{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 3400, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005884732796351466, "grad_norm": 2.994872487021255, "learning_rate": 0.0, "loss": 1.1596, "step": 1 }, { "epoch": 0.0011769465592702932, "grad_norm": 2.9537447939845776, "learning_rate": 2.9411764705882354e-08, "loss": 1.0938, "step": 2 }, { "epoch": 0.0017654198389054397, "grad_norm": 3.3700752663540716, "learning_rate": 5.882352941176471e-08, "loss": 1.1706, "step": 3 }, { "epoch": 0.0023538931185405863, "grad_norm": 3.4735876533384444, "learning_rate": 8.823529411764707e-08, "loss": 1.0894, "step": 4 }, { "epoch": 0.002942366398175733, "grad_norm": 3.307228254868395, "learning_rate": 1.1764705882352942e-07, "loss": 1.0786, "step": 5 }, { "epoch": 0.0035308396778108795, "grad_norm": 2.7987485158403738, "learning_rate": 1.4705882352941178e-07, "loss": 1.1288, "step": 6 }, { "epoch": 0.004119312957446026, "grad_norm": 2.72474299131369, "learning_rate": 1.7647058823529414e-07, "loss": 1.0416, "step": 7 }, { "epoch": 0.004707786237081173, "grad_norm": 3.1381299762091266, "learning_rate": 2.0588235294117647e-07, "loss": 1.0487, "step": 8 }, { "epoch": 0.005296259516716319, "grad_norm": 3.6096319492921176, "learning_rate": 2.3529411764705883e-07, "loss": 1.0446, "step": 9 }, { "epoch": 0.005884732796351466, "grad_norm": 2.94884510474711, "learning_rate": 2.647058823529412e-07, "loss": 1.1094, "step": 10 }, { "epoch": 0.006473206075986612, "grad_norm": 2.775154550404211, "learning_rate": 2.9411764705882356e-07, "loss": 1.0857, "step": 11 }, { "epoch": 0.007061679355621759, "grad_norm": 2.5628535670724597, "learning_rate": 3.235294117647059e-07, "loss": 1.0541, "step": 12 }, { "epoch": 0.007650152635256906, "grad_norm": 2.771853836936518, "learning_rate": 3.529411764705883e-07, "loss": 1.1555, "step": 13 }, { "epoch": 0.008238625914892052, "grad_norm": 3.2318740557518084, "learning_rate": 3.8235294117647064e-07, "loss": 1.0959, "step": 14 }, { "epoch": 0.008827099194527198, "grad_norm": 2.682219629488663, "learning_rate": 4.1176470588235295e-07, "loss": 1.0019, "step": 15 }, { "epoch": 0.009415572474162345, "grad_norm": 3.106351125571997, "learning_rate": 4.4117647058823536e-07, "loss": 1.0949, "step": 16 }, { "epoch": 0.010004045753797491, "grad_norm": 3.163809326034569, "learning_rate": 4.7058823529411767e-07, "loss": 1.1618, "step": 17 }, { "epoch": 0.010592519033432638, "grad_norm": 2.949018349720342, "learning_rate": 5.000000000000001e-07, "loss": 1.0445, "step": 18 }, { "epoch": 0.011180992313067784, "grad_norm": 2.7578856805832057, "learning_rate": 5.294117647058824e-07, "loss": 1.0473, "step": 19 }, { "epoch": 0.011769465592702932, "grad_norm": 3.1496383849576506, "learning_rate": 5.588235294117648e-07, "loss": 0.9371, "step": 20 }, { "epoch": 0.012357938872338077, "grad_norm": 3.1309047118608615, "learning_rate": 5.882352941176471e-07, "loss": 1.0884, "step": 21 }, { "epoch": 0.012946412151973225, "grad_norm": 3.129649151068586, "learning_rate": 6.176470588235295e-07, "loss": 1.1055, "step": 22 }, { "epoch": 0.01353488543160837, "grad_norm": 2.943142998849525, "learning_rate": 6.470588235294118e-07, "loss": 1.0122, "step": 23 }, { "epoch": 0.014123358711243518, "grad_norm": 2.9220325090770576, "learning_rate": 6.764705882352941e-07, "loss": 1.0812, "step": 24 }, { "epoch": 0.014711831990878664, "grad_norm": 2.5688017196244974, "learning_rate": 7.058823529411766e-07, "loss": 1.1769, "step": 25 }, { "epoch": 0.015300305270513811, "grad_norm": 2.5768785350511525, "learning_rate": 7.352941176470589e-07, "loss": 1.0326, "step": 26 }, { "epoch": 0.015888778550148957, "grad_norm": 2.7517243364895694, "learning_rate": 7.647058823529413e-07, "loss": 1.0728, "step": 27 }, { "epoch": 0.016477251829784104, "grad_norm": 2.5644177501152874, "learning_rate": 7.941176470588236e-07, "loss": 1.0467, "step": 28 }, { "epoch": 0.017065725109419252, "grad_norm": 2.428246200487348, "learning_rate": 8.235294117647059e-07, "loss": 1.1208, "step": 29 }, { "epoch": 0.017654198389054396, "grad_norm": 2.296698900778094, "learning_rate": 8.529411764705883e-07, "loss": 1.0798, "step": 30 }, { "epoch": 0.018242671668689543, "grad_norm": 2.4545325826822766, "learning_rate": 8.823529411764707e-07, "loss": 1.0707, "step": 31 }, { "epoch": 0.01883114494832469, "grad_norm": 2.2371849242995014, "learning_rate": 9.117647058823529e-07, "loss": 0.9806, "step": 32 }, { "epoch": 0.019419618227959838, "grad_norm": 2.236696657333955, "learning_rate": 9.411764705882353e-07, "loss": 1.1172, "step": 33 }, { "epoch": 0.020008091507594982, "grad_norm": 2.183269891215553, "learning_rate": 9.705882352941176e-07, "loss": 0.9861, "step": 34 }, { "epoch": 0.02059656478723013, "grad_norm": 1.8950850310331329, "learning_rate": 1.0000000000000002e-06, "loss": 1.0245, "step": 35 }, { "epoch": 0.021185038066865277, "grad_norm": 1.9368342995747545, "learning_rate": 1.0294117647058825e-06, "loss": 0.9309, "step": 36 }, { "epoch": 0.021773511346500424, "grad_norm": 1.9545358147294818, "learning_rate": 1.0588235294117648e-06, "loss": 1.1121, "step": 37 }, { "epoch": 0.02236198462613557, "grad_norm": 2.0762201984762996, "learning_rate": 1.088235294117647e-06, "loss": 1.0779, "step": 38 }, { "epoch": 0.022950457905770716, "grad_norm": 2.0496025830697016, "learning_rate": 1.1176470588235296e-06, "loss": 1.0716, "step": 39 }, { "epoch": 0.023538931185405863, "grad_norm": 1.8139207824622237, "learning_rate": 1.1470588235294117e-06, "loss": 1.0911, "step": 40 }, { "epoch": 0.02412740446504101, "grad_norm": 1.8132643115132272, "learning_rate": 1.1764705882352942e-06, "loss": 0.9672, "step": 41 }, { "epoch": 0.024715877744676155, "grad_norm": 1.7202187427746571, "learning_rate": 1.2058823529411765e-06, "loss": 1.0021, "step": 42 }, { "epoch": 0.025304351024311302, "grad_norm": 1.8145636247375998, "learning_rate": 1.235294117647059e-06, "loss": 1.0317, "step": 43 }, { "epoch": 0.02589282430394645, "grad_norm": 1.8977108888386771, "learning_rate": 1.2647058823529412e-06, "loss": 1.0855, "step": 44 }, { "epoch": 0.026481297583581597, "grad_norm": 1.551136256288999, "learning_rate": 1.2941176470588237e-06, "loss": 1.02, "step": 45 }, { "epoch": 0.02706977086321674, "grad_norm": 1.6489150444646026, "learning_rate": 1.323529411764706e-06, "loss": 0.9181, "step": 46 }, { "epoch": 0.02765824414285189, "grad_norm": 1.6382713826995117, "learning_rate": 1.3529411764705883e-06, "loss": 1.1814, "step": 47 }, { "epoch": 0.028246717422487036, "grad_norm": 1.4923262364803602, "learning_rate": 1.3823529411764708e-06, "loss": 1.034, "step": 48 }, { "epoch": 0.028835190702122183, "grad_norm": 1.504202433912306, "learning_rate": 1.4117647058823531e-06, "loss": 1.0295, "step": 49 }, { "epoch": 0.029423663981757327, "grad_norm": 1.7570857041736776, "learning_rate": 1.4411764705882352e-06, "loss": 0.927, "step": 50 }, { "epoch": 0.030012137261392475, "grad_norm": 1.448154683134747, "learning_rate": 1.4705882352941177e-06, "loss": 0.9229, "step": 51 }, { "epoch": 0.030600610541027622, "grad_norm": 1.438317432530715, "learning_rate": 1.5e-06, "loss": 0.8994, "step": 52 }, { "epoch": 0.031189083820662766, "grad_norm": 1.5752763386417494, "learning_rate": 1.5294117647058826e-06, "loss": 0.8703, "step": 53 }, { "epoch": 0.031777557100297914, "grad_norm": 1.3235402643622718, "learning_rate": 1.5588235294117649e-06, "loss": 0.8938, "step": 54 }, { "epoch": 0.03236603037993306, "grad_norm": 1.5254234296134703, "learning_rate": 1.5882352941176472e-06, "loss": 1.192, "step": 55 }, { "epoch": 0.03295450365956821, "grad_norm": 1.2646148280990845, "learning_rate": 1.6176470588235297e-06, "loss": 0.9694, "step": 56 }, { "epoch": 0.033542976939203356, "grad_norm": 1.412991383472386, "learning_rate": 1.6470588235294118e-06, "loss": 0.9797, "step": 57 }, { "epoch": 0.034131450218838504, "grad_norm": 1.2740445437101842, "learning_rate": 1.676470588235294e-06, "loss": 0.9543, "step": 58 }, { "epoch": 0.034719923498473644, "grad_norm": 1.247916141803905, "learning_rate": 1.7058823529411766e-06, "loss": 1.0163, "step": 59 }, { "epoch": 0.03530839677810879, "grad_norm": 1.3408450126651783, "learning_rate": 1.735294117647059e-06, "loss": 1.0279, "step": 60 }, { "epoch": 0.03589687005774394, "grad_norm": 1.1414296435465283, "learning_rate": 1.7647058823529414e-06, "loss": 0.8584, "step": 61 }, { "epoch": 0.036485343337379086, "grad_norm": 1.1368932946122885, "learning_rate": 1.7941176470588238e-06, "loss": 0.867, "step": 62 }, { "epoch": 0.037073816617014234, "grad_norm": 1.2050311112620502, "learning_rate": 1.8235294117647058e-06, "loss": 0.9618, "step": 63 }, { "epoch": 0.03766228989664938, "grad_norm": 1.2328169134957196, "learning_rate": 1.8529411764705884e-06, "loss": 0.9144, "step": 64 }, { "epoch": 0.03825076317628453, "grad_norm": 1.1535775442736118, "learning_rate": 1.8823529411764707e-06, "loss": 0.8212, "step": 65 }, { "epoch": 0.038839236455919676, "grad_norm": 1.2602354176125725, "learning_rate": 1.9117647058823528e-06, "loss": 0.8947, "step": 66 }, { "epoch": 0.03942770973555482, "grad_norm": 1.1020963571399058, "learning_rate": 1.9411764705882353e-06, "loss": 0.8795, "step": 67 }, { "epoch": 0.040016183015189964, "grad_norm": 0.9629036018317353, "learning_rate": 1.970588235294118e-06, "loss": 0.7908, "step": 68 }, { "epoch": 0.04060465629482511, "grad_norm": 1.141110758674757, "learning_rate": 2.0000000000000003e-06, "loss": 1.0027, "step": 69 }, { "epoch": 0.04119312957446026, "grad_norm": 1.044949910941186, "learning_rate": 2.0294117647058824e-06, "loss": 0.8427, "step": 70 }, { "epoch": 0.041781602854095407, "grad_norm": 1.2446178188120232, "learning_rate": 2.058823529411765e-06, "loss": 0.9351, "step": 71 }, { "epoch": 0.042370076133730554, "grad_norm": 1.1300794799469234, "learning_rate": 2.0882352941176475e-06, "loss": 0.8198, "step": 72 }, { "epoch": 0.0429585494133657, "grad_norm": 1.023485750963968, "learning_rate": 2.1176470588235296e-06, "loss": 0.9004, "step": 73 }, { "epoch": 0.04354702269300085, "grad_norm": 0.9917818662069169, "learning_rate": 2.1470588235294117e-06, "loss": 0.8554, "step": 74 }, { "epoch": 0.04413549597263599, "grad_norm": 1.0977589103911267, "learning_rate": 2.176470588235294e-06, "loss": 0.9213, "step": 75 }, { "epoch": 0.04472396925227114, "grad_norm": 0.9849740590602571, "learning_rate": 2.2058823529411767e-06, "loss": 0.8483, "step": 76 }, { "epoch": 0.045312442531906284, "grad_norm": 0.8611792117365, "learning_rate": 2.2352941176470592e-06, "loss": 0.7704, "step": 77 }, { "epoch": 0.04590091581154143, "grad_norm": 0.9741366964327156, "learning_rate": 2.2647058823529413e-06, "loss": 0.8769, "step": 78 }, { "epoch": 0.04648938909117658, "grad_norm": 0.9544753580159859, "learning_rate": 2.2941176470588234e-06, "loss": 0.8421, "step": 79 }, { "epoch": 0.04707786237081173, "grad_norm": 0.9706022877921006, "learning_rate": 2.323529411764706e-06, "loss": 0.9278, "step": 80 }, { "epoch": 0.047666335650446874, "grad_norm": 0.781758536204594, "learning_rate": 2.3529411764705885e-06, "loss": 0.7525, "step": 81 }, { "epoch": 0.04825480893008202, "grad_norm": 0.925319280804056, "learning_rate": 2.3823529411764705e-06, "loss": 0.8032, "step": 82 }, { "epoch": 0.04884328220971716, "grad_norm": 1.117381715372415, "learning_rate": 2.411764705882353e-06, "loss": 0.7819, "step": 83 }, { "epoch": 0.04943175548935231, "grad_norm": 0.7735831709891251, "learning_rate": 2.4411764705882356e-06, "loss": 0.7314, "step": 84 }, { "epoch": 0.05002022876898746, "grad_norm": 1.0333137705506992, "learning_rate": 2.470588235294118e-06, "loss": 0.9376, "step": 85 }, { "epoch": 0.050608702048622604, "grad_norm": 0.8823378888652399, "learning_rate": 2.5e-06, "loss": 0.8059, "step": 86 }, { "epoch": 0.05119717532825775, "grad_norm": 0.9242675452167171, "learning_rate": 2.5294117647058823e-06, "loss": 0.8825, "step": 87 }, { "epoch": 0.0517856486078929, "grad_norm": 0.9570406084740753, "learning_rate": 2.558823529411765e-06, "loss": 0.8958, "step": 88 }, { "epoch": 0.05237412188752805, "grad_norm": 0.7272362993930886, "learning_rate": 2.5882352941176473e-06, "loss": 0.7145, "step": 89 }, { "epoch": 0.052962595167163194, "grad_norm": 0.9963530560398847, "learning_rate": 2.61764705882353e-06, "loss": 0.9142, "step": 90 }, { "epoch": 0.053551068446798335, "grad_norm": 0.9250964509183709, "learning_rate": 2.647058823529412e-06, "loss": 0.8572, "step": 91 }, { "epoch": 0.05413954172643348, "grad_norm": 0.874218303252959, "learning_rate": 2.6764705882352945e-06, "loss": 0.8261, "step": 92 }, { "epoch": 0.05472801500606863, "grad_norm": 0.7634434043874815, "learning_rate": 2.7058823529411766e-06, "loss": 0.8277, "step": 93 }, { "epoch": 0.05531648828570378, "grad_norm": 0.8976726061291317, "learning_rate": 2.7352941176470595e-06, "loss": 0.8479, "step": 94 }, { "epoch": 0.055904961565338925, "grad_norm": 0.8159830636598591, "learning_rate": 2.7647058823529416e-06, "loss": 0.8035, "step": 95 }, { "epoch": 0.05649343484497407, "grad_norm": 0.7144845762732425, "learning_rate": 2.7941176470588237e-06, "loss": 0.7504, "step": 96 }, { "epoch": 0.05708190812460922, "grad_norm": 0.6897043642265328, "learning_rate": 2.8235294117647062e-06, "loss": 0.6889, "step": 97 }, { "epoch": 0.05767038140424437, "grad_norm": 0.7190425915535908, "learning_rate": 2.8529411764705883e-06, "loss": 0.7397, "step": 98 }, { "epoch": 0.05825885468387951, "grad_norm": 0.707010558326179, "learning_rate": 2.8823529411764704e-06, "loss": 0.7721, "step": 99 }, { "epoch": 0.058847327963514655, "grad_norm": 0.7107478165198532, "learning_rate": 2.9117647058823534e-06, "loss": 0.7525, "step": 100 }, { "epoch": 0.0594358012431498, "grad_norm": 0.7908697342819719, "learning_rate": 2.9411764705882355e-06, "loss": 0.7843, "step": 101 }, { "epoch": 0.06002427452278495, "grad_norm": 0.754276176856247, "learning_rate": 2.970588235294118e-06, "loss": 0.6388, "step": 102 }, { "epoch": 0.0606127478024201, "grad_norm": 0.8219350092729553, "learning_rate": 3e-06, "loss": 0.8278, "step": 103 }, { "epoch": 0.061201221082055245, "grad_norm": 0.6426932377444309, "learning_rate": 3.0294117647058826e-06, "loss": 0.7279, "step": 104 }, { "epoch": 0.06178969436169039, "grad_norm": 0.8497603491106752, "learning_rate": 3.058823529411765e-06, "loss": 0.806, "step": 105 }, { "epoch": 0.06237816764132553, "grad_norm": 0.6770049560338024, "learning_rate": 3.0882352941176476e-06, "loss": 0.7108, "step": 106 }, { "epoch": 0.06296664092096069, "grad_norm": 0.7801051753009922, "learning_rate": 3.1176470588235297e-06, "loss": 0.6663, "step": 107 }, { "epoch": 0.06355511420059583, "grad_norm": 0.7404515672040582, "learning_rate": 3.147058823529412e-06, "loss": 0.8718, "step": 108 }, { "epoch": 0.06414358748023098, "grad_norm": 1.0823248592828054, "learning_rate": 3.1764705882352943e-06, "loss": 0.7272, "step": 109 }, { "epoch": 0.06473206075986612, "grad_norm": 0.7209482555414111, "learning_rate": 3.205882352941177e-06, "loss": 0.7474, "step": 110 }, { "epoch": 0.06532053403950126, "grad_norm": 0.753054957214645, "learning_rate": 3.2352941176470594e-06, "loss": 0.7412, "step": 111 }, { "epoch": 0.06590900731913642, "grad_norm": 0.9084689150091002, "learning_rate": 3.2647058823529415e-06, "loss": 0.8834, "step": 112 }, { "epoch": 0.06649748059877156, "grad_norm": 0.7780870405191123, "learning_rate": 3.2941176470588236e-06, "loss": 0.8004, "step": 113 }, { "epoch": 0.06708595387840671, "grad_norm": 0.6597074917945273, "learning_rate": 3.323529411764706e-06, "loss": 0.8004, "step": 114 }, { "epoch": 0.06767442715804185, "grad_norm": 0.6637563541076776, "learning_rate": 3.352941176470588e-06, "loss": 0.7374, "step": 115 }, { "epoch": 0.06826290043767701, "grad_norm": 0.6372071897145712, "learning_rate": 3.382352941176471e-06, "loss": 0.5855, "step": 116 }, { "epoch": 0.06885137371731215, "grad_norm": 0.6675442809651566, "learning_rate": 3.4117647058823532e-06, "loss": 0.674, "step": 117 }, { "epoch": 0.06943984699694729, "grad_norm": 0.8757630398176597, "learning_rate": 3.4411764705882358e-06, "loss": 0.8153, "step": 118 }, { "epoch": 0.07002832027658244, "grad_norm": 0.6956538721929239, "learning_rate": 3.470588235294118e-06, "loss": 0.7992, "step": 119 }, { "epoch": 0.07061679355621758, "grad_norm": 0.7855340757369211, "learning_rate": 3.5e-06, "loss": 0.7974, "step": 120 }, { "epoch": 0.07120526683585274, "grad_norm": 0.6058397561881602, "learning_rate": 3.529411764705883e-06, "loss": 0.6468, "step": 121 }, { "epoch": 0.07179374011548788, "grad_norm": 0.6487755269611208, "learning_rate": 3.558823529411765e-06, "loss": 0.699, "step": 122 }, { "epoch": 0.07238221339512303, "grad_norm": 0.7901084166783627, "learning_rate": 3.5882352941176475e-06, "loss": 0.8363, "step": 123 }, { "epoch": 0.07297068667475817, "grad_norm": 0.6413357746586161, "learning_rate": 3.6176470588235296e-06, "loss": 0.7179, "step": 124 }, { "epoch": 0.07355915995439333, "grad_norm": 0.6397471945889088, "learning_rate": 3.6470588235294117e-06, "loss": 0.683, "step": 125 }, { "epoch": 0.07414763323402847, "grad_norm": 0.6930580780226645, "learning_rate": 3.6764705882352946e-06, "loss": 0.7395, "step": 126 }, { "epoch": 0.07473610651366361, "grad_norm": 0.8194558307737906, "learning_rate": 3.7058823529411767e-06, "loss": 0.7257, "step": 127 }, { "epoch": 0.07532457979329876, "grad_norm": 0.6470559646760365, "learning_rate": 3.7352941176470593e-06, "loss": 0.7207, "step": 128 }, { "epoch": 0.0759130530729339, "grad_norm": 0.644305421712633, "learning_rate": 3.7647058823529414e-06, "loss": 0.7662, "step": 129 }, { "epoch": 0.07650152635256906, "grad_norm": 0.6623798397772546, "learning_rate": 3.794117647058824e-06, "loss": 0.6942, "step": 130 }, { "epoch": 0.0770899996322042, "grad_norm": 0.7272152003203826, "learning_rate": 3.8235294117647055e-06, "loss": 0.7446, "step": 131 }, { "epoch": 0.07767847291183935, "grad_norm": 0.5745235866547446, "learning_rate": 3.852941176470589e-06, "loss": 0.6564, "step": 132 }, { "epoch": 0.07826694619147449, "grad_norm": 0.7231769501148295, "learning_rate": 3.882352941176471e-06, "loss": 0.8105, "step": 133 }, { "epoch": 0.07885541947110963, "grad_norm": 0.5664739983384839, "learning_rate": 3.911764705882353e-06, "loss": 0.643, "step": 134 }, { "epoch": 0.07944389275074479, "grad_norm": 0.7964097726095265, "learning_rate": 3.941176470588236e-06, "loss": 0.7907, "step": 135 }, { "epoch": 0.08003236603037993, "grad_norm": 0.8052979450190245, "learning_rate": 3.970588235294118e-06, "loss": 0.7199, "step": 136 }, { "epoch": 0.08062083931001508, "grad_norm": 0.8305938189306179, "learning_rate": 4.000000000000001e-06, "loss": 0.7746, "step": 137 }, { "epoch": 0.08120931258965022, "grad_norm": 0.6300750443203864, "learning_rate": 4.029411764705883e-06, "loss": 0.7464, "step": 138 }, { "epoch": 0.08179778586928538, "grad_norm": 0.6454175401063073, "learning_rate": 4.058823529411765e-06, "loss": 0.6488, "step": 139 }, { "epoch": 0.08238625914892052, "grad_norm": 0.7151863740172832, "learning_rate": 4.088235294117647e-06, "loss": 0.7112, "step": 140 }, { "epoch": 0.08297473242855567, "grad_norm": 0.6368151428067, "learning_rate": 4.11764705882353e-06, "loss": 0.6298, "step": 141 }, { "epoch": 0.08356320570819081, "grad_norm": 0.5795336585568585, "learning_rate": 4.147058823529412e-06, "loss": 0.6719, "step": 142 }, { "epoch": 0.08415167898782595, "grad_norm": 0.5527244880823492, "learning_rate": 4.176470588235295e-06, "loss": 0.6066, "step": 143 }, { "epoch": 0.08474015226746111, "grad_norm": 0.5906013311869461, "learning_rate": 4.205882352941177e-06, "loss": 0.6469, "step": 144 }, { "epoch": 0.08532862554709625, "grad_norm": 0.6941099106004949, "learning_rate": 4.235294117647059e-06, "loss": 0.7701, "step": 145 }, { "epoch": 0.0859170988267314, "grad_norm": 0.7750287373629081, "learning_rate": 4.264705882352942e-06, "loss": 0.6785, "step": 146 }, { "epoch": 0.08650557210636654, "grad_norm": 0.8990905023056512, "learning_rate": 4.294117647058823e-06, "loss": 0.8276, "step": 147 }, { "epoch": 0.0870940453860017, "grad_norm": 0.8195109648301305, "learning_rate": 4.323529411764707e-06, "loss": 0.7188, "step": 148 }, { "epoch": 0.08768251866563684, "grad_norm": 0.6266184493531808, "learning_rate": 4.352941176470588e-06, "loss": 0.6949, "step": 149 }, { "epoch": 0.08827099194527198, "grad_norm": 0.873327112196409, "learning_rate": 4.382352941176471e-06, "loss": 0.7464, "step": 150 }, { "epoch": 0.08885946522490713, "grad_norm": 0.6975873469791097, "learning_rate": 4.411764705882353e-06, "loss": 0.7533, "step": 151 }, { "epoch": 0.08944793850454227, "grad_norm": 0.63037613392637, "learning_rate": 4.441176470588235e-06, "loss": 0.6491, "step": 152 }, { "epoch": 0.09003641178417743, "grad_norm": 0.6698002237693914, "learning_rate": 4.4705882352941184e-06, "loss": 0.7077, "step": 153 }, { "epoch": 0.09062488506381257, "grad_norm": 0.5504896566981474, "learning_rate": 4.5e-06, "loss": 0.6395, "step": 154 }, { "epoch": 0.09121335834344772, "grad_norm": 0.5822662171948179, "learning_rate": 4.529411764705883e-06, "loss": 0.5922, "step": 155 }, { "epoch": 0.09180183162308286, "grad_norm": 0.6099798768830045, "learning_rate": 4.558823529411765e-06, "loss": 0.6248, "step": 156 }, { "epoch": 0.09239030490271802, "grad_norm": 0.8431123682828167, "learning_rate": 4.588235294117647e-06, "loss": 0.6014, "step": 157 }, { "epoch": 0.09297877818235316, "grad_norm": 0.7315889045178279, "learning_rate": 4.61764705882353e-06, "loss": 0.7852, "step": 158 }, { "epoch": 0.0935672514619883, "grad_norm": 0.6134109558531912, "learning_rate": 4.647058823529412e-06, "loss": 0.698, "step": 159 }, { "epoch": 0.09415572474162345, "grad_norm": 0.7311491093467011, "learning_rate": 4.676470588235294e-06, "loss": 0.7743, "step": 160 }, { "epoch": 0.0947441980212586, "grad_norm": 0.5388486068773455, "learning_rate": 4.705882352941177e-06, "loss": 0.6209, "step": 161 }, { "epoch": 0.09533267130089375, "grad_norm": 0.7115573732850511, "learning_rate": 4.7352941176470594e-06, "loss": 0.6893, "step": 162 }, { "epoch": 0.09592114458052889, "grad_norm": 0.6216449438743102, "learning_rate": 4.764705882352941e-06, "loss": 0.6802, "step": 163 }, { "epoch": 0.09650961786016404, "grad_norm": 0.5778089546092825, "learning_rate": 4.7941176470588245e-06, "loss": 0.6918, "step": 164 }, { "epoch": 0.09709809113979918, "grad_norm": 0.5201582281880245, "learning_rate": 4.823529411764706e-06, "loss": 0.6032, "step": 165 }, { "epoch": 0.09768656441943432, "grad_norm": 0.6900163016716859, "learning_rate": 4.852941176470589e-06, "loss": 0.6468, "step": 166 }, { "epoch": 0.09827503769906948, "grad_norm": 0.7227422462588133, "learning_rate": 4.882352941176471e-06, "loss": 0.6362, "step": 167 }, { "epoch": 0.09886351097870462, "grad_norm": 0.5700667298145622, "learning_rate": 4.911764705882353e-06, "loss": 0.6396, "step": 168 }, { "epoch": 0.09945198425833977, "grad_norm": 0.6488806538430971, "learning_rate": 4.941176470588236e-06, "loss": 0.673, "step": 169 }, { "epoch": 0.10004045753797491, "grad_norm": 0.6980049678527189, "learning_rate": 4.970588235294118e-06, "loss": 0.7143, "step": 170 }, { "epoch": 0.10062893081761007, "grad_norm": 0.5583095269864528, "learning_rate": 5e-06, "loss": 0.6108, "step": 171 }, { "epoch": 0.10121740409724521, "grad_norm": 0.7752801286702317, "learning_rate": 5.029411764705883e-06, "loss": 0.8498, "step": 172 }, { "epoch": 0.10180587737688035, "grad_norm": 0.5441760909585082, "learning_rate": 5.058823529411765e-06, "loss": 0.668, "step": 173 }, { "epoch": 0.1023943506565155, "grad_norm": 0.570373102360756, "learning_rate": 5.088235294117647e-06, "loss": 0.625, "step": 174 }, { "epoch": 0.10298282393615064, "grad_norm": 0.8056887954753529, "learning_rate": 5.11764705882353e-06, "loss": 0.7911, "step": 175 }, { "epoch": 0.1035712972157858, "grad_norm": 0.613735411994671, "learning_rate": 5.147058823529411e-06, "loss": 0.6378, "step": 176 }, { "epoch": 0.10415977049542094, "grad_norm": 0.7089397506170212, "learning_rate": 5.176470588235295e-06, "loss": 0.7518, "step": 177 }, { "epoch": 0.1047482437750561, "grad_norm": 0.6518371885902531, "learning_rate": 5.205882352941177e-06, "loss": 0.7224, "step": 178 }, { "epoch": 0.10533671705469123, "grad_norm": 0.6814789056852261, "learning_rate": 5.23529411764706e-06, "loss": 0.7776, "step": 179 }, { "epoch": 0.10592519033432639, "grad_norm": 0.6518185715973227, "learning_rate": 5.264705882352941e-06, "loss": 0.7286, "step": 180 }, { "epoch": 0.10651366361396153, "grad_norm": 0.6691945794722888, "learning_rate": 5.294117647058824e-06, "loss": 0.6769, "step": 181 }, { "epoch": 0.10710213689359667, "grad_norm": 0.6663952772930075, "learning_rate": 5.3235294117647064e-06, "loss": 0.6976, "step": 182 }, { "epoch": 0.10769061017323182, "grad_norm": 0.6835756978564436, "learning_rate": 5.352941176470589e-06, "loss": 0.7479, "step": 183 }, { "epoch": 0.10827908345286696, "grad_norm": 0.6665841191098142, "learning_rate": 5.382352941176471e-06, "loss": 0.7249, "step": 184 }, { "epoch": 0.10886755673250212, "grad_norm": 0.6296097301245284, "learning_rate": 5.411764705882353e-06, "loss": 0.7297, "step": 185 }, { "epoch": 0.10945603001213726, "grad_norm": 0.6069402274332087, "learning_rate": 5.441176470588236e-06, "loss": 0.6887, "step": 186 }, { "epoch": 0.11004450329177241, "grad_norm": 0.6668021672039735, "learning_rate": 5.470588235294119e-06, "loss": 0.759, "step": 187 }, { "epoch": 0.11063297657140755, "grad_norm": 0.6436765426915717, "learning_rate": 5.500000000000001e-06, "loss": 0.6918, "step": 188 }, { "epoch": 0.1112214498510427, "grad_norm": 0.5724575795288158, "learning_rate": 5.529411764705883e-06, "loss": 0.6788, "step": 189 }, { "epoch": 0.11180992313067785, "grad_norm": 0.6952810671674433, "learning_rate": 5.558823529411766e-06, "loss": 0.8032, "step": 190 }, { "epoch": 0.11239839641031299, "grad_norm": 0.5770305251987676, "learning_rate": 5.588235294117647e-06, "loss": 0.6725, "step": 191 }, { "epoch": 0.11298686968994814, "grad_norm": 0.581143385799274, "learning_rate": 5.61764705882353e-06, "loss": 0.6856, "step": 192 }, { "epoch": 0.11357534296958328, "grad_norm": 0.5767663342935295, "learning_rate": 5.6470588235294125e-06, "loss": 0.6038, "step": 193 }, { "epoch": 0.11416381624921844, "grad_norm": 0.7065764166700359, "learning_rate": 5.676470588235294e-06, "loss": 0.6493, "step": 194 }, { "epoch": 0.11475228952885358, "grad_norm": 0.5774213793916674, "learning_rate": 5.705882352941177e-06, "loss": 0.6177, "step": 195 }, { "epoch": 0.11534076280848873, "grad_norm": 0.627826939485883, "learning_rate": 5.735294117647059e-06, "loss": 0.7021, "step": 196 }, { "epoch": 0.11592923608812387, "grad_norm": 0.6868763463049803, "learning_rate": 5.764705882352941e-06, "loss": 0.7533, "step": 197 }, { "epoch": 0.11651770936775901, "grad_norm": 0.6335966733759173, "learning_rate": 5.794117647058824e-06, "loss": 0.6871, "step": 198 }, { "epoch": 0.11710618264739417, "grad_norm": 0.5622250814287315, "learning_rate": 5.823529411764707e-06, "loss": 0.5826, "step": 199 }, { "epoch": 0.11769465592702931, "grad_norm": 0.8357111322856082, "learning_rate": 5.852941176470589e-06, "loss": 0.7669, "step": 200 }, { "epoch": 0.11828312920666446, "grad_norm": 0.680984886985973, "learning_rate": 5.882352941176471e-06, "loss": 0.7763, "step": 201 }, { "epoch": 0.1188716024862996, "grad_norm": 0.6439217068644828, "learning_rate": 5.9117647058823534e-06, "loss": 0.6785, "step": 202 }, { "epoch": 0.11946007576593476, "grad_norm": 0.7088650187853981, "learning_rate": 5.941176470588236e-06, "loss": 0.7862, "step": 203 }, { "epoch": 0.1200485490455699, "grad_norm": 0.6581286916279897, "learning_rate": 5.970588235294118e-06, "loss": 0.6281, "step": 204 }, { "epoch": 0.12063702232520504, "grad_norm": 0.5576740057042101, "learning_rate": 6e-06, "loss": 0.6136, "step": 205 }, { "epoch": 0.1212254956048402, "grad_norm": 0.5930924673190601, "learning_rate": 6.029411764705883e-06, "loss": 0.6736, "step": 206 }, { "epoch": 0.12181396888447533, "grad_norm": 0.7084250462237611, "learning_rate": 6.058823529411765e-06, "loss": 0.6047, "step": 207 }, { "epoch": 0.12240244216411049, "grad_norm": 0.7459637233547837, "learning_rate": 6.088235294117647e-06, "loss": 0.686, "step": 208 }, { "epoch": 0.12299091544374563, "grad_norm": 0.5898227407245208, "learning_rate": 6.11764705882353e-06, "loss": 0.6451, "step": 209 }, { "epoch": 0.12357938872338078, "grad_norm": 0.6436070200675544, "learning_rate": 6.147058823529413e-06, "loss": 0.6904, "step": 210 }, { "epoch": 0.12416786200301592, "grad_norm": 0.60483431640519, "learning_rate": 6.176470588235295e-06, "loss": 0.6763, "step": 211 }, { "epoch": 0.12475633528265107, "grad_norm": 0.635468801737919, "learning_rate": 6.205882352941177e-06, "loss": 0.6464, "step": 212 }, { "epoch": 0.12534480856228622, "grad_norm": 0.5131747239718439, "learning_rate": 6.2352941176470595e-06, "loss": 0.556, "step": 213 }, { "epoch": 0.12593328184192137, "grad_norm": 0.6478477001570558, "learning_rate": 6.264705882352942e-06, "loss": 0.7311, "step": 214 }, { "epoch": 0.1265217551215565, "grad_norm": 0.5539134069796678, "learning_rate": 6.294117647058824e-06, "loss": 0.6366, "step": 215 }, { "epoch": 0.12711022840119166, "grad_norm": 0.5472541640098021, "learning_rate": 6.323529411764706e-06, "loss": 0.5338, "step": 216 }, { "epoch": 0.1276987016808268, "grad_norm": 0.5838535685999099, "learning_rate": 6.352941176470589e-06, "loss": 0.5665, "step": 217 }, { "epoch": 0.12828717496046196, "grad_norm": 0.6934121320311707, "learning_rate": 6.38235294117647e-06, "loss": 0.8046, "step": 218 }, { "epoch": 0.1288756482400971, "grad_norm": 0.6499456484404326, "learning_rate": 6.411764705882354e-06, "loss": 0.6573, "step": 219 }, { "epoch": 0.12946412151973224, "grad_norm": 0.6719994619101402, "learning_rate": 6.441176470588236e-06, "loss": 0.6394, "step": 220 }, { "epoch": 0.1300525947993674, "grad_norm": 0.5145216579208479, "learning_rate": 6.470588235294119e-06, "loss": 0.623, "step": 221 }, { "epoch": 0.13064106807900253, "grad_norm": 0.5835581384364646, "learning_rate": 6.5000000000000004e-06, "loss": 0.6193, "step": 222 }, { "epoch": 0.13122954135863768, "grad_norm": 0.7604852072914522, "learning_rate": 6.529411764705883e-06, "loss": 0.7024, "step": 223 }, { "epoch": 0.13181801463827283, "grad_norm": 0.5482528688430366, "learning_rate": 6.5588235294117655e-06, "loss": 0.6741, "step": 224 }, { "epoch": 0.132406487917908, "grad_norm": 0.685577056631993, "learning_rate": 6.588235294117647e-06, "loss": 0.6804, "step": 225 }, { "epoch": 0.13299496119754312, "grad_norm": 0.5640092376118345, "learning_rate": 6.61764705882353e-06, "loss": 0.6188, "step": 226 }, { "epoch": 0.13358343447717827, "grad_norm": 0.6506027301225134, "learning_rate": 6.647058823529412e-06, "loss": 0.7752, "step": 227 }, { "epoch": 0.13417190775681342, "grad_norm": 0.6463248664176454, "learning_rate": 6.676470588235294e-06, "loss": 0.6544, "step": 228 }, { "epoch": 0.13476038103644855, "grad_norm": 0.5217560796883516, "learning_rate": 6.705882352941176e-06, "loss": 0.6242, "step": 229 }, { "epoch": 0.1353488543160837, "grad_norm": 0.5955741207103803, "learning_rate": 6.73529411764706e-06, "loss": 0.7258, "step": 230 }, { "epoch": 0.13593732759571886, "grad_norm": 0.4666478152571619, "learning_rate": 6.764705882352942e-06, "loss": 0.5555, "step": 231 }, { "epoch": 0.13652580087535401, "grad_norm": 0.5502756508112803, "learning_rate": 6.794117647058824e-06, "loss": 0.638, "step": 232 }, { "epoch": 0.13711427415498914, "grad_norm": 0.5698125272988078, "learning_rate": 6.8235294117647065e-06, "loss": 0.6392, "step": 233 }, { "epoch": 0.1377027474346243, "grad_norm": 0.7153385529641074, "learning_rate": 6.852941176470589e-06, "loss": 0.7403, "step": 234 }, { "epoch": 0.13829122071425945, "grad_norm": 0.5710419647396174, "learning_rate": 6.8823529411764715e-06, "loss": 0.5877, "step": 235 }, { "epoch": 0.13887969399389458, "grad_norm": 0.5322730448614496, "learning_rate": 6.911764705882353e-06, "loss": 0.6423, "step": 236 }, { "epoch": 0.13946816727352973, "grad_norm": 0.6076129909227577, "learning_rate": 6.941176470588236e-06, "loss": 0.5974, "step": 237 }, { "epoch": 0.14005664055316489, "grad_norm": 0.6019279944465069, "learning_rate": 6.970588235294118e-06, "loss": 0.633, "step": 238 }, { "epoch": 0.14064511383280004, "grad_norm": 0.5855103854080802, "learning_rate": 7e-06, "loss": 0.6574, "step": 239 }, { "epoch": 0.14123358711243517, "grad_norm": 0.4477508137933101, "learning_rate": 7.029411764705882e-06, "loss": 0.5594, "step": 240 }, { "epoch": 0.14182206039207032, "grad_norm": 0.5933142131625009, "learning_rate": 7.058823529411766e-06, "loss": 0.6141, "step": 241 }, { "epoch": 0.14241053367170547, "grad_norm": 0.7246924782068748, "learning_rate": 7.088235294117648e-06, "loss": 0.7149, "step": 242 }, { "epoch": 0.14299900695134063, "grad_norm": 0.756075427915649, "learning_rate": 7.11764705882353e-06, "loss": 0.7328, "step": 243 }, { "epoch": 0.14358748023097576, "grad_norm": 0.8059855525123575, "learning_rate": 7.1470588235294125e-06, "loss": 0.6915, "step": 244 }, { "epoch": 0.1441759535106109, "grad_norm": 0.5538477579614607, "learning_rate": 7.176470588235295e-06, "loss": 0.6307, "step": 245 }, { "epoch": 0.14476442679024606, "grad_norm": 0.6135737505744773, "learning_rate": 7.205882352941177e-06, "loss": 0.643, "step": 246 }, { "epoch": 0.1453529000698812, "grad_norm": 0.5743784682022813, "learning_rate": 7.235294117647059e-06, "loss": 0.628, "step": 247 }, { "epoch": 0.14594137334951635, "grad_norm": 0.5647716107953951, "learning_rate": 7.264705882352942e-06, "loss": 0.6091, "step": 248 }, { "epoch": 0.1465298466291515, "grad_norm": 0.734835682068792, "learning_rate": 7.294117647058823e-06, "loss": 0.7358, "step": 249 }, { "epoch": 0.14711831990878665, "grad_norm": 0.5373751736451209, "learning_rate": 7.323529411764706e-06, "loss": 0.5779, "step": 250 }, { "epoch": 0.14770679318842178, "grad_norm": 0.533386729821521, "learning_rate": 7.352941176470589e-06, "loss": 0.6446, "step": 251 }, { "epoch": 0.14829526646805694, "grad_norm": 0.5661161374667835, "learning_rate": 7.382352941176472e-06, "loss": 0.5978, "step": 252 }, { "epoch": 0.1488837397476921, "grad_norm": 0.6550641926629402, "learning_rate": 7.4117647058823535e-06, "loss": 0.7469, "step": 253 }, { "epoch": 0.14947221302732722, "grad_norm": 0.5531253814651138, "learning_rate": 7.441176470588236e-06, "loss": 0.6085, "step": 254 }, { "epoch": 0.15006068630696237, "grad_norm": 0.6335458297684669, "learning_rate": 7.4705882352941185e-06, "loss": 0.7335, "step": 255 }, { "epoch": 0.15064915958659753, "grad_norm": 0.5285593558114037, "learning_rate": 7.500000000000001e-06, "loss": 0.5699, "step": 256 }, { "epoch": 0.15123763286623268, "grad_norm": 0.6095402820941969, "learning_rate": 7.529411764705883e-06, "loss": 0.7183, "step": 257 }, { "epoch": 0.1518261061458678, "grad_norm": 0.6215452983111865, "learning_rate": 7.558823529411765e-06, "loss": 0.674, "step": 258 }, { "epoch": 0.15241457942550296, "grad_norm": 0.598612062852326, "learning_rate": 7.588235294117648e-06, "loss": 0.661, "step": 259 }, { "epoch": 0.15300305270513812, "grad_norm": 0.5497708505337561, "learning_rate": 7.617647058823529e-06, "loss": 0.6175, "step": 260 }, { "epoch": 0.15359152598477324, "grad_norm": 0.5687456217848029, "learning_rate": 7.647058823529411e-06, "loss": 0.6538, "step": 261 }, { "epoch": 0.1541799992644084, "grad_norm": 0.6503939520573805, "learning_rate": 7.676470588235294e-06, "loss": 0.5804, "step": 262 }, { "epoch": 0.15476847254404355, "grad_norm": 0.6543559521574499, "learning_rate": 7.705882352941178e-06, "loss": 0.6494, "step": 263 }, { "epoch": 0.1553569458236787, "grad_norm": 0.6004452939366555, "learning_rate": 7.73529411764706e-06, "loss": 0.6538, "step": 264 }, { "epoch": 0.15594541910331383, "grad_norm": 0.5919991060376325, "learning_rate": 7.764705882352941e-06, "loss": 0.6445, "step": 265 }, { "epoch": 0.15653389238294899, "grad_norm": 0.5088520413604722, "learning_rate": 7.794117647058825e-06, "loss": 0.5885, "step": 266 }, { "epoch": 0.15712236566258414, "grad_norm": 0.6473337694024841, "learning_rate": 7.823529411764706e-06, "loss": 0.6433, "step": 267 }, { "epoch": 0.15771083894221927, "grad_norm": 0.6318593343543398, "learning_rate": 7.85294117647059e-06, "loss": 0.7513, "step": 268 }, { "epoch": 0.15829931222185442, "grad_norm": 0.5638718927102779, "learning_rate": 7.882352941176471e-06, "loss": 0.6319, "step": 269 }, { "epoch": 0.15888778550148958, "grad_norm": 0.5040797810284929, "learning_rate": 7.911764705882353e-06, "loss": 0.5646, "step": 270 }, { "epoch": 0.15947625878112473, "grad_norm": 0.5635365513627102, "learning_rate": 7.941176470588236e-06, "loss": 0.5649, "step": 271 }, { "epoch": 0.16006473206075986, "grad_norm": 0.6590199129372637, "learning_rate": 7.97058823529412e-06, "loss": 0.6548, "step": 272 }, { "epoch": 0.160653205340395, "grad_norm": 0.5683639695913217, "learning_rate": 8.000000000000001e-06, "loss": 0.7087, "step": 273 }, { "epoch": 0.16124167862003017, "grad_norm": 0.5246451221640207, "learning_rate": 8.029411764705883e-06, "loss": 0.5631, "step": 274 }, { "epoch": 0.1618301518996653, "grad_norm": 0.5588149001847457, "learning_rate": 8.058823529411766e-06, "loss": 0.519, "step": 275 }, { "epoch": 0.16241862517930045, "grad_norm": 0.5378995288590399, "learning_rate": 8.088235294117648e-06, "loss": 0.5898, "step": 276 }, { "epoch": 0.1630070984589356, "grad_norm": 0.5395266364982481, "learning_rate": 8.11764705882353e-06, "loss": 0.5807, "step": 277 }, { "epoch": 0.16359557173857076, "grad_norm": 0.6732711565757394, "learning_rate": 8.147058823529413e-06, "loss": 0.5888, "step": 278 }, { "epoch": 0.16418404501820588, "grad_norm": 0.6300010173661125, "learning_rate": 8.176470588235295e-06, "loss": 0.6041, "step": 279 }, { "epoch": 0.16477251829784104, "grad_norm": 0.5953110373915832, "learning_rate": 8.205882352941176e-06, "loss": 0.6177, "step": 280 }, { "epoch": 0.1653609915774762, "grad_norm": 0.5797599517555226, "learning_rate": 8.23529411764706e-06, "loss": 0.6077, "step": 281 }, { "epoch": 0.16594946485711135, "grad_norm": 0.5795836458301701, "learning_rate": 8.264705882352941e-06, "loss": 0.6954, "step": 282 }, { "epoch": 0.16653793813674647, "grad_norm": 0.5469985477876862, "learning_rate": 8.294117647058825e-06, "loss": 0.5902, "step": 283 }, { "epoch": 0.16712641141638163, "grad_norm": 0.7371727008992566, "learning_rate": 8.323529411764707e-06, "loss": 0.6683, "step": 284 }, { "epoch": 0.16771488469601678, "grad_norm": 0.5711246881485778, "learning_rate": 