norm": 0.08211137201727235, + "learning_rate": 9.377904183203353e-06, + "loss": 0.0009, + "step": 1072 + }, + { + "epoch": 1.6062874251497006, + "grad_norm": 0.09352507617778973, + "learning_rate": 9.361767547261918e-06, + "loss": 0.0009, + "step": 1073 + }, + { + "epoch": 1.6077844311377245, + "grad_norm": 0.04451673559222652, + "learning_rate": 9.345632579846306e-06, + "loss": 0.0003, + "step": 1074 + }, + { + "epoch": 1.6092814371257484, + "grad_norm": 0.11115723113169731, + "learning_rate": 9.329499323138035e-06, + "loss": 0.0007, + "step": 1075 + }, + { + "epoch": 1.6107784431137726, + "grad_norm": 0.046162578574199704, + "learning_rate": 9.313367819314136e-06, + "loss": 0.0005, + "step": 1076 + }, + { + "epoch": 1.6122754491017965, + "grad_norm": 0.04801631426459736, + "learning_rate": 9.297238110547075e-06, + "loss": 0.0006, + "step": 1077 + }, + { + "epoch": 1.6137724550898205, + "grad_norm": 0.042030541819221175, + "learning_rate": 9.281110239004607e-06, + "loss": 0.0006, + "step": 1078 + }, + { + "epoch": 1.6152694610778444, + "grad_norm": 0.04359620580995584, + "learning_rate": 9.264984246849699e-06, + "loss": 0.0008, + "step": 1079 + }, + { + "epoch": 1.6167664670658684, + "grad_norm": 0.12249935790486885, + "learning_rate": 9.2488601762404e-06, + "loss": 0.0012, + "step": 1080 + }, + { + "epoch": 1.6182634730538923, + "grad_norm": 0.08869067151066333, + "learning_rate": 9.23273806932973e-06, + "loss": 0.0006, + "step": 1081 + }, + { + "epoch": 1.6197604790419162, + "grad_norm": 0.04713954076892878, + "learning_rate": 9.216617968265587e-06, + "loss": 0.0004, + "step": 1082 + }, + { + "epoch": 1.6212574850299402, + "grad_norm": 0.022606337617157864, + "learning_rate": 9.200499915190609e-06, + "loss": 0.0002, + "step": 1083 + }, + { + "epoch": 1.622754491017964, + "grad_norm": 0.044196202525303686, + "learning_rate": 9.184383952242095e-06, + "loss": 0.0004, + "step": 1084 + }, + { + "epoch": 1.624251497005988, + "grad_norm": 0.04585085088431671, + "learning_rate": 9.16827012155187e-06, + "loss": 0.0003, + "step": 1085 + }, + { + "epoch": 1.625748502994012, + "grad_norm": 0.1099988952217994, + "learning_rate": 9.152158465246192e-06, + "loss": 0.0012, + "step": 1086 + }, + { + "epoch": 1.627245508982036, + "grad_norm": 0.21437678680661543, + "learning_rate": 9.136049025445633e-06, + "loss": 0.0024, + "step": 1087 + }, + { + "epoch": 1.6287425149700598, + "grad_norm": 0.07403926272991478, + "learning_rate": 9.11994184426496e-06, + "loss": 0.0005, + "step": 1088 + }, + { + "epoch": 1.6302395209580838, + "grad_norm": 0.059002757939724, + "learning_rate": 9.10383696381305e-06, + "loss": 0.0007, + "step": 1089 + }, + { + "epoch": 1.6317365269461077, + "grad_norm": 0.12675376681424935, + "learning_rate": 9.087734426192753e-06, + "loss": 0.001, + "step": 1090 + }, + { + "epoch": 1.6332335329341316, + "grad_norm": 0.13626773883445795, + "learning_rate": 9.07163427350081e-06, + "loss": 0.0011, + "step": 1091 + }, + { + "epoch": 1.6347305389221556, + "grad_norm": 0.09619385565164779, + "learning_rate": 9.055536547827708e-06, + "loss": 0.0012, + "step": 1092 + }, + { + "epoch": 1.6362275449101795, + "grad_norm": 0.07077546546688278, + "learning_rate": 9.0394412912576e-06, + "loss": 0.0008, + "step": 1093 + }, + { + "epoch": 1.6377245508982035, + "grad_norm": 0.11742541713396401, + "learning_rate": 9.023348545868187e-06, + "loss": 0.0017, + "step": 1094 + }, + { + "epoch": 1.6392215568862274, + "grad_norm": 0.05392358126836917, + "learning_rate": 9.007258353730594e-06, + "loss": 0.0005, + "step": 1095 + }, + { + "epoch": 1.6407185628742516, + "grad_norm": 0.05286986458798405, + "learning_rate": 8.99117075690928e-06, + "loss": 0.0004, + "step": 1096 + }, + { + "epoch": 1.6422155688622755, + "grad_norm": 0.1037025844104848, + "learning_rate": 8.975085797461915e-06, + "loss": 0.0009, + "step": 1097 + }, + { + "epoch": 1.6437125748502994, + "grad_norm": 0.026409960690047918, + "learning_rate": 8.959003517439282e-06, + "loss": 0.0003, + "step": 1098 + }, + { + "epoch": 1.6452095808383234, + "grad_norm": 0.06352856914344773, + "learning_rate": 8.942923958885142e-06, + "loss": 0.0003, + "step": 1099 + }, + { + "epoch": 1.6467065868263473, + "grad_norm": 0.061764806505025756, + "learning_rate": 8.92684716383616e-06, + "loss": 0.0008, + "step": 1100 + }, + { + "epoch": 1.6482035928143712, + "grad_norm": 0.05291688814800548, + "learning_rate": 8.910773174321765e-06, + "loss": 0.0007, + "step": 1101 + }, + { + "epoch": 1.6497005988023952, + "grad_norm": 0.07133556607963744, + "learning_rate": 8.894702032364053e-06, + "loss": 0.0011, + "step": 1102 + }, + { + "epoch": 1.6511976047904193, + "grad_norm": 0.09968542393075754, + "learning_rate": 8.878633779977686e-06, + "loss": 0.0019, + "step": 1103 + }, + { + "epoch": 1.6526946107784433, + "grad_norm": 0.03251881336576024, + "learning_rate": 8.862568459169755e-06, + "loss": 0.0003, + "step": 1104 + }, + { + "epoch": 1.6541916167664672, + "grad_norm": 0.06370247682649403, + "learning_rate": 8.846506111939699e-06, + "loss": 0.0005, + "step": 1105 + }, + { + "epoch": 1.6556886227544911, + "grad_norm": 0.06213918678886638, + "learning_rate": 8.830446780279175e-06, + "loss": 0.0005, + "step": 1106 + }, + { + "epoch": 1.657185628742515, + "grad_norm": 0.10415475247431645, + "learning_rate": 8.814390506171967e-06, + "loss": 0.0013, + "step": 1107 + }, + { + "epoch": 1.658682634730539, + "grad_norm": 0.07744622474255668, + "learning_rate": 8.79833733159386e-06, + "loss": 0.002, + "step": 1108 + }, + { + "epoch": 1.660179640718563, + "grad_norm": 0.04324172934252314, + "learning_rate": 8.78228729851253e-06, + "loss": 0.0004, + "step": 1109 + }, + { + "epoch": 1.6616766467065869, + "grad_norm": 0.1796309695730341, + "learning_rate": 8.766240448887454e-06, + "loss": 0.002, + "step": 1110 + }, + { + "epoch": 1.6631736526946108, + "grad_norm": 0.049210623724537356, + "learning_rate": 8.750196824669765e-06, + "loss": 0.0008, + "step": 1111 + }, + { + "epoch": 1.6646706586826348, + "grad_norm": 0.06374661903865328, + "learning_rate": 8.73415646780219e-06, + "loss": 0.0005, + "step": 1112 + }, + { + "epoch": 1.6661676646706587, + "grad_norm": 0.0678542988687008, + "learning_rate": 8.718119420218894e-06, + "loss": 0.0006, + "step": 1113 + }, + { + "epoch": 1.6676646706586826, + "grad_norm": 0.05131894267995365, + "learning_rate": 8.702085723845402e-06, + "loss": 0.0005, + "step": 1114 + }, + { + "epoch": 1.6691616766467066, + "grad_norm": 0.11138493277856698, + "learning_rate": 8.686055420598474e-06, + "loss": 0.0011, + "step": 1115 + }, + { + "epoch": 1.6706586826347305, + "grad_norm": 0.04388393803892101, + "learning_rate": 8.670028552385994e-06, + "loss": 0.0004, + "step": 1116 + }, + { + "epoch": 1.6721556886227544, + "grad_norm": 0.11972182187525221, + "learning_rate": 8.654005161106877e-06, + "loss": 0.0011, + "step": 1117 + }, + { + "epoch": 1.6736526946107784, + "grad_norm": 0.126003813560071, + "learning_rate": 8.637985288650939e-06, + "loss": 0.0009, + "step": 1118 + }, + { + "epoch": 1.6751497005988023, + "grad_norm": 0.10943077524662276, + "learning_rate": 8.621968976898804e-06, + "loss": 0.0019, + "step": 1119 + }, + { + "epoch": 1.6766467065868262, + "grad_norm": 0.06513537092099632, + "learning_rate": 8.605956267721778e-06, + "loss": 0.0007, + "step": 1120 + }, + { + "epoch": 1.6781437125748502, + "grad_norm": 0.13419673873932153, + "learning_rate": 8.589947202981755e-06, + "loss": 0.0008, + "step": 1121 + }, + { + "epoch": 1.6796407185628741, + "grad_norm": 0.04079284884695222, + "learning_rate": 8.573941824531109e-06, + "loss": 0.0003, + "step": 1122 + }, + { + "epoch": 1.681137724550898, + "grad_norm": 0.0368419755710961, + "learning_rate": 8.557940174212555e-06, + "loss": 0.0003, + "step": 1123 + }, + { + "epoch": 1.6826347305389222, + "grad_norm": 0.11915015080619738, + "learning_rate": 8.541942293859086e-06, + "loss": 0.0015, + "step": 1124 + }, + { + "epoch": 1.6841317365269461, + "grad_norm": 0.11280338061106095, + "learning_rate": 8.52594822529382e-06, + "loss": 0.001, + "step": 1125 + }, + { + "epoch": 1.68562874251497, + "grad_norm": 0.11129477072458856, + "learning_rate": 8.509958010329931e-06, + "loss": 0.0009, + "step": 1126 + }, + { + "epoch": 1.687125748502994, + "grad_norm": 0.11817747887791782, + "learning_rate": 8.49397169077049e-06, + "loss": 0.0011, + "step": 1127 + }, + { + "epoch": 1.688622754491018, + "grad_norm": 0.09555692890154081, + "learning_rate": 8.477989308408409e-06, + "loss": 0.0014, + "step": 1128 + }, + { + "epoch": 1.6901197604790419, + "grad_norm": 0.06104674622378712, + "learning_rate": 8.462010905026298e-06, + "loss": 0.0004, + "step": 1129 + }, + { + "epoch": 1.6916167664670658, + "grad_norm": 0.11401940154366785, + "learning_rate": 8.44603652239636e-06, + "loss": 0.0009, + "step": 1130 + }, + { + "epoch": 1.69311377245509, + "grad_norm": 0.1359921086438187, + "learning_rate": 8.430066202280302e-06, + "loss": 0.0009, + "step": 1131 + }, + { + "epoch": 1.694610778443114, + "grad_norm": 0.07677903643974247, + "learning_rate": 8.414099986429188e-06, + "loss": 0.0005, + "step": 1132 + }, + { + "epoch": 1.6961077844311379, + "grad_norm": 0.04839849020100479, + "learning_rate": 8.398137916583371e-06, + "loss": 0.0003, + "step": 1133 + }, + { + "epoch": 1.6976047904191618, + "grad_norm": 0.12381943481528089, + "learning_rate": 8.382180034472353e-06, + "loss": 0.0008, + "step": 1134 + }, + { + "epoch": 1.6991017964071857, + "grad_norm": 0.07469322717802963, + "learning_rate": 8.366226381814698e-06, + "loss": 0.0007, + "step": 1135 + }, + { + "epoch": 1.7005988023952097, + "grad_norm": 0.06504097124135295, + "learning_rate": 8.350277000317911e-06, + "loss": 0.0006, + "step": 1136 + }, + { + "epoch": 1.7020958083832336, + "grad_norm": 0.05289705787630766, + "learning_rate": 8.334331931678317e-06, + "loss": 0.0005, + "step": 1137 + }, + { + "epoch": 1.7035928143712575, + "grad_norm": 0.08661338942838477, + "learning_rate": 8.31839121758099e-06, + "loss": 0.0015, + "step": 1138 + }, + { + "epoch": 1.7050898203592815, + "grad_norm": 0.08767555555868468, + "learning_rate": 8.302454899699597e-06, + "loss": 0.0006, + "step": 1139 + }, + { + "epoch": 1.7065868263473054, + "grad_norm": 0.09256232285744476, + "learning_rate": 8.286523019696327e-06, + "loss": 0.001, + "step": 1140 + }, + { + "epoch": 1.7080838323353293, + "grad_norm": 0.053608939965644864, + "learning_rate": 8.270595619221762e-06, + "loss": 0.0005, + "step": 1141 + }, + { + "epoch": 1.7095808383233533, + "grad_norm": 0.03018588325618429, + "learning_rate": 8.254672739914771e-06, + "loss": 0.0005, + "step": 1142 + }, + { + "epoch": 1.7110778443113772, + "grad_norm": 0.01313607644598902, + "learning_rate": 8.238754423402411e-06, + "loss": 0.0001, + "step": 1143 + }, + { + "epoch": 1.7125748502994012, + "grad_norm": 0.12588799971374123, + "learning_rate": 8.222840711299798e-06, + "loss": 0.0008, + "step": 1144 + }, + { + "epoch": 1.714071856287425, + "grad_norm": 0.08395380803153646, + "learning_rate": 8.206931645210025e-06, + "loss": 0.0007, + "step": 1145 + }, + { + "epoch": 1.715568862275449, + "grad_norm": 0.05648989201497437, + "learning_rate": 8.191027266724026e-06, + "loss": 0.0005, + "step": 1146 + }, + { + "epoch": 1.717065868263473, + "grad_norm": 0.299079726997726, + "learning_rate": 8.175127617420489e-06, + "loss": 0.0038, + "step": 1147 + }, + { + "epoch": 1.718562874251497, + "grad_norm": 0.2211219838190909, + "learning_rate": 8.15923273886574e-06, + "loss": 0.0017, + "step": 1148 + }, + { + "epoch": 1.7200598802395208, + "grad_norm": 0.13423753642877773, + "learning_rate": 8.14334267261362e-06, + "loss": 0.0013, + "step": 1149 + }, + { + "epoch": 1.7215568862275448, + "grad_norm": 0.021323410927144678, + "learning_rate": 8.127457460205403e-06, + "loss": 0.0002, + "step": 1150 + }, + { + "epoch": 1.7230538922155687, + "grad_norm": 0.10537068848194503, + "learning_rate": 8.111577143169667e-06, + "loss": 0.0009, + "step": 1151 + }, + { + "epoch": 1.7245508982035929, + "grad_norm": 0.16469004106932947, + "learning_rate": 8.095701763022197e-06, + "loss": 0.0013, + "step": 1152 + }, + { + "epoch": 1.7260479041916168, + "grad_norm": 0.17092342034188415, + "learning_rate": 8.079831361265862e-06, + "loss": 0.0017, + "step": 1153 + }, + { + "epoch": 1.7275449101796407, + "grad_norm": 0.06779177154424461, + "learning_rate": 8.063965979390526e-06, + "loss": 0.0006, + "step": 1154 + }, + { + "epoch": 1.7290419161676647, + "grad_norm": 0.1236163086259948, + "learning_rate": 8.048105658872928e-06, + "loss": 0.0019, + "step": 1155 + }, + { + "epoch": 1.7305389221556886, + "grad_norm": 0.12610975535688002, + "learning_rate": 8.032250441176566e-06, + "loss": 0.0028, + "step": 1156 + }, + { + "epoch": 1.7320359281437125, + "grad_norm": 0.18941444937914406, + "learning_rate": 8.01640036775161e-06, + "loss": 0.0012, + "step": 1157 + }, + { + "epoch": 1.7335329341317365, + "grad_norm": 0.0831881723747311, + "learning_rate": 8.000555480034774e-06, + "loss": 0.0006, + "step": 1158 + }, + { + "epoch": 1.7350299401197606, + "grad_norm": 0.11838944511546944, + "learning_rate": 7.984715819449219e-06, + "loss": 0.0009, + "step": 1159 + }, + { + "epoch": 1.7365269461077846, + "grad_norm": 0.03888943100450595, + "learning_rate": 7.968881427404433e-06, + "loss": 0.0004, + "step": 1160 + }, + { + "epoch": 1.7380239520958085, + "grad_norm": 0.04784679025928149, + "learning_rate": 7.953052345296141e-06, + "loss": 0.0004, + "step": 1161 + }, + { + "epoch": 1.7395209580838324, + "grad_norm": 0.09465522756842437, + "learning_rate": 7.937228614506183e-06, + "loss": 0.0011, + "step": 1162 + }, + { + "epoch": 1.7410179640718564, + "grad_norm": 0.02839056527145619, + "learning_rate": 7.921410276402403e-06, + "loss": 0.0002, + "step": 1163 + }, + { + "epoch": 1.7425149700598803, + "grad_norm": 0.11897825771964533, + "learning_rate": 7.905597372338558e-06, + "loss": 0.0015, + "step": 1164 + }, + { + "epoch": 1.7440119760479043, + "grad_norm": 0.13433046182833272, + "learning_rate": 7.889789943654185e-06, + "loss": 0.0014, + "step": 1165 + }, + { + "epoch": 1.7455089820359282, + "grad_norm": 0.042192664781619316, + "learning_rate": 7.87398803167452e-06, + "loss": 0.0003, + "step": 1166 + }, + { + "epoch": 1.7470059880239521, + "grad_norm": 0.09786003136143633, + "learning_rate": 7.858191677710364e-06, + "loss": 0.0018, + "step": 1167 + }, + { + "epoch": 1.748502994011976, + "grad_norm": 0.1278807783300935, + "learning_rate": 7.842400923058e-06, + "loss": 0.0012, + "step": 1168 + }, + { + "epoch": 1.75, + "grad_norm": 0.18490520684949707, + "learning_rate": 7.826615808999066e-06, + "loss": 0.0015, + "step": 1169 + }, + { + "epoch": 1.751497005988024, + "grad_norm": 0.10074807032478746, + "learning_rate": 7.810836376800455e-06, + "loss": 0.0007, + "step": 1170 + }, + { + "epoch": 1.7529940119760479, + "grad_norm": 0.06251541594987227, + "learning_rate": 7.795062667714211e-06, + "loss": 0.0005, + "step": 1171 + }, + { + "epoch": 1.7544910179640718, + "grad_norm": 0.05781097111200272, + "learning_rate": 7.7792947229774e-06, + "loss": 0.0002, + "step": 1172 + }, + { + "epoch": 1.7559880239520957, + "grad_norm": 0.2185198392929054, + "learning_rate": 7.763532583812042e-06, + "loss": 0.0018, + "step": 1173 + }, + { + "epoch": 1.7574850299401197, + "grad_norm": 0.064144359442749, + "learning_rate": 7.747776291424955e-06, + "loss": 0.0004, + "step": 1174 + }, + { + "epoch": 1.7589820359281436, + "grad_norm": 0.08498960587781633, + "learning_rate": 7.732025887007693e-06, + "loss": 0.0008, + "step": 1175 + }, + { + "epoch": 1.7604790419161676, + "grad_norm": 0.1922473098948509, + "learning_rate": 7.716281411736411e-06, + "loss": 0.0016, + "step": 1176 + }, + { + "epoch": 1.7619760479041915, + "grad_norm": 0.0661949673802176, + "learning_rate": 7.700542906771747e-06, + "loss": 0.0004, + "step": 1177 + }, + { + "epoch": 1.7634730538922154, + "grad_norm": 0.10775727570874787, + "learning_rate": 7.684810413258756e-06, + "loss": 0.0015, + "step": 1178 + }, + { + "epoch": 1.7649700598802394, + "grad_norm": 0.041135301287911916, + "learning_rate": 7.669083972326761e-06, + "loss": 0.0004, + "step": 1179 + }, + { + "epoch": 1.7664670658682635, + "grad_norm": 0.07967634346867503, + "learning_rate": 7.653363625089272e-06, + "loss": 0.0009, + "step": 1180 + }, + { + "epoch": 1.7679640718562875, + "grad_norm": 0.06632082900120188, + "learning_rate": 7.637649412643852e-06, + "loss": 0.0015, + "step": 1181 + }, + { + "epoch": 1.7694610778443114, + "grad_norm": 0.0597435675787796, + "learning_rate": 7.621941376072045e-06, + "loss": 0.0005, + "step": 1182 + }, + { + "epoch": 1.7709580838323353, + "grad_norm": 0.10680414994888268, + "learning_rate": 7.606239556439243e-06, + "loss": 0.0008, + "step": 1183 + }, + { + "epoch": 1.7724550898203593, + "grad_norm": 0.08131544743100391, + "learning_rate": 7.590543994794575e-06, + "loss": 0.0005, + "step": 1184 + }, + { + "epoch": 1.7739520958083832, + "grad_norm": 0.14796670864528785, + "learning_rate": 7.574854732170824e-06, + "loss": 0.0011, + "step": 1185 + }, + { + "epoch": 1.7754491017964071, + "grad_norm": 0.05808741649463366, + "learning_rate": 7.559171809584298e-06, + "loss": 0.0004, + "step": 1186 + }, + { + "epoch": 1.7769461077844313, + "grad_norm": 0.0791406942247839, + "learning_rate": 7.543495268034735e-06, + "loss": 0.0009, + "step": 1187 + }, + { + "epoch": 1.7784431137724552, + "grad_norm": 0.047886021234621046, + "learning_rate": 7.527825148505181e-06, + "loss": 0.0003, + "step": 1188 + }, + { + "epoch": 1.7799401197604792, + "grad_norm": 0.1335212254194146, + "learning_rate": 7.512161491961903e-06, + "loss": 