8.35294117647059e-06, "loss": 0.6931, "step": 285 }, { "epoch": 0.1683033579756519, "grad_norm": 0.5789519932982666, "learning_rate": 8.382352941176472e-06, "loss": 0.703, "step": 286 }, { "epoch": 0.16889183125528706, "grad_norm": 0.6523147447429422, "learning_rate": 8.411764705882353e-06, "loss": 0.6755, "step": 287 }, { "epoch": 0.16948030453492222, "grad_norm": 0.5437716288458889, "learning_rate": 8.441176470588237e-06, "loss": 0.6293, "step": 288 }, { "epoch": 0.17006877781455737, "grad_norm": 0.6150481924984925, "learning_rate": 8.470588235294118e-06, "loss": 0.6303, "step": 289 }, { "epoch": 0.1706572510941925, "grad_norm": 0.5539560646862446, "learning_rate": 8.5e-06, "loss": 0.6334, "step": 290 }, { "epoch": 0.17124572437382765, "grad_norm": 0.5750081515158385, "learning_rate": 8.529411764705883e-06, "loss": 0.6086, "step": 291 }, { "epoch": 0.1718341976534628, "grad_norm": 0.5551514468204218, "learning_rate": 8.558823529411765e-06, "loss": 0.6358, "step": 292 }, { "epoch": 0.17242267093309793, "grad_norm": 0.5834833807531306, "learning_rate": 8.588235294117647e-06, "loss": 0.6403, "step": 293 }, { "epoch": 0.1730111442127331, "grad_norm": 0.5971259073654892, "learning_rate": 8.61764705882353e-06, "loss": 0.7345, "step": 294 }, { "epoch": 0.17359961749236824, "grad_norm": 0.5881404885762963, "learning_rate": 8.647058823529413e-06, "loss": 0.5998, "step": 295 }, { "epoch": 0.1741880907720034, "grad_norm": 0.6512998209040352, "learning_rate": 8.676470588235295e-06, "loss": 0.6015, "step": 296 }, { "epoch": 0.17477656405163852, "grad_norm": 0.5311409785596524, "learning_rate": 8.705882352941177e-06, "loss": 0.6969, "step": 297 }, { "epoch": 0.17536503733127368, "grad_norm": 0.5096159090742605, "learning_rate": 8.73529411764706e-06, "loss": 0.5181, "step": 298 }, { "epoch": 0.17595351061090883, "grad_norm": 0.5645393785728908, "learning_rate": 8.764705882352942e-06, "loss": 0.5911, "step": 299 }, { "epoch": 0.17654198389054396, "grad_norm": 0.6253577324417849, "learning_rate": 8.794117647058823e-06, "loss": 0.679, "step": 300 }, { "epoch": 0.1771304571701791, "grad_norm": 0.6283982225555622, "learning_rate": 8.823529411764707e-06, "loss": 0.6033, "step": 301 }, { "epoch": 0.17771893044981427, "grad_norm": 0.5861330491173634, "learning_rate": 8.852941176470588e-06, "loss": 0.7372, "step": 302 }, { "epoch": 0.17830740372944942, "grad_norm": 0.5804161076683354, "learning_rate": 8.88235294117647e-06, "loss": 0.5481, "step": 303 }, { "epoch": 0.17889587700908455, "grad_norm": 0.6386522019438818, "learning_rate": 8.911764705882354e-06, "loss": 0.6897, "step": 304 }, { "epoch": 0.1794843502887197, "grad_norm": 0.5963699092389586, "learning_rate": 8.941176470588237e-06, "loss": 0.6587, "step": 305 }, { "epoch": 0.18007282356835486, "grad_norm": 0.554424586646233, "learning_rate": 8.970588235294119e-06, "loss": 0.6425, "step": 306 }, { "epoch": 0.18066129684798998, "grad_norm": 0.602520350567725, "learning_rate": 9e-06, "loss": 0.6837, "step": 307 }, { "epoch": 0.18124977012762514, "grad_norm": 0.582456459507909, "learning_rate": 9.029411764705884e-06, "loss": 0.623, "step": 308 }, { "epoch": 0.1818382434072603, "grad_norm": 0.5520394658040876, "learning_rate": 9.058823529411765e-06, "loss": 0.5957, "step": 309 }, { "epoch": 0.18242671668689545, "grad_norm": 0.6299163954366428, "learning_rate": 9.088235294117647e-06, "loss": 0.6606, "step": 310 }, { "epoch": 0.18301518996653057, "grad_norm": 0.6096407173075333, "learning_rate": 9.11764705882353e-06, "loss": 0.6036, "step": 311 }, { "epoch": 0.18360366324616573, "grad_norm": 0.6686854647677282, "learning_rate": 9.147058823529412e-06, "loss": 0.7184, "step": 312 }, { "epoch": 0.18419213652580088, "grad_norm": 0.48738754189239236, "learning_rate": 9.176470588235294e-06, "loss": 0.5401, "step": 313 }, { "epoch": 0.18478060980543604, "grad_norm": 0.5335200926095099, "learning_rate": 9.205882352941177e-06, "loss": 0.626, "step": 314 }, { "epoch": 0.18536908308507116, "grad_norm": 0.46310196544425775, "learning_rate": 9.23529411764706e-06, "loss": 0.5366, "step": 315 }, { "epoch": 0.18595755636470632, "grad_norm": 0.5206018434520028, "learning_rate": 9.264705882352942e-06, "loss": 0.5623, "step": 316 }, { "epoch": 0.18654602964434147, "grad_norm": 0.5740965759538678, "learning_rate": 9.294117647058824e-06, "loss": 0.6117, "step": 317 }, { "epoch": 0.1871345029239766, "grad_norm": 0.43860912481668635, "learning_rate": 9.323529411764707e-06, "loss": 0.4957, "step": 318 }, { "epoch": 0.18772297620361175, "grad_norm": 0.5747124640943229, "learning_rate": 9.352941176470589e-06, "loss": 0.6631, "step": 319 }, { "epoch": 0.1883114494832469, "grad_norm": 0.5848167942399802, "learning_rate": 9.382352941176472e-06, "loss": 0.6371, "step": 320 }, { "epoch": 0.18889992276288206, "grad_norm": 0.5132436642107735, "learning_rate": 9.411764705882354e-06, "loss": 0.652, "step": 321 }, { "epoch": 0.1894883960425172, "grad_norm": 0.8908520649837621, "learning_rate": 9.441176470588235e-06, "loss": 0.6134, "step": 322 }, { "epoch": 0.19007686932215234, "grad_norm": 0.4686848779206886, "learning_rate": 9.470588235294119e-06, "loss": 0.5249, "step": 323 }, { "epoch": 0.1906653426017875, "grad_norm": 0.5380931288149583, "learning_rate": 9.5e-06, "loss": 0.6089, "step": 324 }, { "epoch": 0.19125381588142262, "grad_norm": 0.5630589842881866, "learning_rate": 9.529411764705882e-06, "loss": 0.6157, "step": 325 }, { "epoch": 0.19184228916105778, "grad_norm": 0.6697148941054103, "learning_rate": 9.558823529411766e-06, "loss": 0.5782, "step": 326 }, { "epoch": 0.19243076244069293, "grad_norm": 0.5993867910890379, "learning_rate": 9.588235294117649e-06, "loss": 0.6065, "step": 327 }, { "epoch": 0.1930192357203281, "grad_norm": 0.6649536260688523, "learning_rate": 9.61764705882353e-06, "loss": 0.602, "step": 328 }, { "epoch": 0.1936077089999632, "grad_norm": 0.46436692787140027, "learning_rate": 9.647058823529412e-06, "loss": 0.5425, "step": 329 }, { "epoch": 0.19419618227959837, "grad_norm": 0.5839568876679355, "learning_rate": 9.676470588235296e-06, "loss": 0.627, "step": 330 }, { "epoch": 0.19478465555923352, "grad_norm": 0.5574767521044424, "learning_rate": 9.705882352941177e-06, "loss": 0.5932, "step": 331 }, { "epoch": 0.19537312883886865, "grad_norm": 0.518120928099274, "learning_rate": 9.735294117647059e-06, "loss": 0.5903, "step": 332 }, { "epoch": 0.1959616021185038, "grad_norm": 0.5532337681830789, "learning_rate": 9.764705882352942e-06, "loss": 0.6383, "step": 333 }, { "epoch": 0.19655007539813896, "grad_norm": 0.5014462743449162, "learning_rate": 9.794117647058824e-06, "loss": 0.5398, "step": 334 }, { "epoch": 0.1971385486777741, "grad_norm": 0.5669973810665945, "learning_rate": 9.823529411764706e-06, "loss": 0.6081, "step": 335 }, { "epoch": 0.19772702195740924, "grad_norm": 0.5302651413775142, "learning_rate": 9.852941176470589e-06, "loss": 0.544, "step": 336 }, { "epoch": 0.1983154952370444, "grad_norm": 0.47525550943772626, "learning_rate": 9.882352941176472e-06, "loss": 0.5483, "step": 337 }, { "epoch": 0.19890396851667955, "grad_norm": 0.5361914407536916, "learning_rate": 9.911764705882354e-06, "loss": 0.6024, "step": 338 }, { "epoch": 0.19949244179631467, "grad_norm": 0.5275826657513204, "learning_rate": 9.941176470588236e-06, "loss": 0.567, "step": 339 }, { "epoch": 0.20008091507594983, "grad_norm": 0.48270926062185554, "learning_rate": 9.970588235294119e-06, "loss": 0.5884, "step": 340 }, { "epoch": 0.20066938835558498, "grad_norm": 0.5875131217243535, "learning_rate": 1e-05, "loss": 0.6115, "step": 341 }, { "epoch": 0.20125786163522014, "grad_norm": 0.5827033139156935, "learning_rate": 9.999997364901446e-06, "loss": 0.6485, "step": 342 }, { "epoch": 0.20184633491485526, "grad_norm": 0.42051379094719105, "learning_rate": 9.999989459608555e-06, "loss": 0.5101, "step": 343 }, { "epoch": 0.20243480819449042, "grad_norm": 0.4643013677384041, "learning_rate": 9.999976284129664e-06, "loss": 0.5376, "step": 344 }, { "epoch": 0.20302328147412557, "grad_norm": 0.4758528918712667, "learning_rate": 9.99995783847866e-06, "loss": 0.5463, "step": 345 }, { "epoch": 0.2036117547537607, "grad_norm": 0.5885186038223846, "learning_rate": 9.999934122674986e-06, "loss": 0.6098, "step": 346 }, { "epoch": 0.20420022803339585, "grad_norm": 0.6053662074902049, "learning_rate": 9.999905136743635e-06, "loss": 0.6439, "step": 347 }, { "epoch": 0.204788701313031, "grad_norm": 0.49450967147423563, "learning_rate": 9.999870880715163e-06, "loss": 0.5915, "step": 348 }, { "epoch": 0.20537717459266616, "grad_norm": 0.47420841352874804, "learning_rate": 9.999831354625678e-06, "loss": 0.5879, "step": 349 }, { "epoch": 0.2059656478723013, "grad_norm": 0.41100339755175497, "learning_rate": 9.99978655851684e-06, "loss": 0.4916, "step": 350 }, { "epoch": 0.20655412115193644, "grad_norm": 0.4991894864596591, "learning_rate": 9.999736492435867e-06, "loss": 0.5736, "step": 351 }, { "epoch": 0.2071425944315716, "grad_norm": 0.5005547952165861, "learning_rate": 9.999681156435528e-06, "loss": 0.5758, "step": 352 }, { "epoch": 0.20773106771120675, "grad_norm": 0.5305604805698183, "learning_rate": 9.999620550574155e-06, "loss": 0.5568, "step": 353 }, { "epoch": 0.20831954099084188, "grad_norm": 0.5511393391459933, "learning_rate": 9.999554674915621e-06, "loss": 0.6187, "step": 354 }, { "epoch": 0.20890801427047703, "grad_norm": 0.5148421740324445, "learning_rate": 9.999483529529369e-06, "loss": 0.6105, "step": 355 }, { "epoch": 0.2094964875501122, "grad_norm": 0.5630692642146141, "learning_rate": 9.999407114490384e-06, "loss": 0.6379, "step": 356 }, { "epoch": 0.2100849608297473, "grad_norm": 0.5651536912203585, "learning_rate": 9.999325429879215e-06, "loss": 0.7351, "step": 357 }, { "epoch": 0.21067343410938247, "grad_norm": 0.4903250940599811, "learning_rate": 9.999238475781957e-06, "loss": 0.5906, "step": 358 }, { "epoch": 0.21126190738901762, "grad_norm": 0.5433742675936746, "learning_rate": 9.999146252290264e-06, "loss": 0.6132, "step": 359 }, { "epoch": 0.21185038066865278, "grad_norm": 0.5101395527192613, "learning_rate": 9.999048759501343e-06, "loss": 0.612, "step": 360 }, { "epoch": 0.2124388539482879, "grad_norm": 0.5413456134723492, "learning_rate": 9.998945997517957e-06, "loss": 0.6496, "step": 361 }, { "epoch": 0.21302732722792306, "grad_norm": 0.5560640937377428, "learning_rate": 9.99883796644842e-06, "loss": 0.6332, "step": 362 }, { "epoch": 0.2136158005075582, "grad_norm": 0.5850700422636292, "learning_rate": 9.9987246664066e-06, "loss": 0.6925, "step": 363 }, { "epoch": 0.21420427378719334, "grad_norm": 0.5136864937602158, "learning_rate": 9.998606097511921e-06, "loss": 0.5646, "step": 364 }, { "epoch": 0.2147927470668285, "grad_norm": 0.5175247680146347, "learning_rate": 9.99848225988936e-06, "loss": 0.6411, "step": 365 }, { "epoch": 0.21538122034646365, "grad_norm": 0.47971856446089695, "learning_rate": 9.998353153669443e-06, "loss": 0.5666, "step": 366 }, { "epoch": 0.2159696936260988, "grad_norm": 0.5214463780079773, "learning_rate": 9.998218778988257e-06, "loss": 0.5636, "step": 367 }, { "epoch": 0.21655816690573393, "grad_norm": 0.5019301723586079, "learning_rate": 9.998079135987437e-06, "loss": 0.5663, "step": 368 }, { "epoch": 0.21714664018536908, "grad_norm": 0.4519028843713875, "learning_rate": 9.997934224814173e-06, "loss": 0.553, "step": 369 }, { "epoch": 0.21773511346500424, "grad_norm": 0.6009311083282598, "learning_rate": 9.997784045621205e-06, "loss": 0.6306, "step": 370 }, { "epoch": 0.21832358674463936, "grad_norm": 0.5135671572774999, "learning_rate": 9.99762859856683e-06, "loss": 0.6464, "step": 371 }, { "epoch": 0.21891206002427452, "grad_norm": 0.5659301388289658, "learning_rate": 9.997467883814894e-06, "loss": 0.6311, "step": 372 }, { "epoch": 0.21950053330390967, "grad_norm": 0.519304992305958, "learning_rate": 9.997301901534797e-06, "loss": 0.6771, "step": 373 }, { "epoch": 0.22008900658354483, "grad_norm": 0.5397863268546083, "learning_rate": 9.997130651901491e-06, "loss": 0.6441, "step": 374 }, { "epoch": 0.22067747986317995, "grad_norm": 0.5131617036373693, "learning_rate": 9.99695413509548e-06, "loss": 0.5835, "step": 375 }, { "epoch": 0.2212659531428151, "grad_norm": 0.48954478398105455, "learning_rate": 9.99677235130282e-06, "loss": 0.5744, "step": 376 }, { "epoch": 0.22185442642245026, "grad_norm": 0.5582431439864736, "learning_rate": 9.996585300715117e-06, "loss": 0.6484, "step": 377 }, { "epoch": 0.2224428997020854, "grad_norm": 0.6359601272259273, "learning_rate": 9.996392983529528e-06, "loss": 0.6836, "step": 378 }, { "epoch": 0.22303137298172054, "grad_norm": 0.5267655212281486, "learning_rate": 9.996195399948767e-06, "loss": 0.6209, "step": 379 }, { "epoch": 0.2236198462613557, "grad_norm": 0.4272176096218373, "learning_rate": 9.995992550181095e-06, "loss": 0.6152, "step": 380 }, { "epoch": 0.22420831954099085, "grad_norm": 0.5395578348409978, "learning_rate": 9.99578443444032e-06, "loss": 0.6347, "step": 381 }, { "epoch": 0.22479679282062598, "grad_norm": 0.51483865243508, "learning_rate": 9.995571052945807e-06, "loss": 0.6078, "step": 382 }, { "epoch": 0.22538526610026113, "grad_norm": 0.4776791638206742, "learning_rate": 9.995352405922467e-06, "loss": 0.6176, "step": 383 }, { "epoch": 0.2259737393798963, "grad_norm": 0.4416192487864194, "learning_rate": 9.995128493600763e-06, "loss": 0.5304, "step": 384 }, { "epoch": 0.22656221265953141, "grad_norm": 0.5306066004900354, "learning_rate": 9.994899316216709e-06, "loss": 0.6267, "step": 385 }, { "epoch": 0.22715068593916657, "grad_norm": 0.5003543143202668, "learning_rate": 9.994664874011864e-06, "loss": 0.6728, "step": 386 }, { "epoch": 0.22773915921880172, "grad_norm": 0.4749946501051026, "learning_rate": 9.99442516723334e-06, "loss": 0.5427, "step": 387 }, { "epoch": 0.22832763249843688, "grad_norm": 0.4778973458395408, "learning_rate": 9.9941801961338e-06, "loss": 0.5776, "step": 388 }, { "epoch": 0.228916105778072, "grad_norm": 0.527606646551393, "learning_rate": 9.99392996097145e-06, "loss": 0.6371, "step": 389 }, { "epoch": 0.22950457905770716, "grad_norm": 0.5373665720086263, "learning_rate": 9.99367446201005e-06, "loss": 0.6544, "step": 390 }, { "epoch": 0.2300930523373423, "grad_norm": 0.6432534438944446, "learning_rate": 9.993413699518906e-06, "loss": 0.699, "step": 391 }, { "epoch": 0.23068152561697747, "grad_norm": 0.44044098884852334, "learning_rate": 9.993147673772869e-06, "loss": 0.5551, "step": 392 }, { "epoch": 0.2312699988966126, "grad_norm": 0.5378270052740327, "learning_rate": 9.992876385052346e-06, "loss": 0.5167, "step": 393 }, { "epoch": 0.23185847217624775, "grad_norm": 0.5295476799527921, "learning_rate": 9.99259983364328e-06, "loss": 0.6632, "step": 394 }, { "epoch": 0.2324469454558829, "grad_norm": 0.4703333269973618, "learning_rate": 9.992318019837171e-06, "loss": 0.5804, "step": 395 }, { "epoch": 0.23303541873551803, "grad_norm": 0.3980311273849008, "learning_rate": 9.992030943931059e-06, "loss": 0.5349, "step": 396 }, { "epoch": 0.23362389201515318, "grad_norm": 0.583745332380449, "learning_rate": 9.991738606227537e-06, "loss": 0.618, "step": 397 }, { "epoch": 0.23421236529478834, "grad_norm": 0.44609888199130876, "learning_rate": 9.991441007034738e-06, "loss": 0.607, "step": 398 }, { "epoch": 0.2348008385744235, "grad_norm": 0.4670347689113555, "learning_rate": 9.991138146666344e-06, "loss": 0.5023, "step": 399 }, { "epoch": 0.23538931185405862, "grad_norm": 0.41777944014603396, "learning_rate": 9.99083002544158e-06, "loss": 0.5321, "step": 400 }, { "epoch": 0.23597778513369377, "grad_norm": 0.6367913401137651, "learning_rate": 9.990516643685222e-06, "loss": 0.6567, "step": 401 }, { "epoch": 0.23656625841332893, "grad_norm": 0.4963588963379063, "learning_rate": 9.990198001727582e-06, "loss": 0.5059, "step": 402 }, { "epoch": 0.23715473169296405, "grad_norm": 0.6761768440972852, "learning_rate": 9.989874099904524e-06, "loss": 0.5793, "step": 403 }, { "epoch": 0.2377432049725992, "grad_norm": 0.6065838531703244, "learning_rate": 9.989544938557453e-06, "loss": 0.6465, "step": 404 }, { "epoch": 0.23833167825223436, "grad_norm": 0.4422749677277523, "learning_rate": 9.989210518033316e-06, "loss": 0.5562, "step": 405 }, { "epoch": 0.23892015153186952, "grad_norm": 0.4739135686062783, "learning_rate": 9.98887083868461e-06, "loss": 0.5496, "step": 406 }, { "epoch": 0.23950862481150464, "grad_norm": 0.4793319476367638, "learning_rate": 9.988525900869366e-06, "loss": 0.6583, "step": 407 }, { "epoch": 0.2400970980911398, "grad_norm": 0.4212494078293219, "learning_rate": 9.988175704951163e-06, "loss": 0.5074, "step": 408 }, { "epoch": 0.24068557137077495, "grad_norm": 0.5312095966103221, "learning_rate": 9.987820251299121e-06, "loss": 0.5538, "step": 409 }, { "epoch": 0.24127404465041008, "grad_norm": 0.42321703714356995, "learning_rate": 9.987459540287906e-06, "loss": 0.5295, "step": 410 }, { "epoch": 0.24186251793004523, "grad_norm": 0.6147494116708798, "learning_rate": 9.987093572297716e-06, "loss": 0.7262, "step": 411 }, { "epoch": 0.2424509912096804, "grad_norm": 0.5971064451088822, "learning_rate": 9.9867223477143e-06, "loss": 0.6445, "step": 412 }, { "epoch": 0.24303946448931554, "grad_norm": 0.5171841908253608, "learning_rate": 9.98634586692894e-06, "loss": 0.6465, "step": 413 }, { "epoch": 0.24362793776895067, "grad_norm": 0.4644246410888609, "learning_rate": 9.985964130338466e-06, "loss": 0.4861, "step": 414 }, { "epoch": 0.24421641104858582, "grad_norm": 0.5183473140383351, "learning_rate": 9.98557713834524e-06, "loss": 0.6054, "step": 415 }, { "epoch": 0.24480488432822098, "grad_norm": 0.4476644035671486, "learning_rate": 9.985184891357165e-06, "loss": 0.5604, "step": 416 }, { "epoch": 0.2453933576078561, "grad_norm": 0.4817656095954495, "learning_rate": 9.984787389787689e-06, "loss": 0.5382, "step": 417 }, { "epoch": 0.24598183088749126, "grad_norm": 0.4920231131497407, "learning_rate": 9.984384634055793e-06, "loss": 0.6114, "step": 418 }, { "epoch": 0.24657030416712641, "grad_norm": 0.524648525759575, "learning_rate": 9.983976624585996e-06, "loss": 0.5372, "step": 419 }, { "epoch": 0.24715877744676157, "grad_norm": 0.4672409878929941, "learning_rate": 9.983563361808357e-06, "loss": 0.5767, "step": 420 }, { "epoch": 0.2477472507263967, "grad_norm": 0.5489315637083576, "learning_rate": 9.983144846158472e-06, "loss": 0.5875, "step": 421 }, { "epoch": 0.24833572400603185, "grad_norm": 0.47136190271930783, "learning_rate": 9.982721078077474e-06, "loss": 0.5378, "step": 422 }, { "epoch": 0.248924197285667, "grad_norm": 0.4533219504060194, "learning_rate": 9.982292058012027e-06, "loss": 0.5635, "step": 423 }, { "epoch": 0.24951267056530213, "grad_norm": 0.5017399251765493, "learning_rate": 9.981857786414339e-06, "loss": 0.6014, "step": 424 }, { "epoch": 0.2501011438449373, "grad_norm": 0.41982240452550657, "learning_rate": 9.981418263742148e-06, "loss": 0.5614, "step": 425 }, { "epoch": 0.25068961712457244, "grad_norm": 0.6467783260779305, "learning_rate": 9.980973490458728e-06, "loss": 0.5738, "step": 426 }, { "epoch": 0.25127809040420757, "grad_norm": 0.5076484484003937, "learning_rate": 9.98052346703289e-06, "loss": 0.6511, "step": 427 }, { "epoch": 0.25186656368384275, "grad_norm": 0.4604505304663248, "learning_rate": 9.980068193938971e-06, "loss": 0.5447, "step": 428 }, { "epoch": 0.2524550369634779, "grad_norm": 0.4267731137070173, "learning_rate": 9.979607671656852e-06, "loss": 0.588, "step": 429 }, { "epoch": 0.253043510243113, "grad_norm": 0.4711556243081078, "learning_rate": 9.979141900671938e-06, "loss": 0.5206, "step": 430 }, { "epoch": 0.2536319835227482, "grad_norm": 0.5222630201564635, "learning_rate": 9.978670881475173e-06, "loss": 0.5599, "step": 431 }, { "epoch": 0.2542204568023833, "grad_norm": 0.4295033221596563, "learning_rate": 9.978194614563028e-06, "loss": 0.5685, "step": 432 }, { "epoch": 0.25480893008201844, "grad_norm": 0.459658498012405, "learning_rate": 9.97771310043751e-06, "loss": 0.5679, "step": 433 }, { "epoch": 0.2553974033616536, "grad_norm": 0.5347382411817936, "learning_rate": 9.977226339606149e-06, "loss": 0.6384, "step": 434 }, { "epoch": 0.25598587664128875, "grad_norm": 0.5697169530311275, "learning_rate": 9.976734332582012e-06, "loss": 0.6157, "step": 435 }, { "epoch": 0.2565743499209239, "grad_norm": 0.5107955861038092, "learning_rate": 9.976237079883697e-06, "loss": 0.5223, "step": 436 }, { "epoch": 0.25716282320055905, "grad_norm": 0.5540088464648386, "learning_rate": 9.975734582035323e-06, "loss": 0.6691, "step": 437 }, { "epoch": 0.2577512964801942, "grad_norm": 0.4601498375176672, "learning_rate": 9.975226839566545e-06, "loss": 0.5477, "step": 438 }, { "epoch": 0.25833976975982936, "grad_norm": 0.8078854731520843, "learning_rate": 9.974713853012546e-06, "loss": 0.692, "step": 439 }, { "epoch": 0.2589282430394645, "grad_norm": 0.47215826869795924, "learning_rate": 9.97419562291403e-06, "loss": 0.6088, "step": 440 }, { "epoch": 0.2595167163190996, "grad_norm": 0.4292714438279743, "learning_rate": 9.973672149817232e-06, "loss": 0.4983, "step": 441 }, { "epoch": 0.2601051895987348, "grad_norm": 0.4988059625897621, "learning_rate": 9.973143434273916e-06, "loss": 0.6048, "step": 442 }, { "epoch": 0.2606936628783699, "grad_norm": 0.529917540131581, "learning_rate": 9.972609476841368e-06, "loss": 0.5977, "step": 443 }, { "epoch": 0.26128213615800505, "grad_norm": 0.42451082777127214, "learning_rate": 9.972070278082399e-06, "loss": 0.5817, "step": 444 }, { "epoch": 0.26187060943764023, "grad_norm": 0.4516574329218887, "learning_rate": 9.971525838565348e-06, "loss": 0.5927, "step": 445 }, { "epoch": 0.26245908271727536, "grad_norm": 0.6410050409765762, "learning_rate": 9.970976158864074e-06, "loss": 0.5757, "step": 446 }, { "epoch": 0.26304755599691054, "grad_norm": 0.4844338127081837, "learning_rate": 9.970421239557961e-06, "loss": 0.5889, "step": 447 }, { "epoch": 0.26363602927654567, "grad_norm": 0.688515631849798, "learning_rate": 9.969861081231916e-06, "loss": 0.555, "step": 448 }, { "epoch": 0.2642245025561808, "grad_norm": 0.5303389051794632, "learning_rate": 9.96929568447637e-06, "loss": 0.4635, "step": 449 }, { "epoch": 0.264812975835816, "grad_norm": 0.5322658531200801, "learning_rate": 9.96872504988727e-06, "loss": 0.5263, "step": 450 }, { "epoch": 0.2654014491154511, "grad_norm": 1.12850090119112, "learning_rate": 9.96814917806609e-06, "loss": 0.6298, "step": 451 }, { "epoch": 0.26598992239508623, "grad_norm": 0.47470242927668643, "learning_rate": 9.96756806961982e-06, "loss": 0.5932, "step": 452 }, { "epoch": 0.2665783956747214, "grad_norm": 0.4333995461566384, "learning_rate": 9.966981725160972e-06, "loss": 0.5158, "step": 453 }, { "epoch": 0.26716686895435654, "grad_norm": 0.5058393397378746, "learning_rate": 9.966390145307576e-06, "loss": 0.6138, "step": 454 }, { "epoch": 0.26775534223399167, "grad_norm": 0.44052974001218376, "learning_rate": 9.965793330683182e-06, "loss": 0.5777, "step": 455 }, { "epoch": 0.26834381551362685, "grad_norm": 0.475118129278307, "learning_rate": 9.965191281916852e-06, "loss": 0.5809, "step": 456 }, { "epoch": 0.268932288793262, "grad_norm": 0.4551349236468978, "learning_rate": 9.964583999643174e-06, "loss": 0.5921, "step": 457 }, { "epoch": 0.2695207620728971, "grad_norm": 0.47722055658191204, "learning_rate": 9.963971484502247e-06, "loss": 0.5773, "step": 458 }, { "epoch": 0.2701092353525323, "grad_norm": 0.7077698770242086, "learning_rate": 9.96335373713968e-06, "loss": 0.4906, "step": 459 }, { "epoch": 0.2706977086321674, "grad_norm": 0.47844689366562965, "learning_rate": 9.962730758206612e-06, "loss": 0.531, "step": 460 }, { "epoch": 0.2712861819118026, "grad_norm": 0.5309740800927105, "learning_rate": 9.96210254835968e-06, "loss": 0.6009, "step": 461 }, { "epoch": 0.2718746551914377, "grad_norm": 0.6748374004960324, "learning_rate": 9.961469108261044e-06, "loss": 0.6047, "step": 462 }, { "epoch": 0.27246312847107285, "grad_norm": 0.4314114585332593, "learning_rate": 9.960830438578378e-06, "loss": 0.5463, "step": 463 }, { "epoch": 0.27305160175070803, "grad_norm": 0.5714348216452276, "learning_rate": 9.960186539984862e-06, "loss": 0.6149, "step": 464 }, { "epoch": 0.27364007503034316, "grad_norm": 0.5726749515406556, "learning_rate": 9.95953741315919e-06, "loss": 0.6314, "step": 465 }, { "epoch": 0.2742285483099783, "grad_norm": 0.5146881752501714, "learning_rate": 9.95888305878557e-06, "loss": 0.6878, "step": 466 }, { "epoch": 0.27481702158961346, "grad_norm": 0.42714056343895224, "learning_rate": 9.958223477553715e-06, "loss": 0.5506, "step": 467 }, { "epoch": 0.2754054948692486, "grad_norm": 0.43551501555561045, "learning_rate": 9.95755867015885e-06, "loss": 0.5704, "step": 468 }, { "epoch": 0.2759939681488837, "grad_norm": 0.39215508057061704, "learning_rate": 9.95688863730171e-06, "loss": 0.4995, "step": 469 }, { "epoch": 0.2765824414285189, "grad_norm": 0.5424741105900363, "learning_rate": 9.956213379688531e-06, "loss": 0.5735, "step": 470 }, { "epoch": 0.277170914708154, "grad_norm": 0.4888366336999114, "learning_rate": 9.955532898031069e-06, "loss": 0.5732, "step": 471 }, { "epoch": 0.27775938798778915, "grad_norm": 0.5421588582982955, "learning_rate": 9.954847193046571e-06, "loss": 0.6026, "step": 472 }, { "epoch": 0.27834786126742433, "grad_norm": 0.464571834546167, "learning_rate": 9.954156265457801e-06, "loss": 0.6183, "step": 473 }, { "epoch": 0.27893633454705946, "grad_norm": 0.47361492549525386, "learning_rate": 9.953460115993025e-06, "loss": 0.5701, "step": 474 }, { "epoch": 0.27952480782669464, "grad_norm": 0.3933638334191058, "learning_rate": 9.95275874538601e-06, "loss": 0.5089, "step": 475 }, { "epoch": 0.28011328110632977, "grad_norm": 0.46651723325586736, "learning_rate": 9.952052154376027e-06, "loss": 0.4998, "step": 476 }, { "epoch": 0.2807017543859649, "grad_norm": 0.44807842977333046, "learning_rate": 9.951340343707852e-06, "loss": 0.5248, "step": 477 }, { "epoch": 0.2812902276656001, "grad_norm": 0.42159526179167073, "learning_rate": 9.950623314131762e-06, "loss": 0.4772, "step": 478 }, { "epoch": 0.2818787009452352, "grad_norm": 0.44374040512839286, "learning_rate": 9.949901066403536e-06, "loss": 0.5496, "step": 479 }, { "epoch": 0.28246717422487033, "grad_norm": 0.4619307128334352, "learning_rate": 9.949173601284448e-06, "loss": 0.543, "step": 480 }, { "epoch": 0.2830556475045055, "grad_norm": 0.4531772633937499, "learning_rate": 9.948440919541277e-06, "loss": 0.5279, "step": 481 }, { "epoch": 0.28364412078414064, "grad_norm": 0.5514792379382489, "learning_rate": 9.9477030219463e-06, "loss": 0.5669, "step": 482 }, { "epoch": 0.28423259406377577, "grad_norm": 0.4895637709561767, "learning_rate": 9.946959909277285e-06, "loss": 0.5328, "step": 483 }, { "epoch": 0.28482106734341095, "grad_norm": 0.4410044977397799, "learning_rate": 9.946211582317506e-06, "loss": 0.5831, "step": 484 }, { "epoch": 0.2854095406230461, "grad_norm": 0.34611788576599034, "learning_rate": 9.945458041855732e-06, "loss": 0.4591, "step": 485 }, { "epoch": 0.28599801390268126, "grad_norm": 0.7485831812597351, "learning_rate": 9.944699288686217e-06, "loss": 0.5847, "step": 486 }, { "epoch": 0.2865864871823164, "grad_norm": 0.47548438224470546, "learning_rate": 9.943935323608723e-06, "loss": 0.547, "step": 487 }, { "epoch": 0.2871749604619515, "grad_norm": 0.5014924923441143, "learning_rate": 9.943166147428494e-06, "loss": 0.6306, "step": 488 }, { "epoch": 0.2877634337415867, "grad_norm": 0.5269828151950541, "learning_rate": 9.942391760956277e-06, "loss": 0.6487, "step": 489 }, { "epoch": 0.2883519070212218, "grad_norm": 0.5737567147822462, "learning_rate": 9.941612165008302e-06, "loss": 0.6232, "step": 490 }, { "epoch": 0.28894038030085695, "grad_norm": 0.3948749498897012, "learning_rate": 9.940827360406297e-06, "loss": 0.4913, "step": 491 }, { "epoch": 0.28952885358049213, "grad_norm": 0.5339965596478529, "learning_rate": 9.940037347977477e-06, "loss": 0.5741, "step": 492 }, { "epoch": 0.29011732686012726, "grad_norm": 0.518938818796049, "learning_rate": 9.939242128554542e-06, "loss": 0.6651, "step": 493 }, { "epoch": 0.2907058001397624, "grad_norm": 0.3990548667752122, "learning_rate": 9.938441702975689e-06, "loss": 0.5275, "step": 494 }, { "epoch": 0.29129427341939756, "grad_norm": 0.4227437015574823, "learning_rate": 9.937636072084597e-06, "loss": 0.5661, "step": 495 }, { "epoch": 0.2918827466990327, "grad_norm": 0.48262816848381684, "learning_rate": 9.936825236730431e-06, "loss": 0.5948, "step": 496 }, { "epoch": 0.2924712199786678, "grad_norm": 0.5798066249759417, "learning_rate": 9.936009197767847e-06, "loss": 0.6495, "step": 497 }, { "epoch": 0.293059693258303, "grad_norm": 0.6117995587787014, "learning_rate": 9.935187956056977e-06, "loss": 0.646, "step": 498 }, { "epoch": 0.2936481665379381, "grad_norm": 0.43685846828984903, "learning_rate": 9.934361512463447e-06, "loss": 0.4838, "step": 499 }, { "epoch": 0.2942366398175733, "grad_norm": 0.5400092973799753, "learning_rate": 9.933529867858358e-06, "loss": 0.6063, "step": 500 }, { "epoch": 0.29482511309720844, "grad_norm": 0.41309299937251, "learning_rate": 9.932693023118299e-06, "loss": 0.475, "step": 501 }, { "epoch": 0.29541358637684356, "grad_norm": 0.7549315468302409, "learning_rate": 9.931850979125335e-06, "loss": 0.6124, "step": 502 }, { "epoch": 0.29600205965647874, "grad_norm": 0.5418635616994921, "learning_rate": 9.931003736767013e-06, "loss": 0.6121, "step": 503 }, { "epoch": 0.29659053293611387, "grad_norm": 0.42473690323232677, "learning_rate": 9.930151296936364e-06, "loss": 0.5241, "step": 504 }, { "epoch": 0.297179006215749, "grad_norm": 0.45842201958989276, "learning_rate": 9.929293660531889e-06, "loss": 0.5631, "step": 505 }, { "epoch": 0.2977674794953842, "grad_norm": 0.4377821216688983, "learning_rate": 9.92843082845757e-06, "loss": 0.54, "step": 506 }, { "epoch": 0.2983559527750193, "grad_norm": 0.3977310288151529, "learning_rate": 9.92756280162287e-06, "loss": 0.5455, "step": 507 }, { "epoch": 0.29894442605465443, "grad_norm": 0.3951322887977084, "learning_rate": 9.926689580942721e-06, "loss": 0.4964, "step": 508 }, { "epoch": 0.2995328993342896, "grad_norm": 0.4782034217780556, "learning_rate": 9.925811167337533e-06, "loss": 0.5934, "step": 509 }, { "epoch": 0.30012137261392474, "grad_norm": 0.3915994591393619, "learning_rate": 9.924927561733188e-06, "loss": 0.467, "step": 510 }, { "epoch": 0.30070984589355987, "grad_norm": 0.8722327382047171, "learning_rate": 9.924038765061042e-06, "loss": 0.5577, "step": 511 }, { "epoch": 0.30129831917319505, "grad_norm": 0.49776518638695527, "learning_rate": 9.923144778257918e-06, "loss": 0.5929, "step": 512 }, { "epoch": 0.3018867924528302, "grad_norm": 0.4009353664723987, "learning_rate": 9.922245602266119e-06, "loss": 0.5064, "step": 513 }, { "epoch": 0.30247526573246536, "grad_norm": 0.49200385766060595, "learning_rate": 9.921341238033409e-06, "loss": 0.5993, "step": 514 }, { "epoch": 0.3030637390121005, "grad_norm": 0.611390327493375, "learning_rate": 9.920431686513023e-06, "loss": 0.6065, "step": 515 }, { "epoch": 0.3036522122917356, "grad_norm": 0.4878344673478778, "learning_rate": 9.919516948663666e-06, "loss": 0.5393, "step": 516 }, { "epoch": 0.3042406855713708, "grad_norm": 0.4261819552929651, "learning_rate": 9.918597025449505e-06, "loss": 0.4793, "step": 517 }, { "epoch": 0.3048291588510059, "grad_norm": 0.48246550861276816, "learning_rate": 9.917671917840178e-06, "loss": 0.6154, "step": 518 }, { "epoch": 0.30541763213064105, "grad_norm": 0.5112905531916819, "learning_rate": 9.916741626810784e-06, "loss": 0.5778, "step": 519 }, { "epoch": 0.30600610541027623, "grad_norm": 0.4569444498196963, "learning_rate": 9.915806153341885e-06, "loss": 0.5045, "step": 520 }, { "epoch": 0.30659457868991136, "grad_norm": 0.45914624513187696, "learning_rate": 9.91486549841951e-06, "loss": 0.5779, "step": 521 }, { "epoch": 0.3071830519695465, "grad_norm": 0.47788621217387867, "learning_rate": 9.913919663035143e-06, "loss": 0.6054, "step": 522 }, { "epoch": 0.30777152524918167, "grad_norm": 0.47034952514854594, "learning_rate": 9.912968648185734e-06, "loss": 0.5447, "step": 523 }, { "epoch": 0.3083599985288168, "grad_norm": 0.4256968800941425, "learning_rate": 9.912012454873691e-06, "loss": 0.5546, "step": 524 }, { "epoch": 0.308948471808452, "grad_norm": 0.43865336944402566, "learning_rate": 9.911051084106877e-06, "loss": 0.5065, "step": 525 }, { "epoch": 0.3095369450880871, "grad_norm": 0.4099006639813132, "learning_rate": 9.910084536898615e-06, "loss": 0.473, "step": 526 }, { "epoch": 0.3101254183677222, "grad_norm": 0.409502964667668, "learning_rate": 9.909112814267686e-06, "loss": 0.5133, "step": 527 }, { "epoch": 0.3107138916473574, "grad_norm": 0.49391044239531884, "learning_rate": 9.908135917238321e-06, "loss": 0.5794, "step": 528 }, { "epoch": 0.31130236492699254, "grad_norm": 0.4614452520731292, "learning_rate": 9.90715384684021e-06, "loss": 0.5819, "step": 529 }, { "epoch": 0.31189083820662766, "grad_norm": 0.370737226508401, "learning_rate": 9.906166604108494e-06, "loss": 0.5252, "step": 530 }, { "epoch": 0.31247931148626285, "grad_norm": 0.4648096807114507, "learning_rate": 9.905174190083763e-06, "loss": 0.5194, "step": 531 }, { "epoch": 0.31306778476589797, "grad_norm": 0.47856276004853815, "learning_rate": 9.904176605812065e-06, "loss": 0.581, "step": 532 }, { "epoch": 0.3136562580455331, "grad_norm": 0.4266272982658827, "learning_rate": 9.903173852344889e-06, "loss": 0.549, "step": 533 }, { "epoch": 0.3142447313251683, "grad_norm": 0.42975042941932917, "learning_rate": 9.90216593073918e-06, "loss": 0.5158, "step": 534 }, { "epoch": 0.3148332046048034, "grad_norm": 0.436229366365434, "learning_rate": 9.901152842057325e-06, "loss": 0.5927, "step": 535 }, { "epoch": 0.31542167788443853, "grad_norm": 0.5034159497657995, "learning_rate": 9.90013458736716e-06, "loss": 0.5925, "step": 536 }, { "epoch": 0.3160101511640737, "grad_norm": 0.49004970573994505, "learning_rate": 9.899111167741966e-06, "loss": 0.4934, "step": 537 }, { "epoch": 0.31659862444370884, "grad_norm": 0.3981459637844768, "learning_rate": 9.898082584260467e-06, "loss": 0.5552, "step": 538 }, { "epoch": 0.317187097723344, "grad_norm": 0.5422120982363273, "learning_rate": 9.89704883800683e-06, "loss": 0.5511, "step": 539 }, { "epoch": 0.31777557100297915, "grad_norm": 0.4896928414609172, "learning_rate": 9.896009930070666e-06, "loss": 0.6032, "step": 540 }, { "epoch": 0.3183640442826143, "grad_norm": 0.3940392471805379, "learning_rate": 9.894965861547023e-06, "loss": 0.4843, "step": 541 }, { "epoch": 0.31895251756224946, "grad_norm": 0.49508092941441745, "learning_rate": 9.893916633536393e-06, "loss": 0.5321, "step": 542 }, { "epoch": 0.3195409908418846, "grad_norm": 0.47464322948081755, "learning_rate": 9.892862247144701e-06, "loss": 0.5575, "step": 543 }, { "epoch": 0.3201294641215197, "grad_norm": 0.4780121976086582, "learning_rate": 9.891802703483314e-06, "loss": 0.647, "step": 544 }, { "epoch": 0.3207179374011549, "grad_norm": 0.44393705527997696, "learning_rate": 9.890738003669029e-06, "loss": 0.5302, "step": 545 }, { "epoch": 0.32130641068079, "grad_norm": 0.5110071128560698, "learning_rate": 9.889668148824086e-06, "loss": 0.5724, "step": 546 }, { "epoch": 0.32189488396042515, "grad_norm": 0.43932623131683163, "learning_rate": 9.888593140076152e-06, "loss": 0.5104, "step": 547 }, { "epoch": 0.32248335724006033, "grad_norm": 0.6337654883493324, "learning_rate": 9.887512978558329e-06, "loss": 0.6792, "step": 548 }, { "epoch": 0.32307183051969546, "grad_norm": 0.4537515161705662, "learning_rate": 9.88642766540915e-06, "loss": 0.541, "step": 549 }, { "epoch": 0.3236603037993306, "grad_norm": 0.5064209943291982, "learning_rate": 9.885337201772579e-06, "loss": 0.5679, "step": 550 }, { "epoch": 0.32424877707896577, "grad_norm": 0.5394284954182017, "learning_rate": 9.884241588798004e-06, "loss": 0.5529, "step": 551 }, { "epoch": 0.3248372503586009, "grad_norm": 0.4762289965404846, "learning_rate": 9.883140827640248e-06, "loss": 0.6097, "step": 552 }, { "epoch": 0.3254257236382361, "grad_norm": 0.46355272122577346, "learning_rate": 9.882034919459556e-06, "loss": 0.5598, "step": 553 }, { "epoch": 0.3260141969178712, "grad_norm": 0.413173552287594, "learning_rate": 9.880923865421597e-06, "loss": 0.5394, "step": 554 }, { "epoch": 0.32660267019750633, "grad_norm": 0.49824537861441315, "learning_rate": 9.879807666697467e-06, "loss": 0.5577, "step": 555 }, { "epoch": 0.3271911434771415, "grad_norm": 0.43462625642274777, "learning_rate": 9.878686324463683e-06, "loss": 0.5385, "step": 556 }, { "epoch": 0.32777961675677664, "grad_norm": 0.4580850371042482, "learning_rate": 9.877559839902185e-06, "loss": 0.602, "step": 557 }, { "epoch": 0.32836809003641176, "grad_norm": 0.42124954048307856, "learning_rate": 9.87642821420033e-06, "loss": 0.537, "step": 558 }, { "epoch": 0.32895656331604695, "grad_norm": 0.5455372751359436, "learning_rate": 9.875291448550899e-06, "loss": 0.5416, "step": 559 }, { "epoch": 0.3295450365956821, "grad_norm": 0.4366895577252232, "learning_rate": 9.874149544152085e-06, "loss": 0.524, "step": 560 }, { "epoch": 0.3301335098753172, "grad_norm": 0.5038025017511302, "learning_rate": 9.873002502207502e-06, "loss": 0.6255, "step": 561 }, { "epoch": 0.3307219831549524, "grad_norm": 0.44376775026267584, "learning_rate": 9.871850323926178e-06, "loss": 0.582, "step": 562 }, { "epoch": 0.3313104564345875, "grad_norm": 0.45991377414830464, "learning_rate": 9.870693010522552e-06, "loss": 0.5586, "step": 563 }, { "epoch": 0.3318989297142227, "grad_norm": 0.47599664611091486, "learning_rate": 9.869530563216479e-06, "loss": 0.5937, "step": 564 }, { "epoch": 0.3324874029938578, "grad_norm": 0.4256631994860406, "learning_rate": 9.868362983233226e-06, "loss": 0.4759, "step": 565 }, { "epoch": 0.33307587627349294, "grad_norm": 0.4900463868251775, "learning_rate": 9.867190271803466e-06, "loss": 0.5595, "step": 566 }, { "epoch": 0.3336643495531281, "grad_norm": 0.45540987233905306, "learning_rate": 9.866012430163283e-06, "loss": 0.6345, "step": 567 }, { "epoch": 0.33425282283276325, "grad_norm": 0.48385763927811143, "learning_rate": 9.86482945955417e-06, "loss": 0.638, "step": 568 }, { "epoch": 0.3348412961123984, "grad_norm": 0.6259562922980026, "learning_rate": 9.863641361223025e-06, "loss": 0.6515, "step": 569 }, { "epoch": 0.33542976939203356, "grad_norm": 0.5359760331890745, "learning_rate": 9.86244813642215e-06, "loss": 0.6591, "step": 570 }, { "epoch": 0.3360182426716687, "grad_norm": 0.4701002454414917, "learning_rate": 9.861249786409248e-06, "loss": 0.5924, "step": 571 }, { "epoch": 0.3366067159513038, "grad_norm": 0.49568270572451645, "learning_rate": 9.860046312447432e-06, "loss": 0.5901, "step": 572 }, { "epoch": 0.337195189230939, "grad_norm": 0.5179164523222638, "learning_rate": 9.858837715805207e-06, "loss": 0.6082, "step": 573 }, { "epoch": 0.3377836625105741, "grad_norm": 0.4592003064104772, "learning_rate": 9.857623997756485e-06, "loss": 0.5633, "step": 574 }, { "epoch": 0.33837213579020925, "grad_norm": 0.49826504466953164, "learning_rate": 9.85640515958057e-06, "loss": 0.5004, "step": 575 }, { "epoch": 0.33896060906984443, "grad_norm": 0.4644709551434048, "learning_rate": 9.855181202562168e-06, "loss": 0.5603, "step": 576 }, { "epoch": 0.33954908234947956, "grad_norm": 0.5058718370149957, "learning_rate": 9.853952127991374e-06, "loss": 0.6465, "step": 577 }, { "epoch": 0.34013755562911474, "grad_norm": 0.43409067668422147, "learning_rate": 9.852717937163683e-06, "loss": 0.5708, "step": 578 }, { "epoch": 0.34072602890874987, "grad_norm": 0.4705103062916938, "learning_rate": 9.851478631379982e-06, "loss": 0.6112, "step": 579 }, { "epoch": 0.341314502188385, "grad_norm": 0.4473780361302094, "learning_rate": 9.850234211946548e-06, "loss": 0.5437, "step": 580 }, { "epoch": 0.3419029754680202, "grad_norm": 0.5349259704423881, "learning_rate": 9.848984680175049e-06, "loss": 0.5939, "step": 581 }, { "epoch": 0.3424914487476553, "grad_norm": 0.4104678483536991, "learning_rate": 9.847730037382536e-06, "loss": 0.5864, "step": 582 }, { "epoch": 0.34307992202729043, "grad_norm": 0.35925292544998405, "learning_rate": 9.846470284891457e-06, "loss": 0.5197, "step": 583 }, { "epoch": 0.3436683953069256, "grad_norm": 0.49977348637092167, "learning_rate": 9.845205424029639e-06, "loss": 0.6527, "step": 584 }, { "epoch": 0.34425686858656074, "grad_norm": 0.5293911597085432, "learning_rate": 9.843935456130295e-06, "loss": 0.6108, "step": 585 }, { "epoch": 0.34484534186619586, "grad_norm": 0.4228106812651455, "learning_rate": 9.84266038253202e-06, "loss": 0.4895, "step": 586 }, { "epoch": 0.34543381514583105, "grad_norm": 0.45131888167400336, "learning_rate": 9.841380204578795e-06, "loss": 0.497, "step": 587 }, { "epoch": 0.3460222884254662, "grad_norm": 0.5569270657291467, "learning_rate": 9.840094923619977e-06, "loss": 0.5499, "step": 588 }, { "epoch": 0.3466107617051013, "grad_norm": 0.4196053670699418, "learning_rate": 9.8388045410103e-06, "loss": 0.5936, "step": 589 }, { "epoch": 0.3471992349847365, "grad_norm": 0.4881344976127253, "learning_rate": 9.83750905810988e-06, "loss": 0.5171, "step": 590 }, { "epoch": 0.3477877082643716, "grad_norm": 0.5100463063328117, "learning_rate": 9.836208476284208e-06, "loss": 0.604, "step": 591 }, { "epoch": 0.3483761815440068, "grad_norm": 0.43307541810171407, "learning_rate": 9.834902796904148e-06, "loss": 0.4938, "step": 592 }, { "epoch": 0.3489646548236419, "grad_norm": 0.47660507276559166, "learning_rate": 9.833592021345938e-06, "loss": 0.6137, "step": 593 }, { "epoch": 0.34955312810327704, "grad_norm": 0.4332196529974898, "learning_rate": 9.832276150991185e-06, "loss": 0.4466, "step": 594 }, { "epoch": 0.3501416013829122, "grad_norm": 0.49777455300067086, "learning_rate": 9.830955187226872e-06, "loss": 0.5198, "step": 595 }, { "epoch": 0.35073007466254735, "grad_norm": 0.4457611196864995, "learning_rate": 9.829629131445342e-06, "loss": 0.4995, "step": 596 }, { "epoch": 0.3513185479421825, "grad_norm": 0.5040078018253271, "learning_rate": 9.828297985044314e-06, "loss": 0.6323, "step": 597 }, { "epoch": 0.35190702122181766, "grad_norm": 0.4599972596610061, "learning_rate": 9.826961749426867e-06, "loss": 0.5888, "step": 598 }, { "epoch": 0.3524954945014528, "grad_norm": 0.4996750551204354, "learning_rate": 9.825620426001446e-06, "loss": 0.5856, "step": 599 }, { "epoch": 0.3530839677810879, "grad_norm": 0.41698214549231816, "learning_rate": 9.824274016181859e-06, "loss": 0.5451, "step": 600 }, { "epoch": 0.3536724410607231, "grad_norm": 0.454724718145187, "learning_rate": 9.822922521387277e-06, "loss": 0.5485, "step": 601 }, { "epoch": 0.3542609143403582, "grad_norm": 0.4151574938070065, "learning_rate": 9.821565943042225e-06, "loss": 0.5549, "step": 602 }, { "epoch": 0.3548493876199934, "grad_norm": 0.479227168107094, "learning_rate": 9.820204282576594e-06, "loss": 0.5654, "step": 603 }, { "epoch": 0.35543786089962853, "grad_norm": 0.48270013224916236, "learning_rate": 9.818837541425623e-06, "loss": 0.6081, "step": 604 }, { "epoch": 0.35602633417926366, "grad_norm": 0.5148459842837766, "learning_rate": 9.817465721029916e-06, "loss": 0.598, "step": 605 }, { "epoch": 0.35661480745889884, "grad_norm": 0.4768376578314123, "learning_rate": 9.816088822835423e-06, "loss": 0.5075, "step": 606 }, { "epoch": 0.35720328073853397, "grad_norm": 0.48333503808112344, "learning_rate": 9.814706848293452e-06, "loss": 0.534, "step": 607 }, { "epoch": 0.3577917540181691, "grad_norm": 0.4879210935745823, "learning_rate": 9.813319798860654e-06, "loss": 0.5858, "step": 608 }, { "epoch": 0.3583802272978043, "grad_norm": 0.555366282397036, "learning_rate": 9.811927675999035e-06, "loss": 0.6217, "step": 609 }, { "epoch": 0.3589687005774394, "grad_norm": 0.5019732851128141, "learning_rate": 9.81053048117595e-06, "loss": 0.5605, "step": 610 }, { "epoch": 0.35955717385707453, "grad_norm": 0.5056365308195051, "learning_rate": 9.809128215864096e-06, "loss": 0.6769, "step": 611 }, { "epoch": 0.3601456471367097, "grad_norm": 0.4826083758669427, "learning_rate": 9.807720881541516e-06, "loss": 0.5759, "step": 612 }, { "epoch": 0.36073412041634484, "grad_norm": 0.4427759069688132, "learning_rate": 9.806308479691595e-06, "loss": 0.5398, "step": 613 }, { "epoch": 0.36132259369597997, "grad_norm": 0.3981616605127294, "learning_rate": 9.804891011803062e-06, "loss": 0.5156, "step": 614 }, { "epoch": 0.36191106697561515, "grad_norm": 0.4764868818798259, "learning_rate": 9.803468479369981e-06, "loss": 0.5577, "step": 615 }, { "epoch": 0.3624995402552503, "grad_norm": 0.47870638397111515, "learning_rate": 9.802040883891762e-06, "loss": 0.596, "step": 616 }, { "epoch": 0.36308801353488546, "grad_norm": 0.40309526335229034, "learning_rate": 9.800608226873143e-06, "loss": 0.5721, "step": 617 }, { "epoch": 0.3636764868145206, "grad_norm": 0.6118642763587226, "learning_rate": 9.7991705098242e-06, "loss": 0.766, "step": 618 }, { "epoch": 0.3642649600941557, "grad_norm": 0.508877618067198, "learning_rate": 9.797727734260348e-06, "loss": 0.6753, "step": 619 }, { "epoch": 0.3648534333737909, "grad_norm": 0.496057615437513, "learning_rate": 9.796279901702326e-06, "loss": 0.5968, "step": 620 }, { "epoch": 0.365441906653426, "grad_norm": 0.4888195130727514, "learning_rate": 9.794827013676206e-06, "loss": 0.4833, "step": 621 }, { "epoch": 0.36603037993306115, "grad_norm": 0.4937151779448675, "learning_rate": 9.793369071713393e-06, "loss": 0.5137, "step": 622 }, { "epoch": 0.3666188532126963, "grad_norm": 0.4173176335607143, "learning_rate": 9.791906077350613e-06, "loss": 0.5504, "step": 623 }, { "epoch": 0.36720732649233145, "grad_norm": 0.413189639201702, "learning_rate": 9.790438032129919e-06, "loss": 0.5154, "step": 624 }, { "epoch": 0.3677957997719666, "grad_norm": 0.4590828085319845, "learning_rate": 9.788964937598688e-06, "loss": 0.5376, "step": 625 }, { "epoch": 0.36838427305160176, "grad_norm": 0.45820137462437854, "learning_rate": 9.787486795309621e-06, "loss": 0.5823, "step": 626 }, { "epoch": 0.3689727463312369, "grad_norm": 0.49921908085319316, "learning_rate": 9.786003606820742e-06, "loss": 0.478, "step": 627 }, { "epoch": 0.36956121961087207, "grad_norm": 0.45161186650719687, "learning_rate": 9.784515373695382e-06, "loss": 0.5691, "step": 628 }, { "epoch": 0.3701496928905072, "grad_norm": 0.47167131108986504, "learning_rate": 9.783022097502204e-06, "loss": 0.5401, "step": 629 }, { "epoch": 0.3707381661701423, "grad_norm": 0.38686535663782995, "learning_rate": 9.781523779815178e-06, "loss": 0.4701, "step": 630 }, { "epoch": 0.3713266394497775, "grad_norm": 0.3845132731193939, "learning_rate": 9.78002042221359e-06, "loss": 0.4585, "step": 631 }, { "epoch": 0.37191511272941263, "grad_norm": 0.4212608200085171, "learning_rate": 9.778512026282037e-06, "loss": 0.5649, "step": 632 }, { "epoch": 0.37250358600904776, "grad_norm": 0.4245660348669521, "learning_rate": 9.776998593610428e-06, "loss": 0.5297, "step": 633 }, { "epoch": 0.37309205928868294, "grad_norm": 0.8536512448833727, "learning_rate": 9.775480125793981e-06, "loss": 0.4937, "step": 634 }, { "epoch": 0.37368053256831807, "grad_norm": 0.5127506463733048, "learning_rate": 9.773956624433224e-06, "loss": 0.6014, "step": 635 }, { "epoch": 0.3742690058479532, "grad_norm": 0.5559314341955032, "learning_rate": 9.77242809113398e-06, "loss": 0.5898, "step": 636 }, { "epoch": 0.3748574791275884, "grad_norm": 0.434160764331554, "learning_rate": 9.770894527507393e-06, "loss": 0.5333, "step": 637 }, { "epoch": 0.3754459524072235, "grad_norm": 0.40380616470000447, "learning_rate": 9.76935593516989e-06, "loss": 0.545, "step": 638 }, { "epoch": 0.37603442568685863, "grad_norm": 0.463771513762846, "learning_rate": 9.767812315743218e-06, "loss": 0.5061, "step": 639 }, { "epoch": 0.3766228989664938, "grad_norm": 0.38825407780373383, "learning_rate": 9.766263670854403e-06, "loss": 0.5231, "step": 640 }, { "epoch": 0.37721137224612894, "grad_norm": 0.5040288543411928, "learning_rate": 9.764710002135784e-06, "loss": 0.5461, "step": 641 }, { "epoch": 0.3777998455257641, "grad_norm": 0.4514022557400679, "learning_rate": 9.763151311224986e-06, "loss": 0.5637, "step": 642 }, { "epoch": 0.37838831880539925, "grad_norm": 0.4491219334169601, "learning_rate": 9.761587599764932e-06, "loss": 0.5842, "step": 643 }, { "epoch": 0.3789767920850344, "grad_norm": 0.42430849292492495, "learning_rate": 9.760018869403837e-06, "loss": 0.4913, "step": 644 }, { "epoch": 0.37956526536466956, "grad_norm": 0.425955170730525, "learning_rate": 9.7584451217952e-06, "loss": 0.5187, "step": 645 }, { "epoch": 0.3801537386443047, "grad_norm": 0.5861169864813751, "learning_rate": 9.756866358597819e-06, "loss": 0.5103, "step": 646 }, { "epoch": 0.3807422119239398, "grad_norm": 0.44955619888217835, "learning_rate": 9.755282581475769e-06, "loss": 0.6369, "step": 647 }, { "epoch": 0.381330685203575, "grad_norm": 0.45800347591142987, "learning_rate": 9.753693792098412e-06, "loss": 0.5617, "step": 648 }, { "epoch": 0.3819191584832101, "grad_norm": 0.38209486490242933, "learning_rate": 9.752099992140401e-06, "loss": 0.4916, "step": 649 }, { "epoch": 0.38250763176284525, "grad_norm": 0.5031869017042281, "learning_rate": 9.750501183281655e-06, "loss": 0.5663, "step": 650 }, { "epoch": 0.38309610504248043, "grad_norm": 0.43718899978490644, "learning_rate": 9.748897367207391e-06, "loss": 0.5274, "step": 651 }, { "epoch": 0.38368457832211555, "grad_norm": 0.3798894971252191, "learning_rate": 9.747288545608087e-06, "loss": 0.4974, "step": 652 }, { "epoch": 0.3842730516017507, "grad_norm": 0.4539313035888414, "learning_rate": 9.745674720179507e-06, "loss": 0.5608, "step": 653 }, { "epoch": 0.38486152488138586, "grad_norm": 0.38804995123782554, "learning_rate": 9.744055892622689e-06, "loss": 0.5205, "step": 654 }, { "epoch": 0.385449998161021, "grad_norm": 0.4780672259503453, "learning_rate": 9.742432064643936e-06, "loss": 0.5498, "step": 655 }, { "epoch": 0.3860384714406562, "grad_norm": 0.4376454710804824, "learning_rate": 9.74080323795483e-06, "loss": 0.5629, "step": 656 }, { "epoch": 0.3866269447202913, "grad_norm": 0.4317389355743099, "learning_rate": 9.739169414272219e-06, "loss": 0.4915, "step": 657 }, { "epoch": 0.3872154179999264, "grad_norm": 0.46888184511820513, "learning_rate": 9.737530595318216e-06, "loss": 0.5817, "step": 658 }, { "epoch": 0.3878038912795616, "grad_norm": 0.40258541679177307, "learning_rate": 9.735886782820202e-06, "loss": 0.4903, "step": 659 }, { "epoch": 0.38839236455919673, "grad_norm": 0.4688500172378739, "learning_rate": 9.73423797851082e-06, "loss": 0.5619, "step": 660 }, { "epoch": 0.38898083783883186, "grad_norm": 0.3920329094862744, "learning_rate": 9.732584184127973e-06, "loss": 0.4562, "step": 661 }, { "epoch": 0.38956931111846704, "grad_norm": 0.49303396447860465, "learning_rate": 9.730925401414828e-06, "loss": 0.5302, "step": 662 }, { "epoch": 0.39015778439810217, "grad_norm": 0.4522924500439807, "learning_rate": 9.729261632119806e-06, "loss": 0.4686, "step": 663 }, { "epoch": 0.3907462576777373, "grad_norm": 0.6052638689089915, "learning_rate": 9.727592877996585e-06, "loss": 0.563, "step": 664 }, { "epoch": 0.3913347309573725, "grad_norm": 0.47855179374131346, "learning_rate": 9.7259191408041e-06, "loss": 0.5362, "step": 665 }, { "epoch": 0.3919232042370076, "grad_norm": 0.49569529381151867, "learning_rate": 9.724240422306531e-06, "loss": 0.5861, "step": 666 }, { "epoch": 0.3925116775166428, "grad_norm": 0.4292467283060194, "learning_rate": 9.722556724273318e-06, "loss": 0.6147, "step": 667 }, { "epoch": 0.3931001507962779, "grad_norm": 0.43342381500890126, "learning_rate": 9.720868048479146e-06, "loss": 0.5013, "step": 668 }, { "epoch": 0.39368862407591304, "grad_norm": 0.43230323489694816, "learning_rate": 9.719174396703941e-06, "loss": 0.5279, "step": 669 }, { "epoch": 0.3942770973555482, "grad_norm": 0.4962709366525989, "learning_rate": 9.717475770732883e-06, "loss": 0.6166, "step": 670 }, { "epoch": 0.39486557063518335, "grad_norm": 0.40781086512314046, "learning_rate": 9.715772172356388e-06, "loss": 0.5617, "step": 671 }, { "epoch": 0.3954540439148185, "grad_norm": 0.4602796519137617, "learning_rate": 9.714063603370118e-06, "loss": 0.5347, "step": 672 }, { "epoch": 0.39604251719445366, "grad_norm": 0.3655829190552029, "learning_rate": 9.71235006557497e-06, "loss": 0.4563, "step": 673 }, { "epoch": 0.3966309904740888, "grad_norm": 0.43659334930464194, "learning_rate": 9.710631560777082e-06, "loss": 0.5508, "step": 674 }, { "epoch": 0.3972194637537239, "grad_norm": 0.42986001543225816, "learning_rate": 9.708908090787825e-06, "loss": 0.5243, "step": 675 }, { "epoch": 0.3978079370333591, "grad_norm": 0.393079421756463, "learning_rate": 9.707179657423806e-06, "loss": 0.4589, "step": 676 }, { "epoch": 0.3983964103129942, "grad_norm": 0.4995690145043239, "learning_rate": 9.705446262506858e-06, "loss": 0.4904, "step": 677 }, { "epoch": 0.39898488359262935, "grad_norm": 0.512496723197366, "learning_rate": 9.703707907864052e-06, "loss": 0.5876, "step": 678 }, { "epoch": 0.39957335687226453, "grad_norm": 0.4711902896030533, "learning_rate": 9.701964595327679e-06, "loss": 0.5438, "step": 679 }, { "epoch": 0.40016183015189966, "grad_norm": 0.42750151357077026, "learning_rate": 9.700216326735263e-06, "loss": 0.5516, "step": 680 }, { "epoch": 0.40075030343153484, "grad_norm": 0.48127678634359566, "learning_rate": 9.698463103929542e-06, "loss": 0.5807, "step": 681 }, { "epoch": 0.40133877671116996, "grad_norm": 0.38431972107024454, "learning_rate": 9.696704928758488e-06, "loss": 0.4893, "step": 682 }, { "epoch": 0.4019272499908051, "grad_norm": 0.44855988433414945, "learning_rate": 9.694941803075285e-06, "loss": 0.5328, "step": 683 }, { "epoch": 0.4025157232704403, "grad_norm": 0.43455894507106396, "learning_rate": 9.693173728738335e-06, "loss": 0.4895, "step": 684 }, { "epoch": 0.4031041965500754, "grad_norm": 0.44564946713103387, "learning_rate": 9.691400707611258e-06, "loss": 0.4884, "step": 685 }, { "epoch": 0.4036926698297105, "grad_norm": 0.4569674426337324, "learning_rate": 9.689622741562891e-06, "loss": 0.5382, "step": 686 }, { "epoch": 0.4042811431093457, "grad_norm": 0.4277852950685041, "learning_rate": 9.687839832467278e-06, "loss": 0.5323, "step": 687 }, { "epoch": 0.40486961638898084, "grad_norm": 0.47528435829031324, "learning_rate": 9.68605198220368e-06, "loss": 0.5489, "step": 688 }, { "epoch": 0.40545808966861596, "grad_norm": 0.4411391959251179, "learning_rate": 9.684259192656554e-06, "loss": 0.559, "step": 689 }, { "epoch": 0.40604656294825114, "grad_norm": 0.38178007953377674, "learning_rate": 9.682461465715576e-06, "loss": 0.5237, "step": 690 }, { "epoch": 0.40663503622788627, "grad_norm": 0.49563815014353624, "learning_rate": 9.68065880327562e-06, "loss": 0.5345, "step": 691 }, { "epoch": 0.4072235095075214, "grad_norm": 0.4363417890857879, "learning_rate": 9.678851207236764e-06, "loss": 0.5059, "step": 692 }, { "epoch": 0.4078119827871566, "grad_norm": 0.4184602979789486, "learning_rate": 9.677038679504285e-06, "loss": 0.4932, "step": 693 }, { "epoch": 0.4084004560667917, "grad_norm": 0.43401737782795097, "learning_rate": 9.675221221988658e-06, "loss": 0.5445, "step": 694 }, { "epoch": 0.4089889293464269, "grad_norm": 0.4544981890622693, "learning_rate": 9.673398836605554e-06, "loss": 0.586, "step": 695 }, { "epoch": 0.409577402626062, "grad_norm": 0.42850623134681787, "learning_rate": 9.671571525275842e-06, "loss": 0.5431, "step": 696 }, { "epoch": 0.41016587590569714, "grad_norm": 0.4369071976528651, "learning_rate": 9.669739289925578e-06, "loss": 0.5482, "step": 697 }, { "epoch": 0.4107543491853323, "grad_norm": 0.42055330305891947, "learning_rate": 9.667902132486009e-06, "loss": 0.4927, "step": 698 }, { "epoch": 0.41134282246496745, "grad_norm": 0.41256262181037706, "learning_rate": 9.666060054893576e-06, "loss": 0.5252, "step": 699 }, { "epoch": 0.4119312957446026, "grad_norm": 0.43265439402369515, "learning_rate": 9.664213059089896e-06, "loss": 0.5018, "step": 700 }, { "epoch": 0.41251976902423776, "grad_norm": 0.4893914301769793, "learning_rate": 9.66236114702178e-06, "loss": 0.5616, "step": 701 }, { "epoch": 0.4131082423038729, "grad_norm": 0.42725379762200166, "learning_rate": 9.660504320641213e-06, "loss": 0.5109, "step": 702 }, { "epoch": 0.413696715583508, "grad_norm": 0.4455716184216144, "learning_rate": 9.658642581905366e-06, "loss": 0.545, "step": 703 }, { "epoch": 0.4142851888631432, "grad_norm": 0.5133194255159209, "learning_rate": 9.656775932776582e-06, "loss": 0.5882, "step": 704 }, { "epoch": 0.4148736621427783, "grad_norm": 0.43810902238087324, "learning_rate": 9.654904375222384e-06, "loss": 0.5038, "step": 705 }, { "epoch": 0.4154621354224135, "grad_norm": 0.46342151158343564, "learning_rate": 9.653027911215469e-06, "loss": 0.533, "step": 706 }, { "epoch": 0.41605060870204863, "grad_norm": 0.49663096510639726, "learning_rate": 9.651146542733702e-06, "loss": 0.4575, "step": 707 }, { "epoch": 0.41663908198168376, "grad_norm": 0.40231005582942264, "learning_rate": 9.649260271760121e-06, "loss": 0.5477, "step": 708 }, { "epoch": 0.41722755526131894, "grad_norm": 0.5786896695212135, "learning_rate": 9.647369100282928e-06, "loss": 0.5314, "step": 709 }, { "epoch": 0.41781602854095407, "grad_norm": 0.4661963415761389, "learning_rate": 9.645473030295496e-06, "loss": 0.4868, "step": 710 }, { "epoch": 0.4184045018205892, "grad_norm": 0.39232570099078234, "learning_rate": 9.643572063796352e-06, "loss": 0.4472, "step": 711 }, { "epoch": 0.4189929751002244, "grad_norm": 0.4547610530006057, "learning_rate": 9.641666202789195e-06, "loss": 0.5467, "step": 712 }, { "epoch": 0.4195814483798595, "grad_norm": 0.45154710677313026, "learning_rate": 9.639755449282874e-06, "loss": 0.5879, "step": 713 }, { "epoch": 0.4201699216594946, "grad_norm": 0.441957958172087, "learning_rate": 9.6378398052914e-06, "loss": 0.5066, "step": 714 }, { "epoch": 0.4207583949391298, "grad_norm": 0.35914781044116667, "learning_rate": 9.635919272833938e-06, "loss": 0.477, "step": 715 }, { "epoch": 0.42134686821876494, "grad_norm": 0.41531747811405606, "learning_rate": 9.633993853934803e-06, "loss": 0.4753, "step": 716 }, { "epoch": 0.42193534149840006, "grad_norm": 0.5099961019845816, "learning_rate": 9.632063550623465e-06, "loss": 0.5912, "step": 717 }, { "epoch": 0.42252381477803524, "grad_norm": 0.42949776557863634, "learning_rate": 9.630128364934536e-06, "loss": 0.4856, "step": 718 }, { "epoch": 0.42311228805767037, "grad_norm": 0.4583573739719953, "learning_rate": 9.628188298907782e-06, "loss": 0.5042, "step": 719 }, { "epoch": 0.42370076133730555, "grad_norm": 0.46743632273728203, "learning_rate": 9.626243354588107e-06, "loss": 0.549, "step": 720 }, { "epoch": 0.4242892346169407, "grad_norm": 0.41766110717460525, "learning_rate": 9.62429353402556e-06, "loss": 0.4481, "step": 721 }, { "epoch": 0.4248777078965758, "grad_norm": 0.48029187235196674, "learning_rate": 9.622338839275327e-06, "loss": 0.5399, "step": 722 }, { "epoch": 0.425466181176211, "grad_norm": 0.37796856410591423, "learning_rate": 9.620379272397735e-06, "loss": 0.4845, "step": 723 }, { "epoch": 0.4260546544558461, "grad_norm": 0.8825706774664608, "learning_rate": 9.618414835458243e-06, "loss": 0.5742, "step": 724 }, { "epoch": 0.42664312773548124, "grad_norm": 0.40627732208387374, "learning_rate": 9.616445530527448e-06, "loss": 0.4447, "step": 725 }, { "epoch": 0.4272316010151164, "grad_norm": 0.6694217878307238, "learning_rate": 9.614471359681072e-06, "loss": 0.5353, "step": 726 }, { "epoch": 0.42782007429475155, "grad_norm": 0.46090354350845786, "learning_rate": 9.61249232499997e-06, "loss": 0.5489, "step": 727 }, { "epoch": 0.4284085475743867, "grad_norm": 0.4817590515235149, "learning_rate": 9.610508428570122e-06, "loss": 0.5526, "step": 728 }, { "epoch": 0.42899702085402186, "grad_norm": 0.5196556461206975, "learning_rate": 9.608519672482635e-06, "loss": 0.5977, "step": 729 }, { "epoch": 0.429585494133657, "grad_norm": 0.4906809905807339, "learning_rate": 9.606526058833734e-06, "loss": 0.5177, "step": 730 }, { "epoch": 0.4301739674132921, "grad_norm": 0.4126257275653058, "learning_rate": 9.60452758972477e-06, "loss": 0.5831, "step": 731 }, { "epoch": 0.4307624406929273, "grad_norm": 0.5912595854900098, "learning_rate": 9.602524267262202e-06, "loss": 0.6542, "step": 732 }, { "epoch": 0.4313509139725624, "grad_norm": 0.37282234452623014, "learning_rate": 9.600516093557618e-06, "loss": 0.4846, "step": 733 }, { "epoch": 0.4319393872521976, "grad_norm": 0.5097156100089998, "learning_rate": 9.598503070727705e-06, "loss": 0.5741, "step": 734 }, { "epoch": 0.43252786053183273, "grad_norm": 0.4361301494873747, "learning_rate": 9.596485200894275e-06, "loss": 0.5278, "step": 735 }, { "epoch": 0.43311633381146786, "grad_norm": 0.3993107046898724, "learning_rate": 9.594462486184237e-06, "loss": 0.5115, "step": 736 }, { "epoch": 0.43370480709110304, "grad_norm": 0.46328679336495787, "learning_rate": 9.592434928729617e-06, "loss": 0.5484, "step": 737 }, { "epoch": 0.43429328037073817, "grad_norm": 0.530005136996952, "learning_rate": 9.590402530667535e-06, "loss": 0.6341, "step": 738 }, { "epoch": 0.4348817536503733, "grad_norm": 0.46894920559546877, "learning_rate": 9.588365294140225e-06, "loss": 0.5835, "step": 739 }, { "epoch": 0.4354702269300085, "grad_norm": 0.44823436596216315, "learning_rate": 9.58632322129501e-06, "loss": 0.5391, "step": 740 }, { "epoch": 0.4360587002096436, "grad_norm": 0.4282026638412034, "learning_rate": 9.584276314284316e-06, "loss": 0.5382, "step": 741 }, { "epoch": 0.43664717348927873, "grad_norm": 0.44842633597105114, "learning_rate": 9.582224575265664e-06, "loss": 0.4826, "step": 742 }, { "epoch": 0.4372356467689139, "grad_norm": 0.41974518342735145, "learning_rate": 9.58016800640167e-06, "loss": 0.4997, "step": 743 }, { "epoch": 0.43782412004854904, "grad_norm": 0.45711607789805725, "learning_rate": 9.578106609860034e-06, "loss": 0.5804, "step": 744 }, { "epoch": 0.4384125933281842, "grad_norm": 0.4418023701013375, "learning_rate": 9.576040387813553e-06, "loss": 0.5576, "step": 745 }, { "epoch": 0.43900106660781935, "grad_norm": 0.4335030738055147, "learning_rate": 9.573969342440107e-06, "loss": 0.527, "step": 746 }, { "epoch": 0.4395895398874545, "grad_norm": 0.3699152659062541, "learning_rate": 9.571893475922656e-06, "loss": 0.4765, "step": 747 }, { "epoch": 0.44017801316708965, "grad_norm": 0.49341335542975884, "learning_rate": 9.569812790449246e-06, "loss": 0.5252, "step": 748 }, { "epoch": 0.4407664864467248, "grad_norm": 0.373185987213436, "learning_rate": 9.567727288213005e-06, "loss": 0.4813, "step": 749 }, { "epoch": 0.4413549597263599, "grad_norm": 0.43445884192943285, "learning_rate": 9.565636971412131e-06, "loss": 0.5375, "step": 750 }, { "epoch": 0.4419434330059951, "grad_norm": 0.4234093966758493, "learning_rate": 9.563541842249903e-06, "loss": 0.5231, "step": 751 }, { "epoch": 0.4425319062856302, "grad_norm": 0.379378441583178, "learning_rate": 9.561441902934667e-06, "loss": 0.4891, "step": 752 }, { "epoch": 0.44312037956526534, "grad_norm": 0.5112443625820544, "learning_rate": 9.559337155679843e-06, "loss": 0.6684, "step": 753 }, { "epoch": 0.4437088528449005, "grad_norm": 0.43552953573696285, "learning_rate": 9.557227602703918e-06, "loss": 0.5436, "step": 754 }, { "epoch": 0.44429732612453565, "grad_norm": 0.437153565600899, "learning_rate": 9.555113246230443e-06, "loss": 0.4816, "step": 755 }, { "epoch": 0.4448857994041708, "grad_norm": 0.3794796327919567, "learning_rate": 9.552994088488033e-06, "loss": 0.4704, "step": 756 }, { "epoch": 0.44547427268380596, "grad_norm": 0.4452416870556794, "learning_rate": 9.550870131710366e-06, "loss": 0.52, "step": 757 }, { "epoch": 0.4460627459634411, "grad_norm": 0.40182045085207796, "learning_rate": 9.548741378136173e-06, "loss": 0.5386, "step": 758 }, { "epoch": 0.44665121924307627, "grad_norm": 0.4144269482758643, "learning_rate": 9.546607830009247e-06, "loss": 0.5215, "step": 759 }, { "epoch": 0.4472396925227114, "grad_norm": 0.4693408944674734, "learning_rate": 9.544469489578432e-06, "loss": 0.5446, "step": 760 }, { "epoch": 0.4478281658023465, "grad_norm": 0.45290405191136257, "learning_rate": 9.542326359097619e-06, "loss": 0.5575, "step": 761 }, { "epoch": 0.4484166390819817, "grad_norm": 0.43663205774754354, "learning_rate": 9.540178440825757e-06, "loss": 0.5057, "step": 762 }, { "epoch": 0.44900511236161683, "grad_norm": 0.3391340950769974, "learning_rate": 9.538025737026833e-06, "loss": 0.4685, "step": 763 }, { "epoch": 0.44959358564125196, "grad_norm": 0.41382360071414537, "learning_rate": 9.535868249969882e-06, "loss": 0.5245, "step": 764 }, { "epoch": 0.45018205892088714, "grad_norm": 0.41268410410937484, "learning_rate": 9.533705981928984e-06, "loss": 0.5179, "step": 765 }, { "epoch": 0.45077053220052227, "grad_norm": 0.3940993238733123, "learning_rate": 9.531538935183252e-06, "loss": 0.5328, "step": 766 }, { "epoch": 0.4513590054801574, "grad_norm": 0.4328458427878601, "learning_rate": 9.529367112016836e-06, "loss": 0.4848, "step": 767 }, { "epoch": 0.4519474787597926, "grad_norm": 0.5252969187712663, "learning_rate": 9.527190514718929e-06, "loss": 0.5899, "step": 768 }, { "epoch": 0.4525359520394277, "grad_norm": 0.4570944230208852, "learning_rate": 9.525009145583746e-06, "loss": 0.6128, "step": 769 }, { "epoch": 0.45312442531906283, "grad_norm": 0.4188051843311699, "learning_rate": 9.522823006910538e-06, "loss": 0.4652, "step": 770 }, { "epoch": 0.453712898598698, "grad_norm": 0.4463534444452438, "learning_rate": 9.520632101003579e-06, "loss": 0.5839, "step": 771 }, { "epoch": 0.45430137187833314, "grad_norm": 0.47959505285244647, "learning_rate": 9.518436430172176e-06, "loss": 0.5388, "step": 772 }, { "epoch": 0.4548898451579683, "grad_norm": 0.4335550127383612, "learning_rate": 9.516235996730645e-06, "loss": 0.5016, "step": 773 }, { "epoch": 0.45547831843760345, "grad_norm": 0.47382335278330273, "learning_rate": 9.514030802998335e-06, "loss": 0.5422, "step": 774 }, { "epoch": 0.4560667917172386, "grad_norm": 0.42385664935572326, "learning_rate": 9.511820851299605e-06, "loss": 0.4967, "step": 775 }, { "epoch": 0.45665526499687376, "grad_norm": 0.36378648357796156, "learning_rate": 9.509606143963832e-06, "loss": 0.4565, "step": 776 }, { "epoch": 0.4572437382765089, "grad_norm": 0.4916421239858945, "learning_rate": 9.507386683325404e-06, "loss": 0.5609, "step": 777 }, { "epoch": 0.457832211556144, "grad_norm": 0.42958347648519896, "learning_rate": 9.50516247172372e-06, "loss": 0.5217, "step": 778 }, { "epoch": 0.4584206848357792, "grad_norm": 0.4392496919459478, "learning_rate": 9.502933511503187e-06, "loss": 0.5557, "step": 779 }, { "epoch": 0.4590091581154143, "grad_norm": 0.41190851412436785, "learning_rate": 9.500699805013217e-06, "loss": 0.4133, "step": 780 }, { "epoch": 0.45959763139504944, "grad_norm": 0.4748743774137102, "learning_rate": 9.498461354608228e-06, "loss": 0.5496, "step": 781 }, { "epoch": 0.4601861046746846, "grad_norm": 0.3965512273018047, "learning_rate": 9.496218162647629e-06, "loss": 0.5088, "step": 782 }, { "epoch": 0.46077457795431975, "grad_norm": 0.4317006721141422, "learning_rate": 9.493970231495836e-06, "loss": 0.5162, "step": 783 }, { "epoch": 0.46136305123395493, "grad_norm": 0.4323793730259307, "learning_rate": 9.491717563522256e-06, "loss": 0.5407, "step": 784 }, { "epoch": 0.46195152451359006, "grad_norm": 0.4716377137264482, "learning_rate": 9.489460161101291e-06, "loss": 0.5835, "step": 785 }, { "epoch": 0.4625399977932252, "grad_norm": 0.46913816041259915, "learning_rate": 9.487198026612333e-06, "loss": 0.5994, "step": 786 }, { "epoch": 0.46312847107286037, "grad_norm": 0.4609075680224678, "learning_rate": 9.484931162439756e-06, "loss": 0.537, "step": 787 }, { "epoch": 0.4637169443524955, "grad_norm": 0.45007268214704277, "learning_rate": 9.48265957097293e-06, "loss": 0.5409, "step": 788 }, { "epoch": 0.4643054176321306, "grad_norm": 0.4667356973275128, "learning_rate": 9.4803832546062e-06, "loss": 0.5082, "step": 789 }, { "epoch": 0.4648938909117658, "grad_norm": 0.42571752505119964, "learning_rate": 9.478102215738892e-06, "loss": 0.5792, "step": 790 }, { "epoch": 0.46548236419140093, "grad_norm": 0.4477287163503163, "learning_rate": 9.475816456775313e-06, "loss": 0.5563, "step": 791 }, { "epoch": 0.46607083747103606, "grad_norm": 0.5666541766450442, "learning_rate": 9.47352598012474e-06, "loss": 0.4997, "step": 792 }, { "epoch": 0.46665931075067124, "grad_norm": 0.4517298230161649, "learning_rate": 9.471230788201429e-06, "loss": 0.6071, "step": 793 }, { "epoch": 0.46724778403030637, "grad_norm": 0.4682463809576537, "learning_rate": 9.4689308834246e-06, "loss": 0.6029, "step": 794 }, { "epoch": 0.4678362573099415, "grad_norm": 0.4230785933176184, "learning_rate": 9.466626268218446e-06, "loss": 0.5237, "step": 795 }, { "epoch": 0.4684247305895767, "grad_norm": 0.4317062417421069, "learning_rate": 9.46431694501212e-06, "loss": 0.5364, "step": 796 }, { "epoch": 0.4690132038692118, "grad_norm": 0.3886777089609946, "learning_rate": 9.46200291623974e-06, "loss": 0.4805, "step": 797 }, { "epoch": 0.469601677148847, "grad_norm": 0.37554696407882404, "learning_rate": 9.459684184340385e-06, "loss": 0.4955, "step": 798 }, { "epoch": 0.4701901504284821, "grad_norm": 0.4939706511891661, "learning_rate": 9.45736075175809e-06, "loss": 0.5856, "step": 799 }, { "epoch": 0.47077862370811724, "grad_norm": 0.38155526222241504, "learning_rate": 9.45503262094184e-06, "loss": 0.5038, "step": 800 }, { "epoch": 0.4713670969877524, "grad_norm": 0.3981159646589328, "learning_rate": 9.452699794345583e-06, "loss": 0.5148, "step": 801 }, { "epoch": 0.47195557026738755, "grad_norm": 0.4452871710167638, "learning_rate": 9.450362274428207e-06, "loss": 0.5948, "step": 802 }, { "epoch": 0.4725440435470227, "grad_norm": 0.4496688604818848, "learning_rate": 9.44802006365355e-06, "loss": 0.5712, "step": 803 }, { "epoch": 0.47313251682665786, "grad_norm": 0.3960920069089525, "learning_rate": 9.445673164490394e-06, "loss": 0.503, "step": 804 }, { "epoch": 0.473720990106293, "grad_norm": 0.44838522298247985, "learning_rate": 9.443321579412465e-06, "loss": 0.5499, "step": 805 }, { "epoch": 0.4743094633859281, "grad_norm": 0.38629849829336954, "learning_rate": 9.440965310898425e-06, "loss": 0.4956, "step": 806 }, { "epoch": 0.4748979366655633, "grad_norm": 0.43234614896308715, "learning_rate": 9.438604361431876e-06, "loss": 0.5648, "step": 807 }, { "epoch": 0.4754864099451984, "grad_norm": 0.39093333174092076, "learning_rate": 9.436238733501349e-06, "loss": 0.5138, "step": 808 }, { "epoch": 0.47607488322483354, "grad_norm": 0.5097626430131461, "learning_rate": 9.43386842960031e-06, "loss": 0.5717, "step": 809 }, { "epoch": 0.4766633565044687, "grad_norm": 0.40469023301308826, "learning_rate": 9.431493452227154e-06, "loss": 0.4897, "step": 810 }, { "epoch": 0.47725182978410385, "grad_norm": 0.5004671569266327, "learning_rate": 9.429113803885199e-06, "loss": 0.5508, "step": 811 }, { "epoch": 0.47784030306373904, "grad_norm": 0.6083783845541281, "learning_rate": 9.426729487082691e-06, "loss": 0.5285, "step": 812 }, { "epoch": 0.47842877634337416, "grad_norm": 0.4056432525217363, "learning_rate": 9.42434050433279e-06, "loss": 0.5174, "step": 813 }, { "epoch": 0.4790172496230093, "grad_norm": 0.4188261870681317, "learning_rate": 9.421946858153581e-06, "loss": 0.4858, "step": 814 }, { "epoch": 0.47960572290264447, "grad_norm": 0.4483268580539617, "learning_rate": 9.419548551068061e-06, "loss": 0.4615, "step": 815 }, { "epoch": 0.4801941961822796, "grad_norm": 0.46849573636768066, "learning_rate": 9.417145585604139e-06, "loss": 0.5476, "step": 816 }, { "epoch": 0.4807826694619147, "grad_norm": 0.4514444517808957, "learning_rate": 9.414737964294636e-06, "loss": 0.579, "step": 817 }, { "epoch": 0.4813711427415499, "grad_norm": 0.46905374336628736, "learning_rate": 9.41232568967728e-06, "loss": 0.4571, "step": 818 }, { "epoch": 0.48195961602118503, "grad_norm": 0.4126955718623331, "learning_rate": 9.409908764294703e-06, "loss": 0.4642, "step": 819 }, { "epoch": 0.48254808930082016, "grad_norm": 0.42403537648272394, "learning_rate": 9.407487190694439e-06, "loss": 0.5225, "step": 820 }, { "epoch": 0.48313656258045534, "grad_norm": 0.4638740920156364, "learning_rate": 9.405060971428924e-06, "loss": 0.5156, "step": 821 }, { "epoch": 0.48372503586009047, "grad_norm": 0.5006953760062337, "learning_rate": 9.402630109055486e-06, "loss": 0.493, "step": 822 }, { "epoch": 0.48431350913972565, "grad_norm": 0.5517134673590581, "learning_rate": 9.400194606136354e-06, "loss": 0.6052, "step": 823 }, { "epoch": 0.4849019824193608, "grad_norm": 0.37701772024845875, "learning_rate": 9.39775446523864e-06, "loss": 0.4978, "step": 824 }, { "epoch": 0.4854904556989959, "grad_norm": 0.41298206757015726, "learning_rate": 9.39530968893435e-06, "loss": 0.5778, "step": 825 }, { "epoch": 0.4860789289786311, "grad_norm": 0.48696538390146776, "learning_rate": 9.392860279800377e-06, "loss": 0.6043, "step": 826 }, { "epoch": 0.4866674022582662, "grad_norm": 0.4124845763104821, "learning_rate": 9.39040624041849e-06, "loss": 0.5184, "step": 827 }, { "epoch": 0.48725587553790134, "grad_norm": 0.47226467574162545, "learning_rate": 9.387947573375347e-06, "loss": 0.59, "step": 828 }, { "epoch": 0.4878443488175365, "grad_norm": 0.3551416509523428, "learning_rate": 9.38548428126248e-06, "loss": 0.4471, "step": 829 }, { "epoch": 0.48843282209717165, "grad_norm": 0.4337801186653824, "learning_rate": 9.383016366676294e-06, "loss": 0.5237, "step": 830 }, { "epoch": 0.4890212953768068, "grad_norm": 0.40866454053713597, "learning_rate": 9.380543832218069e-06, "loss": 0.4716, "step": 831 }, { "epoch": 0.48960976865644196, "grad_norm": 0.44413808164063, "learning_rate": 9.378066680493953e-06, "loss": 0.5245, "step": 832 }, { "epoch": 0.4901982419360771, "grad_norm": 0.40803899292639517, "learning_rate": 9.375584914114963e-06, "loss": 0.5186, "step": 833 }, { "epoch": 0.4907867152157122, "grad_norm": 0.41467033480063503, "learning_rate": 9.37309853569698e-06, "loss": 0.5118, "step": 834 }, { "epoch": 0.4913751884953474, "grad_norm": 0.44567285788004096, "learning_rate": 9.370607547860741e-06, "loss": 0.5604, "step": 835 }, { "epoch": 0.4919636617749825, "grad_norm": 0.3849770810095028, "learning_rate": 9.368111953231849e-06, "loss": 0.5319, "step": 836 }, { "epoch": 0.4925521350546177, "grad_norm": 0.3897973320361136, "learning_rate": 9.365611754440756e-06, "loss": 0.5531, "step": 837 }, { "epoch": 0.49314060833425283, "grad_norm": 0.35410509464609086, "learning_rate": 9.363106954122772e-06, "loss": 0.4547, "step": 838 }, { "epoch": 0.49372908161388795, "grad_norm": 0.35267609572733416, "learning_rate": 9.360597554918055e-06, "loss": 0.3897, "step": 839 }, { "epoch": 0.49431755489352314, "grad_norm": 0.4244545957256661, "learning_rate": 9.35808355947161e-06, "loss": 0.5, "step": 840 }, { "epoch": 0.49490602817315826, "grad_norm": 0.4147188353562415, "learning_rate": 9.355564970433288e-06, "loss": 0.5185, "step": 841 }, { "epoch": 0.4954945014527934, "grad_norm": 1.2629342724054888, "learning_rate": 9.353041790457783e-06, "loss": 0.5848, "step": 842 }, { "epoch": 0.49608297473242857, "grad_norm": 0.5292273004259016, "learning_rate": 9.350514022204623e-06, "loss": 0.5539, "step": 843 }, { "epoch": 0.4966714480120637, "grad_norm": 0.4920615771593631, "learning_rate": 9.347981668338175e-06, "loss": 0.5168, "step": 844 }, { "epoch": 0.4972599212916988, "grad_norm": 0.4397968784765649, "learning_rate": 9.345444731527642e-06, "loss": 0.5299, "step": 845 }, { "epoch": 0.497848394571334, "grad_norm": 0.452069134599077, "learning_rate": 9.342903214447056e-06, "loss": 0.5594, "step": 846 }, { "epoch": 0.49843686785096913, "grad_norm": 0.41676024743375206, "learning_rate": 9.340357119775274e-06, "loss": 0.5922, "step": 847 }, { "epoch": 0.49902534113060426, "grad_norm": 0.47683766488321555, "learning_rate": 9.33780645019598e-06, "loss": 0.559, "step": 848 }, { "epoch": 0.49961381441023944, "grad_norm": 0.4298853131107885, "learning_rate": 9.335251208397684e-06, "loss": 0.5148, "step": 849 }, { "epoch": 0.5002022876898746, "grad_norm": 0.43090608283188664, "learning_rate": 9.332691397073708e-06, "loss": 0.5134, "step": 850 }, { "epoch": 0.5007907609695097, "grad_norm": 0.5190049984128997, "learning_rate": 9.330127018922195e-06, "loss": 0.4892, "step": 851 }, { "epoch": 0.5013792342491449, "grad_norm": 0.4774605829838334, "learning_rate": 9.3275580766461e-06, "loss": 0.4847, "step": 852 }, { "epoch": 0.5019677075287801, "grad_norm": 0.3968212527116857, "learning_rate": 9.32498457295319e-06, "loss": 0.4951, "step": 853 }, { "epoch": 0.5025561808084151, "grad_norm": 0.4712688421327052, "learning_rate": 9.32240651055604e-06, "loss": 0.5612, "step": 854 }, { "epoch": 0.5031446540880503, "grad_norm": 0.42401788794424283, "learning_rate": 9.319823892172029e-06, "loss": 0.4971, "step": 855 }, { "epoch": 