0.001, + "step": 1189 + }, + { + "epoch": 1.781437125748503, + "grad_norm": 0.03913458823023904, + "learning_rate": 7.496504339354274e-06, + "loss": 0.0006, + "step": 1190 + }, + { + "epoch": 1.782934131736527, + "grad_norm": 0.08672922100866931, + "learning_rate": 7.48085373161465e-06, + "loss": 0.0011, + "step": 1191 + }, + { + "epoch": 1.784431137724551, + "grad_norm": 0.11931210383133138, + "learning_rate": 7.465209709658294e-06, + "loss": 0.0011, + "step": 1192 + }, + { + "epoch": 1.785928143712575, + "grad_norm": 0.05929701873982781, + "learning_rate": 7.449572314383237e-06, + "loss": 0.0006, + "step": 1193 + }, + { + "epoch": 1.7874251497005988, + "grad_norm": 0.04505152324525866, + "learning_rate": 7.4339415866701945e-06, + "loss": 0.0003, + "step": 1194 + }, + { + "epoch": 1.7889221556886228, + "grad_norm": 0.12017810117570427, + "learning_rate": 7.418317567382447e-06, + "loss": 0.0008, + "step": 1195 + }, + { + "epoch": 1.7904191616766467, + "grad_norm": 0.09158715938134543, + "learning_rate": 7.402700297365741e-06, + "loss": 0.0007, + "step": 1196 + }, + { + "epoch": 1.7919161676646707, + "grad_norm": 0.1014983459541966, + "learning_rate": 7.38708981744818e-06, + "loss": 0.0006, + "step": 1197 + }, + { + "epoch": 1.7934131736526946, + "grad_norm": 0.08269083942405764, + "learning_rate": 7.3714861684401025e-06, + "loss": 0.0004, + "step": 1198 + }, + { + "epoch": 1.7949101796407185, + "grad_norm": 0.06644722617806384, + "learning_rate": 7.355889391134011e-06, + "loss": 0.0006, + "step": 1199 + }, + { + "epoch": 1.7964071856287425, + "grad_norm": 0.05523988911350223, + "learning_rate": 7.340299526304422e-06, + "loss": 0.0004, + "step": 1200 + }, + { + "epoch": 1.7979041916167664, + "grad_norm": 0.08351184232628797, + "learning_rate": 7.324716614707794e-06, + "loss": 0.0009, + "step": 1201 + }, + { + "epoch": 1.7994011976047903, + "grad_norm": 0.006774670042731655, + "learning_rate": 7.309140697082402e-06, + "loss": 0.0001, + "step": 1202 + }, + { + "epoch": 1.8008982035928143, + "grad_norm": 0.04928550988537502, + "learning_rate": 7.293571814148241e-06, + "loss": 0.0003, + "step": 1203 + }, + { + "epoch": 1.8023952095808382, + "grad_norm": 0.06584370826534322, + "learning_rate": 7.2780100066069174e-06, + "loss": 0.0003, + "step": 1204 + }, + { + "epoch": 1.8038922155688621, + "grad_norm": 0.18345445829638807, + "learning_rate": 7.262455315141526e-06, + "loss": 0.001, + "step": 1205 + }, + { + "epoch": 1.805389221556886, + "grad_norm": 0.07244929387533793, + "learning_rate": 7.246907780416574e-06, + "loss": 0.0005, + "step": 1206 + }, + { + "epoch": 1.80688622754491, + "grad_norm": 0.09817729337327456, + "learning_rate": 7.231367443077851e-06, + "loss": 0.0007, + "step": 1207 + }, + { + "epoch": 1.8083832335329342, + "grad_norm": 0.05483140894713459, + "learning_rate": 7.215834343752337e-06, + "loss": 0.0004, + "step": 1208 + }, + { + "epoch": 1.8098802395209581, + "grad_norm": 0.03551972805679641, + "learning_rate": 7.2003085230480785e-06, + "loss": 0.0003, + "step": 1209 + }, + { + "epoch": 1.811377245508982, + "grad_norm": 0.06212947300716921, + "learning_rate": 7.184790021554102e-06, + "loss": 0.0004, + "step": 1210 + }, + { + "epoch": 1.812874251497006, + "grad_norm": 0.11663130238724655, + "learning_rate": 7.169278879840304e-06, + "loss": 0.001, + "step": 1211 + }, + { + "epoch": 1.81437125748503, + "grad_norm": 0.1655611404625784, + "learning_rate": 7.153775138457326e-06, + "loss": 0.0014, + "step": 1212 + }, + { + "epoch": 1.8158682634730539, + "grad_norm": 0.016811102246245015, + "learning_rate": 7.138278837936475e-06, + "loss": 0.0001, + "step": 1213 + }, + { + "epoch": 1.8173652694610778, + "grad_norm": 0.08174583277697046, + "learning_rate": 7.122790018789597e-06, + "loss": 0.0008, + "step": 1214 + }, + { + "epoch": 1.818862275449102, + "grad_norm": 0.09229932047043227, + "learning_rate": 7.107308721508992e-06, + "loss": 0.0005, + "step": 1215 + }, + { + "epoch": 1.8203592814371259, + "grad_norm": 0.05703431462629337, + "learning_rate": 7.091834986567278e-06, + "loss": 0.0003, + "step": 1216 + }, + { + "epoch": 1.8218562874251498, + "grad_norm": 0.045543500277733254, + "learning_rate": 7.076368854417319e-06, + "loss": 0.0004, + "step": 1217 + }, + { + "epoch": 1.8233532934131738, + "grad_norm": 0.04621091298397737, + "learning_rate": 7.0609103654920955e-06, + "loss": 0.0006, + "step": 1218 + }, + { + "epoch": 1.8248502994011977, + "grad_norm": 0.12194672605605335, + "learning_rate": 7.045459560204604e-06, + "loss": 0.0008, + "step": 1219 + }, + { + "epoch": 1.8263473053892216, + "grad_norm": 0.04569140798030763, + "learning_rate": 7.030016478947763e-06, + "loss": 0.0003, + "step": 1220 + }, + { + "epoch": 1.8278443113772456, + "grad_norm": 0.09081360546587813, + "learning_rate": 7.014581162094284e-06, + "loss": 0.0003, + "step": 1221 + }, + { + "epoch": 1.8293413173652695, + "grad_norm": 0.07508995179575473, + "learning_rate": 6.999153649996595e-06, + "loss": 0.0007, + "step": 1222 + }, + { + "epoch": 1.8308383233532934, + "grad_norm": 0.0236647586499013, + "learning_rate": 6.983733982986709e-06, + "loss": 0.0002, + "step": 1223 + }, + { + "epoch": 1.8323353293413174, + "grad_norm": 0.0348458963052325, + "learning_rate": 6.968322201376136e-06, + "loss": 0.0003, + "step": 1224 + }, + { + "epoch": 1.8338323353293413, + "grad_norm": 0.06740266996280521, + "learning_rate": 6.952918345455772e-06, + "loss": 0.0005, + "step": 1225 + }, + { + "epoch": 1.8353293413173652, + "grad_norm": 0.1410919073179155, + "learning_rate": 6.937522455495787e-06, + "loss": 0.0025, + "step": 1226 + }, + { + "epoch": 1.8368263473053892, + "grad_norm": 0.050967621048530726, + "learning_rate": 6.92213457174553e-06, + "loss": 0.0003, + "step": 1227 + }, + { + "epoch": 1.8383233532934131, + "grad_norm": 0.054601006610630415, + "learning_rate": 6.906754734433414e-06, + "loss": 0.0006, + "step": 1228 + }, + { + "epoch": 1.839820359281437, + "grad_norm": 0.12698206601810638, + "learning_rate": 6.891382983766828e-06, + "loss": 0.0013, + "step": 1229 + }, + { + "epoch": 1.841317365269461, + "grad_norm": 0.08636527463657546, + "learning_rate": 6.876019359932004e-06, + "loss": 0.0012, + "step": 1230 + }, + { + "epoch": 1.842814371257485, + "grad_norm": 0.05071079998230555, + "learning_rate": 6.860663903093943e-06, + "loss": 0.0003, + "step": 1231 + }, + { + "epoch": 1.8443113772455089, + "grad_norm": 0.018134041251772544, + "learning_rate": 6.8453166533962925e-06, + "loss": 0.0001, + "step": 1232 + }, + { + "epoch": 1.8458083832335328, + "grad_norm": 0.10446778142724229, + "learning_rate": 6.829977650961229e-06, + "loss": 0.0016, + "step": 1233 + }, + { + "epoch": 1.8473053892215567, + "grad_norm": 0.055625311823021856, + "learning_rate": 6.814646935889389e-06, + "loss": 0.0004, + "step": 1234 + }, + { + "epoch": 1.8488023952095807, + "grad_norm": 0.07096498187159628, + "learning_rate": 6.79932454825973e-06, + "loss": 0.0004, + "step": 1235 + }, + { + "epoch": 1.8502994011976048, + "grad_norm": 0.07863020430436107, + "learning_rate": 6.7840105281294485e-06, + "loss": 0.0004, + "step": 1236 + }, + { + "epoch": 1.8517964071856288, + "grad_norm": 0.044616260463696586, + "learning_rate": 6.768704915533852e-06, + "loss": 0.0004, + "step": 1237 + }, + { + "epoch": 1.8532934131736527, + "grad_norm": 0.12621641545858503, + "learning_rate": 6.753407750486284e-06, + "loss": 0.0017, + "step": 1238 + }, + { + "epoch": 1.8547904191616766, + "grad_norm": 0.060596994339849504, + "learning_rate": 6.738119072977996e-06, + "loss": 0.0006, + "step": 1239 + }, + { + "epoch": 1.8562874251497006, + "grad_norm": 0.014807449760783308, + "learning_rate": 6.722838922978046e-06, + "loss": 0.0002, + "step": 1240 + }, + { + "epoch": 1.8577844311377245, + "grad_norm": 0.0663517266563627, + "learning_rate": 6.707567340433212e-06, + "loss": 0.0006, + "step": 1241 + }, + { + "epoch": 1.8592814371257484, + "grad_norm": 0.1209263861163919, + "learning_rate": 6.692304365267859e-06, + "loss": 0.0008, + "step": 1242 + }, + { + "epoch": 1.8607784431137726, + "grad_norm": 0.027554525474874584, + "learning_rate": 6.677050037383858e-06, + "loss": 0.0003, + "step": 1243 + }, + { + "epoch": 1.8622754491017965, + "grad_norm": 0.04775088764673739, + "learning_rate": 6.661804396660479e-06, + "loss": 0.0005, + "step": 1244 + }, + { + "epoch": 1.8637724550898205, + "grad_norm": 0.056886026682939766, + "learning_rate": 6.646567482954263e-06, + "loss": 0.0003, + "step": 1245 + }, + { + "epoch": 1.8652694610778444, + "grad_norm": 0.13424552882664162, + "learning_rate": 6.631339336098956e-06, + "loss": 0.0013, + "step": 1246 + }, + { + "epoch": 1.8667664670658684, + "grad_norm": 0.12731628473965262, + "learning_rate": 6.616119995905375e-06, + "loss": 0.0012, + "step": 1247 + }, + { + "epoch": 1.8682634730538923, + "grad_norm": 0.05387273939125406, + "learning_rate": 6.600909502161315e-06, + "loss": 0.0003, + "step": 1248 + }, + { + "epoch": 1.8697604790419162, + "grad_norm": 0.11655593860565898, + "learning_rate": 6.585707894631441e-06, + "loss": 0.0008, + "step": 1249 + }, + { + "epoch": 1.8712574850299402, + "grad_norm": 0.03130014477728136, + "learning_rate": 6.570515213057192e-06, + "loss": 0.0002, + "step": 1250 + }, + { + "epoch": 1.872754491017964, + "grad_norm": 0.06789407097061075, + "learning_rate": 6.555331497156671e-06, + "loss": 0.0006, + "step": 1251 + }, + { + "epoch": 1.874251497005988, + "grad_norm": 0.1053736315966829, + "learning_rate": 6.5401567866245405e-06, + "loss": 0.0006, + "step": 1252 + }, + { + "epoch": 1.875748502994012, + "grad_norm": 0.042293639897476594, + "learning_rate": 6.5249911211319225e-06, + "loss": 0.0003, + "step": 1253 + }, + { + "epoch": 1.877245508982036, + "grad_norm": 0.08285464714326594, + "learning_rate": 6.509834540326285e-06, + "loss": 0.0004, + "step": 1254 + }, + { + "epoch": 1.8787425149700598, + "grad_norm": 0.11049018140607483, + "learning_rate": 6.494687083831362e-06, + "loss": 0.0009, + "step": 1255 + }, + { + "epoch": 1.8802395209580838, + "grad_norm": 0.01829057968863813, + "learning_rate": 6.47954879124701e-06, + "loss": 0.0001, + "step": 1256 + }, + { + "epoch": 1.8817365269461077, + "grad_norm": 0.033052338927285374, + "learning_rate": 6.4644197021491515e-06, + "loss": 0.0003, + "step": 1257 + }, + { + "epoch": 1.8832335329341316, + "grad_norm": 0.13174251290530642, + "learning_rate": 6.4492998560896395e-06, + "loss": 0.0013, + "step": 1258 + }, + { + "epoch": 1.8847305389221556, + "grad_norm": 0.11307650321461288, + "learning_rate": 6.434189292596158e-06, + "loss": 0.0009, + "step": 1259 + }, + { + "epoch": 1.8862275449101795, + "grad_norm": 0.039672940708096134, + "learning_rate": 6.419088051172133e-06, + "loss": 0.0003, + "step": 1260 + }, + { + "epoch": 1.8877245508982035, + "grad_norm": 0.0947573552544148, + "learning_rate": 6.40399617129661e-06, + "loss": 0.001, + "step": 1261 + }, + { + "epoch": 1.8892215568862274, + "grad_norm": 0.018111227856452812, + "learning_rate": 6.3889136924241704e-06, + "loss": 0.0002, + "step": 1262 + }, + { + "epoch": 1.8907185628742516, + "grad_norm": 0.06351616622351396, + "learning_rate": 6.373840653984811e-06, + "loss": 0.0005, + "step": 1263 + }, + { + "epoch": 1.8922155688622755, + "grad_norm": 0.08247922342515682, + "learning_rate": 6.35877709538385e-06, + "loss": 0.0006, + "step": 1264 + }, + { + "epoch": 1.8937125748502994, + "grad_norm": 0.1613355922056432, + "learning_rate": 6.3437230560018335e-06, + "loss": 0.0011, + "step": 1265 + }, + { + "epoch": 1.8952095808383234, + "grad_norm": 0.17460284441167057, + "learning_rate": 6.3286785751944e-06, + "loss": 0.0033, + "step": 1266 + }, + { + "epoch": 1.8967065868263473, + "grad_norm": 0.06534489203227545, + "learning_rate": 6.313643692292219e-06, + "loss": 0.0005, + "step": 1267 + }, + { + "epoch": 1.8982035928143712, + "grad_norm": 0.08598292764832548, + "learning_rate": 6.298618446600856e-06, + "loss": 0.0009, + "step": 1268 + }, + { + "epoch": 1.8997005988023952, + "grad_norm": 0.07498272399084252, + "learning_rate": 6.2836028774006945e-06, + "loss": 0.0005, + "step": 1269 + }, + { + "epoch": 1.9011976047904193, + "grad_norm": 0.07854073646784032, + "learning_rate": 6.268597023946801e-06, + "loss": 0.001, + "step": 1270 + }, + { + "epoch": 1.9026946107784433, + "grad_norm": 0.08325592956362902, + "learning_rate": 6.253600925468861e-06, + "loss": 0.0009, + "step": 1271 + }, + { + "epoch": 1.9041916167664672, + "grad_norm": 0.06261744481965902, + "learning_rate": 6.238614621171055e-06, + "loss": 0.0004, + "step": 1272 + }, + { + "epoch": 1.9056886227544911, + "grad_norm": 0.15054085004158702, + "learning_rate": 6.223638150231945e-06, + "loss": 0.0008, + "step": 1273 + }, + { + "epoch": 1.907185628742515, + "grad_norm": 0.15292361187679987, + "learning_rate": 6.208671551804404e-06, + "loss": 0.0009, + "step": 1274 + }, + { + "epoch": 1.908682634730539, + "grad_norm": 0.06302399358176151, + "learning_rate": 6.193714865015479e-06, + "loss": 0.0007, + "step": 1275 + }, + { + "epoch": 1.910179640718563, + "grad_norm": 0.08082819893731763, + "learning_rate": 6.178768128966319e-06, + "loss": 0.0011, + "step": 1276 + }, + { + "epoch": 1.9116766467065869, + "grad_norm": 0.10017565699972938, + "learning_rate": 6.163831382732046e-06, + "loss": 0.0005, + "step": 1277 + }, + { + "epoch": 1.9131736526946108, + "grad_norm": 0.1603341983015068, + "learning_rate": 6.148904665361677e-06, + "loss": 0.0008, + "step": 1278 + }, + { + "epoch": 1.9146706586826348, + "grad_norm": 0.07977074549410447, + "learning_rate": 6.133988015878005e-06, + "loss": 0.0003, + "step": 1279 + }, + { + "epoch": 1.9161676646706587, + "grad_norm": 0.08604266347649189, + "learning_rate": 6.119081473277502e-06, + "loss": 0.0015, + "step": 1280 + }, + { + "epoch": 1.9176646706586826, + "grad_norm": 0.07316072392265781, + "learning_rate": 6.104185076530224e-06, + "loss": 0.0019, + "step": 1281 + }, + { + "epoch": 1.9191616766467066, + "grad_norm": 0.07219162163636049, + "learning_rate": 6.0892988645796894e-06, + "loss": 0.0011, + "step": 1282 + }, + { + "epoch": 1.9206586826347305, + "grad_norm": 0.03498121695768265, + "learning_rate": 6.074422876342808e-06, + "loss": 0.0002, + "step": 1283 + }, + { + "epoch": 1.9221556886227544, + "grad_norm": 0.18083663339991857, + "learning_rate": 6.059557150709745e-06, + "loss": 0.0009, + "step": 1284 + }, + { + "epoch": 1.9236526946107784, + "grad_norm": 0.059187615256778826, + "learning_rate": 6.0447017265438464e-06, + "loss": 0.0004, + "step": 1285 + }, + { + "epoch": 1.9251497005988023, + "grad_norm": 0.12966709587781838, + "learning_rate": 6.029856642681528e-06, + "loss": 0.001, + "step": 1286 + }, + { + "epoch": 1.9266467065868262, + "grad_norm": 0.03319551814066764, + "learning_rate": 6.015021937932166e-06, + "loss": 0.0003, + "step": 1287 + }, + { + "epoch": 1.9281437125748502, + "grad_norm": 0.08693549001284917, + "learning_rate": 6.000197651078012e-06, + "loss": 0.0002, + "step": 1288 + }, + { + "epoch": 1.9296407185628741, + "grad_norm": 0.11435904368412542, + "learning_rate": 5.985383820874069e-06, + "loss": 0.0008, + "step": 1289 + }, + { + "epoch": 1.931137724550898, + "grad_norm": 0.03605212305635656, + "learning_rate": 5.970580486048016e-06, + "loss": 0.0002, + "step": 1290 + }, + { + "epoch": 1.9326347305389222, + "grad_norm": 0.07980050042842693, + "learning_rate": 5.955787685300085e-06, + "loss": 0.0005, + "step": 1291 + }, + { + "epoch": 1.9341317365269461, + "grad_norm": 0.11431738205279493, + "learning_rate": 5.941005457302975e-06, + "loss": 0.0007, + "step": 1292 + }, + { + "epoch": 1.93562874251497, + "grad_norm": 0.2420720686661468, + "learning_rate": 5.926233840701747e-06, + "loss": 0.0018, + "step": 1293 + }, + { + "epoch": 1.937125748502994, + "grad_norm": 0.17499882775445486, + "learning_rate": 5.911472874113709e-06, + "loss": 0.0015, + "step": 1294 + }, + { + "epoch": 1.938622754491018, + "grad_norm": 0.08874284785143925, + "learning_rate": 5.896722596128337e-06, + "loss": 0.0004, + "step": 1295 + }, + { + "epoch": 1.9401197604790419, + "grad_norm": 0.0714199007843593, + "learning_rate": 5.881983045307158e-06, + "loss": 0.0003, + "step": 1296 + }, + { + "epoch": 1.9416167664670658, + "grad_norm": 0.011510865246892025, + "learning_rate": 5.867254260183665e-06, + "loss": 0.0001, + "step": 1297 + }, + { + "epoch": 1.94311377245509, + "grad_norm": 0.10193764772516307, + "learning_rate": 5.85253627926319e-06, + "loss": 0.0006, + "step": 1298 + }, + { + "epoch": 1.944610778443114, + "grad_norm": 0.15050408491671105, + "learning_rate": 5.837829141022831e-06, + "loss": 0.0007, + "step": 1299 + }, + { + "epoch": 1.9461077844311379, + "grad_norm": 0.10898094115408612, + "learning_rate": 5.823132883911349e-06, + "loss": 0.0009, + "step": 1300 + }, + { + "epoch": 1.9476047904191618, + "grad_norm": 0.13535481273196698, + "learning_rate": 5.80844754634903e-06, + "loss": 0.001, + "step": 1301 + }, + { + "epoch": 1.9491017964071857, + "grad_norm": 0.10884632784883908, + "learning_rate": 5.793773166727648e-06, + "loss": 0.0012, + "step": 1302 + }, + { + "epoch": 1.9505988023952097, + "grad_norm": 0.06725195023294889, + "learning_rate": 5.779109783410295e-06, + "loss": 0.0004, + "step": 1303 + }, + { + "epoch": 1.9520958083832336, + "grad_norm": 0.03344699498459077, + "learning_rate": 5.764457434731345e-06, + "loss": 0.0003, + "step": 1304 + }, + { + "epoch": 1.9535928143712575, + "grad_norm": 0.07364922411512191, + "learning_rate": 5.749816158996309e-06, + "loss": 0.0012, + "step": 1305 + }, + { + "epoch": 1.9550898203592815, + "grad_norm": 0.1706215633793, + "learning_rate": 5.735185994481748e-06, + "loss": 0.0015, + "step": 