0.5037331273676855, "grad_norm": 0.4842479422303751, "learning_rate": 9.317236720523338e-06, "loss": 0.5274, "step": 856 }, { "epoch": 0.5043216006473206, "grad_norm": 0.4706327468439315, "learning_rate": 9.31464499833695e-06, "loss": 0.5623, "step": 857 }, { "epoch": 0.5049100739269557, "grad_norm": 0.4725920088114532, "learning_rate": 9.312048728344639e-06, "loss": 0.4899, "step": 858 }, { "epoch": 0.5054985472065909, "grad_norm": 0.4649471011455456, "learning_rate": 9.309447913282978e-06, "loss": 0.5322, "step": 859 }, { "epoch": 0.506087020486226, "grad_norm": 0.3865342396164752, "learning_rate": 9.306842555893329e-06, "loss": 0.4776, "step": 860 }, { "epoch": 0.5066754937658612, "grad_norm": 0.4228463106632741, "learning_rate": 9.30423265892184e-06, "loss": 0.5343, "step": 861 }, { "epoch": 0.5072639670454964, "grad_norm": 0.4137761467522454, "learning_rate": 9.301618225119445e-06, "loss": 0.5183, "step": 862 }, { "epoch": 0.5078524403251314, "grad_norm": 0.41300584602030227, "learning_rate": 9.298999257241862e-06, "loss": 0.5273, "step": 863 }, { "epoch": 0.5084409136047666, "grad_norm": 0.48534010474571637, "learning_rate": 9.296375758049588e-06, "loss": 0.4694, "step": 864 }, { "epoch": 0.5090293868844018, "grad_norm": 0.46417700782860727, "learning_rate": 9.29374773030789e-06, "loss": 0.583, "step": 865 }, { "epoch": 0.5096178601640369, "grad_norm": 0.47877049812038086, "learning_rate": 9.291115176786814e-06, "loss": 0.5751, "step": 866 }, { "epoch": 0.510206333443672, "grad_norm": 0.4002610991510008, "learning_rate": 9.288478100261178e-06, "loss": 0.4995, "step": 867 }, { "epoch": 0.5107948067233072, "grad_norm": 0.424836542003035, "learning_rate": 9.285836503510562e-06, "loss": 0.4937, "step": 868 }, { "epoch": 0.5113832800029424, "grad_norm": 0.39386045174659495, "learning_rate": 9.283190389319315e-06, "loss": 0.5424, "step": 869 }, { "epoch": 0.5119717532825775, "grad_norm": 0.5062092324151903, "learning_rate": 9.280539760476543e-06, "loss": 0.5788, "step": 870 }, { "epoch": 0.5125602265622127, "grad_norm": 0.48480528191586575, "learning_rate": 9.277884619776116e-06, "loss": 0.5658, "step": 871 }, { "epoch": 0.5131486998418479, "grad_norm": 0.3781559783008652, "learning_rate": 9.275224970016656e-06, "loss": 0.4858, "step": 872 }, { "epoch": 0.5137371731214829, "grad_norm": 0.45482821087902936, "learning_rate": 9.27256081400154e-06, "loss": 0.5385, "step": 873 }, { "epoch": 0.5143256464011181, "grad_norm": 0.46661102743065347, "learning_rate": 9.26989215453889e-06, "loss": 0.5593, "step": 874 }, { "epoch": 0.5149141196807533, "grad_norm": 0.40151056268319163, "learning_rate": 9.26721899444158e-06, "loss": 0.5234, "step": 875 }, { "epoch": 0.5155025929603884, "grad_norm": 0.4302869776813288, "learning_rate": 9.264541336527228e-06, "loss": 0.5882, "step": 876 }, { "epoch": 0.5160910662400235, "grad_norm": 0.3783512998727196, "learning_rate": 9.26185918361819e-06, "loss": 0.4816, "step": 877 }, { "epoch": 0.5166795395196587, "grad_norm": 0.42370175899672835, "learning_rate": 9.259172538541558e-06, "loss": 0.5389, "step": 878 }, { "epoch": 0.5172680127992938, "grad_norm": 0.40686619308616434, "learning_rate": 9.256481404129165e-06, "loss": 0.513, "step": 879 }, { "epoch": 0.517856486078929, "grad_norm": 0.4257737078955358, "learning_rate": 9.253785783217573e-06, "loss": 0.5304, "step": 880 }, { "epoch": 0.5184449593585642, "grad_norm": 0.5146679415515503, "learning_rate": 9.251085678648072e-06, "loss": 0.5238, "step": 881 }, { "epoch": 0.5190334326381992, "grad_norm": 0.39453928798910015, "learning_rate": 9.248381093266677e-06, "loss": 0.4817, "step": 882 }, { "epoch": 0.5196219059178344, "grad_norm": 0.5036483231798066, "learning_rate": 9.24567202992413e-06, "loss": 0.5631, "step": 883 }, { "epoch": 0.5202103791974696, "grad_norm": 0.43326965587217, "learning_rate": 9.242958491475888e-06, "loss": 0.5616, "step": 884 }, { "epoch": 0.5207988524771047, "grad_norm": 0.4852690240488085, "learning_rate": 9.24024048078213e-06, "loss": 0.6223, "step": 885 }, { "epoch": 0.5213873257567399, "grad_norm": 0.41697215765491924, "learning_rate": 9.237518000707745e-06, "loss": 0.5446, "step": 886 }, { "epoch": 0.521975799036375, "grad_norm": 0.45977852246839335, "learning_rate": 9.234791054122336e-06, "loss": 0.5648, "step": 887 }, { "epoch": 0.5225642723160101, "grad_norm": 0.4464909029553939, "learning_rate": 9.23205964390021e-06, "loss": 0.5407, "step": 888 }, { "epoch": 0.5231527455956453, "grad_norm": 0.3952521604306112, "learning_rate": 9.229323772920383e-06, "loss": 0.5054, "step": 889 }, { "epoch": 0.5237412188752805, "grad_norm": 0.41936412958744035, "learning_rate": 9.22658344406657e-06, "loss": 0.5463, "step": 890 }, { "epoch": 0.5243296921549155, "grad_norm": 0.4161156441259018, "learning_rate": 9.223838660227183e-06, "loss": 0.5417, "step": 891 }, { "epoch": 0.5249181654345507, "grad_norm": 1.0188583717944273, "learning_rate": 9.221089424295338e-06, "loss": 0.5652, "step": 892 }, { "epoch": 0.5255066387141859, "grad_norm": 0.4062864670060824, "learning_rate": 9.218335739168833e-06, "loss": 0.5484, "step": 893 }, { "epoch": 0.5260951119938211, "grad_norm": 0.4884592731198578, "learning_rate": 9.215577607750162e-06, "loss": 0.5381, "step": 894 }, { "epoch": 0.5266835852734562, "grad_norm": 0.3361786367786834, "learning_rate": 9.212815032946507e-06, "loss": 0.4237, "step": 895 }, { "epoch": 0.5272720585530913, "grad_norm": 0.41934492830810105, "learning_rate": 9.210048017669727e-06, "loss": 0.5719, "step": 896 }, { "epoch": 0.5278605318327265, "grad_norm": 0.3794235095219038, "learning_rate": 9.207276564836367e-06, "loss": 0.5044, "step": 897 }, { "epoch": 0.5284490051123616, "grad_norm": 0.44829307654848716, "learning_rate": 9.204500677367648e-06, "loss": 0.4735, "step": 898 }, { "epoch": 0.5290374783919968, "grad_norm": 0.5194290533557633, "learning_rate": 9.201720358189464e-06, "loss": 0.6002, "step": 899 }, { "epoch": 0.529625951671632, "grad_norm": 0.3982830986167171, "learning_rate": 9.198935610232383e-06, "loss": 0.5136, "step": 900 }, { "epoch": 0.530214424951267, "grad_norm": 0.4235317348039236, "learning_rate": 9.196146436431635e-06, "loss": 0.5446, "step": 901 }, { "epoch": 0.5308028982309022, "grad_norm": 0.4221891292874487, "learning_rate": 9.193352839727122e-06, "loss": 0.5146, "step": 902 }, { "epoch": 0.5313913715105374, "grad_norm": 0.4192679555484145, "learning_rate": 9.190554823063403e-06, "loss": 0.51, "step": 903 }, { "epoch": 0.5319798447901725, "grad_norm": 0.42290517639108166, "learning_rate": 9.187752389389702e-06, "loss": 0.5964, "step": 904 }, { "epoch": 0.5325683180698076, "grad_norm": 0.4843431870507824, "learning_rate": 9.18494554165989e-06, "loss": 0.4792, "step": 905 }, { "epoch": 0.5331567913494428, "grad_norm": 0.4078107697520062, "learning_rate": 9.182134282832498e-06, "loss": 0.4944, "step": 906 }, { "epoch": 0.5337452646290779, "grad_norm": 0.44180773834200165, "learning_rate": 9.179318615870703e-06, "loss": 0.5748, "step": 907 }, { "epoch": 0.5343337379087131, "grad_norm": 0.4022398375126535, "learning_rate": 9.176498543742328e-06, "loss": 0.5342, "step": 908 }, { "epoch": 0.5349222111883483, "grad_norm": 0.4343129598765399, "learning_rate": 9.173674069419843e-06, "loss": 0.554, "step": 909 }, { "epoch": 0.5355106844679833, "grad_norm": 0.3712386414681519, "learning_rate": 9.17084519588035e-06, "loss": 0.4657, "step": 910 }, { "epoch": 0.5360991577476185, "grad_norm": 0.4208429931155654, "learning_rate": 9.168011926105598e-06, "loss": 0.5504, "step": 911 }, { "epoch": 0.5366876310272537, "grad_norm": 0.41401883249691096, "learning_rate": 9.165174263081964e-06, "loss": 0.5026, "step": 912 }, { "epoch": 0.5372761043068888, "grad_norm": 0.4052661309907464, "learning_rate": 9.162332209800455e-06, "loss": 0.4338, "step": 913 }, { "epoch": 0.537864577586524, "grad_norm": 0.3815128344932401, "learning_rate": 9.15948576925671e-06, "loss": 0.4803, "step": 914 }, { "epoch": 0.5384530508661591, "grad_norm": 0.41467003834061283, "learning_rate": 9.156634944450987e-06, "loss": 0.5174, "step": 915 }, { "epoch": 0.5390415241457942, "grad_norm": 0.46575345680314767, "learning_rate": 9.15377973838817e-06, "loss": 0.5288, "step": 916 }, { "epoch": 0.5396299974254294, "grad_norm": 0.4686333617004513, "learning_rate": 9.150920154077753e-06, "loss": 0.5331, "step": 917 }, { "epoch": 0.5402184707050646, "grad_norm": 0.3647047451049336, "learning_rate": 9.148056194533858e-06, "loss": 0.4444, "step": 918 }, { "epoch": 0.5408069439846996, "grad_norm": 0.6481323515048117, "learning_rate": 9.145187862775208e-06, "loss": 0.6661, "step": 919 }, { "epoch": 0.5413954172643348, "grad_norm": 0.3815196614881027, "learning_rate": 9.14231516182514e-06, "loss": 0.5177, "step": 920 }, { "epoch": 0.54198389054397, "grad_norm": 0.42108655919569593, "learning_rate": 9.13943809471159e-06, "loss": 0.4563, "step": 921 }, { "epoch": 0.5425723638236052, "grad_norm": 0.43556618593960694, "learning_rate": 9.136556664467102e-06, "loss": 0.4692, "step": 922 }, { "epoch": 0.5431608371032403, "grad_norm": 0.37402079160534846, "learning_rate": 9.133670874128818e-06, "loss": 0.4524, "step": 923 }, { "epoch": 0.5437493103828754, "grad_norm": 0.4475351806102736, "learning_rate": 9.130780726738474e-06, "loss": 0.5131, "step": 924 }, { "epoch": 0.5443377836625106, "grad_norm": 0.4716459112841034, "learning_rate": 9.1278862253424e-06, "loss": 0.5481, "step": 925 }, { "epoch": 0.5449262569421457, "grad_norm": 0.42979671398336194, "learning_rate": 9.124987372991512e-06, "loss": 0.5091, "step": 926 }, { "epoch": 0.5455147302217809, "grad_norm": 0.3716849346023735, "learning_rate": 9.122084172741317e-06, "loss": 0.4639, "step": 927 }, { "epoch": 0.5461032035014161, "grad_norm": 0.416845881360676, "learning_rate": 9.119176627651905e-06, "loss": 0.531, "step": 928 }, { "epoch": 0.5466916767810511, "grad_norm": 0.43054309189881956, "learning_rate": 9.116264740787937e-06, "loss": 0.5198, "step": 929 }, { "epoch": 0.5472801500606863, "grad_norm": 0.39214574559395604, "learning_rate": 9.11334851521866e-06, "loss": 0.5145, "step": 930 }, { "epoch": 0.5478686233403215, "grad_norm": 0.5336214795415093, "learning_rate": 9.110427954017891e-06, "loss": 0.5507, "step": 931 }, { "epoch": 0.5484570966199566, "grad_norm": 0.37757515324885693, "learning_rate": 9.107503060264018e-06, "loss": 0.5592, "step": 932 }, { "epoch": 0.5490455698995917, "grad_norm": 0.3983154978211054, "learning_rate": 9.104573837039992e-06, "loss": 0.5011, "step": 933 }, { "epoch": 0.5496340431792269, "grad_norm": 0.41743494527407216, "learning_rate": 9.101640287433328e-06, "loss": 0.5184, "step": 934 }, { "epoch": 0.550222516458862, "grad_norm": 0.4081423194382402, "learning_rate": 9.098702414536107e-06, "loss": 0.5262, "step": 935 }, { "epoch": 0.5508109897384972, "grad_norm": 0.4613818439575659, "learning_rate": 9.09576022144496e-06, "loss": 0.5086, "step": 936 }, { "epoch": 0.5513994630181324, "grad_norm": 0.3993496811938432, "learning_rate": 9.092813711261075e-06, "loss": 0.54, "step": 937 }, { "epoch": 0.5519879362977674, "grad_norm": 0.4261888346636569, "learning_rate": 9.089862887090191e-06, "loss": 0.517, "step": 938 }, { "epoch": 0.5525764095774026, "grad_norm": 0.4139915340702709, "learning_rate": 9.086907752042593e-06, "loss": 0.5201, "step": 939 }, { "epoch": 0.5531648828570378, "grad_norm": 0.3820360342865949, "learning_rate": 9.08394830923311e-06, "loss": 0.4894, "step": 940 }, { "epoch": 0.5537533561366729, "grad_norm": 0.46053706801908406, "learning_rate": 9.08098456178111e-06, "loss": 0.5684, "step": 941 }, { "epoch": 0.554341829416308, "grad_norm": 0.4257446257325766, "learning_rate": 9.0780165128105e-06, "loss": 0.5474, "step": 942 }, { "epoch": 0.5549303026959432, "grad_norm": 0.4134962274731476, "learning_rate": 9.07504416544972e-06, "loss": 0.4738, "step": 943 }, { "epoch": 0.5555187759755783, "grad_norm": 0.4507215460877517, "learning_rate": 9.072067522831743e-06, "loss": 0.5978, "step": 944 }, { "epoch": 0.5561072492552135, "grad_norm": 0.398208079193817, "learning_rate": 9.069086588094067e-06, "loss": 0.4689, "step": 945 }, { "epoch": 0.5566957225348487, "grad_norm": 0.4984476336436579, "learning_rate": 9.066101364378714e-06, "loss": 0.5051, "step": 946 }, { "epoch": 0.5572841958144839, "grad_norm": 0.3916126165713654, "learning_rate": 9.06311185483223e-06, "loss": 0.4913, "step": 947 }, { "epoch": 0.5578726690941189, "grad_norm": 0.3614643670012527, "learning_rate": 9.060118062605671e-06, "loss": 0.4871, "step": 948 }, { "epoch": 0.5584611423737541, "grad_norm": 0.4126731597816729, "learning_rate": 9.057119990854617e-06, "loss": 0.5175, "step": 949 }, { "epoch": 0.5590496156533893, "grad_norm": 0.38383973999049, "learning_rate": 9.054117642739151e-06, "loss": 0.4828, "step": 950 }, { "epoch": 0.5596380889330244, "grad_norm": 0.3934501926968992, "learning_rate": 9.051111021423868e-06, "loss": 0.4988, "step": 951 }, { "epoch": 0.5602265622126595, "grad_norm": 0.47600418007242995, "learning_rate": 9.048100130077864e-06, "loss": 0.4991, "step": 952 }, { "epoch": 0.5608150354922947, "grad_norm": 0.39882536311246847, "learning_rate": 9.045084971874738e-06, "loss": 0.4613, "step": 953 }, { "epoch": 0.5614035087719298, "grad_norm": 0.3891414369621024, "learning_rate": 9.042065549992584e-06, "loss": 0.4954, "step": 954 }, { "epoch": 0.561991982051565, "grad_norm": 0.4461922407967782, "learning_rate": 9.039041867613994e-06, "loss": 0.5728, "step": 955 }, { "epoch": 0.5625804553312002, "grad_norm": 0.48050784313836925, "learning_rate": 9.036013927926049e-06, "loss": 0.5462, "step": 956 }, { "epoch": 0.5631689286108352, "grad_norm": 0.4052172649263698, "learning_rate": 9.032981734120312e-06, "loss": 0.531, "step": 957 }, { "epoch": 0.5637574018904704, "grad_norm": 0.43024627722117165, "learning_rate": 9.029945289392841e-06, "loss": 0.5103, "step": 958 }, { "epoch": 0.5643458751701056, "grad_norm": 0.5781899309167072, "learning_rate": 9.026904596944163e-06, "loss": 0.5505, "step": 959 }, { "epoch": 0.5649343484497407, "grad_norm": 0.5959889618884591, "learning_rate": 9.023859659979293e-06, "loss": 0.5102, "step": 960 }, { "epoch": 0.5655228217293758, "grad_norm": 0.46974352598039076, "learning_rate": 9.020810481707709e-06, "loss": 0.5561, "step": 961 }, { "epoch": 0.566111295009011, "grad_norm": 0.38310003432121414, "learning_rate": 9.017757065343368e-06, "loss": 0.5167, "step": 962 }, { "epoch": 0.5666997682886461, "grad_norm": 0.4510576102168717, "learning_rate": 9.014699414104693e-06, "loss": 0.571, "step": 963 }, { "epoch": 0.5672882415682813, "grad_norm": 0.41698152365443275, "learning_rate": 9.011637531214566e-06, "loss": 0.491, "step": 964 }, { "epoch": 0.5678767148479165, "grad_norm": 0.5078228323625408, "learning_rate": 9.008571419900334e-06, "loss": 0.6244, "step": 965 }, { "epoch": 0.5684651881275515, "grad_norm": 0.45191364890227786, "learning_rate": 9.005501083393799e-06, "loss": 0.523, "step": 966 }, { "epoch": 0.5690536614071867, "grad_norm": 0.4221923621870389, "learning_rate": 9.002426524931213e-06, "loss": 0.5209, "step": 967 }, { "epoch": 0.5696421346868219, "grad_norm": 0.40604763211893236, "learning_rate": 8.999347747753288e-06, "loss": 0.4596, "step": 968 }, { "epoch": 0.570230607966457, "grad_norm": 0.3999276692039076, "learning_rate": 8.996264755105173e-06, "loss": 0.5418, "step": 969 }, { "epoch": 0.5708190812460922, "grad_norm": 0.3668145170027037, "learning_rate": 8.993177550236464e-06, "loss": 0.5126, "step": 970 }, { "epoch": 0.5714075545257273, "grad_norm": 0.41762659510309286, "learning_rate": 8.990086136401199e-06, "loss": 0.5717, "step": 971 }, { "epoch": 0.5719960278053625, "grad_norm": 0.36840936042632133, "learning_rate": 8.986990516857846e-06, "loss": 0.4723, "step": 972 }, { "epoch": 0.5725845010849976, "grad_norm": 0.43212301224280486, "learning_rate": 8.983890694869312e-06, "loss": 0.5702, "step": 973 }, { "epoch": 0.5731729743646328, "grad_norm": 0.41357095784445247, "learning_rate": 8.980786673702932e-06, "loss": 0.5291, "step": 974 }, { "epoch": 0.573761447644268, "grad_norm": 0.4509010760388777, "learning_rate": 8.977678456630466e-06, "loss": 0.5032, "step": 975 }, { "epoch": 0.574349920923903, "grad_norm": 0.387973150711562, "learning_rate": 8.9745660469281e-06, "loss": 0.5197, "step": 976 }, { "epoch": 0.5749383942035382, "grad_norm": 0.43033062438407693, "learning_rate": 8.97144944787643e-06, "loss": 0.4841, "step": 977 }, { "epoch": 0.5755268674831734, "grad_norm": 0.46314535275008556, "learning_rate": 8.96832866276048e-06, "loss": 0.526, "step": 978 }, { "epoch": 0.5761153407628085, "grad_norm": 0.5413442110272423, "learning_rate": 8.965203694869682e-06, "loss": 0.5397, "step": 979 }, { "epoch": 0.5767038140424436, "grad_norm": 0.6086275789625645, "learning_rate": 8.96207454749787e-06, "loss": 0.5441, "step": 980 }, { "epoch": 0.5772922873220788, "grad_norm": 0.40558442169877007, "learning_rate": 8.958941223943292e-06, "loss": 0.5263, "step": 981 }, { "epoch": 0.5778807606017139, "grad_norm": 0.37803524934880534, "learning_rate": 8.955803727508592e-06, "loss": 0.501, "step": 982 }, { "epoch": 0.5784692338813491, "grad_norm": 0.36072119936670016, "learning_rate": 8.952662061500817e-06, "loss": 0.4626, "step": 983 }, { "epoch": 0.5790577071609843, "grad_norm": 0.43003581043862477, "learning_rate": 8.949516229231407e-06, "loss": 0.5478, "step": 984 }, { "epoch": 0.5796461804406193, "grad_norm": 0.41882150535174784, "learning_rate": 8.946366234016192e-06, "loss": 0.5586, "step": 985 }, { "epoch": 0.5802346537202545, "grad_norm": 0.3511978745644142, "learning_rate": 8.943212079175392e-06, "loss": 0.4542, "step": 986 }, { "epoch": 0.5808231269998897, "grad_norm": 0.4845444895138334, "learning_rate": 8.94005376803361e-06, "loss": 0.6227, "step": 987 }, { "epoch": 0.5814116002795248, "grad_norm": 0.44861573662085247, "learning_rate": 8.936891303919832e-06, "loss": 0.559, "step": 988 }, { "epoch": 0.58200007355916, "grad_norm": 0.36241870069232485, "learning_rate": 8.933724690167417e-06, "loss": 0.4809, "step": 989 }, { "epoch": 0.5825885468387951, "grad_norm": 0.38996770304128536, "learning_rate": 8.930553930114103e-06, "loss": 0.537, "step": 990 }, { "epoch": 0.5831770201184302, "grad_norm": 0.4432300917491085, "learning_rate": 8.927379027101994e-06, "loss": 0.5232, "step": 991 }, { "epoch": 0.5837654933980654, "grad_norm": 0.3808460065835783, "learning_rate": 8.924199984477566e-06, "loss": 0.4811, "step": 992 }, { "epoch": 0.5843539666777006, "grad_norm": 0.3720576996281784, "learning_rate": 8.921016805591654e-06, "loss": 0.4935, "step": 993 }, { "epoch": 0.5849424399573356, "grad_norm": 0.44158394218293484, "learning_rate": 8.917829493799453e-06, "loss": 0.6013, "step": 994 }, { "epoch": 0.5855309132369708, "grad_norm": 0.37765875237771895, "learning_rate": 8.914638052460515e-06, "loss": 0.5041, "step": 995 }, { "epoch": 0.586119386516606, "grad_norm": 0.4050858334074909, "learning_rate": 8.911442484938747e-06, "loss": 0.5025, "step": 996 }, { "epoch": 0.5867078597962412, "grad_norm": 0.5490923513552772, "learning_rate": 8.908242794602401e-06, "loss": 0.6012, "step": 997 }, { "epoch": 0.5872963330758763, "grad_norm": 0.370262490203064, "learning_rate": 8.905038984824079e-06, "loss": 0.4364, "step": 998 }, { "epoch": 0.5878848063555114, "grad_norm": 0.3909271652796007, "learning_rate": 8.901831058980718e-06, "loss": 0.4306, "step": 999 }, { "epoch": 0.5884732796351466, "grad_norm": 0.39905950887912645, "learning_rate": 8.898619020453606e-06, "loss": 0.4764, "step": 1000 }, { "epoch": 0.5890617529147817, "grad_norm": 0.4136932059238519, "learning_rate": 8.895402872628352e-06, "loss": 0.4344, "step": 1001 }, { "epoch": 0.5896502261944169, "grad_norm": 0.4107396367419501, "learning_rate": 8.892182618894906e-06, "loss": 0.4683, "step": 1002 }, { "epoch": 0.590238699474052, "grad_norm": 0.3622041416726245, "learning_rate": 8.88895826264754e-06, "loss": 0.4795, "step": 1003 }, { "epoch": 0.5908271727536871, "grad_norm": 0.3860937653308273, "learning_rate": 8.885729807284855e-06, "loss": 0.4675, "step": 1004 }, { "epoch": 0.5914156460333223, "grad_norm": 0.4227604260234112, "learning_rate": 8.882497256209767e-06, "loss": 0.4802, "step": 1005 }, { "epoch": 0.5920041193129575, "grad_norm": 0.393693560622056, "learning_rate": 8.879260612829517e-06, "loss": 0.4925, "step": 1006 }, { "epoch": 0.5925925925925926, "grad_norm": 0.41496107163627693, "learning_rate": 8.87601988055565e-06, "loss": 0.558, "step": 1007 }, { "epoch": 0.5931810658722277, "grad_norm": 0.420229265478634, "learning_rate": 8.872775062804028e-06, "loss": 0.523, "step": 1008 }, { "epoch": 0.5937695391518629, "grad_norm": 0.43211097016738886, "learning_rate": 8.869526162994814e-06, "loss": 0.4684, "step": 1009 }, { "epoch": 0.594358012431498, "grad_norm": 0.4269026210735302, "learning_rate": 8.866273184552483e-06, "loss": 0.5542, "step": 1010 }, { "epoch": 0.5949464857111332, "grad_norm": 0.38433720022917733, "learning_rate": 8.863016130905795e-06, "loss": 0.3888, "step": 1011 }, { "epoch": 0.5955349589907684, "grad_norm": 0.3665485363470546, "learning_rate": 8.859755005487818e-06, "loss": 0.4571, "step": 1012 }, { "epoch": 0.5961234322704034, "grad_norm": 0.3393154461866472, "learning_rate": 8.856489811735904e-06, "loss": 0.43, "step": 1013 }, { "epoch": 0.5967119055500386, "grad_norm": 0.3578694267078152, "learning_rate": 8.853220553091698e-06, "loss": 0.4689, "step": 1014 }, { "epoch": 0.5973003788296738, "grad_norm": 0.43867175336354935, "learning_rate": 8.849947233001125e-06, "loss": 0.5322, "step": 1015 }, { "epoch": 0.5978888521093089, "grad_norm": 0.41118236495746724, "learning_rate": 8.846669854914395e-06, "loss": 0.5939, "step": 1016 }, { "epoch": 0.598477325388944, "grad_norm": 0.3954931683606735, "learning_rate": 8.843388422285995e-06, "loss": 0.5192, "step": 1017 }, { "epoch": 0.5990657986685792, "grad_norm": 0.36822978368376574, "learning_rate": 8.840102938574682e-06, "loss": 0.4965, "step": 1018 }, { "epoch": 0.5996542719482143, "grad_norm": 0.4149652845555159, "learning_rate": 8.836813407243485e-06, "loss": 0.5251, "step": 1019 }, { "epoch": 0.6002427452278495, "grad_norm": 0.40775612655064375, "learning_rate": 8.833519831759703e-06, "loss": 0.5493, "step": 1020 }, { "epoch": 0.6008312185074847, "grad_norm": 0.403803076282041, "learning_rate": 8.83022221559489e-06, "loss": 0.4593, "step": 1021 }, { "epoch": 0.6014196917871197, "grad_norm": 0.36365121789703597, "learning_rate": 8.826920562224868e-06, "loss": 0.46, "step": 1022 }, { "epoch": 0.6020081650667549, "grad_norm": 0.5762414306441389, "learning_rate": 8.823614875129707e-06, "loss": 0.5226, "step": 1023 }, { "epoch": 0.6025966383463901, "grad_norm": 0.4896592576698148, "learning_rate": 8.820305157793732e-06, "loss": 0.5411, "step": 1024 }, { "epoch": 0.6031851116260253, "grad_norm": 0.43659897299290723, "learning_rate": 8.816991413705515e-06, "loss": 0.5395, "step": 1025 }, { "epoch": 0.6037735849056604, "grad_norm": 0.34289554582415926, "learning_rate": 8.813673646357873e-06, "loss": 0.507, "step": 1026 }, { "epoch": 0.6043620581852955, "grad_norm": 0.4309117529629677, "learning_rate": 8.810351859247868e-06, "loss": 0.5415, "step": 1027 }, { "epoch": 0.6049505314649307, "grad_norm": 0.39039368093668825, "learning_rate": 8.807026055876787e-06, "loss": 0.545, "step": 1028 }, { "epoch": 0.6055390047445658, "grad_norm": 0.43162539672219513, "learning_rate": 8.803696239750163e-06, "loss": 0.5083, "step": 1029 }, { "epoch": 0.606127478024201, "grad_norm": 0.3525932263790567, "learning_rate": 8.800362414377753e-06, "loss": 0.4187, "step": 1030 }, { "epoch": 0.6067159513038362, "grad_norm": 0.46480969049315457, "learning_rate": 8.797024583273536e-06, "loss": 0.5339, "step": 1031 }, { "epoch": 0.6073044245834712, "grad_norm": 0.46372535040805124, "learning_rate": 8.793682749955724e-06, "loss": 0.5206, "step": 1032 }, { "epoch": 0.6078928978631064, "grad_norm": 0.36420818049162273, "learning_rate": 8.790336917946737e-06, "loss": 0.4706, "step": 1033 }, { "epoch": 0.6084813711427416, "grad_norm": 0.46111865210376785, "learning_rate": 8.786987090773214e-06, "loss": 0.5908, "step": 1034 }, { "epoch": 0.6090698444223767, "grad_norm": 0.425579666212456, "learning_rate": 8.783633271966008e-06, "loss": 0.5523, "step": 1035 }, { "epoch": 0.6096583177020118, "grad_norm": 0.3936565064116562, "learning_rate": 8.780275465060173e-06, "loss": 0.4992, "step": 1036 }, { "epoch": 0.610246790981647, "grad_norm": 0.4163653279282347, "learning_rate": 8.776913673594968e-06, "loss": 0.4044, "step": 1037 }, { "epoch": 0.6108352642612821, "grad_norm": 0.4009550155794073, "learning_rate": 8.773547901113862e-06, "loss": 0.4914, "step": 1038 }, { "epoch": 0.6114237375409173, "grad_norm": 0.4769054512426748, "learning_rate": 8.770178151164503e-06, "loss": 0.5262, "step": 1039 }, { "epoch": 0.6120122108205525, "grad_norm": 0.36294760355801275, "learning_rate": 8.766804427298745e-06, "loss": 0.4434, "step": 1040 }, { "epoch": 0.6126006841001875, "grad_norm": 0.4577902587236015, "learning_rate": 8.763426733072624e-06, "loss": 0.4602, "step": 1041 }, { "epoch": 0.6131891573798227, "grad_norm": 0.3884339112649091, "learning_rate": 8.760045072046363e-06, "loss": 0.4893, "step": 1042 }, { "epoch": 0.6137776306594579, "grad_norm": 0.4986429715914976, "learning_rate": 8.756659447784367e-06, "loss": 0.5526, "step": 1043 }, { "epoch": 0.614366103939093, "grad_norm": 0.39330698452291557, "learning_rate": 8.753269863855218e-06, "loss": 0.4581, "step": 1044 }, { "epoch": 0.6149545772187281, "grad_norm": 0.4013121474063716, "learning_rate": 8.74987632383167e-06, "loss": 0.4781, "step": 1045 }, { "epoch": 0.6155430504983633, "grad_norm": 0.453931069577336, "learning_rate": 8.746478831290648e-06, "loss": 0.5271, "step": 1046 }, { "epoch": 0.6161315237779984, "grad_norm": 0.3689865540228218, "learning_rate": 8.74307738981324e-06, "loss": 0.5278, "step": 1047 }, { "epoch": 0.6167199970576336, "grad_norm": 0.37540369645771665, "learning_rate": 8.739672002984707e-06, "loss": 0.4687, "step": 1048 }, { "epoch": 0.6173084703372688, "grad_norm": 0.4561408734692791, "learning_rate": 8.736262674394455e-06, "loss": 0.5266, "step": 1049 }, { "epoch": 0.617896943616904, "grad_norm": 0.38707395055031896, "learning_rate": 8.732849407636052e-06, "loss": 0.5033, "step": 1050 }, { "epoch": 0.618485416896539, "grad_norm": 0.42364899208167617, "learning_rate": 8.729432206307218e-06, "loss": 0.6021, "step": 1051 }, { "epoch": 0.6190738901761742, "grad_norm": 0.3747275465788337, "learning_rate": 8.726011074009813e-06, "loss": 0.501, "step": 1052 }, { "epoch": 0.6196623634558094, "grad_norm": 0.46098537731392075, "learning_rate": 8.722586014349851e-06, "loss": 0.5826, "step": 1053 }, { "epoch": 0.6202508367354445, "grad_norm": 0.42276391315354656, "learning_rate": 8.719157030937475e-06, "loss": 0.5385, "step": 1054 }, { "epoch": 0.6208393100150796, "grad_norm": 0.3672044979368004, "learning_rate": 8.715724127386971e-06, "loss": 0.4708, "step": 1055 }, { "epoch": 0.6214277832947148, "grad_norm": 0.35898288521784905, "learning_rate": 8.712287307316756e-06, "loss": 0.4131, "step": 1056 }, { "epoch": 0.6220162565743499, "grad_norm": 0.4392943212861744, "learning_rate": 8.708846574349372e-06, "loss": 0.5207, "step": 1057 }, { "epoch": 0.6226047298539851, "grad_norm": 0.3823057648987862, "learning_rate": 8.705401932111487e-06, "loss": 0.4753, "step": 1058 }, { "epoch": 0.6231932031336203, "grad_norm": 0.39756042583519685, "learning_rate": 8.70195338423389e-06, "loss": 0.5033, "step": 1059 }, { "epoch": 0.6237816764132553, "grad_norm": 0.4708613867767224, "learning_rate": 8.698500934351488e-06, "loss": 0.5765, "step": 1060 }, { "epoch": 0.6243701496928905, "grad_norm": 0.4008015613597958, "learning_rate": 8.695044586103297e-06, "loss": 0.5045, "step": 1061 }, { "epoch": 0.6249586229725257, "grad_norm": 0.4109931840032277, "learning_rate": 8.691584343132445e-06, "loss": 0.5251, "step": 1062 }, { "epoch": 0.6255470962521608, "grad_norm": 0.3946605813486296, "learning_rate": 8.688120209086165e-06, "loss": 0.4898, "step": 1063 }, { "epoch": 0.6261355695317959, "grad_norm": 0.41063625249413227, "learning_rate": 8.68465218761579e-06, "loss": 0.5164, "step": 1064 }, { "epoch": 0.6267240428114311, "grad_norm": 0.3944534187899232, "learning_rate": 8.681180282376754e-06, "loss": 0.4982, "step": 1065 }, { "epoch": 0.6273125160910662, "grad_norm": 0.3972213633854106, "learning_rate": 8.677704497028579e-06, "loss": 0.4803, "step": 1066 }, { "epoch": 0.6279009893707014, "grad_norm": 0.40852636097195316, "learning_rate": 8.674224835234879e-06, "loss": 0.4711, "step": 1067 }, { "epoch": 0.6284894626503366, "grad_norm": 0.40127155093137956, "learning_rate": 8.670741300663359e-06, "loss": 0.49, "step": 1068 }, { "epoch": 0.6290779359299716, "grad_norm": 0.41976001058099033, "learning_rate": 8.667253896985796e-06, "loss": 0.5222, "step": 1069 }, { "epoch": 0.6296664092096068, "grad_norm": 0.4294536792753699, "learning_rate": 8.663762627878059e-06, "loss": 0.4904, "step": 1070 }, { "epoch": 0.630254882489242, "grad_norm": 0.4351516052598317, "learning_rate": 8.660267497020074e-06, "loss": 0.5765, "step": 1071 }, { "epoch": 0.6308433557688771, "grad_norm": 0.4456340022716743, "learning_rate": 8.656768508095853e-06, "loss": 0.5604, "step": 1072 }, { "epoch": 0.6314318290485122, "grad_norm": 0.3537824222279271, "learning_rate": 8.653265664793466e-06, "loss": 0.4502, "step": 1073 }, { "epoch": 0.6320203023281474, "grad_norm": 0.42640412931973337, "learning_rate": 8.649758970805049e-06, "loss": 0.5122, "step": 1074 }, { "epoch": 0.6326087756077826, "grad_norm": 0.4967311134384535, "learning_rate": 8.646248429826795e-06, "loss": 0.5899, "step": 1075 }, { "epoch": 0.6331972488874177, "grad_norm": 0.40664160188680026, "learning_rate": 8.642734045558952e-06, "loss": 0.5462, "step": 1076 }, { "epoch": 0.6337857221670529, "grad_norm": 0.4032127448552604, "learning_rate": 8.639215821705821e-06, "loss": 0.4819, "step": 1077 }, { "epoch": 0.634374195446688, "grad_norm": 0.4120401547329603, "learning_rate": 8.635693761975747e-06, "loss": 0.5545, "step": 1078 }, { "epoch": 0.6349626687263231, "grad_norm": 0.3996169705440541, "learning_rate": 8.632167870081122e-06, "loss": 0.457, "step": 1079 }, { "epoch": 0.6355511420059583, "grad_norm": 0.4614240455292737, "learning_rate": 8.628638149738372e-06, "loss": 0.5784, "step": 1080 }, { "epoch": 0.6361396152855935, "grad_norm": 0.436286367835478, "learning_rate": 8.625104604667965e-06, "loss": 0.5602, "step": 1081 }, { "epoch": 0.6367280885652286, "grad_norm": 0.3657961142613716, "learning_rate": 8.621567238594392e-06, "loss": 0.497, "step": 1082 }, { "epoch": 0.6373165618448637, "grad_norm": 0.4592604308362406, "learning_rate": 8.618026055246183e-06, "loss": 0.5443, "step": 1083 }, { "epoch": 0.6379050351244989, "grad_norm": 0.4381877507199664, "learning_rate": 8.614481058355877e-06, "loss": 0.5338, "step": 1084 }, { "epoch": 0.638493508404134, "grad_norm": 0.40731019510215616, "learning_rate": 8.610932251660046e-06, "loss": 0.4864, "step": 1085 }, { "epoch": 0.6390819816837692, "grad_norm": 0.42123401656084764, "learning_rate": 8.607379638899272e-06, "loss": 0.5375, "step": 1086 }, { "epoch": 0.6396704549634044, "grad_norm": 0.3926132622088456, "learning_rate": 8.603823223818148e-06, "loss": 0.4995, "step": 1087 }, { "epoch": 0.6402589282430394, "grad_norm": 0.39775927996132, "learning_rate": 8.600263010165275e-06, "loss": 0.4985, "step": 1088 }, { "epoch": 0.6408474015226746, "grad_norm": 0.3593163039687307, "learning_rate": 8.596699001693257e-06, "loss": 0.4487, "step": 1089 }, { "epoch": 0.6414358748023098, "grad_norm": 0.44790716398374475, "learning_rate": 8.593131202158703e-06, "loss": 0.4733, "step": 1090 }, { "epoch": 0.6420243480819449, "grad_norm": 0.42187233328822904, "learning_rate": 8.58955961532221e-06, "loss": 0.5218, "step": 1091 }, { "epoch": 0.64261282136158, "grad_norm": 0.44623888118983823, "learning_rate": 8.585984244948376e-06, "loss": 0.4947, "step": 1092 }, { "epoch": 0.6432012946412152, "grad_norm": 0.42416939513935564, "learning_rate": 8.58240509480578e-06, "loss": 0.5825, "step": 1093 }, { "epoch": 0.6437897679208503, "grad_norm": 0.4382243536363249, "learning_rate": 8.578822168666987e-06, "loss": 0.5643, "step": 1094 }, { "epoch": 0.6443782412004855, "grad_norm": 0.43323744574863365, "learning_rate": 8.575235470308543e-06, "loss": 0.5747, "step": 1095 }, { "epoch": 0.6449667144801207, "grad_norm": 0.4336155648859892, "learning_rate": 8.571645003510971e-06, "loss": 0.4922, "step": 1096 }, { "epoch": 0.6455551877597557, "grad_norm": 0.4302457853380892, "learning_rate": 8.568050772058763e-06, "loss": 0.5479, "step": 1097 }, { "epoch": 0.6461436610393909, "grad_norm": 0.4594264619881639, "learning_rate": 8.564452779740379e-06, "loss": 0.5295, "step": 1098 }, { "epoch": 0.6467321343190261, "grad_norm": 0.5046839169211423, "learning_rate": 8.560851030348248e-06, "loss": 0.5618, "step": 1099 }, { "epoch": 0.6473206075986612, "grad_norm": 0.3970145319041329, "learning_rate": 8.557245527678753e-06, "loss": 0.5247, "step": 1100 }, { "epoch": 0.6479090808782964, "grad_norm": 0.4179720200186194, "learning_rate": 8.553636275532236e-06, "loss": 0.4505, "step": 1101 }, { "epoch": 0.6484975541579315, "grad_norm": 0.38779544805800975, "learning_rate": 8.550023277712995e-06, "loss": 0.4932, "step": 1102 }, { "epoch": 0.6490860274375667, "grad_norm": 0.33472904759500205, "learning_rate": 8.546406538029268e-06, "loss": 0.4682, "step": 1103 }, { "epoch": 0.6496745007172018, "grad_norm": 0.37033614882982147, "learning_rate": 8.542786060293242e-06, "loss": 0.4881, "step": 1104 }, { "epoch": 0.650262973996837, "grad_norm": 0.40238748168046934, "learning_rate": 8.539161848321047e-06, "loss": 0.5285, "step": 1105 }, { "epoch": 0.6508514472764722, "grad_norm": 0.3910202078831875, "learning_rate": 8.535533905932739e-06, "loss": 0.5406, "step": 1106 }, { "epoch": 0.6514399205561072, "grad_norm": 0.47130070168304516, "learning_rate": 8.531902236952316e-06, "loss": 0.5401, "step": 1107 }, { "epoch": 0.6520283938357424, "grad_norm": 0.40857945620605707, "learning_rate": 8.5282668452077e-06, "loss": 0.5118, "step": 1108 }, { "epoch": 0.6526168671153776, "grad_norm": 0.4274998153818867, "learning_rate": 8.524627734530738e-06, "loss": 0.528, "step": 1109 }, { "epoch": 0.6532053403950127, "grad_norm": 0.45946178237043844, "learning_rate": 8.520984908757193e-06, "loss": 0.53, "step": 1110 }, { "epoch": 0.6537938136746478, "grad_norm": 0.4282376923400965, "learning_rate": 8.51733837172675e-06, "loss": 0.5531, "step": 1111 }, { "epoch": 0.654382286954283, "grad_norm": 0.4301901349165709, "learning_rate": 8.513688127283002e-06, "loss": 0.4639, "step": 1112 }, { "epoch": 0.6549707602339181, "grad_norm": 0.4123182129951386, "learning_rate": 8.51003417927345e-06, "loss": 0.5268, "step": 1113 }, { "epoch": 0.6555592335135533, "grad_norm": 0.5230445040283299, "learning_rate": 8.5063765315495e-06, "loss": 0.5263, "step": 1114 }, { "epoch": 0.6561477067931885, "grad_norm": 0.3389713625282946, "learning_rate": 8.502715187966455e-06, "loss": 0.4895, "step": 1115 }, { "epoch": 0.6567361800728235, "grad_norm": 0.4670785445842252, "learning_rate": 8.499050152383519e-06, "loss": 0.5735, "step": 1116 }, { "epoch": 0.6573246533524587, "grad_norm": 0.4331175560760018, "learning_rate": 8.49538142866378e-06, "loss": 0.567, "step": 1117 }, { "epoch": 0.6579131266320939, "grad_norm": 0.43550024489279876, "learning_rate": 8.491709020674221e-06, "loss": 0.5356, "step": 1118 }, { "epoch": 0.658501599911729, "grad_norm": 0.37108638426063684, "learning_rate": 8.488032932285704e-06, "loss": 0.4576, "step": 1119 }, { "epoch": 0.6590900731913641, "grad_norm": 0.3820488165686936, "learning_rate": 8.484353167372968e-06, "loss": 0.4709, "step": 1120 }, { "epoch": 0.6596785464709993, "grad_norm": 0.34838851237522683, "learning_rate": 8.480669729814635e-06, "loss": 0.4466, "step": 1121 }, { "epoch": 0.6602670197506344, "grad_norm": 0.37406475539742146, "learning_rate": 8.476982623493188e-06, "loss": 0.458, "step": 1122 }, { "epoch": 0.6608554930302696, "grad_norm": 0.43107520775908964, "learning_rate": 8.473291852294986e-06, "loss": 0.4589, "step": 1123 }, { "epoch": 0.6614439663099048, "grad_norm": 0.3822747659777743, "learning_rate": 8.469597420110249e-06, "loss": 0.5106, "step": 1124 }, { "epoch": 0.6620324395895398, "grad_norm": 0.4473985786127246, "learning_rate": 8.465899330833051e-06, "loss": 0.5895, "step": 1125 }, { "epoch": 0.662620912869175, "grad_norm": 0.44213228296814927, "learning_rate": 8.462197588361323e-06, "loss": 0.5038, "step": 1126 }, { "epoch": 0.6632093861488102, "grad_norm": 0.740874628074262, "learning_rate": 8.458492196596852e-06, "loss": 0.5822, "step": 1127 }, { "epoch": 0.6637978594284454, "grad_norm": 0.34450647078922936, "learning_rate": 8.454783159445261e-06, "loss": 0.476, "step": 1128 }, { "epoch": 0.6643863327080805, "grad_norm": 0.42960443500153267, "learning_rate": 8.451070480816027e-06, "loss": 0.5097, "step": 1129 }, { "epoch": 0.6649748059877156, "grad_norm": 0.40047071733645834, "learning_rate": 8.447354164622457e-06, "loss": 0.5038, "step": 1130 }, { "epoch": 0.6655632792673508, "grad_norm": 0.38800627468612847, "learning_rate": 8.443634214781693e-06, "loss": 0.569, "step": 1131 }, { "epoch": 0.6661517525469859, "grad_norm": 0.3708274414845186, "learning_rate": 8.439910635214713e-06, "loss": 0.4712, "step": 1132 }, { "epoch": 0.6667402258266211, "grad_norm": 0.40862220255516213, "learning_rate": 8.436183429846314e-06, "loss": 0.5156, "step": 1133 }, { "epoch": 0.6673286991062563, "grad_norm": 0.4809160991613638, "learning_rate": 8.432452602605119e-06, "loss": 0.5852, "step": 1134 }, { "epoch": 0.6679171723858913, "grad_norm": 0.9056467955862555, "learning_rate": 8.428718157423564e-06, "loss": 0.5363, "step": 1135 }, { "epoch": 0.6685056456655265, "grad_norm": 0.40639911160871905, "learning_rate": 8.424980098237904e-06, "loss": 0.4665, "step": 1136 }, { "epoch": 0.6690941189451617, "grad_norm": 0.3905316413808341, "learning_rate": 8.421238428988199e-06, "loss": 0.4583, "step": 1137 }, { "epoch": 0.6696825922247968, "grad_norm": 0.4518507968765558, "learning_rate": 8.417493153618318e-06, "loss": 0.6561, "step": 1138 }, { "epoch": 0.6702710655044319, "grad_norm": 0.41251665169709645, "learning_rate": 8.413744276075928e-06, "loss": 0.5304, "step": 1139 }, { "epoch": 0.6708595387840671, "grad_norm": 0.38892980718411163, "learning_rate": 8.409991800312493e-06, "loss": 0.4995, "step": 1140 }, { "epoch": 0.6714480120637022, "grad_norm": 0.3780661835056546, "learning_rate": 8.40623573028327e-06, "loss": 0.5172, "step": 1141 }, { "epoch": 0.6720364853433374, "grad_norm": 0.3657621585782858, "learning_rate": 8.402476069947309e-06, "loss": 0.48, "step": 1142 }, { "epoch": 0.6726249586229726, "grad_norm": 0.42485056541160876, "learning_rate": 8.398712823267435e-06, "loss": 0.543, "step": 1143 }, { "epoch": 0.6732134319026076, "grad_norm": 0.41402288174338847, "learning_rate": 8.39494599421026e-06, "loss": 0.4584, "step": 1144 }, { "epoch": 0.6738019051822428, "grad_norm": 0.43072674064952005, "learning_rate": 8.39117558674617e-06, "loss": 0.5282, "step": 1145 }, { "epoch": 0.674390378461878, "grad_norm": 0.38482381037976104, "learning_rate": 8.387401604849326e-06, "loss": 0.5112, "step": 1146 }, { "epoch": 0.6749788517415131, "grad_norm": 0.36981613696118515, "learning_rate": 8.38362405249765e-06, "loss": 0.4611, "step": 1147 }, { "epoch": 0.6755673250211482, "grad_norm": 1.5627883637562936, "learning_rate": 8.379842933672833e-06, "loss": 0.5596, "step": 1148 }, { "epoch": 0.6761557983007834, "grad_norm": 0.39563074484823046, "learning_rate": 8.376058252360322e-06, "loss": 0.4882, "step": 1149 }, { "epoch": 0.6767442715804185, "grad_norm": 0.4145331806848749, "learning_rate": 8.372270012549322e-06, "loss": 0.5009, "step": 1150 }, { "epoch": 0.6773327448600537, "grad_norm": 0.4150590977916661, "learning_rate": 8.368478218232787e-06, "loss": 0.5333, "step": 1151 }, { "epoch": 0.6779212181396889, "grad_norm": 0.3662367917662946, "learning_rate": 8.364682873407416e-06, "loss": 0.4667, "step": 1152 }, { "epoch": 0.678509691419324, "grad_norm": 0.42215648972580844, "learning_rate": 8.360883982073653e-06, "loss": 0.5342, "step": 1153 }, { "epoch": 0.6790981646989591, "grad_norm": 0.3684664272157653, "learning_rate": 8.35708154823568e-06, "loss": 0.4612, "step": 1154 }, { "epoch": 0.6796866379785943, "grad_norm": 0.441689008788822, "learning_rate": 8.353275575901411e-06, "loss": 0.5709, "step": 1155 }, { "epoch": 0.6802751112582295, "grad_norm": 0.37347963054060135, "learning_rate": 8.349466069082492e-06, "loss": 0.5082, "step": 1156 }, { "epoch": 0.6808635845378646, "grad_norm": 0.37766758703658443, "learning_rate": 8.345653031794292e-06, "loss": 0.554, "step": 1157 }, { "epoch": 0.6814520578174997, "grad_norm": 0.47013118085821887, "learning_rate": 8.341836468055904e-06, "loss": 0.5584, "step": 1158 }, { "epoch": 0.6820405310971349, "grad_norm": 0.4340437295992818, "learning_rate": 8.338016381890135e-06, "loss": 0.5594, "step": 1159 }, { "epoch": 0.68262900437677, "grad_norm": 0.4327925483093373, "learning_rate": 8.334192777323508e-06, "loss": 0.5117, "step": 1160 }, { "epoch": 0.6832174776564052, "grad_norm": 0.3462986205350464, "learning_rate": 8.330365658386252e-06, "loss": 0.4286, "step": 1161 }, { "epoch": 0.6838059509360404, "grad_norm": 0.45955739172949855, "learning_rate": 8.326535029112302e-06, "loss": 0.5806, "step": 1162 }, { "epoch": 0.6843944242156754, "grad_norm": 0.3342517420277197, "learning_rate": 8.32270089353929e-06, "loss": 0.4489, "step": 1163 }, { "epoch": 0.6849828974953106, "grad_norm": 0.43757194400736454, "learning_rate": 8.31886325570855e-06, "loss": 0.5592, "step": 1164 }, { "epoch": 0.6855713707749458, "grad_norm": 0.4314575382298644, "learning_rate": 8.3150221196651e-06, "loss": 0.4713, "step": 1165 }, { "epoch": 0.6861598440545809, "grad_norm": 0.4238623230577011, "learning_rate": 8.311177489457653e-06, "loss": 0.4664, "step": 1166 }, { "epoch": 0.686748317334216, "grad_norm": 0.42060844545083376, "learning_rate": 8.307329369138595e-06, "loss": 0.5126, "step": 1167 }, { "epoch": 0.6873367906138512, "grad_norm": 0.41155154582101183, "learning_rate": 8.303477762764e-06, "loss": 0.4692, "step": 1168 }, { "epoch": 0.6879252638934863, "grad_norm": 0.5395646462150956, "learning_rate": 8.299622674393615e-06, "loss": 0.5995, "step": 1169 }, { "epoch": 0.6885137371731215, "grad_norm": 0.35242916954845205, "learning_rate": 8.29576410809085e-06, "loss": 0.4784, "step": 1170 }, { "epoch": 0.6891022104527567, "grad_norm": 0.39592422365261937, "learning_rate": 8.291902067922791e-06, "loss": 0.5411, "step": 1171 }, { "epoch": 0.6896906837323917, "grad_norm": 0.3675083832579043, "learning_rate": 8.288036557960175e-06, "loss": 0.4493, "step": 1172 }, { "epoch": 0.6902791570120269, "grad_norm": 0.4110506208645906, "learning_rate": 8.284167582277406e-06, "loss": 0.4808, "step": 1173 }, { "epoch": 0.6908676302916621, "grad_norm": 0.3860974535688809, "learning_rate": 8.280295144952537e-06, "loss": 0.4561, "step": 1174 }, { "epoch": 0.6914561035712972, "grad_norm": 0.4024886227196359, "learning_rate": 8.27641925006727e-06, "loss": 0.4928, "step": 1175 }, { "epoch": 0.6920445768509323, "grad_norm": 0.38194685909667614, "learning_rate": 8.272539901706946e-06, "loss": 0.4562, "step": 1176 }, { "epoch": 0.6926330501305675, "grad_norm": 0.35174060173900373, "learning_rate": 8.268657103960558e-06, "loss": 0.4415, "step": 1177 }, { "epoch": 0.6932215234102026, "grad_norm": 0.39785983097537037, "learning_rate": 8.264770860920722e-06, "loss": 0.4531, "step": 1178 }, { "epoch": 0.6938099966898378, "grad_norm": 0.48139373590333767, "learning_rate": 8.260881176683696e-06, "loss": 0.6043, "step": 1179 }, { "epoch": 0.694398469969473, "grad_norm": 0.4739257473250784, "learning_rate": 8.256988055349357e-06, "loss": 0.5263, "step": 1180 }, { "epoch": 0.6949869432491081, "grad_norm": 0.4218791886829106, "learning_rate": 8.25309150102121e-06, "loss": 0.4692, "step": 1181 }, { "epoch": 0.6955754165287432, "grad_norm": 0.3846752167251243, "learning_rate": 8.249191517806378e-06, "loss": 0.4645, "step": 1182 }, { "epoch": 0.6961638898083784, "grad_norm": 0.435533406722097, "learning_rate": 8.245288109815595e-06, "loss": 0.5282, "step": 1183 }, { "epoch": 0.6967523630880136, "grad_norm": 0.3741414472629643, "learning_rate": 8.241381281163207e-06, "loss": 0.4415, "step": 1184 }, { "epoch": 0.6973408363676487, "grad_norm": 0.41326375581947367, "learning_rate": 8.237471035967168e-06, "loss": 0.4793, "step": 1185 }, { "epoch": 0.6979293096472838, "grad_norm": 0.3715875343639412, "learning_rate": 8.233557378349029e-06, "loss": 0.4687, "step": 1186 }, { "epoch": 0.698517782926919, "grad_norm": 0.37378327359150737, "learning_rate": 8.229640312433938e-06, "loss": 0.4252, "step": 1187 }, { "epoch": 0.6991062562065541, "grad_norm": 0.39400520819125723, "learning_rate": 8.225719842350637e-06, "loss": 0.5526, "step": 1188 }, { "epoch": 0.6996947294861893, "grad_norm": 0.3864967889232373, "learning_rate": 8.221795972231459e-06, "loss": 0.4461, "step": 1189 }, { "epoch": 0.7002832027658245, "grad_norm": 0.39143590174645754, "learning_rate": 8.217868706212313e-06, "loss": 0.5165, "step": 1190 }, { "epoch": 0.7008716760454595, "grad_norm": 0.45704910407730565, "learning_rate": 8.213938048432697e-06, "loss": 0.562, "step": 1191 }, { "epoch": 0.7014601493250947, "grad_norm": 0.41756106721747754, "learning_rate": 8.210004003035677e-06, "loss": 0.4629, "step": 1192 }, { "epoch": 0.7020486226047299, "grad_norm": 0.39982681143758747, "learning_rate": 8.206066574167893e-06, "loss": 0.4613, "step": 1193 }, { "epoch": 0.702637095884365, "grad_norm": 0.42516161445896306, "learning_rate": 8.202125765979549e-06, "loss": 0.5006, "step": 1194 }, { "epoch": 0.7032255691640001, "grad_norm": 0.37835185609345307, "learning_rate": 8.198181582624412e-06, "loss": 0.4238, "step": 1195 }, { "epoch": 0.7038140424436353, "grad_norm": 0.4660544320218266, "learning_rate": 8.194234028259806e-06, "loss": 0.5447, "step": 1196 }, { "epoch": 0.7044025157232704, "grad_norm": 0.4058246081966938, "learning_rate": 8.190283107046613e-06, "loss": 0.4871, "step": 1197 }, { "epoch": 0.7049909890029056, "grad_norm": 0.40335115785908426, "learning_rate": 8.186328823149257e-06, "loss": 0.5723, "step": 1198 }, { "epoch": 0.7055794622825408, "grad_norm": 0.4127376089987144, "learning_rate": 8.182371180735708e-06, "loss": 0.4249, "step": 1199 }, { "epoch": 0.7061679355621758, "grad_norm": 0.4546532252912553, "learning_rate": 8.178410183977479e-06, "loss": 0.4682, "step": 1200 }, { "epoch": 0.706756408841811, "grad_norm": 0.4027087657583689, "learning_rate": 8.174445837049614e-06, "loss": 0.5186, "step": 1201 }, { "epoch": 0.7073448821214462, "grad_norm": 0.3491601153281139, "learning_rate": 8.170478144130695e-06, "loss": 0.4726, "step": 1202 }, { "epoch": 0.7079333554010813, "grad_norm": 0.44065202648685564, "learning_rate": 8.166507109402824e-06, "loss": 0.5076, "step": 1203 }, { "epoch": 0.7085218286807164, "grad_norm": 0.46325133005642577, "learning_rate": 8.162532737051629e-06, "loss": 0.4677, "step": 1204 }, { "epoch": 0.7091103019603516, "grad_norm": 0.3609007373564115, "learning_rate": 8.158555031266255e-06, "loss": 0.4801, "step": 1205 }, { "epoch": 0.7096987752399868, "grad_norm": 0.3928304692336274, "learning_rate": 8.15457399623936e-06, "loss": 0.5563, "step": 1206 }, { "epoch": 0.7102872485196219, "grad_norm": 0.3609436937998834, "learning_rate": 8.150589636167115e-06, "loss": 0.5001, "step": 1207 }, { "epoch": 0.7108757217992571, "grad_norm": 0.4003787271017567, "learning_rate": 8.146601955249187e-06, "loss": 0.5242, "step": 1208 }, { "epoch": 0.7114641950788922, "grad_norm": 0.44233235435740254, "learning_rate": 8.142610957688755e-06, "loss": 0.5325, "step": 1209 }, { "epoch": 0.7120526683585273, "grad_norm": 0.3864353933900008, "learning_rate": 8.138616647692484e-06, "loss": 0.4555, "step": 1210 }, { "epoch": 0.7126411416381625, "grad_norm": 0.39248534340012825, "learning_rate": 8.134619029470535e-06, "loss": 0.4729, "step": 1211 }, { "epoch": 0.7132296149177977, "grad_norm": 0.39254377139530644, "learning_rate": 8.130618107236552e-06, "loss": 0.5034, "step": 1212 }, { "epoch": 0.7138180881974328, "grad_norm": 0.4437346542462819, "learning_rate": 8.12661388520767e-06, "loss": 0.5347, "step": 1213 }, { "epoch": 0.7144065614770679, "grad_norm": 0.3972681118343417, "learning_rate": 8.122606367604497e-06, "loss": 0.5053, "step": 1214 }, { "epoch": 0.7149950347567031, "grad_norm": 0.3785102384603412, "learning_rate": 8.11859555865111e-06, "loss": 0.4573, "step": 1215 }, { "epoch": 0.7155835080363382, "grad_norm": 0.3865909191627533, "learning_rate": 8.114581462575063e-06, "loss": 0.4871, "step": 1216 }, { "epoch": 0.7161719813159734, "grad_norm": 0.4181464618972225, "learning_rate": 8.110564083607371e-06, "loss": 0.5168, "step": 1217 }, { "epoch": 0.7167604545956086, "grad_norm": 0.5240135274116373, "learning_rate": 8.106543425982509e-06, "loss": 0.496, "step": 1218 }, { "epoch": 0.7173489278752436, "grad_norm": 0.3791572294647692, "learning_rate": 8.102519493938408e-06, "loss": 0.5081, "step": 1219 }, { "epoch": 0.7179374011548788, "grad_norm": 0.4655626670941109, "learning_rate": 8.098492291716453e-06, "loss": 0.5022, "step": 1220 }, { "epoch": 0.718525874434514, "grad_norm": 0.36055494902365715, "learning_rate": 8.094461823561473e-06, "loss": 0.4679, "step": 1221 }, { "epoch": 0.7191143477141491, "grad_norm": 0.4348446587264877, "learning_rate": 8.09042809372174e-06, "loss": 0.5163, "step": 1222 }, { "epoch": 0.7197028209937842, "grad_norm": 0.4195074341063246, "learning_rate": 8.086391106448965e-06, "loss": 0.5332, "step": 1223 }, { "epoch": 0.7202912942734194, "grad_norm": 0.45357250382881187, "learning_rate": 8.082350865998292e-06, "loss": 0.5385, "step": 1224 }, { "epoch": 0.7208797675530545, "grad_norm": 0.3460388279944798, "learning_rate": 8.078307376628292e-06, "loss": 0.429, "step": 1225 }, { "epoch": 0.7214682408326897, "grad_norm": 0.38776502505924315, "learning_rate": 8.074260642600963e-06, "loss": 0.5995, "step": 1226 }, { "epoch": 0.7220567141123249, "grad_norm": 0.4868570551379059, "learning_rate": 8.070210668181724e-06, "loss": 0.5465, "step": 1227 }, { "epoch": 0.7226451873919599, "grad_norm": 0.42451009623893365, "learning_rate": 8.066157457639406e-06, "loss": 0.4376, "step": 1228 }, { "epoch": 0.7232336606715951, "grad_norm": 0.38662053898158333, "learning_rate": 8.06210101524625e-06, "loss": 0.4579, "step": 1229 }, { "epoch": 0.7238221339512303, "grad_norm": 0.36954262483493455, "learning_rate": 8.05804134527791e-06, "loss": 0.4893, "step": 1230 }, { "epoch": 0.7244106072308655, "grad_norm": 0.37846071926469077, "learning_rate": 8.05397845201344e-06, "loss": 0.4602, "step": 1231 }, { "epoch": 0.7249990805105005, "grad_norm": 0.414848534529878, "learning_rate": 8.049912339735284e-06, "loss": 0.4756, "step": 1232 }, { "epoch": 0.7255875537901357, "grad_norm": 0.5581086701590566, "learning_rate": 8.045843012729288e-06, "loss": 0.4833, "step": 1233 }, { "epoch": 0.7261760270697709, "grad_norm": 0.396107571389696, "learning_rate": 8.041770475284686e-06, "loss": 0.5944, "step": 1234 }, { "epoch": 0.726764500349406, "grad_norm": 0.39974960821175554, "learning_rate": 8.037694731694085e-06, "loss": 0.4889, "step": 1235 }, { "epoch": 0.7273529736290412, "grad_norm": 0.39533564322272574, "learning_rate": 8.033615786253486e-06, "loss": 0.5174, "step": 1236 }, { "epoch": 0.7279414469086763, "grad_norm": 0.4046671887520688, "learning_rate": 8.029533643262257e-06, "loss": 0.5463, "step": 1237 }, { "epoch": 0.7285299201883114, "grad_norm": 0.4230374288338172, "learning_rate": 8.025448307023135e-06, "loss": 0.5453, "step": 1238 }, { "epoch": 0.7291183934679466, "grad_norm": 0.39878079171616376, "learning_rate": 8.021359781842229e-06, "loss": 0.494, "step": 1239 }, { "epoch": 0.7297068667475818, "grad_norm": 0.45478228062873677, "learning_rate": 8.017268072029003e-06, "loss": 0.4904, "step": 1240 }, { "epoch": 0.7302953400272169, "grad_norm": 0.4501096431230873, "learning_rate": 8.013173181896283e-06, "loss": 0.5129, "step": 1241 }, { "epoch": 0.730883813306852, "grad_norm": 0.40828236793773903, "learning_rate": 8.009075115760243e-06, "loss": 0.4491, "step": 1242 }, { "epoch": 0.7314722865864872, "grad_norm": 0.37574374088991563, "learning_rate": 8.004973877940405e-06, "loss": 0.4756, "step": 1243 }, { "epoch": 0.7320607598661223, "grad_norm": 0.3693832460467644, "learning_rate": 8.000869472759639e-06, "loss": 0.4539, "step": 1244 }, { "epoch": 0.7326492331457575, "grad_norm": 0.42115827688977014, "learning_rate": 7.996761904544146e-06, "loss": 0.5497, "step": 1245 }, { "epoch": 0.7332377064253927, "grad_norm": 0.43358095464737223, "learning_rate": 7.992651177623466e-06, "loss": 0.4787, "step": 1246 }, { "epoch": 0.7338261797050277, "grad_norm": 0.34859132273130533, "learning_rate": 7.988537296330468e-06, "loss": 0.3835, "step": 1247 }, { "epoch": 0.7344146529846629, "grad_norm": 0.4796389988665053, "learning_rate": 7.984420265001346e-06, "loss": 0.5746, "step": 1248 }, { "epoch": 0.7350031262642981, "grad_norm": 0.3854581742422268, "learning_rate": 7.980300087975612e-06, "loss": 0.453, "step": 1249 }, { "epoch": 0.7355915995439332, "grad_norm": 0.40870901177352187, "learning_rate": 7.976176769596095e-06, "loss": 0.4632, "step": 1250 }, { "epoch": 0.7361800728235683, "grad_norm": 0.46883510901296455, "learning_rate": 7.972050314208934e-06, "loss": 0.5286, "step": 1251 }, { "epoch": 0.7367685461032035, "grad_norm": 0.3613200517102609, "learning_rate": 7.967920726163579e-06, "loss": 0.4622, "step": 1252 }, { "epoch": 0.7373570193828386, "grad_norm": 0.4128482083925332, "learning_rate": 7.963788009812775e-06, "loss": 0.4586, "step": 1253 }, { "epoch": 0.7379454926624738, "grad_norm": 0.3589402159186721, "learning_rate": 7.95965216951257e-06, "loss": 0.4238, "step": 1254 }, { "epoch": 0.738533965942109, "grad_norm": 0.41763439894981635, "learning_rate": 7.955513209622304e-06, "loss": 0.53, "step": 1255 }, { "epoch": 0.7391224392217441, "grad_norm": 0.32860031083273605, "learning_rate": 7.951371134504599e-06, "loss": 0.4488, "step": 1256 }, { "epoch": 0.7397109125013792, "grad_norm": 0.40510588937654546, "learning_rate": 7.94722594852537e-06, "loss": 0.5428, "step": 1257 }, { "epoch": 0.7402993857810144, "grad_norm": 0.40407621205057415, "learning_rate": 7.943077656053803e-06, "loss": 0.4975, "step": 1258 }, { "epoch": 0.7408878590606496, "grad_norm": 0.4059577234513558, "learning_rate": 7.938926261462366e-06, "loss": 0.5622, "step": 1259 }, { "epoch": 0.7414763323402846, "grad_norm": 0.44105918564460356, "learning_rate": 7.93477176912679e-06, "loss": 0.5175, "step": 1260 }, { "epoch": 0.7420648056199198, "grad_norm": 0.43195958652259986, "learning_rate": 7.930614183426074e-06, "loss": 0.4795, "step": 1261 }, { "epoch": 0.742653278899555, "grad_norm": 0.46030476617835275, "learning_rate": 7.926453508742478e-06, "loss": 0.5057, "step": 1262 }, { "epoch": 0.7432417521791901, "grad_norm": 0.39262370330199636, "learning_rate": 7.922289749461516e-06, "loss": 0.5272, "step": 1263 }, { "epoch": 0.7438302254588253, "grad_norm": 0.41054822345312814, "learning_rate": 7.918122909971952e-06, "loss": 0.4834, "step": 1264 }, { "epoch": 0.7444186987384604, "grad_norm": 0.42301940111066494, "learning_rate": 7.913952994665805e-06, "loss": 0.5101, "step": 1265 }, { "epoch": 0.7450071720180955, "grad_norm": 0.39661732329881116, "learning_rate": 7.909780007938327e-06, "loss": 0.4432, "step": 1266 }, { "epoch": 0.7455956452977307, "grad_norm": 0.40638129694873915, "learning_rate": 7.90560395418801e-06, "loss": 0.5151, "step": 1267 }, { "epoch": 0.7461841185773659, "grad_norm": 0.3886161512675768, "learning_rate": 7.90142483781658e-06, "loss": 0.4862, "step": 1268 }, { "epoch": 0.746772591857001, "grad_norm": 0.3718298234681098, "learning_rate": 7.89724266322899e-06, "loss": 0.5014, "step": 1269 }, { "epoch": 0.7473610651366361, "grad_norm": 0.384184469840395, "learning_rate": 7.893057434833416e-06, "loss": 0.4405, "step": 1270 }, { "epoch": 0.7479495384162713, "grad_norm": 0.39966314170375067, "learning_rate": 7.888869157041257e-06, "loss": 0.5122, "step": 1271 }, { "epoch": 0.7485380116959064, "grad_norm": 0.4119759779870115, "learning_rate": 7.884677834267119e-06, "loss": 0.4986, "step": 1272 }, { "epoch": 0.7491264849755416, "grad_norm": 0.33847544571740723, "learning_rate": 7.880483470928823e-06, "loss": 0.3635, "step": 1273 }, { "epoch": 0.7497149582551768, "grad_norm": 0.3778117827617605, "learning_rate": 7.876286071447395e-06, "loss": 0.5197, "step": 1274 }, { "epoch": 0.7503034315348118, "grad_norm": 0.36554403603494867, "learning_rate": 7.872085640247057e-06, "loss": 0.4992, "step": 1275 }, { "epoch": 0.750891904814447, "grad_norm": 0.42078119934266944, "learning_rate": 7.86788218175523e-06, "loss": 0.5657, "step": 1276 }, { "epoch": 0.7514803780940822, "grad_norm": 0.3980727208482993, "learning_rate": 7.863675700402527e-06, "loss": 0.467, "step": 1277 }, { "epoch": 0.7520688513737173, "grad_norm": 0.39338243304783166, "learning_rate": 7.85946620062274e-06, "loss": 0.4529, "step": 1278 }, { "epoch": 0.7526573246533524, "grad_norm": 0.5113727388941502, "learning_rate": 7.855253686852855e-06, "loss": 0.5854, "step": 1279 }, { "epoch": 0.7532457979329876, "grad_norm": 0.44167520639965024, "learning_rate": 7.85103816353302e-06, "loss": 0.591, "step": 1280 }, { "epoch": 0.7538342712126227, "grad_norm": 0.4173108363901849, "learning_rate": 7.846819635106569e-06, "loss": 0.5056, "step": 1281 }, { "epoch": 0.7544227444922579, "grad_norm": 0.3997084193712445, "learning_rate": 7.842598106019991e-06, "loss": 0.5306, "step": 1282 }, { "epoch": 0.7550112177718931, "grad_norm": 0.4257636517263024, "learning_rate": 7.838373580722952e-06, "loss": 0.5103, "step": 1283 }, { "epoch": 0.7555996910515282, "grad_norm": 0.39669095997439124, "learning_rate": 7.83414606366826e-06, "loss": 0.5444, "step": 1284 }, { "epoch": 0.7561881643311633, "grad_norm": 0.39834586318511045, "learning_rate": 7.829915559311892e-06, "loss": 0.5262, "step": 1285 }, { "epoch": 0.7567766376107985, "grad_norm": 0.39486629220436775, "learning_rate": 7.82568207211296e-06, "loss": 0.5511, "step": 1286 }, { "epoch": 0.7573651108904337, "grad_norm": 0.4878629622619287, "learning_rate": 7.821445606533731e-06, "loss": 0.5768, "step": 1287 }, { "epoch": 0.7579535841700688, "grad_norm": 0.42260217729604105, "learning_rate": 7.817206167039603e-06, "loss": 0.4797, "step": 1288 }, { "epoch": 0.7585420574497039, "grad_norm": 0.44205524645572164, "learning_rate": 7.812963758099118e-06, "loss": 0.5857, "step": 1289 }, { "epoch": 0.7591305307293391, "grad_norm": 0.3277709817360927, "learning_rate": 7.808718384183934e-06, "loss": 0.4387, "step": 1290 }, { "epoch": 0.7597190040089742, "grad_norm": 0.4097241778149125, "learning_rate": 7.80447004976885e-06, "loss": 0.4834, "step": 1291 }, { "epoch": 0.7603074772886094, "grad_norm": 0.38324285442415124, "learning_rate": 7.800218759331775e-06, "loss": 0.4956, "step": 1292 }, { "epoch": 0.7608959505682446, "grad_norm": 0.3987293556696352, "learning_rate": 7.795964517353734e-06, "loss": 0.5395, "step": 1293 }, { "epoch": 0.7614844238478796, "grad_norm": 0.3690132985202921, "learning_rate": 7.791707328318872e-06, "loss": 0.4888, "step": 1294 }, { "epoch": 0.7620728971275148, "grad_norm": 0.3837346029935728, "learning_rate": 7.787447196714428e-06, "loss": 0.5367, "step": 1295 }, { "epoch": 0.76266137040715, "grad_norm": 0.401490811390697, "learning_rate": 7.783184127030752e-06, "loss": 0.5396, "step": 1296 }, { "epoch": 0.7632498436867851, "grad_norm": 0.4834805008848748, "learning_rate": 7.778918123761287e-06, "loss": 0.454, "step": 1297 }, { "epoch": 0.7638383169664202, "grad_norm": 0.4110438187773769, "learning_rate": 7.774649191402569e-06, "loss": 0.5298, "step": 1298 }, { "epoch": 0.7644267902460554, "grad_norm": 0.3659668480246845, "learning_rate": 7.77037733445422e-06, "loss": 0.4517, "step": 1299 }, { "epoch": 0.7650152635256905, "grad_norm": 0.45128861999270325, "learning_rate": 7.766102557418946e-06, "loss": 0.4574, "step": 1300 }, { "epoch": 0.7656037368053257, "grad_norm": 0.37442890468033474, "learning_rate": 7.76182486480253e-06, "loss": 0.4838, "step": 1301 }, { "epoch": 0.7661922100849609, "grad_norm": 0.39907926622796597, "learning_rate": 7.75754426111383e-06, "loss": 0.4842, "step": 1302 }, { "epoch": 0.7667806833645959, "grad_norm": 0.40587652704587124, "learning_rate": 7.753260750864768e-06, "loss": 0.4628, "step": 1303 }, { "epoch": 0.7673691566442311, "grad_norm": 0.5122679498443146, "learning_rate": 7.748974338570337e-06, "loss": 0.4939, "step": 1304 }, { "epoch": 0.7679576299238663, "grad_norm": 0.3944973897241309, "learning_rate": 7.744685028748582e-06, "loss": 0.4956, "step": 1305 }, { "epoch": 0.7685461032035014, "grad_norm": 0.43217597000317215, "learning_rate": 7.740392825920605e-06, "loss": 0.5109, "step": 1306 }, { "epoch": 0.7691345764831365, "grad_norm": 0.5488780605509551, "learning_rate": 7.736097734610557e-06, "loss": 0.5261, "step": 1307 }, { "epoch": 0.7697230497627717, "grad_norm": 0.3654731327218792, "learning_rate": 7.731799759345633e-06, "loss": 0.4823, "step": 1308 }, { "epoch": 0.7703115230424069, "grad_norm": 0.36586494921542156, "learning_rate": 7.72749890465607e-06, "loss": 0.4817, "step": 1309 }, { "epoch": 0.770899996322042, "grad_norm": 0.38027176668083973, "learning_rate": 7.723195175075136e-06, "loss": 0.5068, "step": 1310 }, { "epoch": 0.7714884696016772, "grad_norm": 0.4126750409331027, "learning_rate": 7.718888575139134e-06, "loss": 0.5129, "step": 1311 }, { "epoch": 0.7720769428813123, "grad_norm": 0.3972932054000036, "learning_rate": 7.714579109387388e-06, "loss": 0.4627, "step": 1312 }, { "epoch": 0.7726654161609474, "grad_norm": 0.42359901461966964, "learning_rate": 7.710266782362248e-06, "loss": 0.4869, "step": 1313 }, { "epoch": 0.7732538894405826, "grad_norm": 0.4130807515111283, "learning_rate": 7.705951598609073e-06, "loss": 0.478, "step": 1314 }, { "epoch": 0.7738423627202178, "grad_norm": 0.4425988570238046, "learning_rate": 7.701633562676238e-06, "loss": 0.4922, "step": 1315 }, { "epoch": 0.7744308359998529, "grad_norm": 0.3729137750265103, "learning_rate": 7.697312679115126e-06, "loss": 0.5261, "step": 1316 }, { "epoch": 0.775019309279488, "grad_norm": 0.3823500698390823, "learning_rate": 7.692988952480114e-06, "loss": 0.4803, "step": 1317 }, { "epoch": 0.7756077825591232, "grad_norm": 0.38019693735119636, "learning_rate": 7.688662387328583e-06, "loss": 0.4694, "step": 1318 }, { "epoch": 0.7761962558387583, "grad_norm": 0.4585487959564218, "learning_rate": 7.684332988220901e-06, "loss": 0.5487, "step": 1319 }, { "epoch": 0.7767847291183935, "grad_norm": 0.36685882040760925, "learning_rate": 7.680000759720428e-06, "loss": 0.4493, "step": 1320 }, { "epoch": 0.7773732023980287, "grad_norm": 0.3935848895440232, "learning_rate": 7.675665706393502e-06, "loss": 0.4896, "step": 1321 }, { "epoch": 0.7779616756776637, "grad_norm": 0.4283947498303326, "learning_rate": 7.671327832809442e-06, "loss": 0.4964, "step": 1322 }, { "epoch": 0.7785501489572989, "grad_norm": 0.3706458322772656, "learning_rate": 7.666987143540537e-06, "loss": 0.4628, "step": 1323 }, { "epoch": 0.7791386222369341, "grad_norm": 0.4151227853673153, "learning_rate": 7.662643643162042e-06, "loss": 0.4975, "step": 1324 }, { "epoch": 0.7797270955165692, "grad_norm": 0.488721651258845, "learning_rate": 7.658297336252181e-06, "loss": 0.6119, "step": 1325 }, { "epoch": 0.7803155687962043, "grad_norm": 0.44101136032168586, "learning_rate": 7.65394822739213e-06, "loss": 0.5634, "step": 1326 }, { "epoch": 0.7809040420758395, "grad_norm": 0.39210047214299276, "learning_rate": 7.649596321166024e-06, "loss": 0.4927, "step": 1327 }, { "epoch": 0.7814925153554746, "grad_norm": 0.42877566985375953, "learning_rate": 7.645241622160944e-06, "loss": 0.5388, "step": 1328 }, { "epoch": 0.7820809886351098, "grad_norm": 0.38505803357724333, "learning_rate": 7.64088413496691e-06, "loss": 0.4648, "step": 1329 }, { "epoch": 0.782669461914745, "grad_norm": 0.4137556574114957, "learning_rate": 7.63652386417689e-06, "loss": 0.4962, "step": 1330 }, { "epoch": 0.78325793519438, "grad_norm": 0.4686827631603675, "learning_rate": 7.63216081438678e-06, "loss": 0.5346, "step": 1331 }, { "epoch": 0.7838464084740152, "grad_norm": 0.38740883635635404, "learning_rate": 7.6277949901954044e-06, "loss": 0.4688, "step": 1332 }, { "epoch": 0.7844348817536504, "grad_norm": 0.6240688136983051, "learning_rate": 7.623426396204516e-06, "loss": 0.5129, "step": 1333 }, { "epoch": 0.7850233550332856, "grad_norm": 0.3525585019625932, "learning_rate": 7.619055037018785e-06, "loss": 0.441, "step": 1334 }, { "epoch": 0.7856118283129206, "grad_norm": 0.36921153353045894, "learning_rate": 7.614680917245796e-06, "loss": 0.453, "step": 1335 }, { "epoch": 0.7862003015925558, "grad_norm": 0.3765552323171462, "learning_rate": 7.6103040414960415e-06, "loss": 0.5047, "step": 1336 }, { "epoch": 0.786788774872191, "grad_norm": 0.3747090005445113, "learning_rate": 7.605924414382926e-06, "loss": 0.4635, "step": 1337 }, { "epoch": 0.7873772481518261, "grad_norm": 0.3638472202844138, "learning_rate": 7.6015420405227456e-06, "loss": 0.4713, "step": 1338 }, { "epoch": 0.7879657214314613, "grad_norm": 0.4433078801590044, "learning_rate": 7.5971569245346945e-06, "loss": 0.4921, "step": 1339 }, { "epoch": 0.7885541947110964, "grad_norm": 0.38490735456748654, "learning_rate": 7.5927690710408606e-06, "loss": 0.4426, "step": 1340 }, { "epoch": 0.7891426679907315, "grad_norm": 0.6965668071338852, "learning_rate": 7.588378484666214e-06, "loss": 0.4586, "step": 1341 }, { "epoch": 0.7897311412703667, "grad_norm": 0.3838025109355767, "learning_rate": 7.583985170038603e-06, "loss": 0.442, "step": 1342 }, { "epoch": 0.7903196145500019, "grad_norm": 0.4239876605248207, "learning_rate": 7.579589131788756e-06, "loss": 0.5062, "step": 1343 }, { "epoch": 0.790908087829637, "grad_norm": 0.4238859630424555, "learning_rate": 7.575190374550272e-06, "loss": 0.5339, "step": 1344 }, { "epoch": 0.7914965611092721, "grad_norm": 0.39277432125962897, "learning_rate": 7.570788902959612e-06, "loss": 0.4455, "step": 1345 }, { "epoch": 0.7920850343889073, "grad_norm": 0.40653141510470425, "learning_rate": 7.566384721656103e-06, "loss": 0.5304, "step": 1346 }, { "epoch": 0.7926735076685424, "grad_norm": 0.3563828197179382, "learning_rate": 7.561977835281926e-06, "loss": 0.45, "step": 1347 }, { "epoch": 0.7932619809481776, "grad_norm": 0.4173321631183835, "learning_rate": 7.55756824848211e-06, "loss": 0.4536, "step": 1348 }, { "epoch": 0.7938504542278128, "grad_norm": 0.40814841706832333, "learning_rate": 7.553155965904535e-06, "loss": 0.479, "step": 1349 }, { "epoch": 0.7944389275074478, "grad_norm": 0.4995972726150013, "learning_rate": 7.5487409921999225e-06, "loss": 0.4804, "step": 1350 }, { "epoch": 0.795027400787083, "grad_norm": 0.4428726894069452, "learning_rate": 7.544323332021826e-06, "loss": 0.5171, "step": 1351 }, { "epoch": 0.7956158740667182, "grad_norm": 0.33290575124570976, "learning_rate": 7.539902990026634e-06, "loss": 0.4836, "step": 1352 }, { "epoch": 0.7962043473463533, "grad_norm": 0.4161535448299597, "learning_rate": 7.535479970873563e-06, "loss": 0.4998, "step": 1353 }, { "epoch": 0.7967928206259884, "grad_norm": 0.45177282304309324, "learning_rate": 7.531054279224646e-06, "loss": 0.5611, "step": 1354 }, { "epoch": 0.7973812939056236, "grad_norm": 0.40019821565849256, "learning_rate": 7.526625919744741e-06, "loss": 0.5461, "step": 1355 }, { "epoch": 0.7979697671852587, "grad_norm": 0.3509811623472628, "learning_rate": 7.522194897101511e-06, "loss": 0.4663, "step": 1356 }, { "epoch": 0.7985582404648939, "grad_norm": 0.3814220231313149, "learning_rate": 7.517761215965429e-06, "loss": 0.4228, "step": 1357 }, { "epoch": 0.7991467137445291, "grad_norm": 0.4028329495708391, "learning_rate": 7.513324881009769e-06, "loss": 0.5154, "step": 1358 }, { "epoch": 0.7997351870241641, "grad_norm": 0.4327098304732517, "learning_rate": 7.508885896910605e-06, "loss": 0.4904, "step": 1359 }, { "epoch": 0.8003236603037993, "grad_norm": 0.35323163684572817, "learning_rate": 7.504444268346798e-06, "loss": 0.4604, "step": 1360 }, { "epoch": 0.8009121335834345, "grad_norm": 0.5481956651209084, "learning_rate": 7.500000000000001e-06, "loss": 0.5068, "step": 1361 }, { "epoch": 0.8015006068630697, "grad_norm": 0.37983199182191435, "learning_rate": 7.49555309655465e-06, "loss": 0.4164, "step": 1362 }, { "epoch": 0.8020890801427047, "grad_norm": 0.3570698448754416, "learning_rate": 7.491103562697954e-06, "loss": 0.3986, "step": 1363 }, { "epoch": 0.8026775534223399, "grad_norm": 0.35245620697548763, "learning_rate": 7.486651403119898e-06, "loss": 0.394, "step": 1364 }, { "epoch": 0.8032660267019751, "grad_norm": 0.38009349985319146, "learning_rate": 7.482196622513233e-06, "loss": 0.4966, "step": 1365 }, { "epoch": 0.8038544999816102, "grad_norm": 0.39887240084307174, "learning_rate": 7.477739225573475e-06, "loss": 0.4626, "step": 1366 }, { "epoch": 0.8044429732612454, "grad_norm": 0.36711866600574206, "learning_rate": 7.473279216998896e-06, "loss": 0.4711, "step": 1367 }, { "epoch": 0.8050314465408805, "grad_norm": 0.4539115792855298, "learning_rate": 7.468816601490518e-06, "loss": 0.6054, "step": 1368 }, { "epoch": 0.8056199198205156, "grad_norm": 0.40163241284764767, "learning_rate": 7.464351383752117e-06, "loss": 0.4955, "step": 1369 }, { "epoch": 0.8062083931001508, "grad_norm": 0.4446087319960841, "learning_rate": 7.459883568490207e-06, "loss": 0.5476, "step": 1370 }, { "epoch": 0.806796866379786, "grad_norm": 0.37760527352991397, "learning_rate": 7.4554131604140425e-06, "loss": 0.4988, "step": 1371 }, { "epoch": 0.807385339659421, "grad_norm": 0.414754613994167, "learning_rate": 7.450940164235607e-06, "loss": 0.5122, "step": 1372 }, { "epoch": 0.8079738129390562, "grad_norm": 0.41464473658999235, "learning_rate": 7.4464645846696186e-06, "loss": 0.4959, "step": 1373 }, { "epoch": 0.8085622862186914, "grad_norm": 0.3741702022186685, "learning_rate": 7.441986426433512e-06, "loss": 0.489, "step": 1374 }, { "epoch": 0.8091507594983265, "grad_norm": 0.468934455685958, "learning_rate": 7.437505694247445e-06, "loss": 0.5777, "step": 1375 }, { "epoch": 0.8097392327779617, "grad_norm": 0.37562712583166713, "learning_rate": 7.4330223928342814e-06, "loss": 0.419, "step": 1376 }, { "epoch": 0.8103277060575969, "grad_norm": 0.38838232947227114, "learning_rate": 7.428536526919603e-06, "loss": 0.4671, "step": 1377 }, { "epoch": 0.8109161793372319, "grad_norm": 0.4183205364504691, "learning_rate": 7.424048101231687e-06, "loss": 0.5268, "step": 1378 }, { "epoch": 0.8115046526168671, "grad_norm": 0.40523973366435756, "learning_rate": 7.419557120501508e-06, "loss": 0.5026, "step": 1379 }, { "epoch": 0.8120931258965023, "grad_norm": 0.37700782250596576, "learning_rate": 7.415063589462742e-06, "loss": 0.4437, "step": 1380 }, { "epoch": 0.8126815991761374, "grad_norm": 0.38744857601231314, "learning_rate": 7.4105675128517456e-06, "loss": 0.4545, "step": 1381 }, { "epoch": 0.8132700724557725, "grad_norm": 0.3628292305525873, "learning_rate": 7.406068895407559e-06, "loss": 0.484, "step": 1382 }, { "epoch": 0.8138585457354077, "grad_norm": 0.37036410334166814, "learning_rate": 7.401567741871905e-06, "loss": 0.5415, "step": 1383 }, { "epoch": 0.8144470190150428, "grad_norm": 0.3946383940276982, "learning_rate": 7.397064056989176e-06, "loss": 0.5113, "step": 1384 }, { "epoch": 0.815035492294678, "grad_norm": 0.43081629742482136, "learning_rate": 7.392557845506433e-06, "loss": 0.4749, "step": 1385 }, { "epoch": 0.8156239655743132, "grad_norm": 0.40913278666276764, "learning_rate": 7.388049112173401e-06, "loss": 0.4789, "step": 1386 }, { "epoch": 0.8162124388539483, "grad_norm": 0.3547633878526757, "learning_rate": 7.383537861742464e-06, "loss": 0.4213, "step": 1387 }, { "epoch": 0.8168009121335834, "grad_norm": 0.4377473574861856, "learning_rate": 7.379024098968653e-06, "loss": 0.4758, "step": 1388 }, { "epoch": 0.8173893854132186, "grad_norm": 0.41763685858678595, "learning_rate": 7.374507828609657e-06, "loss": 0.5488, "step": 1389 }, { "epoch": 0.8179778586928538, "grad_norm": 0.4296408421696759, "learning_rate": 7.369989055425802e-06, "loss": 0.4985, "step": 1390 }, { "epoch": 0.8185663319724888, "grad_norm": 0.44601426492869045, "learning_rate": 7.365467784180051e-06, "loss": 0.5216, "step": 1391 }, { "epoch": 0.819154805252124, "grad_norm": 0.37944187236855187, "learning_rate": 7.3609440196380045e-06, "loss": 0.5015, "step": 1392 }, { "epoch": 0.8197432785317592, "grad_norm": 0.386501222127746, "learning_rate": 7.356417766567887e-06, "loss": 0.4934, "step": 1393 }, { "epoch": 0.8203317518113943, "grad_norm": 0.4733689625970107, "learning_rate": 7.351889029740548e-06, "loss": 0.5933, "step": 1394 }, { "epoch": 0.8209202250910295, "grad_norm": 0.39749627003678895, "learning_rate": 7.347357813929455e-06, "loss": 0.5031, "step": 1395 }, { "epoch": 0.8215086983706646, "grad_norm": 0.410577701829068, "learning_rate": 7.342824123910688e-06, "loss": 0.5191, "step": 1396 }, { "epoch": 0.8220971716502997, "grad_norm": 0.38004530237991196, "learning_rate": 7.3382879644629345e-06, "loss": 0.4912, "step": 1397 }, { "epoch": 0.8226856449299349, "grad_norm": 0.40641908191436343, "learning_rate": 7.3337493403674865e-06, "loss": 0.507, "step": 1398 }, { "epoch": 0.8232741182095701, "grad_norm": 0.3815453510969422, "learning_rate": 7.329208256408234e-06, "loss": 0.5002, "step": 1399 }, { "epoch": 0.8238625914892052, "grad_norm": 0.3854192928615949, "learning_rate": 7.324664717371652e-06, "loss": 0.496, "step": 1400 }, { "epoch": 0.8244510647688403, "grad_norm": 0.34970049659029856, "learning_rate": 7.320118728046818e-06, "loss": 0.4377, "step": 1401 }, { "epoch": 0.8250395380484755, "grad_norm": 0.3551874377699427, "learning_rate": 7.315570293225381e-06, "loss": 0.4671, "step": 1402 }, { "epoch": 0.8256280113281106, "grad_norm": 0.5195038929540736, "learning_rate": 7.311019417701567e-06, "loss": 0.5499, "step": 1403 }, { "epoch": 0.8262164846077458, "grad_norm": 0.3820642743698132, "learning_rate": 7.306466106272183e-06, "loss": 0.5238, "step": 1404 }, { "epoch": 0.826804957887381, "grad_norm": 0.36663842458359497, "learning_rate": 7.301910363736596e-06, "loss": 0.4322, "step": 1405 }, { "epoch": 0.827393431167016, "grad_norm": 0.36140673443955257, "learning_rate": 7.297352194896738e-06, "loss": 0.4922, "step": 1406 }, { "epoch": 0.8279819044466512, "grad_norm": 0.4533775331241254, "learning_rate": 7.292791604557102e-06, "loss": 0.5392, "step": 1407 }, { "epoch": 0.8285703777262864, "grad_norm": 0.40152743404925384, "learning_rate": 7.288228597524728e-06, "loss": 0.5, "step": 1408 }, { "epoch": 0.8291588510059215, "grad_norm": 0.425201100032293, "learning_rate": 7.283663178609204e-06, "loss": 0.5746, "step": 1409 }, { "epoch": 0.8297473242855566, "grad_norm": 0.4559237004059688, "learning_rate": 7.279095352622663e-06, "loss": 0.5697, "step": 1410 }, { "epoch": 0.8303357975651918, "grad_norm": 0.36199492465644556, "learning_rate": 7.274525124379773e-06, "loss": 0.4346, "step": 1411 }, { "epoch": 0.830924270844827, "grad_norm": 0.40503516976475157, "learning_rate": 7.269952498697734e-06, "loss": 0.4811, "step": 1412 }, { "epoch": 0.8315127441244621, "grad_norm": 0.41444847707450033, "learning_rate": 7.265377480396277e-06, "loss": 0.4805, "step": 1413 }, { "epoch": 0.8321012174040973, "grad_norm": 0.3666179605051586, "learning_rate": 7.260800074297648e-06, "loss": 0.4169, "step": 1414 }, { "epoch": 0.8326896906837324, "grad_norm": 0.3762847963339845, "learning_rate": 7.256220285226615e-06, "loss": 0.4814, "step": 1415 }, { "epoch": 0.8332781639633675, "grad_norm": 0.3920737928217588, "learning_rate": 7.251638118010456e-06, "loss": 0.4818, "step": 1416 }, { "epoch": 0.8338666372430027, "grad_norm": 0.3698499312288133, "learning_rate": 7.247053577478955e-06, "loss": 0.4627, "step": 1417 }, { "epoch": 0.8344551105226379, "grad_norm": 0.3615137553188879, "learning_rate": 7.242466668464402e-06, "loss": 0.4521, "step": 1418 }, { "epoch": 0.835043583802273, "grad_norm": 0.3710208993624213, "learning_rate": 7.237877395801576e-06, "loss": 0.4991, "step": 1419 }, { "epoch": 0.8356320570819081, "grad_norm": 0.3622859537086576, "learning_rate": 7.233285764327752e-06, "loss": 0.4396, "step": 1420 }, { "epoch": 0.8362205303615433, "grad_norm": 0.3760484187894439, "learning_rate": 7.2286917788826926e-06, "loss": 0.4721, "step": 1421 }, { "epoch": 0.8368090036411784, "grad_norm": 0.3907388863947612, "learning_rate": 7.224095444308637e-06, "loss": 0.453, "step": 1422 }, { "epoch": 0.8373974769208136, "grad_norm": 0.374045491301499, "learning_rate": 7.219496765450305e-06, "loss": 0.4393, "step": 1423 }, { "epoch": 0.8379859502004487, "grad_norm": 0.3500716407088132, "learning_rate": 7.214895747154884e-06, "loss": 