1306 + }, + { + "epoch": 1.9565868263473054, + "grad_norm": 0.16089826885989925, + "learning_rate": 5.720566979435193e-06, + "loss": 0.0014, + "step": 1307 + }, + { + "epoch": 1.9580838323353293, + "grad_norm": 0.13197070265133728, + "learning_rate": 5.705959152074998e-06, + "loss": 0.0011, + "step": 1308 + }, + { + "epoch": 1.9595808383233533, + "grad_norm": 0.10737878089775195, + "learning_rate": 5.6913625505902966e-06, + "loss": 0.0018, + "step": 1309 + }, + { + "epoch": 1.9610778443113772, + "grad_norm": 0.09551442058751701, + "learning_rate": 5.6767772131408606e-06, + "loss": 0.0006, + "step": 1310 + }, + { + "epoch": 1.9625748502994012, + "grad_norm": 0.1495760754554237, + "learning_rate": 5.662203177857015e-06, + "loss": 0.0018, + "step": 1311 + }, + { + "epoch": 1.964071856287425, + "grad_norm": 0.030762250838208226, + "learning_rate": 5.647640482839543e-06, + "loss": 0.0003, + "step": 1312 + }, + { + "epoch": 1.965568862275449, + "grad_norm": 0.0627412308766797, + "learning_rate": 5.633089166159571e-06, + "loss": 0.0014, + "step": 1313 + }, + { + "epoch": 1.967065868263473, + "grad_norm": 0.08508365773836313, + "learning_rate": 5.618549265858499e-06, + "loss": 0.0008, + "step": 1314 + }, + { + "epoch": 1.968562874251497, + "grad_norm": 0.11170302861015123, + "learning_rate": 5.604020819947853e-06, + "loss": 0.0005, + "step": 1315 + }, + { + "epoch": 1.9700598802395208, + "grad_norm": 0.13346911790654048, + "learning_rate": 5.589503866409238e-06, + "loss": 0.0017, + "step": 1316 + }, + { + "epoch": 1.9715568862275448, + "grad_norm": 0.048207609953502435, + "learning_rate": 5.574998443194206e-06, + "loss": 0.0003, + "step": 1317 + }, + { + "epoch": 1.9730538922155687, + "grad_norm": 0.1790415818027207, + "learning_rate": 5.560504588224162e-06, + "loss": 0.0022, + "step": 1318 + }, + { + "epoch": 1.9745508982035929, + "grad_norm": 0.053657875706085344, + "learning_rate": 5.546022339390271e-06, + "loss": 0.0007, + "step": 1319 + }, + { + "epoch": 1.9760479041916168, + "grad_norm": 0.08409549221939727, + "learning_rate": 5.531551734553354e-06, + "loss": 0.0012, + "step": 1320 + }, + { + "epoch": 1.9775449101796407, + "grad_norm": 0.10964353914589475, + "learning_rate": 5.517092811543801e-06, + "loss": 0.0007, + "step": 1321 + }, + { + "epoch": 1.9790419161676647, + "grad_norm": 0.15383349246769662, + "learning_rate": 5.502645608161451e-06, + "loss": 0.0008, + "step": 1322 + }, + { + "epoch": 1.9805389221556886, + "grad_norm": 0.09629278498120927, + "learning_rate": 5.488210162175508e-06, + "loss": 0.0005, + "step": 1323 + }, + { + "epoch": 1.9820359281437125, + "grad_norm": 0.08906179596805544, + "learning_rate": 5.473786511324439e-06, + "loss": 0.0006, + "step": 1324 + }, + { + "epoch": 1.9835329341317365, + "grad_norm": 0.06933054825478269, + "learning_rate": 5.459374693315876e-06, + "loss": 0.0003, + "step": 1325 + }, + { + "epoch": 1.9850299401197606, + "grad_norm": 0.035068650708494103, + "learning_rate": 5.44497474582651e-06, + "loss": 0.0002, + "step": 1326 + }, + { + "epoch": 1.9865269461077846, + "grad_norm": 0.09362975516235468, + "learning_rate": 5.430586706502014e-06, + "loss": 0.001, + "step": 1327 + }, + { + "epoch": 1.9880239520958085, + "grad_norm": 0.06835759304910806, + "learning_rate": 5.416210612956918e-06, + "loss": 0.0009, + "step": 1328 + }, + { + "epoch": 1.9895209580838324, + "grad_norm": 0.019786881306191252, + "learning_rate": 5.401846502774522e-06, + "loss": 0.0001, + "step": 1329 + }, + { + "epoch": 1.9910179640718564, + "grad_norm": 0.1299238026590814, + "learning_rate": 5.3874944135068e-06, + "loss": 0.0015, + "step": 1330 + }, + { + "epoch": 1.9925149700598803, + "grad_norm": 0.14300823664119922, + "learning_rate": 5.373154382674301e-06, + "loss": 0.0009, + "step": 1331 + }, + { + "epoch": 1.9940119760479043, + "grad_norm": 0.09070831138488968, + "learning_rate": 5.3588264477660525e-06, + "loss": 0.0007, + "step": 1332 + }, + { + "epoch": 1.9955089820359282, + "grad_norm": 0.3279851722644398, + "learning_rate": 5.344510646239447e-06, + "loss": 0.0007, + "step": 1333 + }, + { + "epoch": 1.9970059880239521, + "grad_norm": 0.13371612237597238, + "learning_rate": 5.330207015520176e-06, + "loss": 0.0005, + "step": 1334 + }, + { + "epoch": 1.998502994011976, + "grad_norm": 0.05563881939066664, + "learning_rate": 5.3159155930021e-06, + "loss": 0.0004, + "step": 1335 + }, + { + "epoch": 2.0, + "grad_norm": 0.05524915070416573, + "learning_rate": 5.301636416047166e-06, + "loss": 0.0003, + "step": 1336 + }, + { + "epoch": 2.001497005988024, + "grad_norm": 0.078998616135492, + "learning_rate": 5.287369521985308e-06, + "loss": 0.0005, + "step": 1337 + }, + { + "epoch": 2.002994011976048, + "grad_norm": 0.11087032843271219, + "learning_rate": 5.273114948114346e-06, + "loss": 0.0006, + "step": 1338 + }, + { + "epoch": 2.004491017964072, + "grad_norm": 0.01134812052606357, + "learning_rate": 5.258872731699907e-06, + "loss": 0.0002, + "step": 1339 + }, + { + "epoch": 2.0059880239520957, + "grad_norm": 0.053992660508964285, + "learning_rate": 5.244642909975282e-06, + "loss": 0.0005, + "step": 1340 + }, + { + "epoch": 2.0074850299401197, + "grad_norm": 0.06896338730150291, + "learning_rate": 5.23042552014139e-06, + "loss": 0.0003, + "step": 1341 + }, + { + "epoch": 2.0089820359281436, + "grad_norm": 0.03331639781525915, + "learning_rate": 5.216220599366631e-06, + "loss": 0.0003, + "step": 1342 + }, + { + "epoch": 2.0104790419161676, + "grad_norm": 0.03048656266020426, + "learning_rate": 5.202028184786807e-06, + "loss": 0.0003, + "step": 1343 + }, + { + "epoch": 2.0119760479041915, + "grad_norm": 0.10036632313494923, + "learning_rate": 5.187848313505042e-06, + "loss": 0.0008, + "step": 1344 + }, + { + "epoch": 2.0134730538922154, + "grad_norm": 0.04333086369466123, + "learning_rate": 5.173681022591643e-06, + "loss": 0.0004, + "step": 1345 + }, + { + "epoch": 2.0149700598802394, + "grad_norm": 0.0999560008915293, + "learning_rate": 5.159526349084051e-06, + "loss": 0.0009, + "step": 1346 + }, + { + "epoch": 2.0164670658682633, + "grad_norm": 0.09927472161629344, + "learning_rate": 5.145384329986709e-06, + "loss": 0.0006, + "step": 1347 + }, + { + "epoch": 2.0179640718562872, + "grad_norm": 0.046997627545573885, + "learning_rate": 5.13125500227098e-06, + "loss": 0.0003, + "step": 1348 + }, + { + "epoch": 2.019461077844311, + "grad_norm": 0.11233978950395591, + "learning_rate": 5.117138402875048e-06, + "loss": 0.0011, + "step": 1349 + }, + { + "epoch": 2.020958083832335, + "grad_norm": 0.04511729239488089, + "learning_rate": 5.103034568703819e-06, + "loss": 0.0002, + "step": 1350 + }, + { + "epoch": 2.0224550898203595, + "grad_norm": 0.08572458482338798, + "learning_rate": 5.088943536628843e-06, + "loss": 0.0006, + "step": 1351 + }, + { + "epoch": 2.0239520958083834, + "grad_norm": 0.07982885735619828, + "learning_rate": 5.0748653434881735e-06, + "loss": 0.0009, + "step": 1352 + }, + { + "epoch": 2.0254491017964074, + "grad_norm": 0.05293702214000303, + "learning_rate": 5.060800026086322e-06, + "loss": 0.0004, + "step": 1353 + }, + { + "epoch": 2.0269461077844313, + "grad_norm": 0.10137109013871722, + "learning_rate": 5.046747621194132e-06, + "loss": 0.0004, + "step": 1354 + }, + { + "epoch": 2.0284431137724552, + "grad_norm": 0.17093193553427508, + "learning_rate": 5.032708165548682e-06, + "loss": 0.0007, + "step": 1355 + }, + { + "epoch": 2.029940119760479, + "grad_norm": 0.05939541972194821, + "learning_rate": 5.018681695853219e-06, + "loss": 0.0003, + "step": 1356 + }, + { + "epoch": 2.031437125748503, + "grad_norm": 0.05500060420421118, + "learning_rate": 5.00466824877701e-06, + "loss": 0.0003, + "step": 1357 + }, + { + "epoch": 2.032934131736527, + "grad_norm": 0.04529500584356061, + "learning_rate": 4.99066786095531e-06, + "loss": 0.0004, + "step": 1358 + }, + { + "epoch": 2.034431137724551, + "grad_norm": 0.1107914946140763, + "learning_rate": 4.976680568989203e-06, + "loss": 0.0008, + "step": 1359 + }, + { + "epoch": 2.035928143712575, + "grad_norm": 0.13392674083255784, + "learning_rate": 4.9627064094455594e-06, + "loss": 0.0009, + "step": 1360 + }, + { + "epoch": 2.037425149700599, + "grad_norm": 0.10073833115439051, + "learning_rate": 4.948745418856908e-06, + "loss": 0.0009, + "step": 1361 + }, + { + "epoch": 2.038922155688623, + "grad_norm": 0.06806623307714545, + "learning_rate": 4.934797633721343e-06, + "loss": 0.0003, + "step": 1362 + }, + { + "epoch": 2.0404191616766467, + "grad_norm": 0.08373463162355546, + "learning_rate": 4.92086309050246e-06, + "loss": 0.001, + "step": 1363 + }, + { + "epoch": 2.0419161676646707, + "grad_norm": 0.06828569633730956, + "learning_rate": 4.906941825629203e-06, + "loss": 0.0005, + "step": 1364 + }, + { + "epoch": 2.0434131736526946, + "grad_norm": 0.030176625061175792, + "learning_rate": 4.893033875495831e-06, + "loss": 0.0002, + "step": 1365 + }, + { + "epoch": 2.0449101796407185, + "grad_norm": 0.13266862164504664, + "learning_rate": 4.879139276461779e-06, + "loss": 0.0009, + "step": 1366 + }, + { + "epoch": 2.0464071856287425, + "grad_norm": 0.039122073400303665, + "learning_rate": 4.865258064851579e-06, + "loss": 0.0003, + "step": 1367 + }, + { + "epoch": 2.0479041916167664, + "grad_norm": 0.08056457623508984, + "learning_rate": 4.851390276954779e-06, + "loss": 0.001, + "step": 1368 + }, + { + "epoch": 2.0494011976047903, + "grad_norm": 0.12580066367234927, + "learning_rate": 4.837535949025807e-06, + "loss": 0.0007, + "step": 1369 + }, + { + "epoch": 2.0508982035928143, + "grad_norm": 0.1074019479468816, + "learning_rate": 4.823695117283929e-06, + "loss": 0.0007, + "step": 1370 + }, + { + "epoch": 2.052395209580838, + "grad_norm": 0.14807294780861768, + "learning_rate": 4.809867817913114e-06, + "loss": 0.0012, + "step": 1371 + }, + { + "epoch": 2.053892215568862, + "grad_norm": 0.04323366086908764, + "learning_rate": 4.796054087061955e-06, + "loss": 0.0003, + "step": 1372 + }, + { + "epoch": 2.055389221556886, + "grad_norm": 0.11724138562817155, + "learning_rate": 4.782253960843577e-06, + "loss": 0.0007, + "step": 1373 + }, + { + "epoch": 2.05688622754491, + "grad_norm": 0.0943855227419094, + "learning_rate": 4.768467475335531e-06, + "loss": 0.0007, + "step": 1374 + }, + { + "epoch": 2.058383233532934, + "grad_norm": 0.03896694150333076, + "learning_rate": 4.7546946665797215e-06, + "loss": 0.0003, + "step": 1375 + }, + { + "epoch": 2.059880239520958, + "grad_norm": 0.058883121619071395, + "learning_rate": 4.740935570582276e-06, + "loss": 0.0004, + "step": 1376 + }, + { + "epoch": 2.061377245508982, + "grad_norm": 0.11793570437876384, + "learning_rate": 4.7271902233134946e-06, + "loss": 0.001, + "step": 1377 + }, + { + "epoch": 2.062874251497006, + "grad_norm": 0.15311600066554015, + "learning_rate": 4.713458660707723e-06, + "loss": 0.0006, + "step": 1378 + }, + { + "epoch": 2.06437125748503, + "grad_norm": 0.04019764610619817, + "learning_rate": 4.699740918663271e-06, + "loss": 0.0002, + "step": 1379 + }, + { + "epoch": 2.065868263473054, + "grad_norm": 0.1371190868854585, + "learning_rate": 4.686037033042319e-06, + "loss": 0.0009, + "step": 1380 + }, + { + "epoch": 2.067365269461078, + "grad_norm": 0.19257550530901746, + "learning_rate": 4.672347039670817e-06, + "loss": 0.0022, + "step": 1381 + }, + { + "epoch": 2.068862275449102, + "grad_norm": 0.19645409716014256, + "learning_rate": 4.65867097433841e-06, + "loss": 0.0014, + "step": 1382 + }, + { + "epoch": 2.070359281437126, + "grad_norm": 0.05481103486692159, + "learning_rate": 4.645008872798318e-06, + "loss": 0.0002, + "step": 1383 + }, + { + "epoch": 2.07185628742515, + "grad_norm": 0.05062409226086992, + "learning_rate": 4.63136077076726e-06, + "loss": 0.0004, + "step": 1384 + }, + { + "epoch": 2.0733532934131738, + "grad_norm": 0.06611132196602185, + "learning_rate": 4.617726703925357e-06, + "loss": 0.0004, + "step": 1385 + }, + { + "epoch": 2.0748502994011977, + "grad_norm": 0.12704112517705748, + "learning_rate": 4.604106707916038e-06, + "loss": 0.0007, + "step": 1386 + }, + { + "epoch": 2.0763473053892216, + "grad_norm": 0.11551006878341365, + "learning_rate": 4.590500818345943e-06, + "loss": 0.0012, + "step": 1387 + }, + { + "epoch": 2.0778443113772456, + "grad_norm": 0.21126725853949585, + "learning_rate": 4.576909070784836e-06, + "loss": 0.001, + "step": 1388 + }, + { + "epoch": 2.0793413173652695, + "grad_norm": 0.04271579782342062, + "learning_rate": 4.5633315007655165e-06, + "loss": 0.0004, + "step": 1389 + }, + { + "epoch": 2.0808383233532934, + "grad_norm": 0.035707781270513814, + "learning_rate": 4.549768143783709e-06, + "loss": 0.0002, + "step": 1390 + }, + { + "epoch": 2.0823353293413174, + "grad_norm": 0.08987053581533556, + "learning_rate": 4.5362190352979875e-06, + "loss": 0.0006, + "step": 1391 + }, + { + "epoch": 2.0838323353293413, + "grad_norm": 0.022527899659366517, + "learning_rate": 4.522684210729673e-06, + "loss": 0.0002, + "step": 1392 + }, + { + "epoch": 2.0853293413173652, + "grad_norm": 0.09811043078353192, + "learning_rate": 4.509163705462746e-06, + "loss": 0.0006, + "step": 1393 + }, + { + "epoch": 2.086826347305389, + "grad_norm": 0.10474502941427184, + "learning_rate": 4.495657554843747e-06, + "loss": 0.0012, + "step": 1394 + }, + { + "epoch": 2.088323353293413, + "grad_norm": 0.05740099228948984, + "learning_rate": 4.482165794181702e-06, + "loss": 0.0003, + "step": 1395 + }, + { + "epoch": 2.089820359281437, + "grad_norm": 0.07473234668055903, + "learning_rate": 4.468688458748006e-06, + "loss": 0.0005, + "step": 1396 + }, + { + "epoch": 2.091317365269461, + "grad_norm": 0.034739985538301855, + "learning_rate": 4.455225583776346e-06, + "loss": 0.0003, + "step": 1397 + }, + { + "epoch": 2.092814371257485, + "grad_norm": 0.06386517981095234, + "learning_rate": 4.441777204462603e-06, + "loss": 0.0004, + "step": 1398 + }, + { + "epoch": 2.094311377245509, + "grad_norm": 0.02542788250908643, + "learning_rate": 4.4283433559647615e-06, + "loss": 0.0002, + "step": 1399 + }, + { + "epoch": 2.095808383233533, + "grad_norm": 0.05524457421603561, + "learning_rate": 4.4149240734028305e-06, + "loss": 0.0007, + "step": 1400 + }, + { + "epoch": 2.0973053892215567, + "grad_norm": 0.08145735954485667, + "learning_rate": 4.401519391858716e-06, + "loss": 0.0005, + "step": 1401 + }, + { + "epoch": 2.0988023952095807, + "grad_norm": 0.13001843472521485, + "learning_rate": 4.388129346376177e-06, + "loss": 0.0009, + "step": 1402 + }, + { + "epoch": 2.1002994011976046, + "grad_norm": 0.04328997396656813, + "learning_rate": 4.374753971960695e-06, + "loss": 0.0004, + "step": 1403 + }, + { + "epoch": 2.1017964071856285, + "grad_norm": 0.050812707704053536, + "learning_rate": 4.3613933035794e-06, + "loss": 0.0003, + "step": 1404 + }, + { + "epoch": 2.1032934131736525, + "grad_norm": 0.12459385212805167, + "learning_rate": 4.348047376160977e-06, + "loss": 0.0011, + "step": 1405 + }, + { + "epoch": 2.1047904191616764, + "grad_norm": 0.09492623023746591, + "learning_rate": 4.334716224595571e-06, + "loss": 0.0006, + "step": 1406 + }, + { + "epoch": 2.106287425149701, + "grad_norm": 0.07989452008235769, + "learning_rate": 4.321399883734712e-06, + "loss": 0.0005, + "step": 1407 + }, + { + "epoch": 2.1077844311377247, + "grad_norm": 0.07205507467214367, + "learning_rate": 4.308098388391184e-06, + "loss": 0.0004, + "step": 1408 + }, + { + "epoch": 2.1092814371257487, + "grad_norm": 0.017970693243257126, + "learning_rate": 4.2948117733389896e-06, + "loss": 0.0002, + "step": 1409 + }, + { + "epoch": 2.1107784431137726, + "grad_norm": 0.08565640149022209, + "learning_rate": 4.28154007331321e-06, + "loss": 0.0006, + "step": 1410 + }, + { + "epoch": 2.1122754491017965, + "grad_norm": 0.014993195128250375, + "learning_rate": 4.268283323009941e-06, + "loss": 0.0002, + "step": 1411 + }, + { + "epoch": 2.1137724550898205, + "grad_norm": 0.04396523431843643, + "learning_rate": 4.255041557086202e-06, + "loss": 0.0005, + "step": 1412 + }, + { + "epoch": 2.1152694610778444, + "grad_norm": 0.07072673143683754, + "learning_rate": 4.2418148101598215e-06, + "loss": 0.0004, + "step": 1413 + }, + { + "epoch": 2.1167664670658684, + "grad_norm": 0.055515949400535856, + "learning_rate": 4.228603116809382e-06, + "loss": 0.0006, + "step": 1414 + }, + { + "epoch": 2.1182634730538923, + "grad_norm": 0.05168252557459792, + "learning_rate": 4.215406511574104e-06, + "loss": 0.0005, + "step": 1415 + }, + { + "epoch": 2.1197604790419162, + "grad_norm": 0.11254802035543135, + "learning_rate": 4.202225028953758e-06, + "loss": 0.0013, + "step": 1416 + }, + { + "epoch": 2.12125748502994, + "grad_norm": 0.032013853884326086, + "learning_rate": 4.189058703408596e-06, + "loss": 0.0002, + "step": 1417 + }, + { + "epoch": 2.122754491017964, + "grad_norm": 0.0976156994347326, + "learning_rate": 4.175907569359219e-06, + "loss": 0.0007, + "step": 1418 + }, + { + "epoch": 2.124251497005988, + "grad_norm": 0.05995922146986086, + "learning_rate": 4.162771661186544e-06, + "loss": 0.0004, + "step": 1419 + }, + { + "epoch": 2.125748502994012, + "grad_norm": 0.1156120014758531, + "learning_rate": 4.149651013231651e-06, + "loss": 0.0007, + "step": 1420 + }, + { + "epoch": 2.127245508982036, + "grad_norm": 0.0728234683105191, + "learning_rate": 4.136545659795754e-06, + "loss": 0.0003, + "step": 1421 + }, + { + "epoch": 2.12874251497006, + "grad_norm": 0.04225258816552858, + "learning_rate": 4.123455635140066e-06, + "loss": 0.0002, + "step": 1422 + }, + { + "epoch": 2.1302395209580838, + "grad_norm": 0.07510640021056275, + "learning_rate": 4.110380973485726e-06, + "loss": 0.0007, + "step": 1423 + }, + { + "epoch": 2.1317365269461077, + "grad_norm": 