0.4486, "step": 1424 }, { "epoch": 0.8385744234800838, "grad_norm": 0.3975187010168781, "learning_rate": 7.210292394272029e-06, "loss": 0.4799, "step": 1425 }, { "epoch": 0.839162896759719, "grad_norm": 0.32529575252510573, "learning_rate": 7.205686711653856e-06, "loss": 0.444, "step": 1426 }, { "epoch": 0.8397513700393542, "grad_norm": 0.46883203241140714, "learning_rate": 7.201078704154938e-06, "loss": 0.5263, "step": 1427 }, { "epoch": 0.8403398433189893, "grad_norm": 0.45347997382957317, "learning_rate": 7.1964683766322916e-06, "loss": 0.5461, "step": 1428 }, { "epoch": 0.8409283165986244, "grad_norm": 0.35043951658132594, "learning_rate": 7.191855733945388e-06, "loss": 0.4277, "step": 1429 }, { "epoch": 0.8415167898782596, "grad_norm": 0.394448980687866, "learning_rate": 7.187240780956133e-06, "loss": 0.5107, "step": 1430 }, { "epoch": 0.8421052631578947, "grad_norm": 0.4678033333655182, "learning_rate": 7.182623522528866e-06, "loss": 0.5285, "step": 1431 }, { "epoch": 0.8426937364375299, "grad_norm": 0.38752102048405546, "learning_rate": 7.178003963530367e-06, "loss": 0.5033, "step": 1432 }, { "epoch": 0.843282209717165, "grad_norm": 0.4072868188022264, "learning_rate": 7.173382108829826e-06, "loss": 0.5247, "step": 1433 }, { "epoch": 0.8438706829968001, "grad_norm": 0.507005050835647, "learning_rate": 7.168757963298863e-06, "loss": 0.5906, "step": 1434 }, { "epoch": 0.8444591562764353, "grad_norm": 0.38937746995550215, "learning_rate": 7.164131531811509e-06, "loss": 0.4745, "step": 1435 }, { "epoch": 0.8450476295560705, "grad_norm": 0.3892975544681639, "learning_rate": 7.159502819244206e-06, "loss": 0.4738, "step": 1436 }, { "epoch": 0.8456361028357056, "grad_norm": 0.35244420258561737, "learning_rate": 7.154871830475798e-06, "loss": 0.4812, "step": 1437 }, { "epoch": 0.8462245761153407, "grad_norm": 0.4241143250486472, "learning_rate": 7.1502385703875325e-06, "loss": 0.5019, "step": 1438 }, { "epoch": 0.8468130493949759, "grad_norm": 0.3743747532307553, "learning_rate": 7.145603043863045e-06, "loss": 0.4773, "step": 1439 }, { "epoch": 0.8474015226746111, "grad_norm": 0.3629860369350423, "learning_rate": 7.1409652557883655e-06, "loss": 0.4626, "step": 1440 }, { "epoch": 0.8479899959542462, "grad_norm": 0.4009308148681624, "learning_rate": 7.136325211051905e-06, "loss": 0.4952, "step": 1441 }, { "epoch": 0.8485784692338814, "grad_norm": 0.4137710565473535, "learning_rate": 7.131682914544454e-06, "loss": 0.5178, "step": 1442 }, { "epoch": 0.8491669425135165, "grad_norm": 0.40527321468744754, "learning_rate": 7.127038371159174e-06, "loss": 0.4715, "step": 1443 }, { "epoch": 0.8497554157931516, "grad_norm": 0.42510599802163745, "learning_rate": 7.122391585791598e-06, "loss": 0.5841, "step": 1444 }, { "epoch": 0.8503438890727868, "grad_norm": 0.3804168433779166, "learning_rate": 7.117742563339622e-06, "loss": 0.4556, "step": 1445 }, { "epoch": 0.850932362352422, "grad_norm": 0.3809394150657421, "learning_rate": 7.113091308703498e-06, "loss": 0.476, "step": 1446 }, { "epoch": 0.851520835632057, "grad_norm": 0.4021250788946828, "learning_rate": 7.1084378267858325e-06, "loss": 0.5014, "step": 1447 }, { "epoch": 0.8521093089116922, "grad_norm": 0.41749837755354957, "learning_rate": 7.103782122491577e-06, "loss": 0.5252, "step": 1448 }, { "epoch": 0.8526977821913274, "grad_norm": 0.4476217194361225, "learning_rate": 7.099124200728028e-06, "loss": 0.5381, "step": 1449 }, { "epoch": 0.8532862554709625, "grad_norm": 0.45780753671051666, "learning_rate": 7.09446406640482e-06, "loss": 0.5722, "step": 1450 }, { "epoch": 0.8538747287505977, "grad_norm": 0.40635237424934645, "learning_rate": 7.089801724433918e-06, "loss": 0.5422, "step": 1451 }, { "epoch": 0.8544632020302328, "grad_norm": 0.3649553879051312, "learning_rate": 7.0851371797296134e-06, "loss": 0.4204, "step": 1452 }, { "epoch": 0.8550516753098679, "grad_norm": 0.3840879270524602, "learning_rate": 7.0804704372085205e-06, "loss": 0.4577, "step": 1453 }, { "epoch": 0.8556401485895031, "grad_norm": 0.41751606293737475, "learning_rate": 7.07580150178957e-06, "loss": 0.5649, "step": 1454 }, { "epoch": 0.8562286218691383, "grad_norm": 0.43668301754722955, "learning_rate": 7.071130378394003e-06, "loss": 0.4752, "step": 1455 }, { "epoch": 0.8568170951487734, "grad_norm": 0.4485470349853313, "learning_rate": 7.066457071945371e-06, "loss": 0.5013, "step": 1456 }, { "epoch": 0.8574055684284085, "grad_norm": 0.3761038211065385, "learning_rate": 7.061781587369518e-06, "loss": 0.4858, "step": 1457 }, { "epoch": 0.8579940417080437, "grad_norm": 0.3334150384761586, "learning_rate": 7.057103929594593e-06, "loss": 0.4251, "step": 1458 }, { "epoch": 0.8585825149876788, "grad_norm": 0.3716311525015339, "learning_rate": 7.052424103551031e-06, "loss": 0.4413, "step": 1459 }, { "epoch": 0.859170988267314, "grad_norm": 0.384331132750536, "learning_rate": 7.047742114171553e-06, "loss": 0.4833, "step": 1460 }, { "epoch": 0.8597594615469492, "grad_norm": 0.38087340808497877, "learning_rate": 7.043057966391158e-06, "loss": 0.5021, "step": 1461 }, { "epoch": 0.8603479348265842, "grad_norm": 0.3484447545033256, "learning_rate": 7.0383716651471265e-06, "loss": 0.4826, "step": 1462 }, { "epoch": 0.8609364081062194, "grad_norm": 0.3895464200163944, "learning_rate": 7.033683215379002e-06, "loss": 0.5148, "step": 1463 }, { "epoch": 0.8615248813858546, "grad_norm": 0.3863892656106378, "learning_rate": 7.028992622028596e-06, "loss": 0.5168, "step": 1464 }, { "epoch": 0.8621133546654898, "grad_norm": 0.42921425234424304, "learning_rate": 7.024299890039978e-06, "loss": 0.5049, "step": 1465 }, { "epoch": 0.8627018279451248, "grad_norm": 0.41768719937076526, "learning_rate": 7.019605024359475e-06, "loss": 0.52, "step": 1466 }, { "epoch": 0.86329030122476, "grad_norm": 0.4031172119104426, "learning_rate": 7.014908029935659e-06, "loss": 0.514, "step": 1467 }, { "epoch": 0.8638787745043952, "grad_norm": 0.3263335780181873, "learning_rate": 7.010208911719346e-06, "loss": 0.3996, "step": 1468 }, { "epoch": 0.8644672477840303, "grad_norm": 0.36533742848789297, "learning_rate": 7.005507674663594e-06, "loss": 0.4927, "step": 1469 }, { "epoch": 0.8650557210636655, "grad_norm": 0.4436484677046036, "learning_rate": 7.000804323723691e-06, "loss": 0.601, "step": 1470 }, { "epoch": 0.8656441943433006, "grad_norm": 0.3885811621552288, "learning_rate": 6.996098863857155e-06, "loss": 0.5154, "step": 1471 }, { "epoch": 0.8662326676229357, "grad_norm": 0.3681783689356279, "learning_rate": 6.991391300023725e-06, "loss": 0.5165, "step": 1472 }, { "epoch": 0.8668211409025709, "grad_norm": 0.35739756678508616, "learning_rate": 6.986681637185361e-06, "loss": 0.4413, "step": 1473 }, { "epoch": 0.8674096141822061, "grad_norm": 0.39228041812338843, "learning_rate": 6.981969880306231e-06, "loss": 0.5071, "step": 1474 }, { "epoch": 0.8679980874618412, "grad_norm": 0.3417827800854309, "learning_rate": 6.977256034352713e-06, "loss": 0.4725, "step": 1475 }, { "epoch": 0.8685865607414763, "grad_norm": 0.48715768840404533, "learning_rate": 6.972540104293388e-06, "loss": 0.5118, "step": 1476 }, { "epoch": 0.8691750340211115, "grad_norm": 0.4271461094228543, "learning_rate": 6.967822095099031e-06, "loss": 0.577, "step": 1477 }, { "epoch": 0.8697635073007466, "grad_norm": 0.4141036223929461, "learning_rate": 6.963102011742608e-06, "loss": 0.4225, "step": 1478 }, { "epoch": 0.8703519805803818, "grad_norm": 0.39629306703930706, "learning_rate": 6.958379859199277e-06, "loss": 0.5397, "step": 1479 }, { "epoch": 0.870940453860017, "grad_norm": 0.41002602562937696, "learning_rate": 6.953655642446368e-06, "loss": 0.4751, "step": 1480 }, { "epoch": 0.871528927139652, "grad_norm": 0.405545019496186, "learning_rate": 6.948929366463397e-06, "loss": 0.4807, "step": 1481 }, { "epoch": 0.8721174004192872, "grad_norm": 0.41260364752879036, "learning_rate": 6.9442010362320424e-06, "loss": 0.4844, "step": 1482 }, { "epoch": 0.8727058736989224, "grad_norm": 0.3888099534587087, "learning_rate": 6.939470656736149e-06, "loss": 0.5024, "step": 1483 }, { "epoch": 0.8732943469785575, "grad_norm": 0.3505732262015088, "learning_rate": 6.934738232961728e-06, "loss": 0.4779, "step": 1484 }, { "epoch": 0.8738828202581926, "grad_norm": 0.3803428305121752, "learning_rate": 6.930003769896936e-06, "loss": 0.3893, "step": 1485 }, { "epoch": 0.8744712935378278, "grad_norm": 0.373261195850205, "learning_rate": 6.9252672725320835e-06, "loss": 0.4724, "step": 1486 }, { "epoch": 0.8750597668174629, "grad_norm": 0.38680995073780655, "learning_rate": 6.9205287458596305e-06, "loss": 0.5087, "step": 1487 }, { "epoch": 0.8756482400970981, "grad_norm": 0.3812748424416209, "learning_rate": 6.915788194874167e-06, "loss": 0.4832, "step": 1488 }, { "epoch": 0.8762367133767333, "grad_norm": 0.3598142372212699, "learning_rate": 6.91104562457242e-06, "loss": 0.4196, "step": 1489 }, { "epoch": 0.8768251866563684, "grad_norm": 0.3963839591572361, "learning_rate": 6.906301039953248e-06, "loss": 0.4687, "step": 1490 }, { "epoch": 0.8774136599360035, "grad_norm": 0.37564525447542074, "learning_rate": 6.9015544460176296e-06, "loss": 0.508, "step": 1491 }, { "epoch": 0.8780021332156387, "grad_norm": 0.5035343266681225, "learning_rate": 6.896805847768661e-06, "loss": 0.5192, "step": 1492 }, { "epoch": 0.8785906064952739, "grad_norm": 0.3696821344493171, "learning_rate": 6.892055250211552e-06, "loss": 0.4372, "step": 1493 }, { "epoch": 0.879179079774909, "grad_norm": 0.380938896879502, "learning_rate": 6.887302658353622e-06, "loss": 0.4507, "step": 1494 }, { "epoch": 0.8797675530545441, "grad_norm": 0.35818978402997936, "learning_rate": 6.882548077204285e-06, "loss": 0.4702, "step": 1495 }, { "epoch": 0.8803560263341793, "grad_norm": 0.37748291495500713, "learning_rate": 6.877791511775064e-06, "loss": 0.4836, "step": 1496 }, { "epoch": 0.8809444996138144, "grad_norm": 0.41256595143049335, "learning_rate": 6.873032967079562e-06, "loss": 0.5255, "step": 1497 }, { "epoch": 0.8815329728934496, "grad_norm": 0.4205370098923406, "learning_rate": 6.868272448133472e-06, "loss": 0.4936, "step": 1498 }, { "epoch": 0.8821214461730847, "grad_norm": 0.36563683766562527, "learning_rate": 6.86350995995457e-06, "loss": 0.428, "step": 1499 }, { "epoch": 0.8827099194527198, "grad_norm": 0.372078807050244, "learning_rate": 6.858745507562709e-06, "loss": 0.4861, "step": 1500 }, { "epoch": 0.883298392732355, "grad_norm": 0.3642801425684304, "learning_rate": 6.8539790959798045e-06, "loss": 0.4569, "step": 1501 }, { "epoch": 0.8838868660119902, "grad_norm": 0.36176269357256513, "learning_rate": 6.849210730229846e-06, "loss": 0.4806, "step": 1502 }, { "epoch": 0.8844753392916253, "grad_norm": 0.3395722043821801, "learning_rate": 6.844440415338879e-06, "loss": 0.482, "step": 1503 }, { "epoch": 0.8850638125712604, "grad_norm": 0.3396027337345269, "learning_rate": 6.839668156335e-06, "loss": 0.4602, "step": 1504 }, { "epoch": 0.8856522858508956, "grad_norm": 0.4131658071400458, "learning_rate": 6.834893958248361e-06, "loss": 0.5297, "step": 1505 }, { "epoch": 0.8862407591305307, "grad_norm": 0.39941380946746613, "learning_rate": 6.830117826111155e-06, "loss": 0.4928, "step": 1506 }, { "epoch": 0.8868292324101659, "grad_norm": 0.37224659902169344, "learning_rate": 6.825339764957613e-06, "loss": 0.4315, "step": 1507 }, { "epoch": 0.887417705689801, "grad_norm": 0.37541451129319997, "learning_rate": 6.820559779823999e-06, "loss": 0.4413, "step": 1508 }, { "epoch": 0.8880061789694361, "grad_norm": 0.4482474870876049, "learning_rate": 6.815777875748607e-06, "loss": 0.5003, "step": 1509 }, { "epoch": 0.8885946522490713, "grad_norm": 0.3951254322233524, "learning_rate": 6.810994057771752e-06, "loss": 0.4744, "step": 1510 }, { "epoch": 0.8891831255287065, "grad_norm": 0.3420083518657523, "learning_rate": 6.806208330935766e-06, "loss": 0.4967, "step": 1511 }, { "epoch": 0.8897715988083416, "grad_norm": 0.37326627168674265, "learning_rate": 6.801420700284995e-06, "loss": 0.4616, "step": 1512 }, { "epoch": 0.8903600720879767, "grad_norm": 0.4791464698448905, "learning_rate": 6.7966311708657884e-06, "loss": 0.6698, "step": 1513 }, { "epoch": 0.8909485453676119, "grad_norm": 0.4037139318086035, "learning_rate": 6.7918397477265e-06, "loss": 0.4592, "step": 1514 }, { "epoch": 0.8915370186472471, "grad_norm": 0.4126780648796663, "learning_rate": 6.787046435917483e-06, "loss": 0.5388, "step": 1515 }, { "epoch": 0.8921254919268822, "grad_norm": 0.5004251651160528, "learning_rate": 6.782251240491071e-06, "loss": 0.5648, "step": 1516 }, { "epoch": 0.8927139652065174, "grad_norm": 0.37519980078487725, "learning_rate": 6.77745416650159e-06, "loss": 0.5102, "step": 1517 }, { "epoch": 0.8933024384861525, "grad_norm": 0.38473872883745225, "learning_rate": 6.772655219005349e-06, "loss": 0.5658, "step": 1518 }, { "epoch": 0.8938909117657876, "grad_norm": 0.40446601851607, "learning_rate": 6.767854403060623e-06, "loss": 0.4276, "step": 1519 }, { "epoch": 0.8944793850454228, "grad_norm": 0.4774811000202144, "learning_rate": 6.763051723727663e-06, "loss": 0.5604, "step": 1520 }, { "epoch": 0.895067858325058, "grad_norm": 0.4088552708711032, "learning_rate": 6.758247186068684e-06, "loss": 0.468, "step": 1521 }, { "epoch": 0.895656331604693, "grad_norm": 0.4089007442200555, "learning_rate": 6.753440795147857e-06, "loss": 0.4965, "step": 1522 }, { "epoch": 0.8962448048843282, "grad_norm": 0.46517206568891417, "learning_rate": 6.748632556031306e-06, "loss": 0.5622, "step": 1523 }, { "epoch": 0.8968332781639634, "grad_norm": 0.43857911192055676, "learning_rate": 6.743822473787107e-06, "loss": 0.517, "step": 1524 }, { "epoch": 0.8974217514435985, "grad_norm": 0.3402518702094257, "learning_rate": 6.739010553485276e-06, "loss": 0.463, "step": 1525 }, { "epoch": 0.8980102247232337, "grad_norm": 0.43663980880683745, "learning_rate": 6.734196800197763e-06, "loss": 0.5276, "step": 1526 }, { "epoch": 0.8985986980028688, "grad_norm": 0.36366756729036015, "learning_rate": 6.72938121899846e-06, "loss": 0.4628, "step": 1527 }, { "epoch": 0.8991871712825039, "grad_norm": 0.4553662507241533, "learning_rate": 6.724563814963174e-06, "loss": 0.5433, "step": 1528 }, { "epoch": 0.8997756445621391, "grad_norm": 0.4474811253951376, "learning_rate": 6.719744593169642e-06, "loss": 0.5104, "step": 1529 }, { "epoch": 0.9003641178417743, "grad_norm": 0.3637614665058022, "learning_rate": 6.714923558697511e-06, "loss": 0.4264, "step": 1530 }, { "epoch": 0.9009525911214094, "grad_norm": 0.43533277572241036, "learning_rate": 6.710100716628345e-06, "loss": 0.5155, "step": 1531 }, { "epoch": 0.9015410644010445, "grad_norm": 0.37703127021122346, "learning_rate": 6.7052760720456055e-06, "loss": 0.4471, "step": 1532 }, { "epoch": 0.9021295376806797, "grad_norm": 0.41002628526864804, "learning_rate": 6.700449630034662e-06, "loss": 0.5726, "step": 1533 }, { "epoch": 0.9027180109603148, "grad_norm": 0.451216546736548, "learning_rate": 6.695621395682772e-06, "loss": 0.4564, "step": 1534 }, { "epoch": 0.90330648423995, "grad_norm": 0.42110121255111005, "learning_rate": 6.690791374079086e-06, "loss": 0.4997, "step": 1535 }, { "epoch": 0.9038949575195852, "grad_norm": 0.4181852295679939, "learning_rate": 6.685959570314638e-06, "loss": 0.4995, "step": 1536 }, { "epoch": 0.9044834307992202, "grad_norm": 0.341259661863166, "learning_rate": 6.681125989482337e-06, "loss": 0.443, "step": 1537 }, { "epoch": 0.9050719040788554, "grad_norm": 0.3691055949452499, "learning_rate": 6.67629063667697e-06, "loss": 0.4768, "step": 1538 }, { "epoch": 0.9056603773584906, "grad_norm": 0.3693888418244883, "learning_rate": 6.671453516995187e-06, "loss": 0.4405, "step": 1539 }, { "epoch": 0.9062488506381257, "grad_norm": 0.4040082399892138, "learning_rate": 6.6666146355355074e-06, "loss": 0.4795, "step": 1540 }, { "epoch": 0.9068373239177608, "grad_norm": 0.36941340737290623, "learning_rate": 6.6617739973982985e-06, "loss": 0.39, "step": 1541 }, { "epoch": 0.907425797197396, "grad_norm": 0.38666403227719565, "learning_rate": 6.656931607685787e-06, "loss": 0.5102, "step": 1542 }, { "epoch": 0.9080142704770312, "grad_norm": 0.3344912099329994, "learning_rate": 6.65208747150204e-06, "loss": 0.4141, "step": 1543 }, { "epoch": 0.9086027437566663, "grad_norm": 0.37624274191851914, "learning_rate": 6.647241593952969e-06, "loss": 0.5023, "step": 1544 }, { "epoch": 0.9091912170363015, "grad_norm": 0.34177207484152494, "learning_rate": 6.64239398014632e-06, "loss": 0.4133, "step": 1545 }, { "epoch": 0.9097796903159366, "grad_norm": 0.38352609644958846, "learning_rate": 6.63754463519167e-06, "loss": 0.5441, "step": 1546 }, { "epoch": 0.9103681635955717, "grad_norm": 0.4316139060909725, "learning_rate": 6.6326935642004165e-06, "loss": 0.5674, "step": 1547 }, { "epoch": 0.9109566368752069, "grad_norm": 0.3996310423176003, "learning_rate": 6.627840772285784e-06, "loss": 0.5419, "step": 1548 }, { "epoch": 0.9115451101548421, "grad_norm": 0.3591083245325282, "learning_rate": 6.622986264562804e-06, "loss": 0.4608, "step": 1549 }, { "epoch": 0.9121335834344771, "grad_norm": 0.34713947149081587, "learning_rate": 6.618130046148319e-06, "loss": 0.4502, "step": 1550 }, { "epoch": 0.9127220567141123, "grad_norm": 3.3383955361899833, "learning_rate": 6.613272122160975e-06, "loss": 0.5156, "step": 1551 }, { "epoch": 0.9133105299937475, "grad_norm": 0.6279719317625666, "learning_rate": 6.608412497721216e-06, "loss": 0.5577, "step": 1552 }, { "epoch": 0.9138990032733826, "grad_norm": 0.5831038030690291, "learning_rate": 6.6035511779512764e-06, "loss": 0.5418, "step": 1553 }, { "epoch": 0.9144874765530178, "grad_norm": 0.44202263557957827, "learning_rate": 6.5986881679751805e-06, "loss": 0.5282, "step": 1554 }, { "epoch": 0.915075949832653, "grad_norm": 0.4431103983799578, "learning_rate": 6.593823472918731e-06, "loss": 0.5182, "step": 1555 }, { "epoch": 0.915664423112288, "grad_norm": 0.44398771053913755, "learning_rate": 6.588957097909509e-06, "loss": 0.4805, "step": 1556 }, { "epoch": 0.9162528963919232, "grad_norm": 0.4324373009629152, "learning_rate": 6.584089048076866e-06, "loss": 0.471, "step": 1557 }, { "epoch": 0.9168413696715584, "grad_norm": 0.3860611012172054, "learning_rate": 6.579219328551918e-06, "loss": 0.4472, "step": 1558 }, { "epoch": 0.9174298429511935, "grad_norm": 0.426186223995795, "learning_rate": 6.57434794446754e-06, "loss": 0.472, "step": 1559 }, { "epoch": 0.9180183162308286, "grad_norm": 0.3871592620096749, "learning_rate": 6.569474900958366e-06, "loss": 0.5013, "step": 1560 }, { "epoch": 0.9186067895104638, "grad_norm": 0.37522699658441766, "learning_rate": 6.5646002031607726e-06, "loss": 0.4808, "step": 1561 }, { "epoch": 0.9191952627900989, "grad_norm": 0.34396520094739547, "learning_rate": 6.559723856212885e-06, "loss": 0.4333, "step": 1562 }, { "epoch": 0.9197837360697341, "grad_norm": 0.3763760047039893, "learning_rate": 6.554845865254566e-06, "loss": 0.398, "step": 1563 }, { "epoch": 0.9203722093493693, "grad_norm": 0.4090596368277118, "learning_rate": 6.54996623542741e-06, "loss": 0.45, "step": 1564 }, { "epoch": 0.9209606826290043, "grad_norm": 0.3987904400689906, "learning_rate": 6.545084971874738e-06, "loss": 0.5139, "step": 1565 }, { "epoch": 0.9215491559086395, "grad_norm": 0.3881947757988772, "learning_rate": 6.540202079741594e-06, "loss": 0.4568, "step": 1566 }, { "epoch": 0.9221376291882747, "grad_norm": 0.40255808073085664, "learning_rate": 6.535317564174741e-06, "loss": 0.4304, "step": 1567 }, { "epoch": 0.9227261024679099, "grad_norm": 0.4054047135582287, "learning_rate": 6.53043143032265e-06, "loss": 0.5349, "step": 1568 }, { "epoch": 0.9233145757475449, "grad_norm": 0.3668651360541812, "learning_rate": 6.525543683335497e-06, "loss": 0.4457, "step": 1569 }, { "epoch": 0.9239030490271801, "grad_norm": 0.3821315977012238, "learning_rate": 6.520654328365162e-06, "loss": 0.4954, "step": 1570 }, { "epoch": 0.9244915223068153, "grad_norm": 0.43402107921877114, "learning_rate": 6.515763370565218e-06, "loss": 0.5301, "step": 1571 }, { "epoch": 0.9250799955864504, "grad_norm": 0.43433106374965913, "learning_rate": 6.510870815090926e-06, "loss": 0.5597, "step": 1572 }, { "epoch": 0.9256684688660856, "grad_norm": 0.36429072698135917, "learning_rate": 6.505976667099233e-06, "loss": 0.4264, "step": 1573 }, { "epoch": 0.9262569421457207, "grad_norm": 0.35731711574251873, "learning_rate": 6.501080931748764e-06, "loss": 0.4689, "step": 1574 }, { "epoch": 0.9268454154253558, "grad_norm": 0.34656355789290755, "learning_rate": 6.4961836141998166e-06, "loss": 0.4599, "step": 1575 }, { "epoch": 0.927433888704991, "grad_norm": 0.3990677312660943, "learning_rate": 6.4912847196143565e-06, "loss": 0.4772, "step": 1576 }, { "epoch": 0.9280223619846262, "grad_norm": 0.3982847989824541, "learning_rate": 6.486384253156014e-06, "loss": 0.4864, "step": 1577 }, { "epoch": 0.9286108352642612, "grad_norm": 0.3961043157329895, "learning_rate": 6.481482219990071e-06, "loss": 0.5144, "step": 1578 }, { "epoch": 0.9291993085438964, "grad_norm": 0.3760104991388315, "learning_rate": 6.476578625283465e-06, "loss": 0.4396, "step": 1579 }, { "epoch": 0.9297877818235316, "grad_norm": 0.39944267089454527, "learning_rate": 6.471673474204777e-06, "loss": 0.5417, "step": 1580 }, { "epoch": 0.9303762551031667, "grad_norm": 0.38672909158623964, "learning_rate": 6.466766771924231e-06, "loss": 0.4895, "step": 1581 }, { "epoch": 0.9309647283828019, "grad_norm": 0.3721142923172843, "learning_rate": 6.461858523613684e-06, "loss": 0.4995, "step": 1582 }, { "epoch": 0.931553201662437, "grad_norm": 0.38648517943572336, "learning_rate": 6.456948734446624e-06, "loss": 0.5048, "step": 1583 }, { "epoch": 0.9321416749420721, "grad_norm": 0.38233872275454667, "learning_rate": 6.452037409598161e-06, "loss": 0.4798, "step": 1584 }, { "epoch": 0.9327301482217073, "grad_norm": 0.34978059158623886, "learning_rate": 6.447124554245026e-06, "loss": 0.4095, "step": 1585 }, { "epoch": 0.9333186215013425, "grad_norm": 0.3936348941253742, "learning_rate": 6.442210173565562e-06, "loss": 0.4799, "step": 1586 }, { "epoch": 0.9339070947809776, "grad_norm": 0.38813309787241707, "learning_rate": 6.4372942727397195e-06, "loss": 0.4987, "step": 1587 }, { "epoch": 0.9344955680606127, "grad_norm": 0.4951969897226153, "learning_rate": 6.432376856949053e-06, "loss": 0.5394, "step": 1588 }, { "epoch": 0.9350840413402479, "grad_norm": 0.39250259265652876, "learning_rate": 6.427457931376712e-06, "loss": 0.4852, "step": 1589 }, { "epoch": 0.935672514619883, "grad_norm": 0.3734774714208786, "learning_rate": 6.422537501207435e-06, "loss": 0.5037, "step": 1590 }, { "epoch": 0.9362609878995182, "grad_norm": 0.36412022227397706, "learning_rate": 6.417615571627555e-06, "loss": 0.4645, "step": 1591 }, { "epoch": 0.9368494611791534, "grad_norm": 0.42371223865061974, "learning_rate": 6.412692147824976e-06, "loss": 0.4942, "step": 1592 }, { "epoch": 0.9374379344587885, "grad_norm": 0.4247566556618268, "learning_rate": 6.407767234989181e-06, "loss": 0.5138, "step": 1593 }, { "epoch": 0.9380264077384236, "grad_norm": 0.3706849866016851, "learning_rate": 6.402840838311224e-06, "loss": 0.4644, "step": 1594 }, { "epoch": 0.9386148810180588, "grad_norm": 0.370134437473027, "learning_rate": 6.39791296298372e-06, "loss": 0.4877, "step": 1595 }, { "epoch": 0.939203354297694, "grad_norm": 0.35831173522468185, "learning_rate": 6.392983614200844e-06, "loss": 0.4623, "step": 1596 }, { "epoch": 0.939791827577329, "grad_norm": 0.37624268683050505, "learning_rate": 6.388052797158324e-06, "loss": 0.5165, "step": 1597 }, { "epoch": 0.9403803008569642, "grad_norm": 0.3613185084987495, "learning_rate": 6.3831205170534365e-06, "loss": 0.4531, "step": 1598 }, { "epoch": 0.9409687741365994, "grad_norm": 0.3783739151843012, "learning_rate": 6.378186779084996e-06, "loss": 0.4606, "step": 1599 }, { "epoch": 0.9415572474162345, "grad_norm": 0.36472098660904984, "learning_rate": 6.373251588453362e-06, "loss": 0.4271, "step": 1600 }, { "epoch": 0.9421457206958697, "grad_norm": 0.3425550328785131, "learning_rate": 6.368314950360416e-06, "loss": 0.4619, "step": 1601 }, { "epoch": 0.9427341939755048, "grad_norm": 0.39894064790372835, "learning_rate": 6.363376870009569e-06, "loss": 0.4738, "step": 1602 }, { "epoch": 0.9433226672551399, "grad_norm": 0.4141544993645895, "learning_rate": 6.3584373526057556e-06, "loss": 0.4422, "step": 1603 }, { "epoch": 0.9439111405347751, "grad_norm": 0.39188335929763213, "learning_rate": 6.353496403355419e-06, "loss": 0.4845, "step": 1604 }, { "epoch": 0.9444996138144103, "grad_norm": 0.37387190627620764, "learning_rate": 6.3485540274665134e-06, "loss": 0.4964, "step": 1605 }, { "epoch": 0.9450880870940453, "grad_norm": 0.37381279967936354, "learning_rate": 6.343610230148502e-06, "loss": 0.4456, "step": 1606 }, { "epoch": 0.9456765603736805, "grad_norm": 0.35387799287850447, "learning_rate": 6.3386650166123406e-06, "loss": 0.4447, "step": 1607 }, { "epoch": 0.9462650336533157, "grad_norm": 0.358506178927895, "learning_rate": 6.333718392070477e-06, "loss": 0.4034, "step": 1608 }, { "epoch": 0.9468535069329508, "grad_norm": 0.4195195597830937, "learning_rate": 6.32877036173685e-06, "loss": 0.5533, "step": 1609 }, { "epoch": 0.947441980212586, "grad_norm": 0.41062923774398985, "learning_rate": 6.323820930826879e-06, "loss": 0.4725, "step": 1610 }, { "epoch": 0.9480304534922211, "grad_norm": 0.4057595071839761, "learning_rate": 6.318870104557459e-06, "loss": 0.5407, "step": 1611 }, { "epoch": 0.9486189267718562, "grad_norm": 0.3722571565938002, "learning_rate": 6.313917888146956e-06, "loss": 0.4958, "step": 1612 }, { "epoch": 0.9492074000514914, "grad_norm": 0.40177458414819445, "learning_rate": 6.308964286815203e-06, "loss": 0.4921, "step": 1613 }, { "epoch": 0.9497958733311266, "grad_norm": 0.38048494413467115, "learning_rate": 6.304009305783487e-06, "loss": 0.4803, "step": 1614 }, { "epoch": 0.9503843466107617, "grad_norm": 0.4402669409361972, "learning_rate": 6.299052950274556e-06, "loss": 0.5054, "step": 1615 }, { "epoch": 0.9509728198903968, "grad_norm": 0.32898172081535004, "learning_rate": 6.294095225512604e-06, "loss": 0.4323, "step": 1616 }, { "epoch": 0.951561293170032, "grad_norm": 0.36029086509925407, "learning_rate": 6.289136136723268e-06, "loss": 0.4858, "step": 1617 }, { "epoch": 0.9521497664496671, "grad_norm": 0.38534571338457224, "learning_rate": 6.284175689133624e-06, "loss": 0.5073, "step": 1618 }, { "epoch": 0.9527382397293023, "grad_norm": 0.36585418929435887, "learning_rate": 6.279213887972179e-06, "loss": 0.5199, "step": 1619 }, { "epoch": 0.9533267130089375, "grad_norm": 0.3217190289629232, "learning_rate": 6.274250738468866e-06, "loss": 0.4485, "step": 1620 }, { "epoch": 0.9539151862885726, "grad_norm": 0.394477684563973, "learning_rate": 6.269286245855039e-06, "loss": 0.5413, "step": 1621 }, { "epoch": 0.9545036595682077, "grad_norm": 0.41030938027853997, "learning_rate": 6.264320415363473e-06, "loss": 0.5169, "step": 1622 }, { "epoch": 0.9550921328478429, "grad_norm": 0.3973589548594529, "learning_rate": 6.2593532522283465e-06, "loss": 0.5455, "step": 1623 }, { "epoch": 0.9556806061274781, "grad_norm": 0.3543452258967846, "learning_rate": 6.254384761685245e-06, "loss": 0.4762, "step": 1624 }, { "epoch": 0.9562690794071131, "grad_norm": 0.38259286414666444, "learning_rate": 6.249414948971154e-06, "loss": 0.4875, "step": 1625 }, { "epoch": 0.9568575526867483, "grad_norm": 0.3453086802015334, "learning_rate": 6.244443819324454e-06, "loss": 0.4562, "step": 1626 }, { "epoch": 0.9574460259663835, "grad_norm": 0.3651330940875683, "learning_rate": 6.239471377984907e-06, "loss": 0.4139, "step": 1627 }, { "epoch": 0.9580344992460186, "grad_norm": 0.39200812219287234, "learning_rate": 6.234497630193666e-06, "loss": 0.4471, "step": 1628 }, { "epoch": 0.9586229725256538, "grad_norm": 0.3403657148876542, "learning_rate": 6.229522581193257e-06, "loss": 0.3913, "step": 1629 }, { "epoch": 0.9592114458052889, "grad_norm": 0.36801702870680353, "learning_rate": 6.224546236227575e-06, "loss": 0.4457, "step": 1630 }, { "epoch": 0.959799919084924, "grad_norm": 0.3418839199439611, "learning_rate": 6.219568600541886e-06, "loss": 0.4342, "step": 1631 }, { "epoch": 0.9603883923645592, "grad_norm": 0.4096904325434702, "learning_rate": 6.214589679382815e-06, "loss": 0.5363, "step": 1632 }, { "epoch": 0.9609768656441944, "grad_norm": 0.33524802591819597, "learning_rate": 6.209609477998339e-06, "loss": 0.4279, "step": 1633 }, { "epoch": 0.9615653389238294, "grad_norm": 0.38686435900156324, "learning_rate": 6.204628001637788e-06, "loss": 0.4907, "step": 1634 }, { "epoch": 0.9621538122034646, "grad_norm": 0.41423018059531974, "learning_rate": 6.199645255551835e-06, "loss": 0.4823, "step": 1635 }, { "epoch": 0.9627422854830998, "grad_norm": 0.36757399780775957, "learning_rate": 6.194661244992488e-06, "loss": 0.4925, "step": 1636 }, { "epoch": 0.9633307587627349, "grad_norm": 0.38116401466707134, "learning_rate": 6.189675975213094e-06, "loss": 0.4669, "step": 1637 }, { "epoch": 0.9639192320423701, "grad_norm": 0.39919591856799785, "learning_rate": 6.184689451468323e-06, "loss": 0.4742, "step": 1638 }, { "epoch": 0.9645077053220052, "grad_norm": 0.3977595531509558, "learning_rate": 6.179701679014165e-06, "loss": 0.5874, "step": 1639 }, { "epoch": 0.9650961786016403, "grad_norm": 0.37230981235609767, "learning_rate": 6.174712663107933e-06, "loss": 0.5056, "step": 1640 }, { "epoch": 0.9656846518812755, "grad_norm": 0.4032256051275845, "learning_rate": 6.169722409008244e-06, "loss": 0.5678, "step": 1641 }, { "epoch": 0.9662731251609107, "grad_norm": 0.35861231629149787, "learning_rate": 6.164730921975021e-06, "loss": 0.4138, "step": 1642 }, { "epoch": 0.9668615984405458, "grad_norm": 0.40681443497071973, "learning_rate": 6.159738207269491e-06, "loss": 0.5037, "step": 1643 }, { "epoch": 0.9674500717201809, "grad_norm": 0.3909607431857302, "learning_rate": 6.1547442701541714e-06, "loss": 0.4822, "step": 1644 }, { "epoch": 0.9680385449998161, "grad_norm": 0.40009304732647205, "learning_rate": 6.1497491158928694e-06, "loss": 0.4368, "step": 1645 }, { "epoch": 0.9686270182794513, "grad_norm": 0.3989052109464333, "learning_rate": 6.144752749750671e-06, "loss": 0.574, "step": 1646 }, { "epoch": 0.9692154915590864, "grad_norm": 0.3297623746105097, "learning_rate": 6.1397551769939475e-06, "loss": 0.4018, "step": 1647 }, { "epoch": 0.9698039648387216, "grad_norm": 0.379461278612819, "learning_rate": 6.134756402890335e-06, "loss": 0.4741, "step": 1648 }, { "epoch": 0.9703924381183567, "grad_norm": 0.34788687058177137, "learning_rate": 6.129756432708739e-06, "loss": 0.5047, "step": 1649 }, { "epoch": 0.9709809113979918, "grad_norm": 0.3802252615302962, "learning_rate": 6.124755271719326e-06, "loss": 0.4598, "step": 1650 }, { "epoch": 0.971569384677627, "grad_norm": 0.4176626468569386, "learning_rate": 6.119752925193516e-06, "loss": 0.5247, "step": 1651 }, { "epoch": 0.9721578579572622, "grad_norm": 0.3793776131295033, "learning_rate": 6.11474939840398e-06, "loss": 0.4737, "step": 1652 }, { "epoch": 0.9727463312368972, "grad_norm": 0.342827022781576, "learning_rate": 6.109744696624631e-06, "loss": 0.4225, "step": 1653 }, { "epoch": 0.9733348045165324, "grad_norm": 0.38540705424202276, "learning_rate": 6.104738825130625e-06, "loss": 0.516, "step": 1654 }, { "epoch": 0.9739232777961676, "grad_norm": 0.3849781279082399, "learning_rate": 6.099731789198344e-06, "loss": 0.5418, "step": 1655 }, { "epoch": 0.9745117510758027, "grad_norm": 0.3479803789333442, "learning_rate": 6.094723594105404e-06, "loss": 0.4451, "step": 1656 }, { "epoch": 0.9751002243554379, "grad_norm": 0.4301921187743882, "learning_rate": 6.089714245130639e-06, "loss": 0.5191, "step": 1657 }, { "epoch": 0.975688697635073, "grad_norm": 0.3371866206404496, "learning_rate": 6.0847037475541015e-06, "loss": 0.4025, "step": 1658 }, { "epoch": 0.9762771709147081, "grad_norm": 0.3689886531149989, "learning_rate": 6.079692106657052e-06, "loss": 0.4747, "step": 1659 }, { "epoch": 0.9768656441943433, "grad_norm": 0.3730707049492111, "learning_rate": 6.07467932772196e-06, "loss": 0.4874, "step": 1660 }, { "epoch": 0.9774541174739785, "grad_norm": 0.4133080548182097, "learning_rate": 6.0696654160324875e-06, "loss": 0.5754, "step": 1661 }, { "epoch": 0.9780425907536135, "grad_norm": 0.3917701708417934, "learning_rate": 6.064650376873499e-06, "loss": 0.4699, "step": 1662 }, { "epoch": 0.9786310640332487, "grad_norm": 0.39180506042339885, "learning_rate": 6.059634215531043e-06, "loss": 0.4896, "step": 1663 }, { "epoch": 0.9792195373128839, "grad_norm": 0.35330025866154285, "learning_rate": 6.05461693729235e-06, "loss": 0.4309, "step": 1664 }, { "epoch": 0.979808010592519, "grad_norm": 0.4180436536709424, "learning_rate": 6.049598547445829e-06, "loss": 0.4837, "step": 1665 }, { "epoch": 0.9803964838721542, "grad_norm": 0.3951644677261872, "learning_rate": 6.044579051281063e-06, "loss": 0.4926, "step": 1666 }, { "epoch": 0.9809849571517893, "grad_norm": 0.3637927039375087, "learning_rate": 6.039558454088796e-06, "loss": 0.459, "step": 1667 }, { "epoch": 0.9815734304314244, "grad_norm": 0.34808152940600295, "learning_rate": 6.034536761160938e-06, "loss": 0.4815, "step": 1668 }, { "epoch": 0.9821619037110596, "grad_norm": 0.40474570347687694, "learning_rate": 6.02951397779055e-06, "loss": 0.4764, "step": 1669 }, { "epoch": 0.9827503769906948, "grad_norm": 0.3884181966691102, "learning_rate": 6.024490109271842e-06, "loss": 0.4577, "step": 1670 }, { "epoch": 0.98333885027033, "grad_norm": 0.6705714428244329, "learning_rate": 6.019465160900173e-06, "loss": 0.5049, "step": 1671 }, { "epoch": 0.983927323549965, "grad_norm": 0.37414436009682706, "learning_rate": 6.014439137972034e-06, "loss": 0.4309, "step": 1672 }, { "epoch": 0.9845157968296002, "grad_norm": 0.4632576090786421, "learning_rate": 6.009412045785051e-06, "loss": 0.4945, "step": 1673 }, { "epoch": 0.9851042701092354, "grad_norm": 0.4019010376938184, "learning_rate": 6.00438388963798e-06, "loss": 0.4972, "step": 1674 }, { "epoch": 0.9856927433888705, "grad_norm": 0.41468200597547317, "learning_rate": 5.999354674830694e-06, "loss": 0.5121, "step": 1675 }, { "epoch": 0.9862812166685057, "grad_norm": 0.34617490188789485, "learning_rate": 5.994324406664184e-06, "loss": 0.466, "step": 1676 }, { "epoch": 0.9868696899481408, "grad_norm": 0.384491984376409, "learning_rate": 5.98929309044055e-06, "loss": 0.4893, "step": 1677 }, { "epoch": 0.9874581632277759, "grad_norm": 0.34462540781911005, "learning_rate": 5.984260731463e-06, "loss": 0.4392, "step": 1678 }, { "epoch": 0.9880466365074111, "grad_norm": 0.3673999905565276, "learning_rate": 5.9792273350358354e-06, "loss": 0.4188, "step": 1679 }, { "epoch": 0.9886351097870463, "grad_norm": 0.35950472679752976, "learning_rate": 5.974192906464458e-06, "loss": 0.4717, "step": 1680 }, { "epoch": 0.9892235830666813, "grad_norm": 0.48954609032488206, "learning_rate": 5.9691574510553505e-06, "loss": 0.5204, "step": 1681 }, { "epoch": 0.9898120563463165, "grad_norm": 0.4524177026368722, "learning_rate": 5.964120974116085e-06, "loss": 0.4961, "step": 1682 }, { "epoch": 0.9904005296259517, "grad_norm": 0.33382985562742573, "learning_rate": 5.959083480955304e-06, "loss": 0.4563, "step": 1683 }, { "epoch": 0.9909890029055868, "grad_norm": 0.4146217846981245, "learning_rate": 5.954044976882725e-06, "loss": 0.5032, "step": 1684 }, { "epoch": 0.991577476185222, "grad_norm": 0.31904945635470294, "learning_rate": 5.9490054672091305e-06, "loss": 0.3555, "step": 1685 }, { "epoch": 0.9921659494648571, "grad_norm": 0.38751853226924277, "learning_rate": 5.9439649572463595e-06, "loss": 0.512, "step": 1686 }, { "epoch": 0.9927544227444922, "grad_norm": 0.4894632413484775, "learning_rate": 5.938923452307313e-06, "loss": 0.566, "step": 1687 }, { "epoch": 0.9933428960241274, "grad_norm": 0.3686789929296014, "learning_rate": 5.933880957705932e-06, "loss": 0.4404, "step": 1688 }, { "epoch": 0.9939313693037626, "grad_norm": 0.3743257113281994, "learning_rate": 5.928837478757206e-06, "loss": 0.4926, "step": 1689 }, { "epoch": 0.9945198425833977, "grad_norm": 0.4092715128108068, "learning_rate": 5.92379302077716e-06, "loss": 0.5321, "step": 1690 }, { "epoch": 0.9951083158630328, "grad_norm": 0.39806938266832914, "learning_rate": 5.918747589082853e-06, "loss": 0.45, "step": 1691 }, { "epoch": 0.995696789142668, "grad_norm": 0.4254317916459419, "learning_rate": 5.913701188992367e-06, "loss": 0.4736, "step": 1692 }, { "epoch": 0.9962852624223031, "grad_norm": 0.42849288174905953, "learning_rate": 5.908653825824808e-06, "loss": 0.5033, "step": 1693 }, { "epoch": 0.9968737357019383, "grad_norm": 0.3269925716431061, "learning_rate": 5.903605504900296e-06, "loss": 0.4112, "step": 1694 }, { "epoch": 0.9974622089815735, "grad_norm": 0.41779142198751673, "learning_rate": 5.898556231539959e-06, "loss": 0.5063, "step": 1695 }, { "epoch": 0.9980506822612085, "grad_norm": 0.3603579163480973, "learning_rate": 5.893506011065932e-06, "loss": 0.4888, "step": 1696 }, { "epoch": 0.9986391555408437, "grad_norm": 0.37258304789921276, "learning_rate": 5.888454848801345e-06, "loss": 0.4456, "step": 1697 }, { "epoch": 0.9992276288204789, "grad_norm": 0.3770127126442184, "learning_rate": 5.883402750070321e-06, "loss": 0.4689, "step": 1698 }, { "epoch": 0.9998161021001141, "grad_norm": 0.3941010189766722, "learning_rate": 5.878349720197974e-06, "loss": 0.4949, "step": 1699 }, { "epoch": 1.0, "grad_norm": 0.9566625987605052, "learning_rate": 5.8732957645103946e-06, "loss": 0.5005, "step": 1700 }, { "epoch": 1.000588473279635, "grad_norm": 0.3658212516250792, "learning_rate": 5.8682408883346535e-06, "loss": 0.4736, "step": 1701 }, { "epoch": 1.0011769465592704, "grad_norm": 0.3322445685928292, "learning_rate": 5.863185096998787e-06, "loss": 0.4455, "step": 1702 }, { "epoch": 1.0017654198389054, "grad_norm": 0.35688646995011963, "learning_rate": 5.8581283958317995e-06, "loss": 0.4596, "step": 1703 }, { "epoch": 1.0023538931185405, "grad_norm": 0.3490487208343576, "learning_rate": 5.853070790163652e-06, "loss": 0.4686, "step": 1704 }, { "epoch": 1.0029423663981758, "grad_norm": 0.32748917238474445, "learning_rate": 5.848012285325264e-06, "loss": 0.4254, "step": 1705 }, { "epoch": 1.0035308396778109, "grad_norm": 0.33544767397255304, "learning_rate": 5.842952886648496e-06, "loss": 0.3928, "step": 1706 }, { "epoch": 1.004119312957446, "grad_norm": 0.3402876865391869, "learning_rate": 5.837892599466154e-06, "loss": 0.449, "step": 1707 }, { "epoch": 1.0047077862370812, "grad_norm": 0.2931754588747092, "learning_rate": 5.832831429111981e-06, "loss": 0.353, "step": 1708 }, { "epoch": 1.0052962595167163, "grad_norm": 0.3688045499722168, "learning_rate": 5.82776938092065e-06, "loss": 0.4881, "step": 1709 }, { "epoch": 1.0058847327963514, "grad_norm": 0.3893964785078145, "learning_rate": 5.822706460227758e-06, "loss": 0.476, "step": 1710 }, { "epoch": 1.0064732060759867, "grad_norm": 0.3539367847651312, "learning_rate": 5.817642672369825e-06, "loss": 0.4392, "step": 1711 }, { "epoch": 1.0070616793556217, "grad_norm": 0.32731079566456145, "learning_rate": 5.812578022684282e-06, "loss": 0.437, "step": 1712 }, { "epoch": 1.0076501526352568, "grad_norm": 0.33058570811508436, "learning_rate": 5.807512516509468e-06, "loss": 0.4585, "step": 1713 }, { "epoch": 1.008238625914892, "grad_norm": 0.3355819872258361, "learning_rate": 5.8024461591846294e-06, "loss": 0.4083, "step": 1714 }, { "epoch": 1.0088270991945272, "grad_norm": 0.35101508000697085, "learning_rate": 5.797378956049905e-06, "loss": 0.4398, "step": 1715 }, { "epoch": 1.0094155724741622, "grad_norm": 0.3365351733136168, "learning_rate": 5.7923109124463264e-06, "loss": 0.4176, "step": 1716 }, { "epoch": 1.0100040457537975, "grad_norm": 0.38810268727657393, "learning_rate": 5.787242033715809e-06, "loss": 0.5003, "step": 1717 }, { "epoch": 1.0105925190334326, "grad_norm": 0.34764905827066417, "learning_rate": 5.782172325201155e-06, "loss": 0.4339, "step": 1718 }, { "epoch": 1.0111809923130677, "grad_norm": 0.34949719050529726, "learning_rate": 5.777101792246036e-06, "loss": 0.4236, "step": 1719 }, { "epoch": 1.011769465592703, "grad_norm": 0.3327169636935656, "learning_rate": 5.7720304401949915e-06, "loss": 0.4297, "step": 1720 }, { "epoch": 1.012357938872338, "grad_norm": 0.3394843555099991, "learning_rate": 5.766958274393428e-06, "loss": 0.3755, "step": 1721 }, { "epoch": 1.0129464121519731, "grad_norm": 0.3233976008411781, "learning_rate": 5.761885300187609e-06, "loss": 0.3942, "step": 1722 }, { "epoch": 1.0135348854316084, "grad_norm": 0.38059892452153526, "learning_rate": 5.7568115229246455e-06, "loss": 0.4929, "step": 1723 }, { "epoch": 1.0141233587112435, "grad_norm": 0.36384346381344235, "learning_rate": 5.7517369479525044e-06, "loss": 0.5352, "step": 1724 }, { "epoch": 1.0147118319908786, "grad_norm": 0.3281448935855812, "learning_rate": 5.746661580619985e-06, "loss": 0.3896, "step": 1725 }, { "epoch": 1.0153003052705138, "grad_norm": 0.3553478661449535, "learning_rate": 5.741585426276724e-06, "loss": 0.4634, "step": 1726 }, { "epoch": 1.015888778550149, "grad_norm": 0.33132926741507907, "learning_rate": 5.736508490273189e-06, "loss": 0.4128, "step": 1727 }, { "epoch": 1.0164772518297842, "grad_norm": 0.41921374712080467, "learning_rate": 5.731430777960669e-06, "loss": 0.5021, "step": 1728 }, { "epoch": 1.0170657251094193, "grad_norm": 0.35276133537098436, "learning_rate": 5.726352294691275e-06, "loss": 0.4556, "step": 1729 }, { "epoch": 1.0176541983890544, "grad_norm": 0.3616357833210637, "learning_rate": 5.72127304581793e-06, "loss": 0.495, "step": 1730 }, { "epoch": 1.0182426716686896, "grad_norm": 0.3850902291571627, "learning_rate": 5.716193036694359e-06, "loss": 0.4292, "step": 1731 }, { "epoch": 1.0188311449483247, "grad_norm": 0.3230993527361058, "learning_rate": 5.711112272675092e-06, "loss": 0.4143, "step": 1732 }, { "epoch": 1.0194196182279598, "grad_norm": 0.3207069053899187, "learning_rate": 5.706030759115458e-06, "loss": 0.389, "step": 1733 }, { "epoch": 1.020008091507595, "grad_norm": 0.35081684220391257, "learning_rate": 5.70094850137157e-06, "loss": 0.4537, "step": 1734 }, { "epoch": 1.0205965647872302, "grad_norm": 0.3353133251803993, "learning_rate": 5.695865504800328e-06, "loss": 0.4212, "step": 1735 }, { "epoch": 1.0211850380668652, "grad_norm": 0.38638335643395855, "learning_rate": 5.690781774759412e-06, "loss": 0.429, "step": 1736 }, { "epoch": 1.0217735113465005, "grad_norm": 0.3730433585222727, "learning_rate": 5.685697316607274e-06, "loss": 0.5156, "step": 1737 }, { "epoch": 1.0223619846261356, "grad_norm": 0.36405763249626105, "learning_rate": 5.68061213570313e-06, "loss": 0.4895, "step": 1738 }, { "epoch": 1.0229504579057707, "grad_norm": 0.3937555636326131, "learning_rate": 5.675526237406965e-06, "loss": 0.5052, "step": 1739 }, { "epoch": 1.023538931185406, "grad_norm": 0.39084345815392835, "learning_rate": 5.670439627079517e-06, "loss": 0.484, "step": 1740 }, { "epoch": 1.024127404465041, "grad_norm": 0.37214605788021266, "learning_rate": 5.66535231008227e-06, "loss": 0.4831, "step": 1741 }, { "epoch": 1.024715877744676, "grad_norm": 0.37164154889313644, "learning_rate": 5.660264291777458e-06, "loss": 0.4164, "step": 1742 }, { "epoch": 1.0253043510243114, "grad_norm": 0.36664514926396297, "learning_rate": 5.655175577528053e-06, "loss": 0.4557, "step": 1743 }, { "epoch": 1.0258928243039465, "grad_norm": 0.35570955580879443, "learning_rate": 5.650086172697762e-06, "loss": 0.4526, "step": 1744 }, { "epoch": 1.0264812975835815, "grad_norm": 0.361054776500795, "learning_rate": 5.644996082651018e-06, "loss": 0.4332, "step": 1745 }, { "epoch": 1.0270697708632168, "grad_norm": 0.31799599988352195, "learning_rate": 5.639905312752975e-06, "loss": 0.4119, "step": 1746 }, { "epoch": 1.027658244142852, "grad_norm": 0.3744016332476954, "learning_rate": 5.634813868369504e-06, "loss": 0.4577, "step": 1747 }, { "epoch": 1.028246717422487, "grad_norm": 0.36141814173228826, "learning_rate": 5.629721754867192e-06, "loss": 0.5003, "step": 1748 }, { "epoch": 1.0288351907021223, "grad_norm": 0.36972073806677896, "learning_rate": 5.6246289776133246e-06, "loss": 0.4257, "step": 1749 }, { "epoch": 1.0294236639817573, "grad_norm": 0.362203839798034, "learning_rate": 5.61953554197589e-06, "loss": 0.4636, "step": 1750 }, { "epoch": 1.0300121372613924, "grad_norm": 0.33959697883542883, "learning_rate": 5.614441453323571e-06, "loss": 0.3967, "step": 1751 }, { "epoch": 1.0306006105410277, "grad_norm": 0.31713252869221814, "learning_rate": 5.609346717025738e-06, "loss": 0.3624, "step": 1752 }, { "epoch": 1.0311890838206628, "grad_norm": 0.3412222541246552, "learning_rate": 5.604251338452444e-06, "loss": 0.4261, "step": 1753 }, { "epoch": 1.0317775571002978, "grad_norm": 0.33766887804158086, "learning_rate": 5.5991553229744166e-06, "loss": 0.4243, "step": 1754 }, { "epoch": 1.0323660303799331, "grad_norm": 0.3895331650450467, "learning_rate": 5.59405867596306e-06, "loss": 0.4409, "step": 1755 }, { "epoch": 1.0329545036595682, "grad_norm": 0.3998345847042613, "learning_rate": 5.588961402790439e-06, "loss": 0.4802, "step": 1756 }, { "epoch": 1.0335429769392033, "grad_norm": 0.3314568313197255, "learning_rate": 5.583863508829281e-06, "loss": 0.4016, "step": 1757 }, { "epoch": 1.0341314502188386, "grad_norm": 0.33476035737516585, "learning_rate": 5.578764999452969e-06, "loss": 0.4257, "step": 1758 }, { "epoch": 1.0347199234984736, "grad_norm": 0.3251991334176359, "learning_rate": 5.573665880035532e-06, "loss": 0.4066, "step": 1759 }, { "epoch": 1.0353083967781087, "grad_norm": 0.3708300789363363, "learning_rate": 5.5685661559516395e-06, "loss": 0.5436, "step": 1760 }, { "epoch": 1.035896870057744, "grad_norm": 0.33586439239411636, "learning_rate": 5.5634658325766066e-06, "loss": 0.4251, "step": 1761 }, { "epoch": 1.036485343337379, "grad_norm": 0.29151148627540285, "learning_rate": 5.558364915286372e-06, "loss": 0.3502, "step": 1762 }, { "epoch": 1.0370738166170141, "grad_norm": 0.3802216890446029, "learning_rate": 5.553263409457504e-06, "loss": 0.4448, "step": 1763 }, { "epoch": 1.0376622898966494, "grad_norm": 0.3729849430779352, "learning_rate": 5.548161320467194e-06, "loss": 0.4469, "step": 1764 }, { "epoch": 1.0382507631762845, "grad_norm": 0.3825829334461557, "learning_rate": 5.543058653693241e-06, "loss": 0.4065, "step": 1765 }, { "epoch": 1.0388392364559196, "grad_norm": 0.36537505011907284, "learning_rate": 5.537955414514058e-06, "loss": 0.4488, "step": 1766 }, { "epoch": 1.0394277097355549, "grad_norm": 0.35266228810303585, "learning_rate": 5.532851608308661e-06, "loss": 0.4163, "step": 1767 }, { "epoch": 1.04001618301519, "grad_norm": 0.37295339480602774, "learning_rate": 5.527747240456663e-06, "loss": 0.4273, "step": 1768 }, { "epoch": 1.040604656294825, "grad_norm": 0.38617626842920366, "learning_rate": 5.522642316338268e-06, "loss": 0.4296, "step": 1769 }, { "epoch": 1.0411931295744603, "grad_norm": 0.39362160567837456, "learning_rate": 5.517536841334267e-06, "loss": 0.463, "step": 1770 }, { "epoch": 1.0417816028540954, "grad_norm": 0.3599164825943304, "learning_rate": 5.512430820826035e-06, "loss": 0.4419, "step": 1771 }, { "epoch": 1.0423700761337304, "grad_norm": 0.3216613398868511, "learning_rate": 5.507324260195516e-06, "loss": 0.4224, "step": 1772 }, { "epoch": 1.0429585494133657, "grad_norm": 0.32037027280246716, "learning_rate": 5.5022171648252265e-06, "loss": 0.3901, "step": 1773 }, { "epoch": 1.0435470226930008, "grad_norm": 0.34077668649227366, "learning_rate": 5.497109540098246e-06, "loss": 0.4229, "step": 1774 }, { "epoch": 1.0441354959726359, "grad_norm": 0.3412877395947302, "learning_rate": 5.492001391398214e-06, "loss": 0.4155, "step": 1775 }, { "epoch": 1.0447239692522712, "grad_norm": 0.3590989906666583, "learning_rate": 5.486892724109319e-06, "loss": 0.4183, "step": 1776 }, { "epoch": 1.0453124425319062, "grad_norm": 0.36806298125196923, "learning_rate": 5.4817835436163e-06, "loss": 0.4354, "step": 1777 }, { "epoch": 1.0459009158115413, "grad_norm": 0.34900546969758434, "learning_rate": 5.4766738553044306e-06, "loss": 0.4824, "step": 1778 }, { "epoch": 1.0464893890911766, "grad_norm": 0.3596094518611233, "learning_rate": 5.4715636645595285e-06, "loss": 0.4166, "step": 1779 }, { "epoch": 1.0470778623708117, "grad_norm": 0.35006544985804056, "learning_rate": 5.466452976767933e-06, "loss": 0.4464, "step": 1780 }, { "epoch": 1.047666335650447, "grad_norm": 0.3400062116084847, "learning_rate": 5.46134179731651e-06, "loss": 0.4199, "step": 1781 }, { "epoch": 1.048254808930082, "grad_norm": 0.3667641586077432, "learning_rate": 5.456230131592648e-06, "loss": 0.4287, "step": 1782 }, { "epoch": 1.0488432822097171, "grad_norm": 0.35654492796666193, "learning_rate": 5.45111798498424e-06, "loss": 0.4428, "step": 1783 }, { "epoch": 1.0494317554893524, "grad_norm": 0.33026374519057444, "learning_rate": 5.446005362879692e-06, "loss": 0.4384, "step": 1784 }, { "epoch": 1.0500202287689875, "grad_norm": 0.37732192217167215, "learning_rate": 5.440892270667909e-06, "loss": 0.4603, "step": 1785 }, { "epoch": 1.0506087020486226, "grad_norm": 0.3227402040123192, "learning_rate": 5.435778713738292e-06, "loss": 0.4605, "step": 1786 }, { "epoch": 1.0511971753282578, "grad_norm": 0.33896364515887617, "learning_rate": 5.430664697480731e-06, "loss": 0.3945, "step": 1787 }, { "epoch": 1.051785648607893, "grad_norm": 0.33975459096651395, "learning_rate": 5.425550227285601e-06, "loss": 0.3658, "step": 1788 }, { "epoch": 1.052374121887528, "grad_norm": 0.38715468664845404, "learning_rate": 5.420435308543757e-06, "loss": 0.4849, "step": 1789 }, { "epoch": 1.0529625951671633, "grad_norm": 0.38901238720112513, "learning_rate": 5.41531994664652e-06, "loss": 0.5129, "step": 1790 }, { "epoch": 1.0535510684467984, "grad_norm": 0.37431896713214413, "learning_rate": 5.41020414698569e-06, "loss": 0.4501, "step": 1791 }, { "epoch": 1.0541395417264334, "grad_norm": 0.35857546045195693, "learning_rate": 5.405087914953515e-06, "loss": 0.4471, "step": 1792 }, { "epoch": 1.0547280150060687, "grad_norm": 0.3678789690679994, "learning_rate": 5.399971255942708e-06, "loss": 0.4286, "step": 1793 }, { "epoch": 1.0553164882857038, "grad_norm": 0.37047952932872147, "learning_rate": 5.394854175346432e-06, "loss": 0.4612, "step": 1794 }, { "epoch": 1.0559049615653389, "grad_norm": 0.35308586923868124, "learning_rate": 5.389736678558288e-06, "loss": 0.4603, "step": 1795 }, { "epoch": 1.0564934348449742, "grad_norm": 0.3524274541159605, "learning_rate": 5.3846187709723195e-06, "loss": 0.4716, "step": 1796 }, { "epoch": 1.0570819081246092, "grad_norm": 0.3713377655736763, "learning_rate": 5.379500457983006e-06, "loss": 0.4855, "step": 1797 }, { "epoch": 1.0576703814042443, "grad_norm": 0.3834894727683265, "learning_rate": 5.374381744985248e-06, "loss": 0.4599, "step": 1798 }, { "epoch": 1.0582588546838796, "grad_norm": 0.3095076194693259, "learning_rate": 5.36926263737437e-06, "loss": 0.4026, "step": 1799 }, { "epoch": 1.0588473279635147, "grad_norm": 0.37389285121452126, "learning_rate": 5.364143140546116e-06, "loss": 0.5008, "step": 1800 }, { "epoch": 1.0594358012431497, "grad_norm": 0.38866201508335785, "learning_rate": 5.359023259896638e-06, "loss": 0.4655, "step": 1801 }, { "epoch": 1.060024274522785, "grad_norm": 0.36451265344255984, "learning_rate": 5.3539030008224906e-06, "loss": 0.4547, "step": 1802 }, { "epoch": 1.06061274780242, "grad_norm": 0.3595159504309244, "learning_rate": 5.348782368720627e-06, "loss": 0.5118, "step": 1803 }, { "epoch": 1.0612012210820552, "grad_norm": 0.34838928010668097, "learning_rate": 5.3436613689883975e-06, "loss": 0.4387, "step": 1804 }, { "epoch": 1.0617896943616905, "grad_norm": 0.4027964615585603, "learning_rate": 5.338540007023538e-06, "loss": 0.4362, "step": 1805 }, { "epoch": 1.0623781676413255, "grad_norm": 0.42298420858389424, "learning_rate": 5.333418288224164e-06, "loss": 0.5892, "step": 1806 }, { "epoch": 1.0629666409209606, "grad_norm": 0.3825552657805558, "learning_rate": 5.328296217988771e-06, "loss": 0.4641, "step": 1807 }, { "epoch": 1.063555114200596, "grad_norm": 0.36230566779209655, "learning_rate": 5.323173801716222e-06, "loss": 0.4282, "step": 1808 }, { "epoch": 1.064143587480231, "grad_norm": 0.3747384841218869, "learning_rate": 5.318051044805745e-06, "loss": 0.4797, "step": 1809 }, { "epoch": 1.064732060759866, "grad_norm": 0.36319889782012504, "learning_rate": 5.3129279526569284e-06, "loss": 0.4205, "step": 1810 }, { "epoch": 1.0653205340395013, "grad_norm": 0.397081594985582, "learning_rate": 5.3078045306697154e-06, "loss": 0.4919, "step": 1811 }, { "epoch": 1.0659090073191364, "grad_norm": 0.3503721075904531, "learning_rate": 5.302680784244391e-06, "loss": 0.4277, "step": 1812 }, { "epoch": 1.0664974805987715, "grad_norm": 0.36082936508445945, "learning_rate": 5.297556718781588e-06, "loss": 0.4667, "step": 1813 }, { "epoch": 1.0670859538784068, "grad_norm": 0.3458738959434867, "learning_rate": 5.292432339682273e-06, "loss": 0.4276, "step": 1814 }, { "epoch": 1.0676744271580418, "grad_norm": 0.43367615729419384, "learning_rate": 5.287307652347743e-06, "loss": 0.5473, "step": 1815 }, { "epoch": 1.068262900437677, "grad_norm": 0.3637885166239896, "learning_rate": 5.282182662179623e-06, "loss": 0.388, "step": 1816 }, { "epoch": 1.0688513737173122, "grad_norm": 0.35495979980302966, "learning_rate": 5.27705737457985e-06, "loss": 0.4082, "step": 1817 }, { "epoch": 1.0694398469969473, "grad_norm": 0.3786993121315247, "learning_rate": 5.271931794950683e-06, "loss": 0.4686, "step": 1818 }, { "epoch": 1.0700283202765823, "grad_norm": 0.3239591830115375, "learning_rate": 5.266805928694685e-06, "loss": 0.425, "step": 1819 }, { "epoch": 1.0706167935562176, "grad_norm": 0.4211372895742353, "learning_rate": 5.2616797812147205e-06, "loss": 0.4865, "step": 1820 }, { "epoch": 1.0712052668358527, "grad_norm": 0.35672214305262323, "learning_rate": 5.2565533579139484e-06, "loss": 0.4623, "step": 1821 }, { "epoch": 1.0717937401154878, "grad_norm": 0.34242496963976465, "learning_rate": 5.251426664195824e-06, "loss": 0.4283, "step": 1822 }, { "epoch": 1.072382213395123, "grad_norm": 0.3838146356385725, "learning_rate": 5.246299705464085e-06, "loss": 0.4602, "step": 1823 }, { "epoch": 1.0729706866747581, "grad_norm": 0.3301031640150361, "learning_rate": 5.241172487122746e-06, "loss": 0.4035, "step": 1824 }, { "epoch": 1.0735591599543932, "grad_norm": 0.36372108954387444, "learning_rate": 5.236045014576098e-06, "loss": 0.4964, "step": 1825 }, { "epoch": 1.0741476332340285, "grad_norm": 0.349087960806114, "learning_rate": 5.230917293228699e-06, "loss": 0.4095, "step": 1826 }, { "epoch": 1.0747361065136636, "grad_norm": 0.35247987976055833, "learning_rate": 5.225789328485369e-06, "loss": 0.4753, "step": 1827 }, { "epoch": 1.0753245797932989, "grad_norm": 0.3954806867069935, "learning_rate": 5.220661125751186e-06, "loss": 0.4602, "step": 1828 }, { "epoch": 1.075913053072934, "grad_norm": 0.38303828561726816, "learning_rate": 5.2155326904314795e-06, "loss": 0.448, "step": 1829 }, { "epoch": 1.076501526352569, "grad_norm": 0.3349921022745784, "learning_rate": 5.210404027931818e-06, "loss": 0.4213, "step": 1830 }, { "epoch": 1.077089999632204, "grad_norm": 0.3760182859373173, "learning_rate": 5.205275143658018e-06, "loss": 0.4592, "step": 1831 }, { "epoch": 1.0776784729118394, "grad_norm": 0.38867208345791526, "learning_rate": 5.200146043016123e-06, "loss": 0.4462, "step": 1832 }, { "epoch": 1.0782669461914745, "grad_norm": 0.35786878417565193, "learning_rate": 5.1950167314124085e-06, "loss": 0.4459, "step": 1833 }, { "epoch": 1.0788554194711097, "grad_norm": 0.3432112411166149, "learning_rate": 5.189887214253372e-06, "loss": 0.4073, "step": 1834 }, { "epoch": 1.0794438927507448, "grad_norm": 0.43542022500163613, "learning_rate": 5.184757496945726e-06, "loss": 0.5066, "step": 1835 }, { "epoch": 1.0800323660303799, "grad_norm": 0.36625478032725456, "learning_rate": 5.1796275848963926e-06, "loss": 0.4484, "step": 1836 }, { "epoch": 1.0806208393100152, "grad_norm": 0.380570475248749, "learning_rate": 5.174497483512506e-06, "loss": 0.4906, "step": 1837 }, { "epoch": 1.0812093125896503, "grad_norm": 0.4016217246652954, "learning_rate": 5.169367198201391e-06, "loss": 0.465, "step": 1838 }, { "epoch": 1.0817977858692853, "grad_norm": 0.3617636190858625, "learning_rate": 5.164236734370574e-06, "loss": 0.4506, "step": 1839 }, { "epoch": 1.0823862591489206, "grad_norm": 0.3531449794061846, "learning_rate": 5.1591060974277625e-06, "loss": 0.4685, "step": 1840 }, { "epoch": 1.0829747324285557, "grad_norm": 0.3830955774401539, "learning_rate": 5.153975292780852e-06, "loss": 0.481, "step": 1841 }, { "epoch": 1.0835632057081908, "grad_norm": 0.3721010996852401, "learning_rate": 5.148844325837914e-06, "loss": 0.4622, "step": 1842 }, { "epoch": 1.084151678987826, "grad_norm": 0.3987719013534152, "learning_rate": 5.1437132020071886e-06, "loss": 0.5081, "step": 1843 }, { "epoch": 1.0847401522674611, "grad_norm": 0.4075785017236376, "learning_rate": 5.138581926697083e-06, "loss": 0.4492, "step": 1844 }, { "epoch": 1.0853286255470962, "grad_norm": 0.3771651621273601, "learning_rate": 5.133450505316162e-06, "loss": 0.3993, "step": 1845 }, { "epoch": 1.0859170988267315, "grad_norm": 0.3586540353920735, "learning_rate": 5.12831894327315e-06, "loss": 0.4368, "step": 1846 }, { "epoch": 1.0865055721063666, "grad_norm": 0.3517759075398951, "learning_rate": 5.123187245976912e-06, "loss": 0.4304, "step": 1847 }, { "epoch": 1.0870940453860016, "grad_norm": 0.3399176863145899, "learning_rate": 5.118055418836459e-06, "loss": 0.4391, "step": 1848 }, { "epoch": 1.087682518665637, "grad_norm": 0.3943505242783799, "learning_rate": 5.112923467260941e-06, "loss": 0.4821, "step": 1849 }, { "epoch": 1.088270991945272, "grad_norm": 0.38170731559762633, "learning_rate": 5.107791396659637e-06, "loss": 0.4376, "step": 1850 }, { "epoch": 1.088859465224907, "grad_norm": 0.38628500370450963, "learning_rate": 5.102659212441953e-06, "loss": 0.5155, "step": 1851 }, { "epoch": 1.0894479385045424, "grad_norm": 0.3507454369495054, "learning_rate": 5.09752692001741e-06, "loss": 0.4459, "step": 1852 }, { "epoch": 1.0900364117841774, "grad_norm": 0.4243057793069008, "learning_rate": 5.09239452479565e-06, "loss": 0.5462, "step": 1853 }, { "epoch": 1.0906248850638125, "grad_norm": 0.39156701069091954, "learning_rate": 5.087262032186418e-06, "loss": 0.4508, "step": 1854 }, { "epoch": 1.0912133583434478, "grad_norm": 0.3935824185814078, "learning_rate": 5.0821294475995644e-06, "loss": 0.4184, "step": 1855 }, { "epoch": 1.0918018316230829, "grad_norm": 0.340697892213785, "learning_rate": 5.0769967764450345e-06, "loss": 0.4495, "step": 1856 }, { "epoch": 1.092390304902718, "grad_norm": 0.38854566751096425, "learning_rate": 5.071864024132868e-06, "loss": 0.4315, "step": 1857 }, { "epoch": 1.0929787781823532, "grad_norm": 0.3390382956668191, "learning_rate": 5.0667311960731856e-06, "loss": 0.4406, "step": 1858 }, { "epoch": 1.0935672514619883, "grad_norm": 0.34191353497189636, "learning_rate": 5.061598297676192e-06, "loss": 0.4588, "step": 1859 }, { "epoch": 1.0941557247416234, "grad_norm": 0.3653974278343729, "learning_rate": 5.056465334352166e-06, "loss": 0.452, "step": 1860 }, { "epoch": 1.0947441980212587, "grad_norm": 0.36709521148686036, "learning_rate": 5.05133231151145e-06, "loss": 0.444, "step": 1861 }, { "epoch": 1.0953326713008937, "grad_norm": 0.37681886339730364, "learning_rate": 5.046199234564455e-06, "loss": 0.4599, "step": 1862 }, { "epoch": 1.0959211445805288, "grad_norm": 0.3602591809299864, "learning_rate": 5.041066108921645e-06, "loss": 0.5013, "step": 1863 }, { "epoch": 1.096509617860164, "grad_norm": 0.40190482732599253, "learning_rate": 5.035932939993536e-06, "loss": 0.4327, "step": 1864 }, { "epoch": 1.0970980911397992, "grad_norm": 0.36054189737661757, "learning_rate": 5.030799733190694e-06, "loss": 0.4474, "step": 1865 }, { "epoch": 1.0976865644194342, "grad_norm": 0.38886673092346935, "learning_rate": 5.0256664939237186e-06, "loss": 0.4591, "step": 1866 }, { "epoch": 1.0982750376990695, "grad_norm": 0.34301675545784105, "learning_rate": 5.020533227603245e-06, "loss": 0.3712, "step": 1867 }, { "epoch": 1.0988635109787046, "grad_norm": 0.35588995352899055, "learning_rate": 5.015399939639941e-06, "loss": 0.4588, "step": 1868 }, { "epoch": 1.0994519842583397, "grad_norm": 0.3500468735413223, "learning_rate": 5.010266635444495e-06, "loss": 0.4353, "step": 1869 }, { "epoch": 1.100040457537975, "grad_norm": 0.4609661906653302, "learning_rate": 5.005133320427609e-06, "loss": 0.4149, "step": 1870 }, { "epoch": 1.10062893081761, "grad_norm": 0.3525857437057691, "learning_rate": 5e-06, "loss": 0.4429, "step": 1871 }, { "epoch": 1.1012174040972451, "grad_norm": 0.3278609533322834, "learning_rate": 4.994866679572393e-06, "loss": 0.4228, "step": 1872 }, { "epoch": 1.1018058773768804, "grad_norm": 0.40044557355527777, "learning_rate": 4.989733364555507e-06, "loss": 0.5057, "step": 1873 }, { "epoch": 1.1023943506565155, "grad_norm": 0.3578866584622812, "learning_rate": 4.98460006036006e-06, "loss": 0.4147, "step": 1874 }, { "epoch": 1.1029828239361505, "grad_norm": 0.36656842425054864, "learning_rate": 4.979466772396757e-06, "loss": 0.4085, "step": 1875 }, { "epoch": 1.1035712972157858, "grad_norm": 0.37273018970039573, "learning_rate": 4.974333506076283e-06, "loss": 0.4795, "step": 1876 }, { "epoch": 1.104159770495421, "grad_norm": 0.3816458622104856, "learning_rate": 4.9692002668093075e-06, "loss": 0.4597, "step": 1877 }, { "epoch": 1.104748243775056, "grad_norm": 0.317818390867814, "learning_rate": 4.964067060006465e-06, "loss": 0.3701, "step": 1878 }, { "epoch": 1.1053367170546913, "grad_norm": 0.3647055410533946, "learning_rate": 4.958933891078356e-06, "loss": 0.4382, "step": 1879 }, { "epoch": 1.1059251903343263, "grad_norm": 0.36296731880178357, "learning_rate": 4.953800765435547e-06, "loss": 0.4531, "step": 1880 }, { "epoch": 1.1065136636139616, "grad_norm": 0.3717264394634242, "learning_rate": 4.948667688488552e-06, "loss": 0.4366, "step": 1881 }, { "epoch": 1.1071021368935967, "grad_norm": 0.34948049978943746, "learning_rate": 4.943534665647835e-06, "loss": 0.4199, "step": 1882 }, { "epoch": 1.1076906101732318, "grad_norm": 0.39786973156508915, "learning_rate": 4.9384017023238085e-06, "loss": 0.4284, "step": 1883 }, { "epoch": 1.1082790834528669, "grad_norm": 0.3312095643781963, "learning_rate": 4.933268803926816e-06, "loss": 0.4012, "step": 1884 }, { "epoch": 1.1088675567325021, "grad_norm": 0.33150033841099275, "learning_rate": 4.928135975867134e-06, "loss": 0.4232, "step": 1885 }, { "epoch": 1.1094560300121372, "grad_norm": 0.31683055724708786, "learning_rate": 4.923003223554967e-06, "loss": 0.393, "step": 1886 }, { "epoch": 1.1100445032917725, "grad_norm": 0.37556445815073386, "learning_rate": 4.917870552400438e-06, "loss": 0.4268, "step": 1887 }, { "epoch": 1.1106329765714076, "grad_norm": 0.3182211234995804, "learning_rate": 4.9127379678135825e-06, "loss": 0.4533, "step": 1888 }, { "epoch": 1.1112214498510427, "grad_norm": 0.38108983484799863, "learning_rate": 4.907605475204352e-06, "loss": 0.4931, "step": 1889 }, { "epoch": 1.111809923130678, "grad_norm": 0.3665593246393094, "learning_rate": 4.902473079982592e-06, "loss": 0.4057, "step": 1890 }, { "epoch": 1.112398396410313, "grad_norm": 0.38112032124783707, "learning_rate": 4.8973407875580485e-06, "loss": 0.4353, "step": 1891 }, { "epoch": 1.112986869689948, "grad_norm": 0.386708501077322, "learning_rate": 4.892208603340364e-06, "loss": 0.5098, "step": 1892 }, { "epoch": 1.1135753429695834, "grad_norm": 0.4024293965515955, "learning_rate": 4.887076532739061e-06, "loss": 0.4876, "step": 1893 }, { "epoch": 1.1141638162492185, "grad_norm": 0.34593940909878795, "learning_rate": 4.881944581163543e-06, "loss": 0.4106, "step": 1894 }, { "epoch": 1.1147522895288535, "grad_norm": 0.35146909378963154, "learning_rate": 4.876812754023092e-06, "loss": 0.4495, "step": 1895 }, { "epoch": 1.1153407628084888, "grad_norm": 0.3799103979419781, "learning_rate": 4.871681056726853e-06, "loss": 0.4312, "step": 1896 }, { "epoch": 1.1159292360881239, "grad_norm": 0.34043063198844953, "learning_rate": 4.866549494683839e-06, "loss": 0.3955, "step": 1897 }, { "epoch": 1.116517709367759, "grad_norm": 0.43354169608267884, "learning_rate": 4.861418073302919e-06, "loss": 0.4514, "step": 1898 }, { "epoch": 1.1171061826473943, "grad_norm": 0.3749707603984969, "learning_rate": 4.856286797992814e-06, "loss": 0.4322, "step": 1899 }, { "epoch": 1.1176946559270293, "grad_norm": 0.3631945706994865, "learning_rate": 4.8511556741620876e-06, "loss": 0.4117, "step": 1900 }, { "epoch": 1.1182831292066644, "grad_norm": 0.3719709241039531, "learning_rate": 4.846024707219149e-06, "loss": 0.4489, "step": 1901 }, { "epoch": 1.1188716024862997, "grad_norm": 0.34804073618877807, "learning_rate": 4.84089390257224e-06, "loss": 0.3986, "step": 1902 }, { "epoch": 1.1194600757659348, "grad_norm": 0.3600238810730489, "learning_rate": 4.835763265629428e-06, "loss": 0.4767, "step": 1903 }, { "epoch": 1.1200485490455698, "grad_norm": 0.4071865698165178, "learning_rate": 4.830632801798611e-06, "loss": 0.5072, "step": 1904 }, { "epoch": 1.1206370223252051, "grad_norm": 0.34110190151376457, "learning_rate": 4.825502516487497e-06, "loss": 0.3783, "step": 1905 }, { "epoch": 1.1212254956048402, "grad_norm": 0.37130866514642763, "learning_rate": 4.820372415103608e-06, "loss": 0.4094, "step": 1906 }, { "epoch": 1.1218139688844753, "grad_norm": 0.3434188716628505, "learning_rate": 4.815242503054277e-06, "loss": 0.4756, "step": 1907 }, { "epoch": 1.1224024421641106, "grad_norm": 0.3806286994197065, "learning_rate": 4.810112785746631e-06, "loss": 0.5102, "step": 1908 }, { "epoch": 1.1229909154437456, "grad_norm": 0.38359874977804115, "learning_rate": 4.804983268587593e-06, "loss": 0.5002, "step": 1909 }, { "epoch": 1.1235793887233807, "grad_norm": 0.374456573673944, "learning_rate": 4.799853956983879e-06, "loss": 0.4197, "step": 1910 }, { "epoch": 1.124167862003016, "grad_norm": 0.4061814412889254, "learning_rate": 4.794724856341985e-06, "loss": 0.4751, "step": 1911 }, { "epoch": 1.124756335282651, "grad_norm": 0.34482904968873346, "learning_rate": 4.789595972068184e-06, "loss": 0.401, "step": 1912 }, { "epoch": 1.1253448085622861, "grad_norm": 0.3333012268789964, "learning_rate": 4.784467309568524e-06, "loss": 0.4472, "step": 1913 }, { "epoch": 1.1259332818419214, "grad_norm": 0.35419327298279557, "learning_rate": 4.779338874248815e-06, "loss": 0.3835, "step": 1914 }, { "epoch": 1.1265217551215565, "grad_norm": 0.34993322750428363, "learning_rate": 4.774210671514632e-06, "loss": 0.461, "step": 1915 }, { "epoch": 1.1271102284011916, "grad_norm": 0.3734477016028051, "learning_rate": 4.7690827067713035e-06, "loss": 0.4399, "step": 1916 }, { "epoch": 1.1276987016808269, "grad_norm": 0.3360022604945586, "learning_rate": 4.7639549854239045e-06, "loss": 0.4549, "step": 1917 }, { "epoch": 1.128287174960462, "grad_norm": 0.40635132505130317, "learning_rate": 4.758827512877256e-06, "loss": 0.4698, "step": 1918 }, { "epoch": 1.128875648240097, "grad_norm": 0.6420149556328387, "learning_rate": 4.753700294535916e-06, "loss": 0.4591, "step": 1919 }, { "epoch": 1.1294641215197323, "grad_norm": 0.35923659435662103, "learning_rate": 4.748573335804175e-06, "loss": 0.4373, "step": 1920 }, { "epoch": 1.1300525947993674, "grad_norm": 0.3756858121708275, "learning_rate": 4.7434466420860515e-06, "loss": 0.427, "step": 1921 }, { "epoch": 1.1306410680790024, "grad_norm": 0.32196199099798484, "learning_rate": 4.738320218785281e-06, "loss": 0.4258, "step": 1922 }, { "epoch": 1.1312295413586377, "grad_norm": 0.3716365309270254, "learning_rate": 4.733194071305315e-06, "loss": 0.3934, "step": 1923 }, { "epoch": 1.1318180146382728, "grad_norm": 0.37430482757808814, "learning_rate": 4.728068205049316e-06, "loss": 0.411, "step": 1924 }, { "epoch": 1.1324064879179079, "grad_norm": 0.3606919199369192, "learning_rate": 4.7229426254201504e-06, "loss": 0.4494, "step": 1925 }, { "epoch": 1.1329949611975432, "grad_norm": 0.3257959499125212, "learning_rate": 4.717817337820378e-06, "loss": 0.3839, "step": 1926 }, { "epoch": 1.1335834344771782, "grad_norm": 0.41078535526267773, "learning_rate": 4.712692347652257e-06, "loss": 0.482, "step": 1927 }, { "epoch": 1.1341719077568135, "grad_norm": 0.3936234482049514, "learning_rate": 4.7075676603177285e-06, "loss": 0.5459, "step": 1928 }, { "epoch": 1.1347603810364486, "grad_norm": 0.406257145287673, "learning_rate": 4.702443281218413e-06, "loss": 0.4659, "step": 1929 }, { "epoch": 1.1353488543160837, "grad_norm": 0.4034420179017298, "learning_rate": 4.69731921575561e-06, "loss": 0.4587, "step": 1930 }, { "epoch": 1.1359373275957187, "grad_norm": 0.35050247569285975, "learning_rate": 4.692195469330286e-06, "loss": 0.4096, "step": 1931 }, { "epoch": 1.136525800875354, "grad_norm": 0.3627162324210158, "learning_rate": 4.687072047343071e-06, "loss": 0.4596, "step": 1932 }, { "epoch": 1.1371142741549891, "grad_norm": 0.36107849216762594, "learning_rate": 4.681948955194256e-06, "loss": 0.4251, "step": 1933 }, { "epoch": 1.1377027474346244, "grad_norm": 0.3423375502597773, "learning_rate": 4.676826198283779e-06, "loss": 0.4118, "step": 1934 }, { "epoch": 1.1382912207142595, "grad_norm": 0.32530626896381654, "learning_rate": 4.671703782011229e-06, "loss": 0.4239, "step": 1935 }, { "epoch": 1.1388796939938945, "grad_norm": 0.398181334555799, "learning_rate": 4.666581711775836e-06, "loss": 0.4957, "step": 1936 }, { "epoch": 1.1394681672735296, "grad_norm": 0.4173020145609555, "learning_rate": 4.661459992976463e-06, "loss": 0.4966, "step": 1937 }, { "epoch": 1.140056640553165, "grad_norm": 0.37293857562859284, "learning_rate": 4.6563386310116025e-06, "loss": 0.4308, "step": 1938 }, { "epoch": 1.1406451138328, "grad_norm": 0.3375353410025068, "learning_rate": 4.651217631279374e-06, "loss": 0.3904, "step": 1939 }, { "epoch": 1.1412335871124353, "grad_norm": 0.33544108249343463, "learning_rate": 4.646096999177511e-06, "loss": 0.381, "step": 1940 }, { "epoch": 1.1418220603920703, "grad_norm": 0.3550270503600891, "learning_rate": 4.640976740103363e-06, "loss": 0.4164, "step": 1941 }, { "epoch": 1.1424105336717054, "grad_norm": 0.3881922129787188, "learning_rate": 4.635856859453884e-06, "loss": 0.4716, "step": 1942 }, { "epoch": 1.1429990069513407, "grad_norm": 0.3809775219874404, "learning_rate": 4.630737362625631e-06, "loss": 0.4546, "step": 1943 }, { "epoch": 1.1435874802309758, "grad_norm": 0.4050718520195097, "learning_rate": 4.625618255014754e-06, "loss": 0.4868, "step": 1944 }, { "epoch": 1.1441759535106109, "grad_norm": 0.4065167330590048, "learning_rate": 4.620499542016996e-06, "loss": 0.535, "step": 1945 }, { "epoch": 1.1447644267902461, "grad_norm": 0.34462202789145396, "learning_rate": 4.615381229027681e-06, "loss": 0.4506, "step": 1946 }, { "epoch": 1.1453529000698812, "grad_norm": 0.3679420958331272, "learning_rate": 4.610263321441713e-06, "loss": 0.4577, "step": 1947 }, { "epoch": 1.1459413733495163, "grad_norm": 0.3811227084499074, "learning_rate": 4.605145824653569e-06, "loss": 0.4909, "step": 1948 }, { "epoch": 1.1465298466291516, "grad_norm": 0.3682939304416824, "learning_rate": 4.6000287440572925e-06, "loss": 0.4172, "step": 1949 }, { "epoch": 1.1471183199087867, "grad_norm": 0.3814263424229679, "learning_rate": 4.594912085046486e-06, "loss": 0.437, "step": 1950 }, { "epoch": 1.1477067931884217, "grad_norm": 0.4540307270929214, "learning_rate": 4.589795853014313e-06, "loss": 0.4063, "step": 1951 }, { "epoch": 1.148295266468057, "grad_norm": 0.38578737763903953, "learning_rate": 4.584680053353481e-06, "loss": 0.4882, "step": 1952 }, { "epoch": 1.148883739747692, "grad_norm": 0.34321389862865276, "learning_rate": 4.579564691456245e-06, "loss": 0.4483, "step": 1953 }, { "epoch": 1.1494722130273272, "grad_norm": 0.3748543618529673, "learning_rate": 4.5744497727144e-06, "loss": 0.432, "step": 1954 }, { "epoch": 1.1500606863069625, "grad_norm": 0.34570446614003986, "learning_rate": 4.569335302519271e-06, "loss": 0.4257, "step": 1955 }, { "epoch": 1.1506491595865975, "grad_norm": 0.3647518156434874, "learning_rate": 4.564221286261709e-06, "loss": 0.4507, "step": 1956 }, { "epoch": 1.1512376328662326, "grad_norm": 0.40355934969525226, "learning_rate": 4.5591077293320925e-06, "loss": 0.4707, "step": 1957 }, { "epoch": 1.151826106145868, "grad_norm": 0.35927915665898513, "learning_rate": 4.55399463712031e-06, "loss": 0.4376, "step": 1958 }, { "epoch": 1.152414579425503, "grad_norm": 0.4074605150645239, "learning_rate": 4.548882015015761e-06, "loss": 0.5159, "step": 1959 }, { "epoch": 1.153003052705138, "grad_norm": 0.45193097006592553, "learning_rate": 4.543769868407354e-06, "loss": 0.4397, "step": 1960 }, { "epoch": 1.1535915259847733, "grad_norm": 0.3545581982100278, "learning_rate": 4.53865820268349e-06, "loss": 0.3827, "step": 1961 }, { "epoch": 1.1541799992644084, "grad_norm": 0.3382945720733097, "learning_rate": 4.533547023232068e-06, "loss": 0.3885, "step": 1962 }, { "epoch": 1.1547684725440435, "grad_norm": 0.35073171719895585, "learning_rate": 4.528436335440473e-06, "loss": 0.3665, "step": 1963 }, { "epoch": 1.1553569458236788, "grad_norm": 0.364224714904575, "learning_rate": 4.52332614469557e-06, "loss": 0.4716, "step": 1964 }, { "epoch": 1.1559454191033138, "grad_norm": 0.3942229739910492, "learning_rate": 4.5182164563837015e-06, "loss": 0.4663, "step": 1965 }, { "epoch": 1.156533892382949, "grad_norm": 0.3693210981413879, "learning_rate": 4.513107275890682e-06, "loss": 0.4106, "step": 1966 }, { "epoch": 1.1571223656625842, "grad_norm": 0.3335771483429536, "learning_rate": 4.507998608601787e-06, "loss": 0.3895, "step": 1967 }, { "epoch": 1.1577108389422193, "grad_norm": 0.3579367346032733, "learning_rate": 4.502890459901755e-06, "loss": 0.4293, "step": 1968 }, { "epoch": 1.1582993122218543, "grad_norm": 0.4090934781806266, "learning_rate": 4.497782835174775e-06, "loss": 0.4907, "step": 1969 }, { "epoch": 1.1588877855014896, "grad_norm": 0.40922198505637664, "learning_rate": 4.492675739804486e-06, "loss": 0.4578, "step": 1970 }, { "epoch": 1.1594762587811247, "grad_norm": 0.3775017257711927, "learning_rate": 4.4875691791739655e-06, "loss": 0.4413, "step": 1971 }, { "epoch": 1.1600647320607598, "grad_norm": 0.40620800016816305, "learning_rate": 4.4824631586657334e-06, "loss": 0.4343, "step": 1972 }, { "epoch": 1.160653205340395, "grad_norm": 0.3715850692743274, "learning_rate": 4.477357683661734e-06, "loss": 0.3978, "step": 1973 }, { "epoch": 1.1612416786200301, "grad_norm": 0.3998142463624254, "learning_rate": 4.472252759543338e-06, "loss": 0.4695, "step": 1974 }, { "epoch": 1.1618301518996652, "grad_norm": 0.30879984067765814, "learning_rate": 4.4671483916913405e-06, "loss": 0.3752, "step": 1975 }, { "epoch": 1.1624186251793005, "grad_norm": 0.36062935560218745, "learning_rate": 4.462044585485944e-06, "loss": 0.4585, "step": 1976 }, { "epoch": 1.1630070984589356, "grad_norm": 0.5200197165804196, "learning_rate": 4.456941346306761e-06, "loss": 0.5276, "step": 1977 }, { "epoch": 1.1635955717385706, "grad_norm": 0.3814056246032228, "learning_rate": 4.451838679532808e-06, "loss": 0.4391, "step": 1978 }, { "epoch": 1.164184045018206, "grad_norm": 0.33459461569922583, "learning_rate": 4.446736590542497e-06, "loss": 0.435, "step": 1979 }, { "epoch": 1.164772518297841, "grad_norm": 0.33745677658857925, "learning_rate": 4.441635084713629e-06, "loss": 0.432, "step": 1980 }, { "epoch": 1.1653609915774763, "grad_norm": 0.35889208254335914, "learning_rate": 4.436534167423395e-06, "loss": 0.4787, "step": 1981 }, { "epoch": 1.1659494648571114, "grad_norm": 0.32645361898598846, "learning_rate": 4.431433844048361e-06, "loss": 0.3711, "step": 1982 }, { "epoch": 1.1665379381367464, "grad_norm": 0.38335147172172684, "learning_rate": 4.42633411996447e-06, "loss": 0.4347, "step": 1983 }, { "epoch": 1.1671264114163815, "grad_norm": 0.3465330074148761, "learning_rate": 4.4212350005470325e-06, "loss": 0.4666, "step": 1984 }, { "epoch": 1.1677148846960168, "grad_norm": 0.36872549004413946, "learning_rate": 4.41613649117072e-06, "loss": 0.4217, "step": 1985 }, { "epoch": 1.1683033579756519, "grad_norm": 0.3697739222384773, "learning_rate": 4.411038597209562e-06, "loss": 0.3976, "step": 1986 }, { "epoch": 1.1688918312552872, "grad_norm": 0.36154997961203494, "learning_rate": 4.405941324036942e-06, "loss": 0.4577, "step": 1987 }, { "epoch": 1.1694803045349222, "grad_norm": 0.36592979625260474, "learning_rate": 4.400844677025585e-06, "loss": 0.4581, "step": 1988 }, { "epoch": 1.1700687778145573, "grad_norm": 0.3690724755183014, "learning_rate": 4.395748661547558e-06, "loss": 0.4342, "step": 1989 }, { "epoch": 1.1706572510941924, "grad_norm": 0.32404821041598164, "learning_rate": 4.390653282974264e-06, "loss": 0.3804, "step": 1990 }, { "epoch": 1.1712457243738277, "grad_norm": 0.3617121477327275, "learning_rate": 4.3855585466764305e-06, "loss": 