0.11472823224242246, + "learning_rate": 4.097321709013725e-06, + "loss": 0.0009, + "step": 1424 + }, + { + "epoch": 2.1332335329341316, + "grad_norm": 0.04334119016243774, + "learning_rate": 4.084277875864776e-06, + "loss": 0.0002, + "step": 1425 + }, + { + "epoch": 2.1347305389221556, + "grad_norm": 0.03141663452074846, + "learning_rate": 4.071249508139275e-06, + "loss": 0.0003, + "step": 1426 + }, + { + "epoch": 2.1362275449101795, + "grad_norm": 0.06277111693317153, + "learning_rate": 4.058236639897171e-06, + "loss": 0.0004, + "step": 1427 + }, + { + "epoch": 2.1377245508982035, + "grad_norm": 0.016403479959175616, + "learning_rate": 4.0452393051579e-06, + "loss": 0.0001, + "step": 1428 + }, + { + "epoch": 2.1392215568862274, + "grad_norm": 0.07804687214286263, + "learning_rate": 4.032257537900286e-06, + "loss": 0.0006, + "step": 1429 + }, + { + "epoch": 2.1407185628742513, + "grad_norm": 0.07304022000926429, + "learning_rate": 4.019291372062452e-06, + "loss": 0.0009, + "step": 1430 + }, + { + "epoch": 2.1422155688622753, + "grad_norm": 0.04838062003714075, + "learning_rate": 4.006340841541748e-06, + "loss": 0.0004, + "step": 1431 + }, + { + "epoch": 2.143712574850299, + "grad_norm": 0.012920730564890766, + "learning_rate": 3.993405980194627e-06, + "loss": 0.0001, + "step": 1432 + }, + { + "epoch": 2.1452095808383236, + "grad_norm": 0.02060350725486784, + "learning_rate": 3.980486821836599e-06, + "loss": 0.0002, + "step": 1433 + }, + { + "epoch": 2.1467065868263475, + "grad_norm": 0.1090178395185067, + "learning_rate": 3.967583400242109e-06, + "loss": 0.0012, + "step": 1434 + }, + { + "epoch": 2.1482035928143715, + "grad_norm": 0.08624476103626773, + "learning_rate": 3.954695749144466e-06, + "loss": 0.0004, + "step": 1435 + }, + { + "epoch": 2.1497005988023954, + "grad_norm": 0.02414965108037553, + "learning_rate": 3.941823902235751e-06, + "loss": 0.0002, + "step": 1436 + }, + { + "epoch": 2.1511976047904193, + "grad_norm": 0.01602205069339523, + "learning_rate": 3.928967893166721e-06, + "loss": 0.0001, + "step": 1437 + }, + { + "epoch": 2.1526946107784433, + "grad_norm": 0.15078451246023045, + "learning_rate": 3.916127755546743e-06, + "loss": 0.0008, + "step": 1438 + }, + { + "epoch": 2.154191616766467, + "grad_norm": 0.09937561764681273, + "learning_rate": 3.903303522943679e-06, + "loss": 0.0005, + "step": 1439 + }, + { + "epoch": 2.155688622754491, + "grad_norm": 0.2778830186344834, + "learning_rate": 3.890495228883814e-06, + "loss": 0.0013, + "step": 1440 + }, + { + "epoch": 2.157185628742515, + "grad_norm": 0.1014703272227765, + "learning_rate": 3.877702906851765e-06, + "loss": 0.0011, + "step": 1441 + }, + { + "epoch": 2.158682634730539, + "grad_norm": 0.03910173601850433, + "learning_rate": 3.8649265902903935e-06, + "loss": 0.0003, + "step": 1442 + }, + { + "epoch": 2.160179640718563, + "grad_norm": 0.0758231387962015, + "learning_rate": 3.852166312600715e-06, + "loss": 0.0008, + "step": 1443 + }, + { + "epoch": 2.161676646706587, + "grad_norm": 0.0307677226066544, + "learning_rate": 3.839422107141826e-06, + "loss": 0.0002, + "step": 1444 + }, + { + "epoch": 2.163173652694611, + "grad_norm": 0.09752284917524724, + "learning_rate": 3.826694007230792e-06, + "loss": 0.0008, + "step": 1445 + }, + { + "epoch": 2.1646706586826348, + "grad_norm": 0.05457238739954349, + "learning_rate": 3.813982046142581e-06, + "loss": 0.0004, + "step": 1446 + }, + { + "epoch": 2.1661676646706587, + "grad_norm": 0.07303152617434286, + "learning_rate": 3.8012862571099673e-06, + "loss": 0.0005, + "step": 1447 + }, + { + "epoch": 2.1676646706586826, + "grad_norm": 0.016540715207413653, + "learning_rate": 3.788606673323447e-06, + "loss": 0.0002, + "step": 1448 + }, + { + "epoch": 2.1691616766467066, + "grad_norm": 0.0802709349097072, + "learning_rate": 3.7759433279311487e-06, + "loss": 0.0005, + "step": 1449 + }, + { + "epoch": 2.1706586826347305, + "grad_norm": 0.034645966560659984, + "learning_rate": 3.7632962540387575e-06, + "loss": 0.0003, + "step": 1450 + }, + { + "epoch": 2.1721556886227544, + "grad_norm": 0.008870689066133116, + "learning_rate": 3.7506654847094105e-06, + "loss": 0.0002, + "step": 1451 + }, + { + "epoch": 2.1736526946107784, + "grad_norm": 0.07045440620146966, + "learning_rate": 3.738051052963624e-06, + "loss": 0.0004, + "step": 1452 + }, + { + "epoch": 2.1751497005988023, + "grad_norm": 0.028807941828131546, + "learning_rate": 3.7254529917792025e-06, + "loss": 0.0002, + "step": 1453 + }, + { + "epoch": 2.1766467065868262, + "grad_norm": 0.04049065453847004, + "learning_rate": 3.712871334091154e-06, + "loss": 0.0004, + "step": 1454 + }, + { + "epoch": 2.17814371257485, + "grad_norm": 0.07895027308565043, + "learning_rate": 3.7003061127915973e-06, + "loss": 0.0005, + "step": 1455 + }, + { + "epoch": 2.179640718562874, + "grad_norm": 0.07783525776749904, + "learning_rate": 3.6877573607296935e-06, + "loss": 0.0007, + "step": 1456 + }, + { + "epoch": 2.181137724550898, + "grad_norm": 0.05173489235073667, + "learning_rate": 3.67522511071154e-06, + "loss": 0.0005, + "step": 1457 + }, + { + "epoch": 2.182634730538922, + "grad_norm": 0.04498118384069647, + "learning_rate": 3.6627093955000916e-06, + "loss": 0.0006, + "step": 1458 + }, + { + "epoch": 2.184131736526946, + "grad_norm": 0.09963326809445305, + "learning_rate": 3.6502102478150814e-06, + "loss": 0.0006, + "step": 1459 + }, + { + "epoch": 2.18562874251497, + "grad_norm": 0.020623105889192212, + "learning_rate": 3.6377277003329238e-06, + "loss": 0.0002, + "step": 1460 + }, + { + "epoch": 2.187125748502994, + "grad_norm": 0.0842055840763743, + "learning_rate": 3.625261785686649e-06, + "loss": 0.0006, + "step": 1461 + }, + { + "epoch": 2.1886227544910177, + "grad_norm": 0.05511784114857564, + "learning_rate": 3.6128125364657806e-06, + "loss": 0.0003, + "step": 1462 + }, + { + "epoch": 2.190119760479042, + "grad_norm": 0.09680779392796715, + "learning_rate": 3.6003799852163e-06, + "loss": 0.0006, + "step": 1463 + }, + { + "epoch": 2.191616766467066, + "grad_norm": 0.3290860685579756, + "learning_rate": 3.5879641644405183e-06, + "loss": 0.0016, + "step": 1464 + }, + { + "epoch": 2.19311377245509, + "grad_norm": 0.010421333234893785, + "learning_rate": 3.575565106597014e-06, + "loss": 0.0001, + "step": 1465 + }, + { + "epoch": 2.194610778443114, + "grad_norm": 0.15309336269175572, + "learning_rate": 3.563182844100541e-06, + "loss": 0.0011, + "step": 1466 + }, + { + "epoch": 2.196107784431138, + "grad_norm": 0.02091267883568856, + "learning_rate": 3.5508174093219427e-06, + "loss": 0.0001, + "step": 1467 + }, + { + "epoch": 2.197604790419162, + "grad_norm": 0.06446281945069611, + "learning_rate": 3.5384688345880826e-06, + "loss": 0.0004, + "step": 1468 + }, + { + "epoch": 2.1991017964071857, + "grad_norm": 0.052567869194541594, + "learning_rate": 3.5261371521817247e-06, + "loss": 0.0004, + "step": 1469 + }, + { + "epoch": 2.2005988023952097, + "grad_norm": 0.049578551613729545, + "learning_rate": 3.513822394341493e-06, + "loss": 0.0006, + "step": 1470 + }, + { + "epoch": 2.2020958083832336, + "grad_norm": 0.033783781354055624, + "learning_rate": 3.501524593261756e-06, + "loss": 0.0002, + "step": 1471 + }, + { + "epoch": 2.2035928143712575, + "grad_norm": 0.11505653536880667, + "learning_rate": 3.4892437810925484e-06, + "loss": 0.0008, + "step": 1472 + }, + { + "epoch": 2.2050898203592815, + "grad_norm": 0.05951967201522422, + "learning_rate": 3.476979989939506e-06, + "loss": 0.0002, + "step": 1473 + }, + { + "epoch": 2.2065868263473054, + "grad_norm": 0.1056924616624737, + "learning_rate": 3.4647332518637454e-06, + "loss": 0.0005, + "step": 1474 + }, + { + "epoch": 2.2080838323353293, + "grad_norm": 0.06831981985885952, + "learning_rate": 3.4525035988818232e-06, + "loss": 0.0007, + "step": 1475 + }, + { + "epoch": 2.2095808383233533, + "grad_norm": 0.08416246640850832, + "learning_rate": 3.440291062965608e-06, + "loss": 0.0004, + "step": 1476 + }, + { + "epoch": 2.211077844311377, + "grad_norm": 0.009399176800184218, + "learning_rate": 3.428095676042242e-06, + "loss": 0.0001, + "step": 1477 + }, + { + "epoch": 2.212574850299401, + "grad_norm": 0.06148287830535551, + "learning_rate": 3.41591746999402e-06, + "loss": 0.0005, + "step": 1478 + }, + { + "epoch": 2.214071856287425, + "grad_norm": 0.15044803640835408, + "learning_rate": 3.403756476658323e-06, + "loss": 0.0006, + "step": 1479 + }, + { + "epoch": 2.215568862275449, + "grad_norm": 0.039856311845892195, + "learning_rate": 3.3916127278275444e-06, + "loss": 0.0003, + "step": 1480 + }, + { + "epoch": 2.217065868263473, + "grad_norm": 0.03298698632561797, + "learning_rate": 3.379486255248975e-06, + "loss": 0.0004, + "step": 1481 + }, + { + "epoch": 2.218562874251497, + "grad_norm": 0.05989471368709794, + "learning_rate": 3.3673770906247595e-06, + "loss": 0.0003, + "step": 1482 + }, + { + "epoch": 2.220059880239521, + "grad_norm": 0.05031206889739225, + "learning_rate": 3.355285265611784e-06, + "loss": 0.0006, + "step": 1483 + }, + { + "epoch": 2.2215568862275448, + "grad_norm": 0.039637111322910513, + "learning_rate": 3.3432108118216046e-06, + "loss": 0.0004, + "step": 1484 + }, + { + "epoch": 2.2230538922155687, + "grad_norm": 0.02790884081359175, + "learning_rate": 3.3311537608203747e-06, + "loss": 0.0002, + "step": 1485 + }, + { + "epoch": 2.2245508982035926, + "grad_norm": 0.03455227197831783, + "learning_rate": 3.31911414412873e-06, + "loss": 0.0003, + "step": 1486 + }, + { + "epoch": 2.2260479041916166, + "grad_norm": 0.06635476538309946, + "learning_rate": 3.3070919932217517e-06, + "loss": 0.0003, + "step": 1487 + }, + { + "epoch": 2.2275449101796405, + "grad_norm": 0.14675486412500174, + "learning_rate": 3.2950873395288454e-06, + "loss": 0.0014, + "step": 1488 + }, + { + "epoch": 2.229041916167665, + "grad_norm": 0.1360738382194748, + "learning_rate": 3.2831002144336776e-06, + "loss": 0.0007, + "step": 1489 + }, + { + "epoch": 2.230538922155689, + "grad_norm": 0.05674441854621314, + "learning_rate": 3.2711306492740914e-06, + "loss": 0.0004, + "step": 1490 + }, + { + "epoch": 2.2320359281437128, + "grad_norm": 0.022436227700209574, + "learning_rate": 3.2591786753420187e-06, + "loss": 0.0001, + "step": 1491 + }, + { + "epoch": 2.2335329341317367, + "grad_norm": 0.03489642218056514, + "learning_rate": 3.247244323883415e-06, + "loss": 0.0003, + "step": 1492 + }, + { + "epoch": 2.2350299401197606, + "grad_norm": 0.2050412352519987, + "learning_rate": 3.2353276260981447e-06, + "loss": 0.0014, + "step": 1493 + }, + { + "epoch": 2.2365269461077846, + "grad_norm": 0.04952916174910589, + "learning_rate": 3.2234286131399415e-06, + "loss": 0.0003, + "step": 1494 + }, + { + "epoch": 2.2380239520958085, + "grad_norm": 0.15678094109556026, + "learning_rate": 3.2115473161162936e-06, + "loss": 0.0011, + "step": 1495 + }, + { + "epoch": 2.2395209580838324, + "grad_norm": 0.0359162039363321, + "learning_rate": 3.1996837660883763e-06, + "loss": 0.0003, + "step": 1496 + }, + { + "epoch": 2.2410179640718564, + "grad_norm": 0.12068999773483105, + "learning_rate": 3.1878379940709726e-06, + "loss": 0.0007, + "step": 1497 + }, + { + "epoch": 2.2425149700598803, + "grad_norm": 0.07153246649925311, + "learning_rate": 3.1760100310323815e-06, + "loss": 0.0013, + "step": 1498 + }, + { + "epoch": 2.2440119760479043, + "grad_norm": 0.02898441751748169, + "learning_rate": 3.1641999078943566e-06, + "loss": 0.0002, + "step": 1499 + }, + { + "epoch": 2.245508982035928, + "grad_norm": 0.11063959172977524, + "learning_rate": 3.1524076555320007e-06, + "loss": 0.0007, + "step": 1500 + }, + { + "epoch": 2.247005988023952, + "grad_norm": 0.10311560949120008, + "learning_rate": 3.1406333047737012e-06, + "loss": 0.0003, + "step": 1501 + }, + { + "epoch": 2.248502994011976, + "grad_norm": 0.09788152701017722, + "learning_rate": 3.128876886401049e-06, + "loss": 0.0003, + "step": 1502 + }, + { + "epoch": 2.25, + "grad_norm": 0.15031384544339235, + "learning_rate": 3.1171384311487487e-06, + "loss": 0.0012, + "step": 1503 + }, + { + "epoch": 2.251497005988024, + "grad_norm": 0.12304740690296884, + "learning_rate": 3.105417969704544e-06, + "loss": 0.0008, + "step": 1504 + }, + { + "epoch": 2.252994011976048, + "grad_norm": 0.10203703250907652, + "learning_rate": 3.093715532709147e-06, + "loss": 0.0005, + "step": 1505 + }, + { + "epoch": 2.254491017964072, + "grad_norm": 0.10315509055548146, + "learning_rate": 3.0820311507561395e-06, + "loss": 0.0007, + "step": 1506 + }, + { + "epoch": 2.2559880239520957, + "grad_norm": 0.1340732142733853, + "learning_rate": 3.0703648543919053e-06, + "loss": 0.0006, + "step": 1507 + }, + { + "epoch": 2.2574850299401197, + "grad_norm": 0.05868722622806031, + "learning_rate": 3.058716674115544e-06, + "loss": 0.0002, + "step": 1508 + }, + { + "epoch": 2.2589820359281436, + "grad_norm": 0.03557542415759387, + "learning_rate": 3.047086640378798e-06, + "loss": 0.0002, + "step": 1509 + }, + { + "epoch": 2.2604790419161676, + "grad_norm": 0.07662241536677104, + "learning_rate": 3.0354747835859686e-06, + "loss": 0.0004, + "step": 1510 + }, + { + "epoch": 2.2619760479041915, + "grad_norm": 0.04174555140373921, + "learning_rate": 3.0238811340938347e-06, + "loss": 0.0004, + "step": 1511 + }, + { + "epoch": 2.2634730538922154, + "grad_norm": 0.11334824278824039, + "learning_rate": 3.0123057222115835e-06, + "loss": 0.001, + "step": 1512 + }, + { + "epoch": 2.2649700598802394, + "grad_norm": 0.07557692002090421, + "learning_rate": 3.000748578200716e-06, + "loss": 0.0007, + "step": 1513 + }, + { + "epoch": 2.2664670658682633, + "grad_norm": 0.07097266734420278, + "learning_rate": 2.989209732274979e-06, + "loss": 0.0003, + "step": 1514 + }, + { + "epoch": 2.2679640718562872, + "grad_norm": 0.09527091457006696, + "learning_rate": 2.97768921460028e-06, + "loss": 0.0005, + "step": 1515 + }, + { + "epoch": 2.269461077844311, + "grad_norm": 0.0815127360640386, + "learning_rate": 2.9661870552946114e-06, + "loss": 0.0005, + "step": 1516 + }, + { + "epoch": 2.270958083832335, + "grad_norm": 0.0802974444453441, + "learning_rate": 2.9547032844279824e-06, + "loss": 0.0005, + "step": 1517 + }, + { + "epoch": 2.272455089820359, + "grad_norm": 0.04150750081228997, + "learning_rate": 2.943237932022307e-06, + "loss": 0.0002, + "step": 1518 + }, + { + "epoch": 2.2739520958083834, + "grad_norm": 0.031089773582056647, + "learning_rate": 2.9317910280513705e-06, + "loss": 0.0002, + "step": 1519 + }, + { + "epoch": 2.2754491017964074, + "grad_norm": 0.06361121560474334, + "learning_rate": 2.9203626024407185e-06, + "loss": 0.0004, + "step": 1520 + }, + { + "epoch": 2.2769461077844313, + "grad_norm": 0.05424769651819266, + "learning_rate": 2.9089526850675875e-06, + "loss": 0.0004, + "step": 1521 + }, + { + "epoch": 2.2784431137724552, + "grad_norm": 0.048908206558848324, + "learning_rate": 2.8975613057608297e-06, + "loss": 0.0008, + "step": 1522 + }, + { + "epoch": 2.279940119760479, + "grad_norm": 0.12535647237158476, + "learning_rate": 2.8861884943008335e-06, + "loss": 0.0005, + "step": 1523 + }, + { + "epoch": 2.281437125748503, + "grad_norm": 0.13595998922503674, + "learning_rate": 2.874834280419452e-06, + "loss": 0.0012, + "step": 1524 + }, + { + "epoch": 2.282934131736527, + "grad_norm": 0.0193676454435138, + "learning_rate": 2.863498693799901e-06, + "loss": 0.0002, + "step": 1525 + }, + { + "epoch": 2.284431137724551, + "grad_norm": 0.09642186052650302, + "learning_rate": 2.8521817640767225e-06, + "loss": 0.0004, + "step": 1526 + }, + { + "epoch": 2.285928143712575, + "grad_norm": 0.08690858061005316, + "learning_rate": 2.8408835208356667e-06, + "loss": 0.0006, + "step": 1527 + }, + { + "epoch": 2.287425149700599, + "grad_norm": 0.09629388756889663, + "learning_rate": 2.8296039936136355e-06, + "loss": 0.0005, + "step": 1528 + }, + { + "epoch": 2.288922155688623, + "grad_norm": 0.02465511736806249, + "learning_rate": 2.818343211898612e-06, + "loss": 0.0002, + "step": 1529 + }, + { + "epoch": 2.2904191616766467, + "grad_norm": 0.09994990956621684, + "learning_rate": 2.807101205129553e-06, + "loss": 0.0009, + "step": 1530 + }, + { + "epoch": 2.2919161676646707, + "grad_norm": 0.08966911947040915, + "learning_rate": 2.795878002696354e-06, + "loss": 0.0004, + "step": 1531 + }, + { + "epoch": 2.2934131736526946, + "grad_norm": 0.008082694473274558, + "learning_rate": 2.7846736339397362e-06, + "loss": 0.0001, + "step": 1532 + }, + { + "epoch": 2.2949101796407185, + "grad_norm": 0.037330647180128684, + "learning_rate": 2.773488128151186e-06, + "loss": 0.0002, + "step": 1533 + }, + { + "epoch": 2.2964071856287425, + "grad_norm": 0.15808722242025455, + "learning_rate": 2.7623215145728866e-06, + "loss": 0.001, + "step": 1534 + }, + { + "epoch": 2.2979041916167664, + "grad_norm": 0.10096871768612158, + "learning_rate": 2.7511738223976126e-06, + "loss": 0.0009, + "step": 1535 + }, + { + "epoch": 2.2994011976047903, + "grad_norm": 0.06280335460620405, + "learning_rate": 2.740045080768694e-06, + "loss": 0.0003, + "step": 1536 + }, + { + "epoch": 2.3008982035928143, + "grad_norm": 0.07328063719390991, + "learning_rate": 2.7289353187798982e-06, + "loss": 0.0004, + "step": 1537 + }, + { + "epoch": 2.302395209580838, + "grad_norm": 0.07595297277452129, + "learning_rate": 2.7178445654753916e-06, + "loss": 0.0005, + "step": 1538 + }, + { + "epoch": 2.303892215568862, + "grad_norm": 0.11974836266924484, + "learning_rate": 2.706772849849636e-06, + "loss": 0.0009, + "step": 1539 + }, + { + "epoch": 2.305389221556886, + "grad_norm": 0.0076400019711527585, + "learning_rate": 2.6957202008473238e-06, + "loss": 0.0001, + "step": 1540 + }, + { + "epoch": 2.30688622754491, + "grad_norm": 0.09483349610761937, + "learning_rate": 2.6846866473633126e-06, + "loss": 0.0005, + "step": 1541 + }, + { + "epoch": 2.308383233532934, + "grad_norm": 0.011742806672084745, + "learning_rate": 2.673672218242518e-06, + "loss": 0.0001, + "step": 1542 + }, + { + "epoch": 2.309880239520958, + "grad_norm": 0.027406703176830523, + "learning_rate": 2.662676942279879e-06, + "loss": 0.0002, + "step": 1543 + }, + { + "epoch": 2.3113772455089823, + "grad_norm": 0.04835437820138478, + "learning_rate": 2.651700848220253e-06, + "loss": 0.0003, + "step": 1544 + }, + { + "epoch": 2.312874251497006, + "grad_norm": 0.06402404819188177, + "learning_rate": 2.640743964758349e-06, + "loss": 0.0003, + "step": 1545 + }, + { + "epoch": 2.31437125748503, + "grad_norm": 0.06160368088836378, + "learning_rate": 2.629806320538665e-06, + "loss": 0.0002, + "step": 1546 + }, + { + "epoch": 2.315868263473054, + "grad_norm": 0.08575582684771083, + "learning_rate": 2.6188879441553818e-06, + "loss": 0.0005, + "step": 1547 + }, + { + "epoch": 2.317365269461078, + "grad_norm": 0.06630142016737109, + "learning_rate": 2.60798886415233e-06, + "loss": 0.0004, + "step": 1548 + }, + { + "epoch": 2.318862275449102, + "grad_norm": 0.12139207820757994, + "learning_rate": 2.5971091090228817e-06, + "loss": 0.0018, + "step": 1549 + }, + { + "epoch": 2.320359281437126, + "grad_norm": 0.06552332582410145, + "learning_rate": 2.5862487072098898e-06, + "loss": 0.0005, + "step": 1550 + }, + { + "epoch": 2.32185628742515, + "grad_norm": 0.04409020071690396, + "learning_rate": 2.5754076871056144e-06, + "loss": 0.0002, + "step": 1551 + }, + { + "epoch": 2.3233532934131738, + "grad_norm": 0.10364683798236486, + "learning_rate": 2.564586077051641e-06, + "loss": 0.0003, + "step": 1552 + }, + { + "epoch": 2.3248502994011977, + "grad_norm": 0.155282637159344, + "learning_rate": 2.5537839053388225e-06, + "loss": 0.0009, + "step": 1553 + }, + { + "epoch": 2.3263473053892216, + "grad_norm": 0.06569277881888244, + "learning_rate": 2.543001200207178e-06, + "loss": 0.0003, + "step": 1554 + }, + { + "epoch": 2.3278443113772456, + "grad_norm": 0.07710485032834485, + "learning_rate": 2.5322379898458527e-06, + "loss": 0.0005, + "step": 1555 + }, + { + "epoch": 2.3293413173652695, + "grad_norm": 0.07966793603491328, + "learning_rate": 2.5214943023930138e-06, + "loss": 0.0005, + "step": 1556 + }, + { + "epoch": 2.3308383233532934, + "grad_norm": 0.049815482522710106, + "learning_rate": 2.5107701659357963e-06, + "loss": 0.0002, + "step": 1557 + }, + { + "epoch": 2.3323353293413174, + "grad_norm": 0.11800603839640363, + "learning_rate": 2.5000656085102215e-06, + "loss": 0.0006, + "step": 1558 + }, + { + "epoch": 2.3338323353293413, + "grad_norm": 0.01937251400307602, + "learning_rate": 2.4893806581011215e-06, + "loss": 0.0002, + "step": 1559 + }, + { + "epoch": 2.3353293413173652, + "grad_norm": 0.05368059532814563, + "learning_rate": 2.4787153426420794e-06, + "loss": 0.0003, + "step": 1560 + }, + { + "epoch": 2.336826347305389, + "grad_norm": 0.14182083918266286, + "learning_rate": 2.46806969001534e-06, + "loss": 0.0009, + "step": 1561 + }, + { + "epoch": 2.338323353293413, + "grad_norm": 0.026485749870856008, + "learning_rate": 2.4574437280517437e-06, + "loss": 0.0003, + "step": 1562 + }, + { + "epoch": 2.339820359281437, + "grad_norm": 0.08886356114902899, + "learning_rate": 2.446837484530655e-06, + "loss": 0.0005, + "step": 1563 + }, + { + "epoch": 2.341317365269461, + "grad_norm": 0.02014395320911489, + "learning_rate": 2.436250987179889e-06, + "loss": 0.0001, + "step": 1564 + }, + { + "epoch": 2.342814371257485, + "grad_norm": 0.021864827664785275, + "learning_rate": 2.4256842636756372e-06, + "loss": 0.0002, + "step": 1565 + }, + { + "epoch": 2.344311377245509, + "grad_norm": 0.02967983647931681, + "learning_rate": 2.415137341642397e-06, + "loss": 0.0001, + "step": 1566 + }, + { + "epoch": 2.345808383233533, + "grad_norm": 0.06117231385528192, + "learning_rate": 2.4046102486529033e-06, + "loss": 0.0004, + "step": 1567 + }, + { + "epoch": 2.3473053892215567, + "grad_norm": 0.019912926187557994, + "learning_rate": 2.394103012228045e-06, + "loss": 0.0002, + "step": 1568 + }, + { + "epoch": 2.3488023952095807, + "grad_norm": 0.024535582706188355, + "learning_rate": 2.3836156598368055e-06, + "loss": 0.0003, + "step": 1569 + }, + { + "epoch": 2.3502994011976046, + "grad_norm": 0.03376993608094255, + "learning_rate": 2.373148218896182e-06, + "loss": 0.0003, + "step": 1570 + }, + { + "epoch": 2.3517964071856285, + "grad_norm": 0.0384213575217128, + "learning_rate": 2.362700716771118e-06, + "loss": 0.0005, + "step": 1571 + }, + { + "epoch": 2.3532934131736525, + "grad_norm": 0.046833073744106835, + "learning_rate": 2.3522731807744317e-06, + "loss": 0.0005, + "step": 1572 + }, + { + "epoch": 2.3547904191616764, + "grad_norm": 0.015881510742633775, + "learning_rate": 2.341865638166747e-06, + "loss": 0.0001, + "step": 1573 + }, + { + "epoch": 2.3562874251497004, + "grad_norm": 0.23245452164430105, + "learning_rate": 2.331478116156416e-06, + "loss": 0.0025, + "step": 1574 + }, + { + "epoch": 2.3577844311377247, + "grad_norm": 0.04796842296752511, + "learning_rate": 2.3211106418994512e-06, + "loss": 0.0002, + "step": 1575 + }, + { + "epoch": 2.3592814371257487, + "grad_norm": 0.11392590399503923, + "learning_rate": 2.3107632424994532e-06, + "loss": 0.0015, + "step": 1576 + }, + { + "epoch": 2.3607784431137726, + "grad_norm": 0.01691641893665238, + "learning_rate": 2.300435945007541e-06, + "loss": 0.0001, + "step": 1577 + }, + { + "epoch": 2.3622754491017965, + "grad_norm": 0.02404128261366343, + "learning_rate": 2.290128776422292e-06, + "loss": 0.0002, + "step": 1578 + }, + { + "epoch": 2.3637724550898205, + "grad_norm": 0.06174460774186989, + "learning_rate": 2.2798417636896396e-06, + "loss": 0.0012, + "step": 1579 + }, + { + "epoch": 2.3652694610778444, + "grad_norm": 0.08570446822273553, + "learning_rate": 2.269574933702846e-06, + "loss": 0.0004, + "step": 1580 + }, + { + "epoch": 2.3667664670658684, + "grad_norm": 0.0856077332705348, + "learning_rate": 2.259328313302395e-06, + "loss": 0.0005, + "step": 1581 + }, + { + "epoch": 2.3682634730538923, + "grad_norm": 0.07142666648166325, + "learning_rate": 2.249101929275944e-06, + "loss": 0.0009, + "step": 1582 + }, + { + "epoch": 2.3697604790419162, + "grad_norm": 0.025286357606488962, + "learning_rate": 2.2388958083582423e-06, + "loss": 0.0002, + "step": 1583 + }, + { + "epoch": 2.37125748502994, + "grad_norm": 0.045952040023591774, + "learning_rate": 2.228709977231065e-06, + "loss": 0.0003, + "step": 1584 + }, + { + "epoch": 2.372754491017964, + "grad_norm": 0.10954972663341103, + "learning_rate": 2.2185444625231557e-06, + "loss": 0.0007, + "step": 1585 + }, + { + "epoch": 2.374251497005988, + "grad_norm": 0.10645468456591504, + "learning_rate": 2.2083992908101236e-06, + "loss": 0.001, + "step": 1586 + }, + { + "epoch": 2.375748502994012, + "grad_norm": 0.030096418586315985, + "learning_rate": 2.1982744886144157e-06, + "loss": 0.0004, + "step": 1587 + }, + { + "epoch": 2.377245508982036, + "grad_norm": 0.09053718049959521, + "learning_rate": 2.1881700824052157e-06, + "loss": 0.0004, + "step": 1588 + }, + { + "epoch": 2.37874251497006, + "grad_norm": 0.04106460054597611, + "learning_rate": 2.1780860985983853e-06, + "loss": 0.0002, + "step": 1589 + }, + { + "epoch": 2.3802395209580838, + "grad_norm": 0.017813026852897713, + "learning_rate": 2.1680225635564113e-06, + "loss": 0.0002, + "step": 1590 + }, + { + "epoch": 2.3817365269461077, + "grad_norm": 0.03383373318851492, + "learning_rate": 2.1579795035882955e-06, + "loss": 0.0003, + "step": 1591 + }, + { + "epoch": 2.3832335329341316, + "grad_norm": 0.05074488961782638, + "learning_rate": 2.1479569449495406e-06, + "loss": 0.0004, + "step": 1592 + }, + { + "epoch": 2.3847305389221556, + "grad_norm": 0.01182856111698649, + "learning_rate": 2.1379549138420265e-06, + "loss": 0.0001, + "step": 1593 + }, + { + "epoch": 2.3862275449101795, + "grad_norm": 0.26118958396749364, + "learning_rate": 2.127973436413989e-06, + "loss": 0.0018, + "step": 1594 + }, + { + "epoch": 2.3877245508982035, + "grad_norm": 0.05278864289511616, + "learning_rate": 2.1180125387599193e-06, + "loss": 0.0003, + "step": 1595 + }, + { + "epoch": 2.3892215568862274, + "grad_norm": 0.030355323786991278, + "learning_rate": 2.108072246920506e-06, + "loss": 0.0002, + "step": 1596 + }, + { + "epoch": 2.3907185628742513, + "grad_norm": 0.1099369494009969, + "learning_rate": 2.0981525868825815e-06, + "loss": 0.0006, + "step": 1597 + }, + { + "epoch": 2.3922155688622753, + "grad_norm": 0.1203596546841281, + "learning_rate": 2.0882535845790188e-06, + "loss": 0.0007, + "step": 1598 + }, + { + "epoch": 2.3937125748502996, + "grad_norm": 0.1093149086475125, + "learning_rate": 2.078375265888707e-06, + "loss": 0.0011, + "step": 1599 + }, + { + "epoch": 2.3952095808383236, + "grad_norm": 0.10583209779661093, + "learning_rate": 2.0685176566364507e-06, + "loss": 0.0006, + "step": 1600 + }, + { + "epoch": 2.3967065868263475, + "grad_norm": 0.047336815910952124, + "learning_rate": 2.0586807825929123e-06, + "loss": 0.0002, + "step": 1601 + }, + { + "epoch": 2.3982035928143715, + "grad_norm": 0.09904610970762708, + "learning_rate": 2.0488646694745583e-06, + "loss": 0.0012, + "step": 1602 + }, + { + "epoch": 2.3997005988023954, + "grad_norm": 0.09513445282250857, + "learning_rate": 2.0390693429435626e-06, + "loss": 0.0004, + "step": 1603 + }, + { + "epoch": 2.4011976047904193, + "grad_norm": 0.020422046778283895, + "learning_rate": 2.029294828607773e-06, + "loss": 0.0001, + "step": 1604 + }, + { + "epoch": 2.4026946107784433, + "grad_norm": 0.14061343688126954, + "learning_rate": 2.01954115202062e-06, + "loss": 0.0012, + "step": 1605 + }, + { + "epoch": 2.404191616766467, + "grad_norm": 0.04251232089992607, + "learning_rate": 2.0098083386810596e-06, + "loss": 0.0002, + "step": 1606 + }, + { + "epoch": 2.405688622754491, + "grad_norm": 0.05855588269503358, + "learning_rate": 2.0000964140335044e-06, + "loss": 0.0004, + "step": 1607 + }, + { + "epoch": 2.407185628742515, + "grad_norm": 0.04678849099056471, + "learning_rate": 1.990405403467757e-06, + "loss": 0.0002, + "step": 1608 + }, + { + "epoch": 2.408682634730539, + "grad_norm": 0.15089209688583194, + "learning_rate": 1.980735332318954e-06, + "loss": 0.0005, + "step": 1609 + }, + { + "epoch": 2.410179640718563, + "grad_norm": 0.04533706932582074, + "learning_rate": 1.9710862258674714e-06, + "loss": 0.0002, + "step": 1610 + }, + { + "epoch": 2.411676646706587, + "grad_norm": 0.05696073871947218, + "learning_rate": 1.9614581093388964e-06, + "loss": 0.0004, + "step": 1611 + }, + { + "epoch": 2.413173652694611, + "grad_norm": 0.044373523736829394, + "learning_rate": 1.951851007903931e-06, + "loss": 0.0003, + "step": 1612 + }, + { + "epoch": 2.4146706586826348, + "grad_norm": 0.1647299050829959, + "learning_rate": 1.9422649466783427e-06, + "loss": 0.0009, + "step": 1613 + }, + { + "epoch": 2.4161676646706587, + "grad_norm": 0.08092462607166259, + "learning_rate": 1.9326999507228916e-06, + "loss": 0.0004, + "step": 1614 + }, + { + "epoch": 2.4176646706586826, + "grad_norm": 0.08726562462534888, + "learning_rate": 1.923156045043265e-06, + "loss": 0.0007, + "step": 1615 + }, + { + "epoch": 2.4191616766467066, + "grad_norm": 0.0684185106844218, + "learning_rate": 1.9136332545900215e-06, + "loss": 0.0013, + "step": 1616 + }, + { + "epoch": 2.4206586826347305, + "grad_norm": 0.028069249948623098, + "learning_rate": 1.9041316042585113e-06, + "loss": 0.0002, + "step": 1617 + }, + { + "epoch": 2.4221556886227544, + "grad_norm": 0.024791281262007244, + "learning_rate": 1.8946511188888194e-06, + "loss": 0.0002, + "step": 1618 + }, + { + "epoch": 2.4236526946107784, + "grad_norm": 0.07988190689532107, + "learning_rate": 1.8851918232657007e-06, + "loss": 0.0005, + "step": 1619 + }, + { + "epoch": 2.4251497005988023, + "grad_norm": 0.027087754543380946, + "learning_rate": 1.8757537421185145e-06, + "loss": 0.0002, + "step": 1620 + }, + { + "epoch": 2.4266467065868262, + "grad_norm": 0.030964818451742402, + "learning_rate": 1.8663369001211552e-06, + "loss": 0.0002, + "step": 1621 + }, + { + "epoch": 2.42814371257485, + "grad_norm": 0.042357471381083395, + "learning_rate": 1.8569413218919986e-06, + "loss": 0.0002, + "step": 1622 + }, + { + "epoch": 2.429640718562874, + "grad_norm": 0.050604316712189235, + "learning_rate": 1.8475670319938266e-06, + "loss": 0.0005, + "step": 1623 + }, + { + "epoch": 2.431137724550898, + "grad_norm": 0.05075746766744105, + "learning_rate": 1.8382140549337678e-06, + "loss": 0.0004, + "step": 1624 + }, + { + "epoch": 2.432634730538922, + "grad_norm": 0.06111785771892014, + "learning_rate": 1.8288824151632311e-06, + "loss": 0.0003, + "step": 1625 + }, + { + "epoch": 2.434131736526946, + "grad_norm": 0.07080928006997747, + "learning_rate": 1.8195721370778474e-06, + "loss": 0.0007, + "step": 1626 + }, + { + "epoch": 2.43562874251497, + "grad_norm": 0.027992654625571187, + "learning_rate": 1.810283245017398e-06, + "loss": 0.0002, + "step": 1627 + }, + { + "epoch": 2.437125748502994, + "grad_norm": 0.03162118910496192, + "learning_rate": 1.8010157632657544e-06, + "loss": 0.0002, + "step": 1628 + }, + { + "epoch": 2.4386227544910177, + "grad_norm": 0.0662498938932939, + "learning_rate": 1.791769716050823e-06, + "loss": 0.0005, + "step": 1629 + }, + { + "epoch": 2.440119760479042, + "grad_norm": 0.06063032513396334, + "learning_rate": 1.782545127544466e-06, + "loss": 0.0015, + "step": 1630 + }, + { + "epoch": 2.441616766467066, + "grad_norm": 0.031943908561759334, + "learning_rate": 1.7733420218624486e-06, + "loss": 0.0002, + "step": 1631 + }, + { + "epoch": 2.44311377245509, + "grad_norm": 0.10126704114641819, + "learning_rate": 1.7641604230643738e-06, + "loss": 0.0011, + "step": 1632 + }, + { + "epoch": 2.444610778443114, + "grad_norm": 0.03485971575240973, + "learning_rate": 1.7550003551536177e-06, + "loss": 0.0003, + "step": 1633 + }, + { + "epoch": 2.446107784431138, + "grad_norm": 0.05038807899416904, + "learning_rate": 1.7458618420772777e-06, + "loss": 0.0004, + "step": 1634 + }, + { + "epoch": 2.447604790419162, + "grad_norm": 0.034427102319247505, + "learning_rate": 1.7367449077260823e-06, + "loss": 0.0001, + "step": 1635 + }, + { + "epoch": 2.4491017964071857, + "grad_norm": 0.040602438643352785, + "learning_rate": 1.7276495759343647e-06, + "loss": 0.0003, + "step": 1636 + }, + { + "epoch": 2.4505988023952097, + "grad_norm": 0.0300674690331923, + "learning_rate": 1.7185758704799738e-06, + "loss": 0.0002, + "step": 1637 + }, + { + "epoch": 2.4520958083832336, + "grad_norm": 0.04412630684877554, + "learning_rate": 1.709523815084222e-06, + "loss": 0.0002, + "step": 1638 + }, + { + "epoch": 2.4535928143712575, + "grad_norm": 0.09352208849724938, + "learning_rate": 1.700493433411825e-06, + "loss": 0.0008, + "step": 1639 + }, + { + "epoch": 2.4550898203592815, + "grad_norm": 0.09616189622897968, + "learning_rate": 1.6914847490708296e-06, + "loss": 0.0006, + "step": 1640 + }, + { + "epoch": 2.4565868263473054, + "grad_norm": 0.04164673607392628, + "learning_rate": 1.6824977856125745e-06, + "loss": 0.0003, + "step": 1641 + }, + { + "epoch": 2.4580838323353293, + "grad_norm": 0.02769372221761515, + "learning_rate": 1.6735325665315927e-06, + "loss": 0.0002, + "step": 1642 + }, + { + "epoch": 2.4595808383233533, + "grad_norm": 0.03785938337316224, + "learning_rate": 1.6645891152655903e-06, + "loss": 0.0002, + "step": 1643 + }, + { + "epoch": 2.461077844311377, + "grad_norm": 0.07580082376824222, + "learning_rate": 1.6556674551953555e-06, + "loss": 0.0009, + "step": 1644 + }, + { + "epoch": 2.462574850299401, + "grad_norm": 0.03577911495005629, + "learning_rate": 1.6467676096447083e-06, + "loss": 0.0001, + "step": 1645 + }, + { + "epoch": 2.464071856287425, + "grad_norm": 0.04888563503909771, + "learning_rate": 1.637889601880448e-06, + "loss": 0.0003, + "step": 1646 + }, + { + "epoch": 2.465568862275449, + "grad_norm": 0.028629948760254195, + "learning_rate": 1.6290334551122678e-06, + "loss": 0.0002, + "step": 1647 + }, + { + "epoch": 2.467065868263473, + "grad_norm": 0.09745710567212712, + "learning_rate": 1.6201991924927252e-06, + "loss": 0.0008, + "step": 1648 + }, + { + "epoch": 2.468562874251497, + "grad_norm": 0.007786179248720061, + "learning_rate": 1.6113868371171592e-06, + "loss": 0.0001, + "step": 1649 + }, + { + "epoch": 2.470059880239521, + "grad_norm": 0.09307414131421253, + "learning_rate": 1.6025964120236337e-06, + "loss": 0.0003, + "step": 1650 + }, + { + "epoch": 2.4715568862275448, + "grad_norm": 0.08000883258976127, + "learning_rate": 1.5938279401928925e-06, + "loss": 0.0007, + "step": 1651 + }, + { + "epoch": 2.4730538922155687, + "grad_norm": 0.08529740066647153, + "learning_rate": 1.58508144454827e-06, + "loss": 0.0006, + "step": 1652 + }, + { + "epoch": 2.4745508982035926, + "grad_norm": 0.1732944945698624, + "learning_rate": 1.576356947955665e-06, + "loss": 0.0007, + "step": 1653 + }, + { + "epoch": 2.4760479041916166, + "grad_norm": 0.05910999084092352, + "learning_rate": 1.5676544732234477e-06, + "loss": 0.0004, + "step": 1654 + }, + { + "epoch": 2.477544910179641, + "grad_norm": 0.03736554597645004, + "learning_rate": 1.5589740431024347e-06, + "loss": 0.0003, + "step": 1655 + }, + { + "epoch": 2.479041916167665, + "grad_norm": 0.05815438383385881, + "learning_rate": 1.5503156802857998e-06, + "loss": 0.0003, + "step": 1656 + }, + { + "epoch": 2.480538922155689, + "grad_norm": 0.009680132499123993, + "learning_rate": 1.5416794074090258e-06, + "loss": 0.0001, + "step": 1657 + }, + { + "epoch": 2.4820359281437128, + "grad_norm": 0.043831899612723345, + "learning_rate": 1.5330652470498574e-06, + "loss": 0.0003, + "step": 1658 + }, + { + "epoch": 2.4835329341317367, + "grad_norm": 0.0208540727997247, + "learning_rate": 1.5244732217282132e-06, + "loss": 0.0001, + "step": 1659 + }, + { + "epoch": 2.4850299401197606, + "grad_norm": 0.02609799149850617, + "learning_rate": 1.5159033539061607e-06, + "loss": 0.0002, + "step": 1660 + }, + { + "epoch": 2.4865269461077846, + "grad_norm": 0.012809457260665172, + "learning_rate": 1.507355665987833e-06, + "loss": 0.0001, + "step": 1661 + }, + { + "epoch": 2.4880239520958085, + "grad_norm": 0.03092643008061081, + "learning_rate": 1.4988301803193773e-06, + "loss": 0.0002, + "step": 1662 + }, + { + "epoch": 2.4895209580838324, + "grad_norm": 0.11021032771407564, + "learning_rate": 1.4903269191889068e-06, + "loss": 0.0008, + "step": 1663 + }, + { + "epoch": 2.4910179640718564, + "grad_norm": 0.14645604895626768, + "learning_rate": 1.4818459048264179e-06, + "loss": 0.0008, + "step": 1664 + }, + { + "epoch": 2.4925149700598803, + "grad_norm": 0.08473709608733745, + "learning_rate": 1.4733871594037629e-06, + "loss": 0.0005, + "step": 1665 + }, + { + "epoch": 2.4940119760479043, + "grad_norm": 0.28509040224007653, + "learning_rate": 1.464950705034568e-06, + "loss": 0.0012, + "step": 1666 + }, + { + "epoch": 2.495508982035928, + "grad_norm": 0.024381609398717058, + "learning_rate": 1.4565365637741868e-06, + "loss": 0.0001, + "step": 1667 + }, + { + "epoch": 2.497005988023952, + "grad_norm": 0.09142786032282342, + "learning_rate": 1.4481447576196372e-06, + "loss": 0.0013, + "step": 1668 + }, + { + "epoch": 2.498502994011976, + "grad_norm": 0.09847239340356413, + "learning_rate": 1.439775308509549e-06, + "loss": 0.0007, + "step": 1669 + }, + { + "epoch": 2.5, + "grad_norm": 0.03477841977609638, + "learning_rate": 1.4314282383241097e-06, + "loss": 0.0002, + "step": 1670 + }, + { + "epoch": 2.501497005988024, + "grad_norm": 0.04628201955871934, + "learning_rate": 1.423103568884986e-06, + "loss": 0.0003, + "step": 1671 + }, + { + "epoch": 2.502994011976048, + "grad_norm": 0.09307706909626917, + "learning_rate": 1.4148013219552992e-06, + "loss": 0.0007, + "step": 1672 + }, + { + "epoch": 2.504491017964072, + "grad_norm": 0.07777395668753506, + "learning_rate": 1.406521519239543e-06, + "loss": 0.0004, + "step": 1673 + }, + { + "epoch": 2.5059880239520957, + "grad_norm": 0.038725094614758614, + "learning_rate": 1.3982641823835385e-06, + "loss": 0.0002, + "step": 1674 + }, + { + "epoch": 2.5074850299401197, + "grad_norm": 0.09347821247400721, + "learning_rate": 1.3900293329743707e-06, + "loss": 0.0008, + "step": 1675 + }, + { + "epoch": 2.5089820359281436, + "grad_norm": 0.011973171492511713, + "learning_rate": 1.381816992540338e-06, + "loss": 0.0001, + "step": 1676 + }, + { + "epoch": 2.5104790419161676, + "grad_norm": 0.03349515304688116, + "learning_rate": 1.3736271825508973e-06, + "loss": 0.0002, + "step": 1677 + }, + { + "epoch": 2.5119760479041915, + "grad_norm": 0.01810852300179121, + "learning_rate": 1.3654599244166e-06, + "loss": 0.0001, + "step": 1678 + }, + { + "epoch": 2.5134730538922154, + "grad_norm": 0.04446933710711642, + "learning_rate": 1.3573152394890398e-06, + "loss": 0.0003, + "step": 1679 + }, + { + "epoch": 2.5149700598802394, + "grad_norm": 0.05815871825675632, + "learning_rate": 1.3491931490608023e-06, + "loss": 0.0002, + "step": 1680 + }, + { + "epoch": 2.5164670658682633, + "grad_norm": 0.13415081479279004, + "learning_rate": 1.341093674365398e-06, + "loss": 0.0017, + "step": 1681 + }, + { + "epoch": 2.5179640718562872, + "grad_norm": 0.026418150113077105, + "learning_rate": 1.333016836577219e-06, + "loss": 0.0002, + "step": 1682 + }, + { + "epoch": 2.519461077844311, + "grad_norm": 0.05637402946123024, + "learning_rate": 1.3249626568114716e-06, + "loss": 0.0004, + "step": 1683 + }, + { + "epoch": 2.520958083832335, + "grad_norm": 0.03094599306924757, + "learning_rate": 1.3169311561241371e-06, + "loss": 0.0002, + "step": 1684 + }, + { + "epoch": 2.522455089820359, + "grad_norm": 0.027335429413912742, + "learning_rate": 1.308922355511899e-06, + "loss": 0.0002, + "step": 1685 + }, + { + "epoch": 2.523952095808383, + "grad_norm": 0.1488861125220278, + "learning_rate": 1.300936275912098e-06, + "loss": 0.001, + "step": 1686 + }, + { + "epoch": 2.525449101796407, + "grad_norm": 0.05818953193532018, + "learning_rate": 1.2929729382026757e-06, + "loss": 0.0004, + "step": 1687 + }, + { + "epoch": 2.5269461077844313, + "grad_norm": 0.07088482664758781, + "learning_rate": 1.285032363202121e-06, + "loss": 0.0007, + "step": 1688 + }, + { + "epoch": 2.5284431137724552, + "grad_norm": 0.09878385939626243, + "learning_rate": 1.277114571669411e-06, + "loss": 0.0005, + "step": 1689 + }, + { + "epoch": 2.529940119760479, + "grad_norm": 0.08834537093128993, + "learning_rate": 1.2692195843039668e-06, + "loss": 0.0005, + "step": 1690 + }, + { + "epoch": 2.531437125748503, + "grad_norm": 0.028045651170751838, + "learning_rate": 1.2613474217455856e-06, + "loss": 0.0004, + "step": 1691 + }, + { + "epoch": 2.532934131736527, + "grad_norm": 0.08064194210729632, + "learning_rate": 1.2534981045743987e-06, + "loss": 0.0003, + "step": 1692 + }, + { + "epoch": 2.534431137724551, + "grad_norm": 0.06878912490099846, + "learning_rate": 1.2456716533108094e-06, + "loss": 0.0003, + "step": 1693 + }, + { + "epoch": 2.535928143712575, + "grad_norm": 0.09475527318363501, + "learning_rate": 1.2378680884154437e-06, + "loss": 0.0009, + "step": 1694 + }, + { + "epoch": 2.537425149700599, + "grad_norm": 0.06988351944817416, + "learning_rate": 1.230087430289103e-06, + "loss": 0.0005, + "step": 1695 + }, + { + "epoch": 2.538922155688623, + "grad_norm": 0.06025644551142922, + "learning_rate": 1.2223296992726885e-06, + "loss": 0.0003, + "step": 1696 + }, + { + "epoch": 2.5404191616766467, + "grad_norm": 0.05586088137246275, + "learning_rate": 1.2145949156471781e-06, + "loss": 0.0003, + "step": 1697 + }, + { + "epoch": 2.5419161676646707, + "grad_norm": 0.019554709115490496, + "learning_rate": 1.2068830996335512e-06, + "loss": 0.0001, + "step": 1698 + }, + { + "epoch": 2.5434131736526946, + "grad_norm": 0.060450587058907566, + "learning_rate": 1.1991942713927452e-06, + "loss": 0.0003, + "step": 1699 + }, + { + "epoch": 2.5449101796407185, + "grad_norm": 0.05151733193244522, + "learning_rate": 1.1915284510255998e-06, + "loss": 0.0003, + "step": 1700 + }, + { + "epoch": 2.5464071856287425, + "grad_norm": 0.08154849069751298, + "learning_rate": 1.1838856585728042e-06, + "loss": 0.0008, + "step": 1701 + }, + { + "epoch": 2.5479041916167664, + "grad_norm": 0.025258520621501784, + "learning_rate": 1.1762659140148524e-06, + "loss": 0.0002, + "step": 1702 + }, + { + "epoch": 2.5494011976047903, + "grad_norm": 0.08809425431554686, + "learning_rate": 1.1686692372719721e-06, + "loss": 0.0006, + "step": 1703 + }, + { + "epoch": 2.5508982035928143, + "grad_norm": 0.09231116438170844, + "learning_rate": 1.1610956482040992e-06, + "loss": 0.0005, + "step": 1704 + }, + { + "epoch": 2.552395209580838, + "grad_norm": 0.020971139154724327, + "learning_rate": 1.1535451666108022e-06, + "loss": 0.0002, + "step": 1705 + }, + { + "epoch": 2.553892215568862, + "grad_norm": 0.07893486730149059, + "learning_rate": 1.1460178122312404e-06, + "loss": 0.0006, + "step": 1706 + }, + { + "epoch": 2.555389221556886, + "grad_norm": 0.019260558530191367, + "learning_rate": 1.1385136047441214e-06, + "loss": 0.0001, + "step": 1707 + }, + { + "epoch": 2.55688622754491, + "grad_norm": 0.022801882208881713, + "learning_rate": 1.1310325637676233e-06, + "loss": 0.0002, + "step": 1708 + }, + { + "epoch": 2.5583832335329344, + "grad_norm": 0.006273256765097044, + "learning_rate": 1.1235747088593773e-06, + "loss": 0.0001, + "step": 1709 + }, + { + "epoch": 2.5598802395209583, + "grad_norm": 0.025272286951359548, + "learning_rate": 1.1161400595163885e-06, + "loss": 0.0002, + "step": 1710 + }, + { + "epoch": 2.5613772455089823, + "grad_norm": 0.06144529667834819, + "learning_rate": 1.1087286351749981e-06, + "loss": 0.0005, + "step": 1711 + }, + { + "epoch": 2.562874251497006, + "grad_norm": 0.018647360603314567, + "learning_rate": 1.1013404552108365e-06, + "loss": 0.0001, + "step": 1712 + }, + { + "epoch": 2.56437125748503, + "grad_norm": 0.040545615112091975, + "learning_rate": 1.0939755389387552e-06, + "loss": 0.0003, + "step": 1713 + }, + { + "epoch": 2.565868263473054, + "grad_norm": 0.024320530278532197, + "learning_rate": 1.0866339056127995e-06, + "loss": 0.0002, + "step": 1714 + }, + { + "epoch": 2.567365269461078, + "grad_norm": 0.092450238455231, + "learning_rate": 1.0793155744261352e-06, + "loss": 0.0005, + "step": 1715 + }, + { + "epoch": 2.568862275449102, + "grad_norm": 0.05655287368538696, + "learning_rate": 1.0720205645110193e-06, + "loss": 0.0004, + "step": 1716 + }, + { + "epoch": 2.570359281437126, + "grad_norm": 0.08988286767798637, + "learning_rate": 1.064748894938734e-06, + "loss": 0.0007, + "step": 1717 + }, + { + "epoch": 2.57185628742515, + "grad_norm": 0.07500638380677978, + "learning_rate": 1.0575005847195442e-06, + "loss": 0.0004, + "step": 1718 + }, + { + "epoch": 2.5733532934131738, + "grad_norm": 0.0925835650680482, + "learning_rate": 1.0502756528026526e-06, + "loss": 0.0005, + "step": 1719 + }, + { + "epoch": 2.5748502994011977, + "grad_norm": 0.06643680023971794, + "learning_rate": 1.0430741180761306e-06, + "loss": 0.0004, + "step": 1720 + }, + { + "epoch": 2.5763473053892216, + "grad_norm": 0.07290249607472542, + "learning_rate": 1.0358959993668981e-06, + "loss": 0.0005, + "step": 1721 + }, + { + "epoch": 2.5778443113772456, + "grad_norm": 0.01775705006856804, + "learning_rate": 1.0287413154406478e-06, + "loss": 0.0003, + "step": 1722 + }, + { + "epoch": 2.5793413173652695, + "grad_norm": 0.08273329531565872, + "learning_rate": 1.0216100850018117e-06, + "loss": 0.0006, + "step": 1723 + }, + { + "epoch": 2.5808383233532934, + "grad_norm": 0.028537281972692816, + "learning_rate": 1.0145023266935073e-06, + "loss": 0.0003, + "step": 1724 + }, + { + "epoch": 2.5823353293413174, + "grad_norm": 0.10369851400849742, + "learning_rate": 1.0074180590974847e-06, + "loss": 0.0012, + "step": 1725 + }, + { + "epoch": 2.5838323353293413, + "grad_norm": 0.08409029120428357, + "learning_rate": 1.0003573007340928e-06, + "loss": 0.0005, + "step": 1726 + }, + { + "epoch": 2.5853293413173652, + "grad_norm": 0.021599660351613818, + "learning_rate": 9.93320070062207e-07, + "loss": 0.0002, + "step": 1727 + }, + { + "epoch": 2.586826347305389, + "grad_norm": 0.03987969894778265, + "learning_rate": 9.863063854792055e-07, + "loss": 0.0004, + "step": 1728 + }, + { + "epoch": 2.588323353293413, + "grad_norm": 0.0763711713078856, + "learning_rate": 9.793162653209053e-07, + "loss": 0.0005, + "step": 1729 + }, + { + "epoch": 2.589820359281437, + "grad_norm": 0.06750300685949333, + "learning_rate": 9.72349727861519e-07, + "loss": 0.0002, + "step": 1730 + }, + { + "epoch": 2.591317365269461, + "grad_norm": 0.1192723175715262, + "learning_rate": 9.654067913136079e-07, + "loss": 0.0011, + "step": 1731 + }, + { + "epoch": 2.592814371257485, + "grad_norm": 0.06166676002980035, + "learning_rate": 9.584874738280315e-07, + "loss": 0.0004, + "step": 1732 + }, + { + "epoch": 2.594311377245509, + "grad_norm": 0.08400165230894974, + "learning_rate": 9.515917934939089e-07, + "loss": 0.0005, + "step": 1733 + }, + { + "epoch": 2.595808383233533, + "grad_norm": 0.02748157877354551, + "learning_rate": 9.447197683385579e-07, + "loss": 0.0002, + "step": 1734 + }, + { + "epoch": 2.5973053892215567, + "grad_norm": 0.060188118708937415, + "learning_rate": 9.378714163274571e-07, + "loss": 0.0004, + "step": 1735 + }, + { + "epoch": 2.5988023952095807, + "grad_norm": 0.07502732321886861, + "learning_rate": 9.310467553641967e-07, + "loss": 0.0011, + "step": 1736 + }, + { + "epoch": 2.6002994011976046, + "grad_norm": 0.05755762200466036, + "learning_rate": 9.242458032904311e-07, + "loss": 0.001, + "step": 1737 + }, + { + "epoch": 2.6017964071856285, + "grad_norm": 0.049212105593266754, + "learning_rate": 9.174685778858306e-07, + "loss": 0.0003, + "step": 1738 + }, + { + "epoch": 2.6032934131736525, + "grad_norm": 0.08540752717152685, + "learning_rate": 9.107150968680445e-07, + "loss": 0.0009, + "step": 1739 + }, + { + "epoch": 2.6047904191616764, + "grad_norm": 0.08879092487810307, + "learning_rate": 9.0398537789264e-07, + "loss": 0.0005, + "step": 1740 + }, + { + "epoch": 2.6062874251497004, + "grad_norm": 0.059349429063367105, + "learning_rate": 8.972794385530658e-07, + "loss": 0.0003, + "step": 1741 + }, + { + "epoch": 2.6077844311377243, + "grad_norm": 0.08005756215182938, + "learning_rate": 8.905972963806042e-07, + "loss": 0.0004, + "step": 1742 + }, + { + "epoch": 2.6092814371257482, + "grad_norm": 0.10352902840191881, + "learning_rate": 8.839389688443234e-07, + "loss": 0.0009, + "step": 1743 + }, + { + "epoch": 2.6107784431137726, + "grad_norm": 0.0167977465176402, + "learning_rate": 8.773044733510338e-07, + "loss": 0.0001, + "step": 1744 + }, + { + "epoch": 2.6122754491017965, + "grad_norm": 0.04717208793979273, + "learning_rate": 8.706938272452414e-07, + "loss": 0.0003, + "step": 1745 + }, + { + "epoch": 2.6137724550898205, + "grad_norm": 0.15766515627181438, + "learning_rate": 8.641070478091063e-07, + "loss": 0.0015, + "step": 1746 + }, + { + "epoch": 2.6152694610778444, + "grad_norm": 0.05933348420600678, + "learning_rate": 8.575441522623906e-07, + "loss": 0.0005, + "step": 1747 + }, + { + "epoch": 2.6167664670658684, + "grad_norm": 0.16483126615536, + "learning_rate": 8.510051577624179e-07, + "loss": 0.0008, + "step": 1748 + }, + { + "epoch": 2.6182634730538923, + "grad_norm": 0.04305673654571666, + "learning_rate": 8.444900814040258e-07, + "loss": 0.0002, + "step": 1749 + }, + { + "epoch": 2.6197604790419162, + "grad_norm": 0.0866720060609538, + "learning_rate": 8.379989402195255e-07, + "loss": 0.0004, + "step": 1750 + }, + { + "epoch": 2.62125748502994, + "grad_norm": 0.04999929716635118, + "learning_rate": 8.315317511786547e-07, + "loss": 0.0003, + "step": 1751 + }, + { + "epoch": 2.622754491017964, + "grad_norm": 0.11556476223978758, + "learning_rate": 8.250885311885315e-07, + "loss": 0.0004, + "step": 1752 + }, + { + "epoch": 2.624251497005988, + "grad_norm": 0.138338003099327, + "learning_rate": 8.186692970936127e-07, + "loss": 0.0008, + "step": 1753 + }, + { + "epoch": 2.625748502994012, + "grad_norm": 0.09873045242936929, + "learning_rate": 8.1227406567565e-07, + "loss": 0.0005, + "step": 1754 + }, + { + "epoch": 2.627245508982036, + "grad_norm": 0.07887672568470554, + "learning_rate": 8.059028536536407e-07, + "loss": 0.0006, + "step": 1755 + }, + { + "epoch": 2.62874251497006, + "grad_norm": 0.07738930354523124, + "learning_rate": 7.995556776837964e-07, + "loss": 0.0005, + "step": 1756 + }, + { + "epoch": 2.6302395209580838, + "grad_norm": 0.035685651132636415, + "learning_rate": 7.932325543594821e-07, + "loss": 0.0002, + "step": 1757 + }, + { + "epoch": 2.6317365269461077, + "grad_norm": 0.11039538575586764, + "learning_rate": 7.86933500211191e-07, + "loss": 0.001, + "step": 1758 + }, + { + "epoch": 2.6332335329341316, + "grad_norm": 0.049192018098694695, + "learning_rate": 7.806585317064852e-07, + "loss": 0.0004, + "step": 1759 + }, + { + "epoch": 2.6347305389221556, + "grad_norm": 0.1291838197111577, + "learning_rate": 7.744076652499655e-07, + "loss": 0.0005, + "step": 1760 + }, + { + "epoch": 2.6362275449101795, + "grad_norm": 0.08020464788569276, + "learning_rate": 7.681809171832177e-07, + "loss": 0.0009, + "step": 1761 + }, + { + "epoch": 2.6377245508982035, + "grad_norm": 0.1341437369451868, + "learning_rate": 7.619783037847783e-07, + "loss": 0.001, + "step": 1762 + }, + { + "epoch": 2.6392215568862274, + "grad_norm": 0.09583181308733467, + "learning_rate": 7.557998412700929e-07, + "loss": 0.0006, + "step": 1763 + }, + { + "epoch": 2.6407185628742518, + "grad_norm": 0.06724799411288579, + "learning_rate": 7.496455457914576e-07, + "loss": 0.0006, + "step": 1764 + }, + { + "epoch": 2.6422155688622757, + "grad_norm": 0.009341167893632577, + "learning_rate": 7.435154334380001e-07, + "loss": 0.0001, + "step": 1765 + }, + { + "epoch": 2.6437125748502996, + "grad_norm": 0.005557298748028384, + "learning_rate": 7.374095202356235e-07, + "loss": 0.0001, + "step": 1766 + }, + { + "epoch": 2.6452095808383236, + "grad_norm": 0.06652628053372737, + "learning_rate": 7.313278221469621e-07, + "loss": 0.0004, + "step": 1767 + }, + { + "epoch": 2.6467065868263475, + "grad_norm": 0.054699524495385746, + "learning_rate": 7.252703550713557e-07, + "loss": 0.0003, + "step": 1768 + }, + { + "epoch": 2.6482035928143715, + "grad_norm": 0.03120350774247115, + "learning_rate": 7.192371348447835e-07, + "loss": 0.0002, + "step": 1769 + }, + { + "epoch": 2.6497005988023954, + "grad_norm": 0.04231080059683171, + "learning_rate": 7.132281772398486e-07, + "loss": 0.0002, + "step": 1770 + }, + { + "epoch": 2.6511976047904193, + "grad_norm": 0.2930775130526807, + "learning_rate": 7.072434979657139e-07, + "loss": 0.0008, + "step": 1771 + }, + { + "epoch": 2.6526946107784433, + "grad_norm": 0.06888790537173271, + "learning_rate": 7.012831126680808e-07, + "loss": 0.0006, + "step": 1772 + }, + { + "epoch": 2.654191616766467, + "grad_norm": 0.10410134244700359, + "learning_rate": 6.953470369291349e-07, + "loss": 0.0008, + "step": 1773 + }, + { + "epoch": 2.655688622754491, + "grad_norm": 0.03504060316417563, + "learning_rate": 6.894352862675069e-07, + "loss": 0.0003, + "step": 1774 + }, + { + "epoch": 2.657185628742515, + "grad_norm": 0.07689984954156745, + "learning_rate": 6.835478761382442e-07, + "loss": 0.0005, + "step": 1775 + }, + { + "epoch": 2.658682634730539, + "grad_norm": 0.11005571108502613, + "learning_rate": 6.776848219327469e-07, + "loss": 0.0004, + "step": 1776 + }, + { + "epoch": 2.660179640718563, + "grad_norm": 0.06839160088182221, + "learning_rate": 6.718461389787534e-07, + "loss": 0.0005, + "step": 1777 + }, + { + "epoch": 2.661676646706587, + "grad_norm": 0.1288662945687303, + "learning_rate": 6.660318425402846e-07, + "loss": 0.0011, + "step": 1778 + }, + { + "epoch": 2.663173652694611, + "grad_norm": 0.07138445232695517, + "learning_rate": 6.602419478176037e-07, + "loss": 0.0005, + "step": 1779 + }, + { + "epoch": 2.6646706586826348, + "grad_norm": 0.014327931523637798, + "learning_rate": 6.544764699471906e-07, + "loss": 0.0001, + "step": 1780 + }, + { + "epoch": 2.6661676646706587, + "grad_norm": 0.27063193324926954, + "learning_rate": 6.487354240016807e-07, + "loss": 0.0014, + "step": 1781 + }, + { + "epoch": 2.6676646706586826, + "grad_norm": 0.023863898462553983, + "learning_rate": 6.430188249898473e-07, + "loss": 0.0002, + "step": 1782 + }, + { + "epoch": 2.6691616766467066, + "grad_norm": 0.031269105141780405, + "learning_rate": 6.373266878565454e-07, + "loss": 0.0002, + "step": 1783 + }, + { + "epoch": 2.6706586826347305, + "grad_norm": 0.01870608880096565, + "learning_rate": 6.316590274826828e-07, + "loss": 0.0002, + "step": 1784 + }, + { + "epoch": 2.6721556886227544, + "grad_norm": 0.07554082317016642, + "learning_rate": 6.260158586851761e-07, + "loss": 0.0004, + "step": 1785 + }, + { + "epoch": 2.6736526946107784, + "grad_norm": 0.031568014750176135, + "learning_rate": 6.203971962169153e-07, + "loss": 0.0002, + "step": 1786 + }, + { + "epoch": 2.6751497005988023, + "grad_norm": 0.07930375079821568, + "learning_rate": 6.148030547667272e-07, + "loss": 0.0005, + "step": 1787 + }, + { + "epoch": 2.6766467065868262, + "grad_norm": 0.03552785332381245, + "learning_rate": 6.092334489593233e-07, + "loss": 0.0002, + "step": 1788 + }, + { + "epoch": 2.67814371257485, + "grad_norm": 0.031851214767568116, + "learning_rate": 6.036883933552806e-07, + "loss": 0.0002, + "step": 1789 + }, + { + "epoch": 2.679640718562874, + "grad_norm": 0.04962775237902228, + "learning_rate": 5.981679024509935e-07, + "loss": 0.0002, + "step": 1790 + }, + { + "epoch": 2.681137724550898, + "grad_norm": 0.04407787101186428, + "learning_rate": 5.926719906786348e-07, + "loss": 0.0004, + "step": 1791 + }, + { + "epoch": 2.682634730538922, + "grad_norm": 0.07555818249783224, + "learning_rate": 5.872006724061208e-07, + "loss": 0.0016, + "step": 1792 + }, + { + "epoch": 2.684131736526946, + "grad_norm": 0.02537315182664898, + "learning_rate": 5.817539619370727e-07, + "loss": 0.0002, + "step": 1793 + }, + { + "epoch": 2.68562874251497, + "grad_norm": 0.018001211033930535, + "learning_rate": 5.763318735107826e-07, + "loss": 0.0001, + "step": 1794 + }, + { + "epoch": 2.687125748502994, + "grad_norm": 0.053445457775856076, + "learning_rate": 5.709344213021706e-07, + "loss": 0.0005, + "step": 1795 + }, + { + "epoch": 2.6886227544910177, + "grad_norm": 0.05535585142367004, + "learning_rate": 5.655616194217517e-07, + "loss": 0.0003, + "step": 1796 + }, + { + "epoch": 2.6901197604790417, + "grad_norm": 0.04892377036746475, + "learning_rate": 5.602134819155969e-07, + "loss": 0.0003, + "step": 1797 + }, + { + "epoch": 2.6916167664670656, + "grad_norm": 0.020075394906988262, + "learning_rate": 5.548900227652987e-07, + "loss": 0.0001, + "step": 1798 + }, + { + "epoch": 2.69311377245509, + "grad_norm": 0.07645566349863471, + "learning_rate": 5.495912558879312e-07, + "loss": 0.0003, + "step": 1799 + }, + { + "epoch": 2.694610778443114, + "grad_norm": 0.05819348736751214, + "learning_rate": 5.44317195136016e-07, + "loss": 0.0003, + "step": 1800 + }, + { + "epoch": 2.696107784431138, + "grad_norm": 0.06457619663171396, + "learning_rate": 5.390678542974903e-07, + "loss": 0.0006, + "step": 1801 + }, + { + "epoch": 2.697604790419162, + "grad_norm": 0.05017770166227597, + "learning_rate": 5.33843247095659e-07, + "loss": 0.0004, + "step": 1802 + }, + { + "epoch": 2.6991017964071857, + "grad_norm": 0.06428649128731666, + "learning_rate": 5.286433871891706e-07, + "loss": 0.0003, + "step": 1803 + }, + { + "epoch": 2.7005988023952097, + "grad_norm": 0.021010239698939993, + "learning_rate": 5.234682881719766e-07, + "loss": 0.0002, + "step": 1804 + }, + { + "epoch": 2.7020958083832336, + "grad_norm": 0.03672140143952641, + "learning_rate": 5.183179635732937e-07, + "loss": 0.0002, + "step": 1805 + }, + { + "epoch": 2.7035928143712575, + "grad_norm": 0.024124471357117365, + "learning_rate": 5.131924268575739e-07, + "loss": 0.0001, + "step": 1806 + }, + { + "epoch": 2.7050898203592815, + "grad_norm": 0.08077960062382916, + "learning_rate": 5.080916914244671e-07, + "loss": 0.0008, + "step": 1807 + }, + { + "epoch": 2.7065868263473054, + "grad_norm": 0.10237393145848973, + "learning_rate": 5.030157706087812e-07, + "loss": 0.0006, + "step": 1808 + }, + { + "epoch": 2.7080838323353293, + "grad_norm": 0.07472554370165986, + "learning_rate": 4.979646776804559e-07, + "loss": 0.0002, + "step": 1809 + }, + { + "epoch": 2.7095808383233533, + "grad_norm": 0.05873174733684072, + "learning_rate": 4.929384258445213e-07, + "loss": 0.0003, + "step": 1810 + }, + { + "epoch": 2.711077844311377, + "grad_norm": 0.014282050561278747, + "learning_rate": 4.879370282410633e-07, + "loss": 0.0001, + "step": 1811 + }, + { + "epoch": 2.712574850299401, + "grad_norm": 0.06788628860680303, + "learning_rate": 4.829604979451997e-07, + "loss": 0.0003, + "step": 1812 + }, + { + "epoch": 2.714071856287425, + "grad_norm": 0.12438103093166794, + "learning_rate": 4.780088479670275e-07, + "loss": 0.0006, + "step": 1813 + }, + { + "epoch": 2.715568862275449, + "grad_norm": 0.10531200528785536, + "learning_rate": 4.7308209125160654e-07, + "loss": 0.0004, + "step": 1814 + }, + { + "epoch": 2.717065868263473, + "grad_norm": 0.1130560374315321, + "learning_rate": 4.68180240678916e-07, + "loss": 0.0009, + "step": 1815 + }, + { + "epoch": 2.718562874251497, + "grad_norm": 0.02886495639077637, + "learning_rate": 4.6330330906382394e-07, + "loss": 0.0002, + "step": 1816 + }, + { + "epoch": 2.720059880239521, + "grad_norm": 0.02889119584742746, + "learning_rate": 4.5845130915605165e-07, + "loss": 0.0002, + "step": 1817 + }, + { + "epoch": 2.7215568862275448, + "grad_norm": 0.03501511903364185, + "learning_rate": 4.536242536401403e-07, + "loss": 0.0002, + "step": 1818 + }, + { + "epoch": 2.7230538922155687, + "grad_norm": 0.08233337203210803, + "learning_rate": 4.488221551354266e-07, + "loss": 0.0005, + "step": 1819 + }, + { + "epoch": 2.724550898203593, + "grad_norm": 0.056588559658037704, + "learning_rate": 4.440450261959905e-07, + "loss": 0.0006, + "step": 1820 + }, + { + "epoch": 2.726047904191617, + "grad_norm": 0.07016765453031978, + "learning_rate": 4.3929287931064544e-07, + "loss": 0.0003, + "step": 1821 + }, + { + "epoch": 2.727544910179641, + "grad_norm": 0.0983340591609416, + "learning_rate": 4.34565726902888e-07, + "loss": 0.0006, + "step": 1822 + }, + { + "epoch": 2.729041916167665, + "grad_norm": 0.012487690080718029, + "learning_rate": 4.298635813308727e-07, + "loss": 0.0001, + "step": 1823 + }, + { + "epoch": 2.730538922155689, + "grad_norm": 0.11737318912857932, + "learning_rate": 4.2518645488738296e-07, + "loss": 0.0012, + "step": 1824 + }, + { + "epoch": 2.7320359281437128, + "grad_norm": 0.059519890359665956, + "learning_rate": 4.2053435979978794e-07, + "loss": 0.0004, + "step": 1825 + }, + { + "epoch": 2.7335329341317367, + "grad_norm": 0.09479516367922022, + "learning_rate": 4.1590730823002466e-07, + "loss": 0.0005, + "step": 1826 + }, + { + "epoch": 2.7350299401197606, + "grad_norm": 0.03731826461903811, + "learning_rate": 4.113053122745536e-07, + "loss": 0.0003, + "step": 1827 + }, + { + "epoch": 2.7365269461077846, + "grad_norm": 0.08798270041960435, + "learning_rate": 4.0672838396433436e-07, + "loss": 0.0005, + "step": 1828 + }, + { + "epoch": 2.7380239520958085, + "grad_norm": 0.12804962939354766, + "learning_rate": 4.0217653526479663e-07, + "loss": 0.0008, + "step": 1829 + }, + { + "epoch": 2.7395209580838324, + "grad_norm": 0.12923264052846514, + "learning_rate": 3.9764977807579484e-07, + "loss": 0.0007, + "step": 1830 + }, + { + "epoch": 2.7410179640718564, + "grad_norm": 0.05668518180462198, + "learning_rate": 3.931481242315993e-07, + "loss": 0.0003, + "step": 1831 + }, + { + "epoch": 2.7425149700598803, + "grad_norm": 0.12235579377751511, + "learning_rate": 3.886715855008405e-07, + "loss": 0.0005, + "step": 1832 + }, + { + "epoch": 2.7440119760479043, + "grad_norm": 0.03866968344424915, + "learning_rate": 3.842201735865003e-07, + "loss": 0.0003, + "step": 1833 + }, + { + "epoch": 2.745508982035928, + "grad_norm": 0.11192895051075528, + "learning_rate": 3.797939001258677e-07, + "loss": 0.0005, + "step": 1834 + }, + { + "epoch": 2.747005988023952, + "grad_norm": 0.08805745160822599, + "learning_rate": 3.753927766905108e-07, + "loss": 0.0007, + "step": 1835 + }, + { + "epoch": 2.748502994011976, + "grad_norm": 0.013429115927834216, + "learning_rate": 3.710168147862547e-07, + "loss": 0.0002, + "step": 1836 + }, + { + "epoch": 2.75, + "grad_norm": 0.04732482701187792, + "learning_rate": 3.6666602585313404e-07, + "loss": 0.0003, + "step": 1837 + }, + { + "epoch": 2.751497005988024, + "grad_norm": 0.14685436958806133, + "learning_rate": 3.6234042126538584e-07, + "loss": 0.0011, + "step": 1838 + }, + { + "epoch": 2.752994011976048, + "grad_norm": 0.015481759689920974, + "learning_rate": 3.5804001233139983e-07, + "loss": 0.0001, + "step": 1839 + }, + { + "epoch": 2.754491017964072, + "grad_norm": 0.09709654767571617, + "learning_rate": 3.5376481029369967e-07, + "loss": 0.0006, + "step": 1840 + }, + { + "epoch": 2.7559880239520957, + "grad_norm": 0.09473213466103927, + "learning_rate": 3.495148263289117e-07, + "loss": 0.0011, + "step": 1841 + }, + { + "epoch": 2.7574850299401197, + "grad_norm": 0.029982568028086066, + "learning_rate": 3.452900715477314e-07, + "loss": 0.0002, + "step": 1842 + }, + { + "epoch": 2.7589820359281436, + "grad_norm": 0.037697726669431046, + "learning_rate": 3.4109055699490526e-07, + "loss": 0.0002, + "step": 1843 + }, + { + "epoch": 2.7604790419161676, + "grad_norm": 0.0178343543145911, + "learning_rate": 3.3691629364918323e-07, + "loss": 0.0001, + "step": 1844 + }, + { + "epoch": 2.7619760479041915, + "grad_norm": 0.0868752163044962, + "learning_rate": 3.327672924233116e-07, + "loss": 0.0007, + "step": 1845 + }, + { + "epoch": 2.7634730538922154, + "grad_norm": 0.06282759355626048, + "learning_rate": 3.2864356416398957e-07, + "loss": 0.0003, + "step": 1846 + }, + { + "epoch": 2.7649700598802394, + "grad_norm": 0.023830143819205972, + "learning_rate": 3.2454511965184453e-07, + "loss": 0.0001, + "step": 1847 + }, + { + "epoch": 2.7664670658682633, + "grad_norm": 0.14715683380007932, + "learning_rate": 3.204719696014069e-07, + "loss": 0.0005, + "step": 1848 + }, + { + "epoch": 2.7679640718562872, + "grad_norm": 0.02531860275827708, + "learning_rate": 3.164241246610789e-07, + "loss": 0.0003, + "step": 1849 + }, + { + "epoch": 2.769461077844311, + "grad_norm": 0.0726235081299919, + "learning_rate": 3.1240159541310764e-07, + "loss": 0.0007, + "step": 1850 + }, + { + "epoch": 2.770958083832335, + "grad_norm": 0.10837249282039897, + "learning_rate": 3.0840439237356025e-07, + "loss": 0.0013, + "step": 1851 + }, + { + "epoch": 2.772455089820359, + "grad_norm": 0.046374751147364626, + "learning_rate": 3.0443252599228865e-07, + "loss": 0.0004, + "step": 1852 + }, + { + "epoch": 2.773952095808383, + "grad_norm": 0.04955860777056503, + "learning_rate": 3.004860066529125e-07, + "loss": 0.0003, + "step": 1853 + }, + { + "epoch": 2.775449101796407, + "grad_norm": 0.04175690267753176, + "learning_rate": 2.96564844672782e-07, + "loss": 0.0002, + "step": 1854 + }, + { + "epoch": 2.7769461077844313, + "grad_norm": 0.01101306053826792, + "learning_rate": 2.9266905030296164e-07, + "loss": 0.0001, + "step": 1855 + }, + { + "epoch": 2.7784431137724552, + "grad_norm": 0.061584971883424364, + "learning_rate": 2.887986337281934e-07, + "loss": 0.0005, + "step": 1856 + }, + { + "epoch": 2.779940119760479, + "grad_norm": 0.038380836869109265, + "learning_rate": 2.8495360506687444e-07, + "loss": 0.0002, + "step": 1857 + }, + { + "epoch": 2.781437125748503, + "grad_norm": 0.036721797893085474, + "learning_rate": 2.811339743710328e-07, + "loss": 0.0001, + "step": 1858 + }, + { + "epoch": 2.782934131736527, + "grad_norm": 0.10482041789183853, + "learning_rate": 2.7733975162629655e-07, + "loss": 0.0004, + "step": 1859 + }, + { + "epoch": 2.784431137724551, + "grad_norm": 0.06852021780595845, + "learning_rate": 2.735709467518699e-07, + "loss": 0.0003, + "step": 1860 + }, + { + "epoch": 2.785928143712575, + "grad_norm": 0.007301911146011518, + "learning_rate": 2.6982756960050925e-07, + "loss": 0.0001, + "step": 1861 + }, + { + "epoch": 2.787425149700599, + "grad_norm": 0.0928237098781843, + "learning_rate": 2.661096299584953e-07, + "loss": 0.0005, + "step": 1862 + }, + { + "epoch": 2.788922155688623, + "grad_norm": 0.037133885536317905, + "learning_rate": 2.624171375456064e-07, + "loss": 0.0002, + "step": 1863 + }, + { + "epoch": 2.7904191616766467, + "grad_norm": 0.10006499255445611, + "learning_rate": 2.58750102015094e-07, + "loss": 0.0007, + "step": 1864 + }, + { + "epoch": 2.7919161676646707, + "grad_norm": 0.04256690893240319, + "learning_rate": 2.5510853295366067e-07, + "loss": 0.0003, + "step": 1865 + }, + { + "epoch": 2.7934131736526946, + "grad_norm": 0.0684594948441174, + "learning_rate": 2.514924398814278e-07, + "loss": 0.0003, + "step": 1866 + }, + { + "epoch": 2.7949101796407185, + "grad_norm": 0.09626127387021133, + "learning_rate": 2.479018322519189e-07, + "loss": 0.0006, + "step": 1867 + }, + { + "epoch": 2.7964071856287425, + "grad_norm": 0.01741566020372036, + "learning_rate": 2.4433671945203076e-07, + "loss": 0.0001, + "step": 1868 + }, + { + "epoch": 2.7979041916167664, + "grad_norm": 0.07213965555140582, + "learning_rate": 2.4079711080200683e-07, + "loss": 0.0005, + "step": 1869 + }, + { + "epoch": 2.7994011976047903, + "grad_norm": 0.06007506249091625, + "learning_rate": 2.3728301555541845e-07, + "loss": 0.0003, + "step": 1870 + }, + { + "epoch": 2.8008982035928143, + "grad_norm": 0.06125607539660623, + "learning_rate": 2.3379444289913344e-07, + "loss": 0.0003, + "step": 1871 + }, + { + "epoch": 2.802395209580838, + "grad_norm": 0.015232735958729315, + "learning_rate": 2.3033140195329985e-07, + "loss": 0.0001, + "step": 1872 + }, + { + "epoch": 2.803892215568862, + "grad_norm": 0.03010125402384828, + "learning_rate": 2.2689390177131787e-07, + "loss": 0.0002, + "step": 1873 + }, + { + "epoch": 2.805389221556886, + "grad_norm": 0.05324858893449288, + "learning_rate": 2.2348195133981344e-07, + "loss": 0.0003, + "step": 1874 + }, + { + "epoch": 2.80688622754491, + "grad_norm": 0.026758821808148425, + "learning_rate": 2.200955595786214e-07, + "loss": 0.0002, + "step": 1875 + }, + { + "epoch": 2.8083832335329344, + "grad_norm": 0.04605830316424714, + "learning_rate": 2.1673473534075895e-07, + "loss": 0.0004, + "step": 1876 + }, + { + "epoch": 2.8098802395209583, + "grad_norm": 0.04122472187922105, + "learning_rate": 2.1339948741240013e-07, + "loss": 0.0002, + "step": 1877 + }, + { + "epoch": 2.8113772455089823, + "grad_norm": 0.026915377348818586, + "learning_rate": 2.1008982451285687e-07, + "loss": 0.0003, + "step": 1878 + }, + { + "epoch": 2.812874251497006, + "grad_norm": 0.030265578451253612, + "learning_rate": 2.0680575529455017e-07, + "loss": 0.0004, + "step": 1879 + }, + { + "epoch": 2.81437125748503, + "grad_norm": 0.07654024267241427, + "learning_rate": 2.035472883430001e-07, + "loss": 0.0005, + "step": 1880 + }, + { + "epoch": 2.815868263473054, + "grad_norm": 0.02468323185102218, + "learning_rate": 2.0031443217678582e-07, + "loss": 0.0003, + "step": 1881 + }, + { + "epoch": 2.817365269461078, + "grad_norm": 0.14330870714264016, + "learning_rate": 1.9710719524753896e-07, + "loss": 0.0008, + "step": 1882 + }, + { + "epoch": 2.818862275449102, + "grad_norm": 0.06963422959188056, + "learning_rate": 1.939255859399114e-07, + "loss": 0.0004, + "step": 1883 + }, + { + "epoch": 2.820359281437126, + "grad_norm": 0.03669033456219445, + "learning_rate": 1.9076961257155745e-07, + "loss": 0.0002, + "step": 1884 + }, + { + "epoch": 2.82185628742515, + "grad_norm": 0.08034552402519764, + "learning_rate": 1.876392833931151e-07, + "loss": 0.0008, + "step": 1885 + }, + { + "epoch": 2.8233532934131738, + "grad_norm": 0.011926808859062795, + "learning_rate": 1.8453460658817367e-07, + "loss": 0.0001, + "step": 1886 + }, + { + "epoch": 2.8248502994011977, + "grad_norm": 0.07355376029482306, + "learning_rate": 1.814555902732673e-07, + "loss": 0.0006, + "step": 1887 + }, + { + "epoch": 2.8263473053892216, + "grad_norm": 0.029801616810496537, + "learning_rate": 1.7840224249784154e-07, + "loss": 0.0002, + "step": 1888 + }, + { + "epoch": 2.8278443113772456, + "grad_norm": 0.12698015721022915, + "learning_rate": 1.7537457124423896e-07, + "loss": 0.0004, + "step": 1889 + }, + { + "epoch": 2.8293413173652695, + "grad_norm": 0.06530833972448825, + "learning_rate": 1.723725844276747e-07, + "loss": 0.0004, + "step": 1890 + }, + { + "epoch": 2.8308383233532934, + "grad_norm": 0.12703576094207422, + "learning_rate": 1.693962898962187e-07, + "loss": 0.0005, + "step": 1891 + }, + { + "epoch": 2.8323353293413174, + "grad_norm": 0.06299450353431471, + "learning_rate": 1.664456954307736e-07, + "loss": 0.0004, + "step": 1892 + }, + { + "epoch": 2.8338323353293413, + "grad_norm": 0.0345823141248644, + "learning_rate": 1.635208087450524e-07, + "loss": 0.0002, + "step": 1893 + }, + { + "epoch": 2.8353293413173652, + "grad_norm": 0.07171038665684742, + "learning_rate": 1.6062163748556402e-07, + "loss": 0.0005, + "step": 1894 + }, + { + "epoch": 2.836826347305389, + "grad_norm": 0.040539395882641424, + "learning_rate": 1.5774818923158687e-07, + "loss": 0.0004, + "step": 1895 + }, + { + "epoch": 2.838323353293413, + "grad_norm": 0.01721308734367943, + "learning_rate": 1.5490047149515186e-07, + "loss": 0.0001, + "step": 1896 + }, + { + "epoch": 2.839820359281437, + "grad_norm": 0.03712780760613839, + "learning_rate": 1.5207849172102827e-07, + "loss": 0.0002, + "step": 1897 + }, + { + "epoch": 2.841317365269461, + "grad_norm": 0.017202811427165027, + "learning_rate": 1.492822572866892e-07, + "loss": 0.0001, + "step": 1898 + }, + { + "epoch": 2.842814371257485, + "grad_norm": 0.08062656343801933, + "learning_rate": 1.4651177550231043e-07, + "loss": 0.0004, + "step": 1899 + }, + { + "epoch": 2.844311377245509, + "grad_norm": 0.03769317002353988, + "learning_rate": 1.4376705361073828e-07, + "loss": 0.0004, + "step": 1900 + }, + { + "epoch": 2.845808383233533, + "grad_norm": 0.0595900137707606, + "learning_rate": 1.410480987874774e-07, + "loss": 0.0004, + "step": 1901 + }, + { + "epoch": 2.8473053892215567, + "grad_norm": 0.0456203631893289, + "learning_rate": 1.3835491814066847e-07, + "loss": 0.0003, + "step": 1902 + }, + { + "epoch": 2.8488023952095807, + "grad_norm": 0.06469454614345947, + "learning_rate": 1.3568751871107068e-07, + "loss": 0.0003, + "step": 1903 + }, + { + "epoch": 2.8502994011976046, + "grad_norm": 0.14575247322591645, + "learning_rate": 1.3304590747204582e-07, + "loss": 0.0018, + "step": 1904 + }, + { + "epoch": 2.8517964071856285, + "grad_norm": 0.09067521256579207, + "learning_rate": 1.3043009132953533e-07, + "loss": 0.0009, + "step": 1905 + }, + { + "epoch": 2.8532934131736525, + "grad_norm": 0.013493249897349183, + "learning_rate": 1.2784007712204562e-07, + "loss": 0.0001, + "step": 1906 + }, + { + "epoch": 2.8547904191616764, + "grad_norm": 0.11691368942179577, + "learning_rate": 1.2527587162062927e-07, + "loss": 0.0009, + "step": 1907 + }, + { + "epoch": 2.8562874251497004, + "grad_norm": 0.037850088975618104, + "learning_rate": 1.2273748152886844e-07, + "loss": 0.0003, + "step": 1908 + }, + { + "epoch": 2.8577844311377243, + "grad_norm": 0.1141808276161799, + "learning_rate": 1.2022491348285482e-07, + "loss": 0.0006, + "step": 1909 + }, + { + "epoch": 2.8592814371257482, + "grad_norm": 0.02925451773697013, + "learning_rate": 1.1773817405117405e-07, + "loss": 0.0002, + "step": 1910 + }, + { + "epoch": 2.8607784431137726, + "grad_norm": 0.05017111181234729, + "learning_rate": 1.1527726973488918e-07, + "loss": 0.0003, + "step": 1911 + }, + { + "epoch": 2.8622754491017965, + "grad_norm": 0.04552663871537052, + "learning_rate": 1.1284220696752168e-07, + "loss": 0.0002, + "step": 1912 + }, + { + "epoch": 2.8637724550898205, + "grad_norm": 0.1413296347951028, + "learning_rate": 1.1043299211503489e-07, + "loss": 0.0007, + "step": 1913 + }, + { + "epoch": 2.8652694610778444, + "grad_norm": 0.01983528978302256, + "learning_rate": 1.080496314758217e-07, + "loss": 0.0001, + "step": 1914 + }, + { + "epoch": 2.8667664670658684, + "grad_norm": 0.10833415515320645, + "learning_rate": 1.0569213128068024e-07, + "loss": 0.0003, + "step": 1915 + }, + { + "epoch": 2.8682634730538923, + "grad_norm": 0.04839502505058121, + "learning_rate": 1.0336049769280377e-07, + "loss": 0.0004, + "step": 1916 + }, + { + "epoch": 2.8697604790419162, + "grad_norm": 0.018562019071949903, + "learning_rate": 1.0105473680776301e-07, + "loss": 0.0001, + "step": 1917 + }, + { + "epoch": 2.87125748502994, + "grad_norm": 0.029694138913736658, + "learning_rate": 9.877485465349057e-08, + "loss": 0.0003, + "step": 1918 + }, + { + "epoch": 2.872754491017964, + "grad_norm": 0.04427064392263348, + "learning_rate": 9.652085719026205e-08, + "loss": 0.0005, + "step": 1919 + }, + { + "epoch": 2.874251497005988, + "grad_norm": 0.07693934381799596, + "learning_rate": 9.429275031068608e-08, + "loss": 0.0006, + "step": 1920 + }, + { + "epoch": 2.875748502994012, + "grad_norm": 0.09527796665446048, + "learning_rate": 9.209053983968209e-08, + "loss": 0.0007, + "step": 1921 + }, + { + "epoch": 2.877245508982036, + "grad_norm": 0.058776929014960504, + "learning_rate": 8.991423153447143e-08, + "loss": 0.0003, + "step": 1922 + }, + { + "epoch": 2.87874251497006, + "grad_norm": 0.03288997471143238, + "learning_rate": 8.776383108455966e-08, + "loss": 0.0002, + "step": 1923 + }, + { + "epoch": 2.8802395209580838, + "grad_norm": 0.05757102855593304, + "learning_rate": 8.563934411171981e-08, + "loss": 0.0002, + "step": 1924 + }, + { + "epoch": 2.8817365269461077, + "grad_norm": 0.0387920150897497, + "learning_rate": 8.354077616998247e-08, + "loss": 0.0003, + "step": 1925 + }, + { + "epoch": 2.8832335329341316, + "grad_norm": 0.1232602412566441, + "learning_rate": 8.14681327456146e-08, + "loss": 0.0008, + "step": 1926 + }, + { + "epoch": 2.8847305389221556, + "grad_norm": 0.09510268541356108, + "learning_rate": 7.942141925711078e-08, + "loss": 0.0008, + "step": 1927 + }, + { + "epoch": 2.8862275449101795, + "grad_norm": 0.034899399930684916, + "learning_rate": 7.740064105517753e-08, + "loss": 0.0002, + "step": 1928 + }, + { + "epoch": 2.8877245508982035, + "grad_norm": 0.08538587733278336, + "learning_rate": 7.540580342272008e-08, + "loss": 0.0005, + "step": 1929 + }, + { + "epoch": 2.8892215568862274, + "grad_norm": 0.036141222171045685, + "learning_rate": 7.343691157482347e-08, + "loss": 0.0003, + "step": 1930 + }, + { + "epoch": 2.8907185628742518, + "grad_norm": 0.048494060524911266, + "learning_rate": 7.149397065874697e-08, + "loss": 0.0004, + "step": 1931 + }, + { + "epoch": 2.8922155688622757, + "grad_norm": 0.04348773500793645, + "learning_rate": 6.957698575390415e-08, + "loss": 0.0003, + "step": 1932 + }, + { + "epoch": 2.8937125748502996, + "grad_norm": 0.019348415703767035, + "learning_rate": 6.768596187185394e-08, + "loss": 0.0001, + "step": 1933 + }, + { + "epoch": 2.8952095808383236, + "grad_norm": 0.07990531211236819, + "learning_rate": 6.582090395628294e-08, + "loss": 0.0005, + "step": 1934 + }, + { + "epoch": 2.8967065868263475, + "grad_norm": 0.028337210973887937, + "learning_rate": 6.398181688299753e-08, + "loss": 0.0003, + "step": 1935 + }, + { + "epoch": 2.8982035928143715, + "grad_norm": 0.029288080250697478, + "learning_rate": 6.21687054599085e-08, + "loss": 0.0001, + "step": 1936 + }, + { + "epoch": 2.8997005988023954, + "grad_norm": 0.0868655909482861, + "learning_rate": 6.038157442701642e-08, + "loss": 0.0004, + "step": 1937 + }, + { + "epoch": 2.9011976047904193, + "grad_norm": 0.07224510015715654, + "learning_rate": 5.862042845640403e-08, + "loss": 0.0005, + "step": 1938 + }, + { + "epoch": 2.9026946107784433, + "grad_norm": 0.014587726019568977, + "learning_rate": 5.688527215222062e-08, + "loss": 0.0001, + "step": 1939 + }, + { + "epoch": 2.904191616766467, + "grad_norm": 0.0913841204532074, + "learning_rate": 5.517611005067092e-08, + "loss": 0.0007, + "step": 1940 + }, + { + "epoch": 2.905688622754491, + "grad_norm": 0.02053119590322276, + "learning_rate": 5.3492946620005146e-08, + "loss": 0.0004, + "step": 1941 + }, + { + "epoch": 2.907185628742515, + "grad_norm": 0.048364628173414546, + "learning_rate": 5.183578626050012e-08, + "loss": 0.0003, + "step": 1942 + }, + { + "epoch": 2.908682634730539, + "grad_norm": 0.11064582121615582, + "learning_rate": 5.020463330445813e-08, + "loss": 0.0007, + "step": 1943 + }, + { + "epoch": 2.910179640718563, + "grad_norm": 0.02823565491289188, + "learning_rate": 4.8599492016189184e-08, + "loss": 0.0003, + "step": 1944 + }, + { + "epoch": 2.911676646706587, + "grad_norm": 0.08021570823602606, + "learning_rate": 4.7020366592001046e-08, + "loss": 0.0005, + "step": 1945 + }, + { + "epoch": 2.913173652694611, + "grad_norm": 0.06441990843786155, + "learning_rate": 4.546726116018696e-08, + "loss": 0.0004, + "step": 1946 + }, + { + "epoch": 2.9146706586826348, + "grad_norm": 0.07772901840599725, + "learning_rate": 4.394017978101905e-08, + "loss": 0.0004, + "step": 1947 + }, + { + "epoch": 2.9161676646706587, + "grad_norm": 0.012281701982242639, + "learning_rate": 4.243912644673276e-08, + "loss": 0.0002, + "step": 1948 + }, + { + "epoch": 2.9176646706586826, + "grad_norm": 0.0893888921305495, + "learning_rate": 4.0964105081519046e-08, + "loss": 0.0004, + "step": 1949 + }, + { + "epoch": 2.9191616766467066, + "grad_norm": 0.007291028241031858, + "learning_rate": 3.951511954151554e-08, + "loss": 0.0001, + "step": 1950 + }, + { + "epoch": 2.9206586826347305, + "grad_norm": 0.04531637743292414, + "learning_rate": 3.809217361479212e-08, + "loss": 0.0004, + "step": 1951 + }, + { + "epoch": 2.9221556886227544, + "grad_norm": 0.10016872974204939, + "learning_rate": 3.669527102134418e-08, + "loss": 0.0006, + "step": 1952 + }, + { + "epoch": 2.9236526946107784, + "grad_norm": 0.10136592233723844, + "learning_rate": 3.532441541308384e-08, + "loss": 0.0009, + "step": 1953 + }, + { + "epoch": 2.9251497005988023, + "grad_norm": 0.099063980397416, + "learning_rate": 3.397961037382547e-08, + "loss": 0.0007, + "step": 1954 + }, + { + "epoch": 2.9266467065868262, + "grad_norm": 0.0494457607619648, + "learning_rate": 3.266085941928454e-08, + "loss": 0.0004, + "step": 1955 + }, + { + "epoch": 2.92814371257485, + "grad_norm": 0.10780498117653974, + "learning_rate": 3.136816599705883e-08, + "loss": 0.0006, + "step": 1956 + }, + { + "epoch": 2.929640718562874, + "grad_norm": 0.0658842139801036, + "learning_rate": 3.0101533486626146e-08, + "loss": 0.0005, + "step": 1957 + }, + { + "epoch": 2.931137724550898, + "grad_norm": 0.04093767820581611, + "learning_rate": 2.8860965199334346e-08, + "loss": 0.0002, + "step": 1958 + }, + { + "epoch": 2.932634730538922, + "grad_norm": 0.05070774966783894, + "learning_rate": 2.7646464378390248e-08, + "loss": 0.0004, + "step": 1959 + }, + { + "epoch": 2.934131736526946, + "grad_norm": 0.01789433543858203, + "learning_rate": 2.6458034198852954e-08, + "loss": 0.0001, + "step": 1960 + }, + { + "epoch": 2.93562874251497, + "grad_norm": 0.04336770967449903, + "learning_rate": 2.5295677767627202e-08, + "loss": 0.0002, + "step": 1961 + }, + { + "epoch": 2.937125748502994, + "grad_norm": 0.0865926476879689, + "learning_rate": 2.415939812345003e-08, + "loss": 0.0003, + "step": 1962 + }, + { + "epoch": 2.9386227544910177, + "grad_norm": 0.09911546956818736, + "learning_rate": 2.304919823688856e-08, + "loss": 0.0005, + "step": 1963 + }, + { + "epoch": 2.9401197604790417, + "grad_norm": 0.07993502202450768, + "learning_rate": 2.1965081010328903e-08, + "loss": 0.0007, + "step": 1964 + }, + { + "epoch": 2.9416167664670656, + "grad_norm": 0.04287570725916258, + "learning_rate": 2.0907049277971714e-08, + "loss": 0.0002, + "step": 1965 + }, + { + "epoch": 2.94311377245509, + "grad_norm": 0.041199397205273956, + "learning_rate": 1.987510580581664e-08, + "loss": 0.0003, + "step": 1966 + }, + { + "epoch": 2.944610778443114, + "grad_norm": 0.12923583448830503, + "learning_rate": 1.8869253291669e-08, + "loss": 0.0007, + "step": 1967 + }, + { + "epoch": 2.946107784431138, + "grad_norm": 0.03545184716551114, + "learning_rate": 1.7889494365118665e-08, + "loss": 0.0002, + "step": 1968 + }, + { + "epoch": 2.947604790419162, + "grad_norm": 0.05849241385118926, + "learning_rate": 1.6935831587541197e-08, + "loss": 0.0003, + "step": 1969 + }, + { + "epoch": 2.9491017964071857, + "grad_norm": 0.11944886195630253, + "learning_rate": 1.6008267452090054e-08, + "loss": 0.0014, + "step": 1970 + }, + { + "epoch": 2.9505988023952097, + "grad_norm": 0.023633602514857067, + "learning_rate": 1.5106804383688834e-08, + "loss": 0.0002, + "step": 1971 + }, + { + "epoch": 2.9520958083832336, + "grad_norm": 0.06252240701759237, + "learning_rate": 1.4231444739023492e-08, + "loss": 0.0005, + "step": 1972 + }, + { + "epoch": 2.9535928143712575, + "grad_norm": 0.08288723195823507, + "learning_rate": 1.338219080654013e-08, + "loss": 0.0004, + "step": 1973 + }, + { + "epoch": 2.9550898203592815, + "grad_norm": 0.02470100516563367, + "learning_rate": 1.2559044806437215e-08, + "loss": 0.0002, + "step": 1974 + }, + { + "epoch": 2.9565868263473054, + "grad_norm": 0.035384313511806896, + "learning_rate": 1.1762008890657816e-08, + "loss": 0.0002, + "step": 1975 + }, + { + "epoch": 2.9580838323353293, + "grad_norm": 0.0726241495411833, + "learning_rate": 1.099108514288627e-08, + "loss": 0.0005, + "step": 1976 + }, + { + "epoch": 2.9595808383233533, + "grad_norm": 0.05304485849070656, + "learning_rate": 1.024627557854152e-08, + "loss": 0.0002, + "step": 1977 + }, + { + "epoch": 2.961077844311377, + "grad_norm": 0.11794290635555527, + "learning_rate": 9.527582144773785e-09, + "loss": 0.0006, + "step": 1978 + }, + { + "epoch": 2.962574850299401, + "grad_norm": 0.012966785537484036, + "learning_rate": 8.835006720457895e-09, + "loss": 0.0001, + "step": 1979 + }, + { + "epoch": 2.964071856287425, + "grad_norm": 0.05050410993770117, + "learning_rate": 8.16855111618775e-09, + "loss": 0.0003, + "step": 1980 + }, + { + "epoch": 2.965568862275449, + "grad_norm": 0.0697386602951542, + "learning_rate": 7.528217074272982e-09, + "loss": 0.0004, + "step": 1981 + }, + { + "epoch": 2.967065868263473, + "grad_norm": 0.0324564659125771, + "learning_rate": 6.914006268732287e-09, + "loss": 0.0003, + "step": 1982 + }, + { + "epoch": 2.968562874251497, + "grad_norm": 0.03373070903115685, + "learning_rate": 6.325920305292332e-09, + "loss": 0.0002, + "step": 1983 + }, + { + "epoch": 2.970059880239521, + "grad_norm": 0.011788534817898908, + "learning_rate": 5.763960721382189e-09, + "loss": 0.0002, + "step": 1984 + }, + { + "epoch": 2.9715568862275448, + "grad_norm": 0.08221326208806214, + "learning_rate": 5.228128986127789e-09, + "loss": 0.0006, + "step": 1985 + }, + { + "epoch": 2.9730538922155687, + "grad_norm": 0.1270884872336597, + "learning_rate": 4.718426500349705e-09, + "loss": 0.0005, + "step": 1986 + }, + { + "epoch": 2.974550898203593, + "grad_norm": 0.058075307647141146, + "learning_rate": 4.2348545965575914e-09, + "loss": 0.0002, + "step": 1987 + }, + { + "epoch": 2.976047904191617, + "grad_norm": 0.03512480614793991, + "learning_rate": 3.7774145389524176e-09, + "loss": 0.0003, + "step": 1988 + }, + { + "epoch": 2.977544910179641, + "grad_norm": 0.07371078403449538, + "learning_rate": 3.3461075234131336e-09, + "loss": 0.0001, + "step": 1989 + }, + { + "epoch": 2.979041916167665, + "grad_norm": 0.022109669320123617, + "learning_rate": 2.9409346775033376e-09, + "loss": 0.0001, + "step": 1990 + }, + { + "epoch": 2.980538922155689, + "grad_norm": 0.048672724122854005, + "learning_rate": 2.561897060462393e-09, + "loss": 0.0004, + "step": 1991 + }, + { + "epoch": 2.9820359281437128, + "grad_norm": 0.01931500903511454, + "learning_rate": 2.208995663206537e-09, + "loss": 0.0002, + "step": 1992 + }, + { + "epoch": 2.9835329341317367, + "grad_norm": 0.04749199082195791, + "learning_rate": 1.8822314083222216e-09, + "loss": 0.0004, + "step": 1993 + }, + { + "epoch": 2.9850299401197606, + "grad_norm": 0.09729739241328733, + "learning_rate": 1.5816051500661122e-09, + "loss": 0.001, + "step": 1994 + }, + { + "epoch": 2.9865269461077846, + "grad_norm": 0.044947007048880785, + "learning_rate": 1.3071176743628677e-09, + "loss": 0.0003, + "step": 1995 + }, + { + "epoch": 2.9880239520958085, + "grad_norm": 0.13515173564413327, + "learning_rate": 1.0587696988029195e-09, + "loss": 0.0008, + "step": 1996 + }, + { + "epoch": 2.9895209580838324, + "grad_norm": 0.06726018374004353, + "learning_rate": 8.365618726413616e-10, + "loss": 0.0005, + "step": 1997 + }, + { + "epoch": 2.9910179640718564, + "grad_norm": 0.1032600661241297, + "learning_rate": 6.404947767935099e-10, + "loss": 0.0011, + "step": 1998 + }, + { + "epoch": 2.9925149700598803, + "grad_norm": 0.08635105614698947, + "learning_rate": 4.705689238360123e-10, + "loss": 0.0009, + "step": 1999 + }, + { + "epoch": 2.9940119760479043, + "grad_norm": 0.10265564812162384, + "learning_rate": 3.267847580046279e-10, + "loss": 0.0004, + "step": 2000 + }, + { + "epoch": 2.995508982035928, + "grad_norm": 0.040089587915776735, + "learning_rate": 2.091426551931175e-10, + "loss": 0.0002, + "step": 2001 + }, + { + "epoch": 2.997005988023952, + "grad_norm": 0.05526462283205673, + "learning_rate": 1.1764292295213254e-10, + "loss": 0.0002, + "step": 2002 + }, + { + "epoch": 2.998502994011976, + "grad_norm": 0.046950315243127365, + "learning_rate": 5.2285800488105495e-11, + "loss": 0.0003, + "step": 2003 + }, + { + "epoch": 3.0, + "grad_norm": 0.02255507610632034, + "learning_rate": 1.3071458665470105e-11, + "loss": 0.0001, + "step": 2004 + }, { "epoch": 3.0, - "step": 441, - "total_flos": 34413652910080.0, - "train_loss": 0.10652120225131512, - "train_runtime": 8437.5347, - "train_samples_per_second": 0.836, - "train_steps_per_second": 0.052 + "step": 2004, + "total_flos": 160061552422912.0, + "train_loss": 0.02935138660071818, + "train_runtime": 23628.3662, + "train_samples_per_second": 1.356, + "train_steps_per_second": 0.085 } ], "logging_steps": 1.0, - "max_steps": 441, + "max_steps": 2004, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 50000, - "total_flos": 34413652910080.0, + "total_flos": 160061552422912.0, "train_batch_size": 8, "trial_name": null, "trial_params": null