0.4592, "step": 1991 }, { "epoch": 1.1718341976534628, "grad_norm": 0.3858636136626514, "learning_rate": 4.380464458024113e-06, "loss": 0.5001, "step": 1992 }, { "epoch": 1.172422670933098, "grad_norm": 0.34727242400617103, "learning_rate": 4.375371022386677e-06, "loss": 0.4182, "step": 1993 }, { "epoch": 1.1730111442127331, "grad_norm": 0.37975471524726406, "learning_rate": 4.370278245132811e-06, "loss": 0.5227, "step": 1994 }, { "epoch": 1.1735996174923682, "grad_norm": 0.3545449243036624, "learning_rate": 4.365186131630498e-06, "loss": 0.4091, "step": 1995 }, { "epoch": 1.1741880907720035, "grad_norm": 0.3220981395843049, "learning_rate": 4.360094687247028e-06, "loss": 0.3872, "step": 1996 }, { "epoch": 1.1747765640516385, "grad_norm": 0.35859873389775093, "learning_rate": 4.355003917348985e-06, "loss": 0.4575, "step": 1997 }, { "epoch": 1.1753650373312736, "grad_norm": 0.3791669725675476, "learning_rate": 4.34991382730224e-06, "loss": 0.4801, "step": 1998 }, { "epoch": 1.175953510610909, "grad_norm": 0.37030049994808745, "learning_rate": 4.344824422471948e-06, "loss": 0.466, "step": 1999 }, { "epoch": 1.176541983890544, "grad_norm": 0.3397626919244967, "learning_rate": 4.339735708222544e-06, "loss": 0.456, "step": 2000 }, { "epoch": 1.177130457170179, "grad_norm": 0.38074855299170257, "learning_rate": 4.334647689917734e-06, "loss": 0.4303, "step": 2001 }, { "epoch": 1.1777189304498143, "grad_norm": 0.3766825351939364, "learning_rate": 4.329560372920486e-06, "loss": 0.4118, "step": 2002 }, { "epoch": 1.1783074037294494, "grad_norm": 0.3558666785560033, "learning_rate": 4.324473762593037e-06, "loss": 0.4036, "step": 2003 }, { "epoch": 1.1788958770090845, "grad_norm": 0.3856751145439741, "learning_rate": 4.319387864296872e-06, "loss": 0.4884, "step": 2004 }, { "epoch": 1.1794843502887198, "grad_norm": 0.32162146480726383, "learning_rate": 4.314302683392729e-06, "loss": 0.3597, "step": 2005 }, { "epoch": 1.1800728235683549, "grad_norm": 0.37757240845820017, "learning_rate": 4.309218225240591e-06, "loss": 0.4322, "step": 2006 }, { "epoch": 1.18066129684799, "grad_norm": 0.3865013076313436, "learning_rate": 4.304134495199675e-06, "loss": 0.4736, "step": 2007 }, { "epoch": 1.1812497701276252, "grad_norm": 0.3542273877642132, "learning_rate": 4.2990514986284325e-06, "loss": 0.4697, "step": 2008 }, { "epoch": 1.1818382434072603, "grad_norm": 0.33294889663312766, "learning_rate": 4.293969240884545e-06, "loss": 0.4083, "step": 2009 }, { "epoch": 1.1824267166868954, "grad_norm": 0.3642353104627034, "learning_rate": 4.288887727324911e-06, "loss": 0.4848, "step": 2010 }, { "epoch": 1.1830151899665307, "grad_norm": 0.37046585173217933, "learning_rate": 4.283806963305644e-06, "loss": 0.4802, "step": 2011 }, { "epoch": 1.1836036632461657, "grad_norm": 0.34441414129915543, "learning_rate": 4.278726954182073e-06, "loss": 0.4292, "step": 2012 }, { "epoch": 1.1841921365258008, "grad_norm": 0.37171490664900964, "learning_rate": 4.273647705308726e-06, "loss": 0.453, "step": 2013 }, { "epoch": 1.184780609805436, "grad_norm": 0.34748510254822457, "learning_rate": 4.2685692220393324e-06, "loss": 0.4703, "step": 2014 }, { "epoch": 1.1853690830850712, "grad_norm": 0.3644603981401473, "learning_rate": 4.263491509726812e-06, "loss": 0.4251, "step": 2015 }, { "epoch": 1.1859575563647062, "grad_norm": 0.3593573227123771, "learning_rate": 4.258414573723277e-06, "loss": 0.4177, "step": 2016 }, { "epoch": 1.1865460296443415, "grad_norm": 0.36589523099600724, "learning_rate": 4.253338419380016e-06, "loss": 0.4278, "step": 2017 }, { "epoch": 1.1871345029239766, "grad_norm": 0.35667899626363475, "learning_rate": 4.2482630520474955e-06, "loss": 0.5026, "step": 2018 }, { "epoch": 1.1877229762036117, "grad_norm": 0.3213050330373795, "learning_rate": 4.243188477075354e-06, "loss": 0.4013, "step": 2019 }, { "epoch": 1.188311449483247, "grad_norm": 0.3512438675462199, "learning_rate": 4.238114699812394e-06, "loss": 0.4416, "step": 2020 }, { "epoch": 1.188899922762882, "grad_norm": 0.3546582309947171, "learning_rate": 4.233041725606573e-06, "loss": 0.4499, "step": 2021 }, { "epoch": 1.189488396042517, "grad_norm": 0.46100229585958985, "learning_rate": 4.227969559805009e-06, "loss": 0.4628, "step": 2022 }, { "epoch": 1.1900768693221524, "grad_norm": 0.36922376603693413, "learning_rate": 4.222898207753966e-06, "loss": 0.43, "step": 2023 }, { "epoch": 1.1906653426017875, "grad_norm": 0.40019560014870625, "learning_rate": 4.217827674798845e-06, "loss": 0.4401, "step": 2024 }, { "epoch": 1.1912538158814225, "grad_norm": 0.370949280982328, "learning_rate": 4.212757966284191e-06, "loss": 0.4979, "step": 2025 }, { "epoch": 1.1918422891610578, "grad_norm": 0.34922701627883196, "learning_rate": 4.207689087553676e-06, "loss": 0.4095, "step": 2026 }, { "epoch": 1.192430762440693, "grad_norm": 0.36926121855478483, "learning_rate": 4.202621043950096e-06, "loss": 0.4449, "step": 2027 }, { "epoch": 1.1930192357203282, "grad_norm": 0.38823106806191615, "learning_rate": 4.1975538408153705e-06, "loss": 0.5244, "step": 2028 }, { "epoch": 1.1936077089999633, "grad_norm": 0.3587101366981455, "learning_rate": 4.192487483490532e-06, "loss": 0.3876, "step": 2029 }, { "epoch": 1.1941961822795983, "grad_norm": 0.3817494670782773, "learning_rate": 4.187421977315718e-06, "loss": 0.4462, "step": 2030 }, { "epoch": 1.1947846555592334, "grad_norm": 0.3934958381384815, "learning_rate": 4.182357327630175e-06, "loss": 0.4013, "step": 2031 }, { "epoch": 1.1953731288388687, "grad_norm": 0.3548902325333919, "learning_rate": 4.177293539772242e-06, "loss": 0.4132, "step": 2032 }, { "epoch": 1.1959616021185038, "grad_norm": 0.36472793431876516, "learning_rate": 4.17223061907935e-06, "loss": 0.4503, "step": 2033 }, { "epoch": 1.196550075398139, "grad_norm": 0.3779683677941678, "learning_rate": 4.167168570888019e-06, "loss": 0.4319, "step": 2034 }, { "epoch": 1.1971385486777741, "grad_norm": 0.3953038143766201, "learning_rate": 4.162107400533846e-06, "loss": 0.5167, "step": 2035 }, { "epoch": 1.1977270219574092, "grad_norm": 0.35505554922858856, "learning_rate": 4.157047113351504e-06, "loss": 0.456, "step": 2036 }, { "epoch": 1.1983154952370443, "grad_norm": 0.38106626767482954, "learning_rate": 4.151987714674737e-06, "loss": 0.4369, "step": 2037 }, { "epoch": 1.1989039685166796, "grad_norm": 0.35160312274844785, "learning_rate": 4.146929209836348e-06, "loss": 0.3947, "step": 2038 }, { "epoch": 1.1994924417963146, "grad_norm": 0.35527060073468164, "learning_rate": 4.141871604168201e-06, "loss": 0.4036, "step": 2039 }, { "epoch": 1.20008091507595, "grad_norm": 0.33242653388502585, "learning_rate": 4.136814903001215e-06, "loss": 0.4073, "step": 2040 }, { "epoch": 1.200669388355585, "grad_norm": 0.34935640807900453, "learning_rate": 4.131759111665349e-06, "loss": 0.4438, "step": 2041 }, { "epoch": 1.20125786163522, "grad_norm": 0.3691217325312803, "learning_rate": 4.126704235489606e-06, "loss": 0.4292, "step": 2042 }, { "epoch": 1.2018463349148552, "grad_norm": 0.3944122380643103, "learning_rate": 4.121650279802027e-06, "loss": 0.4507, "step": 2043 }, { "epoch": 1.2024348081944904, "grad_norm": 0.3545522934578816, "learning_rate": 4.11659724992968e-06, "loss": 0.4318, "step": 2044 }, { "epoch": 1.2030232814741255, "grad_norm": 0.39156102454042696, "learning_rate": 4.111545151198657e-06, "loss": 0.4802, "step": 2045 }, { "epoch": 1.2036117547537608, "grad_norm": 0.36334407695416493, "learning_rate": 4.106493988934069e-06, "loss": 0.4462, "step": 2046 }, { "epoch": 1.2042002280333959, "grad_norm": 0.7312951338685294, "learning_rate": 4.101443768460042e-06, "loss": 0.5039, "step": 2047 }, { "epoch": 1.204788701313031, "grad_norm": 0.3255499749018088, "learning_rate": 4.0963944950997046e-06, "loss": 0.4392, "step": 2048 }, { "epoch": 1.2053771745926662, "grad_norm": 0.43653388662724346, "learning_rate": 4.091346174175193e-06, "loss": 0.4593, "step": 2049 }, { "epoch": 1.2059656478723013, "grad_norm": 0.3561672914636117, "learning_rate": 4.0862988110076345e-06, "loss": 0.4817, "step": 2050 }, { "epoch": 1.2065541211519364, "grad_norm": 0.3439131497521877, "learning_rate": 4.081252410917148e-06, "loss": 0.4248, "step": 2051 }, { "epoch": 1.2071425944315717, "grad_norm": 0.4212897616747787, "learning_rate": 4.076206979222842e-06, "loss": 0.5029, "step": 2052 }, { "epoch": 1.2077310677112068, "grad_norm": 0.3344711653583945, "learning_rate": 4.071162521242796e-06, "loss": 0.4391, "step": 2053 }, { "epoch": 1.2083195409908418, "grad_norm": 0.3993222607091772, "learning_rate": 4.06611904229407e-06, "loss": 0.4243, "step": 2054 }, { "epoch": 1.2089080142704771, "grad_norm": 0.3615989506669775, "learning_rate": 4.061076547692689e-06, "loss": 0.4503, "step": 2055 }, { "epoch": 1.2094964875501122, "grad_norm": 0.34859739737794815, "learning_rate": 4.056035042753641e-06, "loss": 0.4629, "step": 2056 }, { "epoch": 1.2100849608297473, "grad_norm": 0.3891581117869114, "learning_rate": 4.050994532790871e-06, "loss": 0.4283, "step": 2057 }, { "epoch": 1.2106734341093826, "grad_norm": 0.3488535482444155, "learning_rate": 4.045955023117276e-06, "loss": 0.4087, "step": 2058 }, { "epoch": 1.2112619073890176, "grad_norm": 0.3429347198605702, "learning_rate": 4.040916519044697e-06, "loss": 0.4154, "step": 2059 }, { "epoch": 1.2118503806686527, "grad_norm": 0.35936812298627174, "learning_rate": 4.035879025883916e-06, "loss": 0.469, "step": 2060 }, { "epoch": 1.212438853948288, "grad_norm": 0.36902174120886216, "learning_rate": 4.03084254894465e-06, "loss": 0.4275, "step": 2061 }, { "epoch": 1.213027327227923, "grad_norm": 0.35703067164309316, "learning_rate": 4.025807093535544e-06, "loss": 0.4104, "step": 2062 }, { "epoch": 1.2136158005075581, "grad_norm": 0.3748335582510641, "learning_rate": 4.020772664964166e-06, "loss": 0.4686, "step": 2063 }, { "epoch": 1.2142042737871934, "grad_norm": 0.3972444485112336, "learning_rate": 4.015739268537002e-06, "loss": 0.4651, "step": 2064 }, { "epoch": 1.2147927470668285, "grad_norm": 0.3532226867912143, "learning_rate": 4.010706909559452e-06, "loss": 0.4641, "step": 2065 }, { "epoch": 1.2153812203464636, "grad_norm": 0.34977964835311204, "learning_rate": 4.005675593335818e-06, "loss": 0.3853, "step": 2066 }, { "epoch": 1.2159696936260989, "grad_norm": 0.38382100006445774, "learning_rate": 4.000645325169307e-06, "loss": 0.4562, "step": 2067 }, { "epoch": 1.216558166905734, "grad_norm": 0.41318465416600764, "learning_rate": 3.995616110362021e-06, "loss": 0.3895, "step": 2068 }, { "epoch": 1.217146640185369, "grad_norm": 0.3569646751635026, "learning_rate": 3.99058795421495e-06, "loss": 0.4088, "step": 2069 }, { "epoch": 1.2177351134650043, "grad_norm": 0.37242965497703373, "learning_rate": 3.9855608620279675e-06, "loss": 0.3926, "step": 2070 }, { "epoch": 1.2183235867446394, "grad_norm": 0.3539080452960208, "learning_rate": 3.980534839099829e-06, "loss": 0.4203, "step": 2071 }, { "epoch": 1.2189120600242744, "grad_norm": 0.326149606319327, "learning_rate": 3.975509890728159e-06, "loss": 0.3832, "step": 2072 }, { "epoch": 1.2195005333039097, "grad_norm": 0.3558229765496754, "learning_rate": 3.970486022209451e-06, "loss": 0.4183, "step": 2073 }, { "epoch": 1.2200890065835448, "grad_norm": 0.34314965166967415, "learning_rate": 3.965463238839063e-06, "loss": 0.4304, "step": 2074 }, { "epoch": 1.2206774798631799, "grad_norm": 0.4820527826942609, "learning_rate": 3.960441545911205e-06, "loss": 0.4963, "step": 2075 }, { "epoch": 1.2212659531428152, "grad_norm": 0.3683862099266719, "learning_rate": 3.9554209487189375e-06, "loss": 0.3809, "step": 2076 }, { "epoch": 1.2218544264224502, "grad_norm": 0.3437576942038157, "learning_rate": 3.950401452554171e-06, "loss": 0.4005, "step": 2077 }, { "epoch": 1.2224428997020853, "grad_norm": 0.33023246675316786, "learning_rate": 3.945383062707652e-06, "loss": 0.3877, "step": 2078 }, { "epoch": 1.2230313729817206, "grad_norm": 0.36263657897231033, "learning_rate": 3.940365784468958e-06, "loss": 0.4506, "step": 2079 }, { "epoch": 1.2236198462613557, "grad_norm": 0.37392817805833667, "learning_rate": 3.935349623126502e-06, "loss": 0.3855, "step": 2080 }, { "epoch": 1.224208319540991, "grad_norm": 0.360800503335886, "learning_rate": 3.930334583967514e-06, "loss": 0.4089, "step": 2081 }, { "epoch": 1.224796792820626, "grad_norm": 0.3563360292457591, "learning_rate": 3.925320672278043e-06, "loss": 0.4207, "step": 2082 }, { "epoch": 1.225385266100261, "grad_norm": 0.38527118322945025, "learning_rate": 3.920307893342949e-06, "loss": 0.4662, "step": 2083 }, { "epoch": 1.2259737393798962, "grad_norm": 0.3825011365658156, "learning_rate": 3.9152962524459e-06, "loss": 0.4393, "step": 2084 }, { "epoch": 1.2265622126595315, "grad_norm": 0.34762972635551254, "learning_rate": 3.910285754869362e-06, "loss": 0.4093, "step": 2085 }, { "epoch": 1.2271506859391665, "grad_norm": 0.34179231006191174, "learning_rate": 3.905276405894598e-06, "loss": 0.4238, "step": 2086 }, { "epoch": 1.2277391592188018, "grad_norm": 0.3388674926507682, "learning_rate": 3.9002682108016585e-06, "loss": 0.4218, "step": 2087 }, { "epoch": 1.228327632498437, "grad_norm": 0.3405546984122325, "learning_rate": 3.8952611748693774e-06, "loss": 0.4149, "step": 2088 }, { "epoch": 1.228916105778072, "grad_norm": 0.4563332527795296, "learning_rate": 3.890255303375371e-06, "loss": 0.529, "step": 2089 }, { "epoch": 1.229504579057707, "grad_norm": 0.40797045722195574, "learning_rate": 3.885250601596023e-06, "loss": 0.528, "step": 2090 }, { "epoch": 1.2300930523373423, "grad_norm": 0.3601333416016302, "learning_rate": 3.8802470748064855e-06, "loss": 0.4192, "step": 2091 }, { "epoch": 1.2306815256169774, "grad_norm": 0.3824928528153359, "learning_rate": 3.875244728280676e-06, "loss": 0.4742, "step": 2092 }, { "epoch": 1.2312699988966127, "grad_norm": 0.3495152512269485, "learning_rate": 3.870243567291263e-06, "loss": 0.4335, "step": 2093 }, { "epoch": 1.2318584721762478, "grad_norm": 0.39120261217571606, "learning_rate": 3.8652435971096666e-06, "loss": 0.483, "step": 2094 }, { "epoch": 1.2324469454558828, "grad_norm": 0.34486398483182007, "learning_rate": 3.860244823006055e-06, "loss": 0.3903, "step": 2095 }, { "epoch": 1.233035418735518, "grad_norm": 0.3529020164480294, "learning_rate": 3.855247250249331e-06, "loss": 0.4663, "step": 2096 }, { "epoch": 1.2336238920151532, "grad_norm": 0.3491396367319095, "learning_rate": 3.850250884107133e-06, "loss": 0.3789, "step": 2097 }, { "epoch": 1.2342123652947883, "grad_norm": 0.40877869225475993, "learning_rate": 3.84525572984583e-06, "loss": 0.4683, "step": 2098 }, { "epoch": 1.2348008385744236, "grad_norm": 0.3913758136098801, "learning_rate": 3.840261792730511e-06, "loss": 0.4886, "step": 2099 }, { "epoch": 1.2353893118540586, "grad_norm": 0.35685093449350025, "learning_rate": 3.83526907802498e-06, "loss": 0.417, "step": 2100 }, { "epoch": 1.2359777851336937, "grad_norm": 0.3660209478473284, "learning_rate": 3.8302775909917585e-06, "loss": 0.4622, "step": 2101 }, { "epoch": 1.236566258413329, "grad_norm": 0.38081794498125876, "learning_rate": 3.82528733689207e-06, "loss": 0.4849, "step": 2102 }, { "epoch": 1.237154731692964, "grad_norm": 0.34923639664967365, "learning_rate": 3.820298320985836e-06, "loss": 0.3916, "step": 2103 }, { "epoch": 1.2377432049725992, "grad_norm": 0.36114083237039146, "learning_rate": 3.8153105485316795e-06, "loss": 0.4309, "step": 2104 }, { "epoch": 1.2383316782522344, "grad_norm": 0.43425148292664295, "learning_rate": 3.8103240247869077e-06, "loss": 0.4712, "step": 2105 }, { "epoch": 1.2389201515318695, "grad_norm": 0.3489526915470937, "learning_rate": 3.8053387550075126e-06, "loss": 0.4278, "step": 2106 }, { "epoch": 1.2395086248115046, "grad_norm": 0.3634558388435432, "learning_rate": 3.8003547444481675e-06, "loss": 0.4559, "step": 2107 }, { "epoch": 1.2400970980911399, "grad_norm": 0.34138702017679096, "learning_rate": 3.795371998362214e-06, "loss": 0.4176, "step": 2108 }, { "epoch": 1.240685571370775, "grad_norm": 0.3319496981095481, "learning_rate": 3.790390522001662e-06, "loss": 0.4231, "step": 2109 }, { "epoch": 1.24127404465041, "grad_norm": 0.3565046023628405, "learning_rate": 3.785410320617187e-06, "loss": 0.4204, "step": 2110 }, { "epoch": 1.2418625179300453, "grad_norm": 0.36724800791612094, "learning_rate": 3.7804313994581143e-06, "loss": 0.4269, "step": 2111 }, { "epoch": 1.2424509912096804, "grad_norm": 0.3539300996467344, "learning_rate": 3.775453763772426e-06, "loss": 0.4285, "step": 2112 }, { "epoch": 1.2430394644893155, "grad_norm": 0.3355278355162667, "learning_rate": 3.770477418806744e-06, "loss": 0.4006, "step": 2113 }, { "epoch": 1.2436279377689508, "grad_norm": 0.3861904373341187, "learning_rate": 3.765502369806334e-06, "loss": 0.4605, "step": 2114 }, { "epoch": 1.2442164110485858, "grad_norm": 0.3686011206803316, "learning_rate": 3.760528622015094e-06, "loss": 0.4937, "step": 2115 }, { "epoch": 1.244804884328221, "grad_norm": 0.36333853901936025, "learning_rate": 3.755556180675547e-06, "loss": 0.4325, "step": 2116 }, { "epoch": 1.2453933576078562, "grad_norm": 0.39530158804520243, "learning_rate": 3.7505850510288455e-06, "loss": 0.4598, "step": 2117 }, { "epoch": 1.2459818308874913, "grad_norm": 0.35860602937481156, "learning_rate": 3.7456152383147555e-06, "loss": 0.4546, "step": 2118 }, { "epoch": 1.2465703041671263, "grad_norm": 0.363730538529918, "learning_rate": 3.7406467477716535e-06, "loss": 0.4636, "step": 2119 }, { "epoch": 1.2471587774467616, "grad_norm": 0.39229692856984366, "learning_rate": 3.7356795846365268e-06, "loss": 0.4675, "step": 2120 }, { "epoch": 1.2477472507263967, "grad_norm": 0.35848869807388, "learning_rate": 3.730713754144961e-06, "loss": 0.4042, "step": 2121 }, { "epoch": 1.2483357240060318, "grad_norm": 0.3830901608656173, "learning_rate": 3.7257492615311346e-06, "loss": 0.4451, "step": 2122 }, { "epoch": 1.248924197285667, "grad_norm": 0.3672054715851043, "learning_rate": 3.720786112027822e-06, "loss": 0.4103, "step": 2123 }, { "epoch": 1.2495126705653021, "grad_norm": 0.3621437614722083, "learning_rate": 3.715824310866376e-06, "loss": 0.4238, "step": 2124 }, { "epoch": 1.2501011438449372, "grad_norm": 0.3655040405176081, "learning_rate": 3.7108638632767314e-06, "loss": 0.4317, "step": 2125 }, { "epoch": 1.2506896171245725, "grad_norm": 0.33157257529879697, "learning_rate": 3.705904774487396e-06, "loss": 0.3929, "step": 2126 }, { "epoch": 1.2512780904042076, "grad_norm": 0.3411465524966507, "learning_rate": 3.7009470497254446e-06, "loss": 0.4031, "step": 2127 }, { "epoch": 1.2518665636838429, "grad_norm": 0.3666618260816964, "learning_rate": 3.6959906942165137e-06, "loss": 0.4692, "step": 2128 }, { "epoch": 1.252455036963478, "grad_norm": 0.36321922260173084, "learning_rate": 3.6910357131847986e-06, "loss": 0.4721, "step": 2129 }, { "epoch": 1.253043510243113, "grad_norm": 0.3619712512442801, "learning_rate": 3.6860821118530445e-06, "loss": 0.4864, "step": 2130 }, { "epoch": 1.253631983522748, "grad_norm": 0.32789738512591293, "learning_rate": 3.68112989544254e-06, "loss": 0.3575, "step": 2131 }, { "epoch": 1.2542204568023834, "grad_norm": 0.33339918975689986, "learning_rate": 3.6761790691731207e-06, "loss": 0.3694, "step": 2132 }, { "epoch": 1.2548089300820184, "grad_norm": 0.39284942260329203, "learning_rate": 3.6712296382631505e-06, "loss": 0.523, "step": 2133 }, { "epoch": 1.2553974033616537, "grad_norm": 0.35700263662269127, "learning_rate": 3.6662816079295236e-06, "loss": 0.4408, "step": 2134 }, { "epoch": 1.2559858766412888, "grad_norm": 0.3671197473413335, "learning_rate": 3.6613349833876607e-06, "loss": 0.5062, "step": 2135 }, { "epoch": 1.2565743499209239, "grad_norm": 0.391958313047766, "learning_rate": 3.6563897698514984e-06, "loss": 0.4355, "step": 2136 }, { "epoch": 1.257162823200559, "grad_norm": 0.3845584183643659, "learning_rate": 3.6514459725334874e-06, "loss": 0.4817, "step": 2137 }, { "epoch": 1.2577512964801942, "grad_norm": 0.3824840216239987, "learning_rate": 3.646503596644583e-06, "loss": 0.5035, "step": 2138 }, { "epoch": 1.2583397697598293, "grad_norm": 0.351032057133313, "learning_rate": 3.6415626473942465e-06, "loss": 0.4465, "step": 2139 }, { "epoch": 1.2589282430394646, "grad_norm": 0.4232657362224137, "learning_rate": 3.6366231299904326e-06, "loss": 0.4341, "step": 2140 }, { "epoch": 1.2595167163190997, "grad_norm": 0.3996146962928062, "learning_rate": 3.6316850496395863e-06, "loss": 0.4932, "step": 2141 }, { "epoch": 1.2601051895987347, "grad_norm": 0.3487597707597974, "learning_rate": 3.6267484115466402e-06, "loss": 0.4302, "step": 2142 }, { "epoch": 1.2606936628783698, "grad_norm": 0.34956100657290307, "learning_rate": 3.6218132209150047e-06, "loss": 0.4712, "step": 2143 }, { "epoch": 1.261282136158005, "grad_norm": 0.30063403569034575, "learning_rate": 3.616879482946565e-06, "loss": 0.3634, "step": 2144 }, { "epoch": 1.2618706094376402, "grad_norm": 0.3889044216656717, "learning_rate": 3.6119472028416776e-06, "loss": 0.4636, "step": 2145 }, { "epoch": 1.2624590827172755, "grad_norm": 0.4328753698712123, "learning_rate": 3.607016385799158e-06, "loss": 0.553, "step": 2146 }, { "epoch": 1.2630475559969105, "grad_norm": 0.39922095948094377, "learning_rate": 3.602087037016281e-06, "loss": 0.4498, "step": 2147 }, { "epoch": 1.2636360292765456, "grad_norm": 0.3620505421469442, "learning_rate": 3.5971591616887776e-06, "loss": 0.4093, "step": 2148 }, { "epoch": 1.2642245025561807, "grad_norm": 0.3446955248085071, "learning_rate": 3.5922327650108203e-06, "loss": 0.4295, "step": 2149 }, { "epoch": 1.264812975835816, "grad_norm": 0.3645163325636404, "learning_rate": 3.587307852175025e-06, "loss": 0.4578, "step": 2150 }, { "epoch": 1.265401449115451, "grad_norm": 0.4002143888486798, "learning_rate": 3.5823844283724464e-06, "loss": 0.4478, "step": 2151 }, { "epoch": 1.2659899223950863, "grad_norm": 0.3979032806611202, "learning_rate": 3.577462498792566e-06, "loss": 0.4551, "step": 2152 }, { "epoch": 1.2665783956747214, "grad_norm": 0.3356620641491436, "learning_rate": 3.5725420686232903e-06, "loss": 0.4024, "step": 2153 }, { "epoch": 1.2671668689543565, "grad_norm": 0.3591179226635225, "learning_rate": 3.5676231430509488e-06, "loss": 0.443, "step": 2154 }, { "epoch": 1.2677553422339916, "grad_norm": 0.3435117224238979, "learning_rate": 3.5627057272602817e-06, "loss": 0.4038, "step": 2155 }, { "epoch": 1.2683438155136268, "grad_norm": 0.3589242719276517, "learning_rate": 3.557789826434439e-06, "loss": 0.4342, "step": 2156 }, { "epoch": 1.268932288793262, "grad_norm": 0.406876471557213, "learning_rate": 3.5528754457549754e-06, "loss": 0.467, "step": 2157 }, { "epoch": 1.2695207620728972, "grad_norm": 0.37549474571107805, "learning_rate": 3.5479625904018407e-06, "loss": 0.4486, "step": 2158 }, { "epoch": 1.2701092353525323, "grad_norm": 0.36193822918447827, "learning_rate": 3.5430512655533774e-06, "loss": 0.3868, "step": 2159 }, { "epoch": 1.2706977086321674, "grad_norm": 0.33932891406723037, "learning_rate": 3.538141476386317e-06, "loss": 0.4163, "step": 2160 }, { "epoch": 1.2712861819118026, "grad_norm": 0.3474904346830683, "learning_rate": 3.5332332280757706e-06, "loss": 0.4663, "step": 2161 }, { "epoch": 1.2718746551914377, "grad_norm": 0.39599170458791716, "learning_rate": 3.528326525795224e-06, "loss": 0.4643, "step": 2162 }, { "epoch": 1.2724631284710728, "grad_norm": 0.3735455353777395, "learning_rate": 3.5234213747165368e-06, "loss": 0.4703, "step": 2163 }, { "epoch": 1.273051601750708, "grad_norm": 0.7131855789101237, "learning_rate": 3.5185177800099314e-06, "loss": 0.4945, "step": 2164 }, { "epoch": 1.2736400750303432, "grad_norm": 0.3614859786567769, "learning_rate": 3.513615746843987e-06, "loss": 0.3971, "step": 2165 }, { "epoch": 1.2742285483099782, "grad_norm": 0.3623596687924315, "learning_rate": 3.508715280385644e-06, "loss": 0.438, "step": 2166 }, { "epoch": 1.2748170215896135, "grad_norm": 0.3368880417874529, "learning_rate": 3.503816385800185e-06, "loss": 0.393, "step": 2167 }, { "epoch": 1.2754054948692486, "grad_norm": 0.34696057620881365, "learning_rate": 3.498919068251237e-06, "loss": 0.4784, "step": 2168 }, { "epoch": 1.2759939681488837, "grad_norm": 0.39488256895121254, "learning_rate": 3.494023332900768e-06, "loss": 0.4692, "step": 2169 }, { "epoch": 1.276582441428519, "grad_norm": 0.33717331849555915, "learning_rate": 3.4891291849090757e-06, "loss": 0.4722, "step": 2170 }, { "epoch": 1.277170914708154, "grad_norm": 0.3651831672494591, "learning_rate": 3.484236629434783e-06, "loss": 0.533, "step": 2171 }, { "epoch": 1.277759387987789, "grad_norm": 0.37527330180965707, "learning_rate": 3.479345671634839e-06, "loss": 0.4014, "step": 2172 }, { "epoch": 1.2783478612674244, "grad_norm": 0.34777919655250766, "learning_rate": 3.474456316664504e-06, "loss": 0.425, "step": 2173 }, { "epoch": 1.2789363345470595, "grad_norm": 0.37807251777203216, "learning_rate": 3.469568569677352e-06, "loss": 0.442, "step": 2174 }, { "epoch": 1.2795248078266948, "grad_norm": 0.32093158381097375, "learning_rate": 3.46468243582526e-06, "loss": 0.3814, "step": 2175 }, { "epoch": 1.2801132811063298, "grad_norm": 0.3904241617480749, "learning_rate": 3.4597979202584077e-06, "loss": 0.4758, "step": 2176 }, { "epoch": 1.280701754385965, "grad_norm": 0.34146423107033774, "learning_rate": 3.4549150281252635e-06, "loss": 0.4482, "step": 2177 }, { "epoch": 1.2812902276656, "grad_norm": 0.34387184399602094, "learning_rate": 3.4500337645725922e-06, "loss": 0.4193, "step": 2178 }, { "epoch": 1.2818787009452353, "grad_norm": 0.3573393587768351, "learning_rate": 3.4451541347454355e-06, "loss": 0.4033, "step": 2179 }, { "epoch": 1.2824671742248703, "grad_norm": 0.3895188838195966, "learning_rate": 3.440276143787116e-06, "loss": 0.4975, "step": 2180 }, { "epoch": 1.2830556475045056, "grad_norm": 0.35124738874873557, "learning_rate": 3.4353997968392295e-06, "loss": 0.4113, "step": 2181 }, { "epoch": 1.2836441207841407, "grad_norm": 0.36084232068185107, "learning_rate": 3.4305250990416373e-06, "loss": 0.4585, "step": 2182 }, { "epoch": 1.2842325940637758, "grad_norm": 0.3121880714013649, "learning_rate": 3.4256520555324613e-06, "loss": 0.3963, "step": 2183 }, { "epoch": 1.2848210673434108, "grad_norm": 0.34750101756720847, "learning_rate": 3.4207806714480845e-06, "loss": 0.4365, "step": 2184 }, { "epoch": 1.2854095406230461, "grad_norm": 0.34684781548504073, "learning_rate": 3.415910951923137e-06, "loss": 0.447, "step": 2185 }, { "epoch": 1.2859980139026812, "grad_norm": 0.3592567095841563, "learning_rate": 3.4110429020904924e-06, "loss": 0.4153, "step": 2186 }, { "epoch": 1.2865864871823165, "grad_norm": 0.3970184887835358, "learning_rate": 3.406176527081271e-06, "loss": 0.506, "step": 2187 }, { "epoch": 1.2871749604619516, "grad_norm": 0.3645877001454868, "learning_rate": 3.4013118320248215e-06, "loss": 0.4271, "step": 2188 }, { "epoch": 1.2877634337415866, "grad_norm": 0.4027043523822163, "learning_rate": 3.3964488220487252e-06, "loss": 0.5198, "step": 2189 }, { "epoch": 1.2883519070212217, "grad_norm": 0.3658363740054663, "learning_rate": 3.3915875022787857e-06, "loss": 0.4767, "step": 2190 }, { "epoch": 1.288940380300857, "grad_norm": 0.3256378003111717, "learning_rate": 3.386727877839027e-06, "loss": 0.4037, "step": 2191 }, { "epoch": 1.289528853580492, "grad_norm": 0.3447808212461055, "learning_rate": 3.3818699538516827e-06, "loss": 0.4043, "step": 2192 }, { "epoch": 1.2901173268601274, "grad_norm": 0.3635978625638613, "learning_rate": 3.3770137354371977e-06, "loss": 0.43, "step": 2193 }, { "epoch": 1.2907058001397624, "grad_norm": 0.3268909922301729, "learning_rate": 3.372159227714218e-06, "loss": 0.3781, "step": 2194 }, { "epoch": 1.2912942734193975, "grad_norm": 0.3266176036922121, "learning_rate": 3.3673064357995844e-06, "loss": 0.3996, "step": 2195 }, { "epoch": 1.2918827466990326, "grad_norm": 0.36153481948730126, "learning_rate": 3.3624553648083323e-06, "loss": 0.4448, "step": 2196 }, { "epoch": 1.2924712199786679, "grad_norm": 0.35709463010564707, "learning_rate": 3.357606019853682e-06, "loss": 0.4339, "step": 2197 }, { "epoch": 1.293059693258303, "grad_norm": 0.3282623120987051, "learning_rate": 3.352758406047033e-06, "loss": 0.3945, "step": 2198 }, { "epoch": 1.2936481665379382, "grad_norm": 0.38027485226596086, "learning_rate": 3.347912528497962e-06, "loss": 0.4832, "step": 2199 }, { "epoch": 1.2942366398175733, "grad_norm": 0.38552199189444397, "learning_rate": 3.343068392314216e-06, "loss": 0.4278, "step": 2200 }, { "epoch": 1.2948251130972084, "grad_norm": 0.34091517415473643, "learning_rate": 3.3382260026017027e-06, "loss": 0.4277, "step": 2201 }, { "epoch": 1.2954135863768435, "grad_norm": 0.34055870863750665, "learning_rate": 3.3333853644644946e-06, "loss": 0.4101, "step": 2202 }, { "epoch": 1.2960020596564787, "grad_norm": 0.4017568136866498, "learning_rate": 3.3285464830048143e-06, "loss": 0.5228, "step": 2203 }, { "epoch": 1.2965905329361138, "grad_norm": 0.3512855086682918, "learning_rate": 3.3237093633230323e-06, "loss": 0.3845, "step": 2204 }, { "epoch": 1.297179006215749, "grad_norm": 0.371151006208715, "learning_rate": 3.3188740105176655e-06, "loss": 0.4055, "step": 2205 }, { "epoch": 1.2977674794953842, "grad_norm": 0.46153976113523254, "learning_rate": 3.3140404296853635e-06, "loss": 0.5101, "step": 2206 }, { "epoch": 1.2983559527750193, "grad_norm": 0.3835400327481173, "learning_rate": 3.3092086259209144e-06, "loss": 0.4554, "step": 2207 }, { "epoch": 1.2989444260546543, "grad_norm": 0.38587783401124, "learning_rate": 3.3043786043172286e-06, "loss": 0.4329, "step": 2208 }, { "epoch": 1.2995328993342896, "grad_norm": 0.3824924201483367, "learning_rate": 3.2995503699653385e-06, "loss": 0.4163, "step": 2209 }, { "epoch": 1.3001213726139247, "grad_norm": 0.4164963671995823, "learning_rate": 3.294723927954395e-06, "loss": 0.5048, "step": 2210 }, { "epoch": 1.30070984589356, "grad_norm": 0.35415305034012323, "learning_rate": 3.289899283371657e-06, "loss": 0.4717, "step": 2211 }, { "epoch": 1.301298319173195, "grad_norm": 0.4096155003309245, "learning_rate": 3.2850764413024895e-06, "loss": 0.456, "step": 2212 }, { "epoch": 1.3018867924528301, "grad_norm": 0.37876409616227563, "learning_rate": 3.2802554068303595e-06, "loss": 0.4132, "step": 2213 }, { "epoch": 1.3024752657324654, "grad_norm": 0.37535208442736034, "learning_rate": 3.275436185036827e-06, "loss": 0.4548, "step": 2214 }, { "epoch": 1.3030637390121005, "grad_norm": 0.37820060845235653, "learning_rate": 3.270618781001541e-06, "loss": 0.4671, "step": 2215 }, { "epoch": 1.3036522122917356, "grad_norm": 0.31489275793224436, "learning_rate": 3.2658031998022368e-06, "loss": 0.3785, "step": 2216 }, { "epoch": 1.3042406855713709, "grad_norm": 0.37597426311194254, "learning_rate": 3.260989446514726e-06, "loss": 0.4337, "step": 2217 }, { "epoch": 1.304829158851006, "grad_norm": 0.34148342932989595, "learning_rate": 3.2561775262128935e-06, "loss": 0.4165, "step": 2218 }, { "epoch": 1.305417632130641, "grad_norm": 0.45055158384595645, "learning_rate": 3.2513674439686945e-06, "loss": 0.4975, "step": 2219 }, { "epoch": 1.3060061054102763, "grad_norm": 0.4440749028235631, "learning_rate": 3.2465592048521446e-06, "loss": 0.5041, "step": 2220 }, { "epoch": 1.3065945786899114, "grad_norm": 0.3915694389103389, "learning_rate": 3.241752813931316e-06, "loss": 0.479, "step": 2221 }, { "epoch": 1.3071830519695464, "grad_norm": 0.38090060963216993, "learning_rate": 3.236948276272337e-06, "loss": 0.474, "step": 2222 }, { "epoch": 1.3077715252491817, "grad_norm": 0.4385914465388644, "learning_rate": 3.2321455969393783e-06, "loss": 0.4612, "step": 2223 }, { "epoch": 1.3083599985288168, "grad_norm": 0.3673317562144963, "learning_rate": 3.2273447809946524e-06, "loss": 0.4358, "step": 2224 }, { "epoch": 1.3089484718084519, "grad_norm": 0.39310763141136545, "learning_rate": 3.22254583349841e-06, "loss": 0.4122, "step": 2225 }, { "epoch": 1.3095369450880872, "grad_norm": 0.4143156589903175, "learning_rate": 3.217748759508931e-06, "loss": 0.5509, "step": 2226 }, { "epoch": 1.3101254183677222, "grad_norm": 0.3702196294690117, "learning_rate": 3.2129535640825176e-06, "loss": 0.4589, "step": 2227 }, { "epoch": 1.3107138916473575, "grad_norm": 0.331125489919597, "learning_rate": 3.2081602522734987e-06, "loss": 0.4254, "step": 2228 }, { "epoch": 1.3113023649269926, "grad_norm": 0.37839835115270326, "learning_rate": 3.2033688291342124e-06, "loss": 0.4429, "step": 2229 }, { "epoch": 1.3118908382066277, "grad_norm": 0.3516910816178103, "learning_rate": 3.1985792997150065e-06, "loss": 0.4944, "step": 2230 }, { "epoch": 1.3124793114862627, "grad_norm": 0.3715023874507549, "learning_rate": 3.1937916690642356e-06, "loss": 0.4547, "step": 2231 }, { "epoch": 1.313067784765898, "grad_norm": 0.3784079483062194, "learning_rate": 3.1890059422282504e-06, "loss": 0.466, "step": 2232 }, { "epoch": 1.313656258045533, "grad_norm": 0.34106952846517635, "learning_rate": 3.1842221242513944e-06, "loss": 0.4417, "step": 2233 }, { "epoch": 1.3142447313251684, "grad_norm": 0.35775928611893915, "learning_rate": 3.1794402201760022e-06, "loss": 0.4043, "step": 2234 }, { "epoch": 1.3148332046048035, "grad_norm": 0.3981900031680666, "learning_rate": 3.1746602350423895e-06, "loss": 0.4909, "step": 2235 }, { "epoch": 1.3154216778844385, "grad_norm": 0.43422873887058205, "learning_rate": 3.1698821738888463e-06, "loss": 0.465, "step": 2236 }, { "epoch": 1.3160101511640736, "grad_norm": 0.3840197024431024, "learning_rate": 3.16510604175164e-06, "loss": 0.4105, "step": 2237 }, { "epoch": 1.316598624443709, "grad_norm": 0.3330046920127202, "learning_rate": 3.1603318436650014e-06, "loss": 0.4211, "step": 2238 }, { "epoch": 1.317187097723344, "grad_norm": 0.37086183365607023, "learning_rate": 3.155559584661123e-06, "loss": 0.4757, "step": 2239 }, { "epoch": 1.3177755710029793, "grad_norm": 0.3727763002500521, "learning_rate": 3.150789269770155e-06, "loss": 0.3897, "step": 2240 }, { "epoch": 1.3183640442826143, "grad_norm": 0.34476202064996464, "learning_rate": 3.1460209040201967e-06, "loss": 0.4202, "step": 2241 }, { "epoch": 1.3189525175622494, "grad_norm": 0.38164838262928497, "learning_rate": 3.141254492437292e-06, "loss": 0.4638, "step": 2242 }, { "epoch": 1.3195409908418845, "grad_norm": 0.37265371090428767, "learning_rate": 3.13649004004543e-06, "loss": 0.4462, "step": 2243 }, { "epoch": 1.3201294641215198, "grad_norm": 0.37428249192749047, "learning_rate": 3.1317275518665296e-06, "loss": 0.4341, "step": 2244 }, { "epoch": 1.3207179374011548, "grad_norm": 0.4356259876243016, "learning_rate": 3.12696703292044e-06, "loss": 0.499, "step": 2245 }, { "epoch": 1.3213064106807901, "grad_norm": 0.3385931612695915, "learning_rate": 3.1222084882249375e-06, "loss": 0.4066, "step": 2246 }, { "epoch": 1.3218948839604252, "grad_norm": 0.3369020918367389, "learning_rate": 3.1174519227957156e-06, "loss": 0.4017, "step": 2247 }, { "epoch": 1.3224833572400603, "grad_norm": 0.37797120870442297, "learning_rate": 3.11269734164638e-06, "loss": 0.4174, "step": 2248 }, { "epoch": 1.3230718305196953, "grad_norm": 0.3355218888357253, "learning_rate": 3.107944749788449e-06, "loss": 0.4247, "step": 2249 }, { "epoch": 1.3236603037993306, "grad_norm": 0.3730776105136381, "learning_rate": 3.1031941522313414e-06, "loss": 0.4007, "step": 2250 }, { "epoch": 1.3242487770789657, "grad_norm": 0.413854721632191, "learning_rate": 3.098445553982372e-06, "loss": 0.4513, "step": 2251 }, { "epoch": 1.324837250358601, "grad_norm": 0.3533324443070663, "learning_rate": 3.093698960046754e-06, "loss": 0.4409, "step": 2252 }, { "epoch": 1.325425723638236, "grad_norm": 0.3383080694857471, "learning_rate": 3.088954375427582e-06, "loss": 0.4181, "step": 2253 }, { "epoch": 1.3260141969178711, "grad_norm": 0.3687999267207765, "learning_rate": 3.084211805125835e-06, "loss": 0.5126, "step": 2254 }, { "epoch": 1.3266026701975062, "grad_norm": 0.37777754159787313, "learning_rate": 3.0794712541403716e-06, "loss": 0.4398, "step": 2255 }, { "epoch": 1.3271911434771415, "grad_norm": 0.36078486455680764, "learning_rate": 3.0747327274679173e-06, "loss": 0.4217, "step": 2256 }, { "epoch": 1.3277796167567766, "grad_norm": 0.34955361055203005, "learning_rate": 3.069996230103066e-06, "loss": 0.4668, "step": 2257 }, { "epoch": 1.3283680900364119, "grad_norm": 0.3588710982882473, "learning_rate": 3.0652617670382745e-06, "loss": 0.4448, "step": 2258 }, { "epoch": 1.328956563316047, "grad_norm": 0.35572542704178717, "learning_rate": 3.0605293432638516e-06, "loss": 0.4175, "step": 2259 }, { "epoch": 1.329545036595682, "grad_norm": 0.3544952356773136, "learning_rate": 3.055798963767959e-06, "loss": 0.4315, "step": 2260 }, { "epoch": 1.330133509875317, "grad_norm": 0.3353910367032978, "learning_rate": 3.0510706335366034e-06, "loss": 0.4854, "step": 2261 }, { "epoch": 1.3307219831549524, "grad_norm": 0.3486725932537452, "learning_rate": 3.0463443575536324e-06, "loss": 0.4462, "step": 2262 }, { "epoch": 1.3313104564345875, "grad_norm": 0.34202201447501346, "learning_rate": 3.041620140800724e-06, "loss": 0.4295, "step": 2263 }, { "epoch": 1.3318989297142227, "grad_norm": 0.32191394911045673, "learning_rate": 3.0368979882573926e-06, "loss": 0.4402, "step": 2264 }, { "epoch": 1.3324874029938578, "grad_norm": 0.3331419198812526, "learning_rate": 3.0321779049009714e-06, "loss": 0.4557, "step": 2265 }, { "epoch": 1.3330758762734929, "grad_norm": 0.3927127350309912, "learning_rate": 3.0274598957066132e-06, "loss": 0.4013, "step": 2266 }, { "epoch": 1.3336643495531282, "grad_norm": 0.358932945977655, "learning_rate": 3.0227439656472878e-06, "loss": 0.477, "step": 2267 }, { "epoch": 1.3342528228327633, "grad_norm": 0.3548833613904612, "learning_rate": 3.018030119693771e-06, "loss": 0.4278, "step": 2268 }, { "epoch": 1.3348412961123983, "grad_norm": 0.3732834691897786, "learning_rate": 3.01331836281464e-06, "loss": 0.432, "step": 2269 }, { "epoch": 1.3354297693920336, "grad_norm": 0.4127843530270197, "learning_rate": 3.0086086999762755e-06, "loss": 0.4992, "step": 2270 }, { "epoch": 1.3360182426716687, "grad_norm": 0.35638445698614274, "learning_rate": 3.0039011361428466e-06, "loss": 0.4033, "step": 2271 }, { "epoch": 1.3366067159513038, "grad_norm": 0.33141120865990625, "learning_rate": 2.9991956762763098e-06, "loss": 0.4046, "step": 2272 }, { "epoch": 1.337195189230939, "grad_norm": 0.3760238911692819, "learning_rate": 2.9944923253364066e-06, "loss": 0.44, "step": 2273 }, { "epoch": 1.3377836625105741, "grad_norm": 0.37937364380477906, "learning_rate": 2.989791088280655e-06, "loss": 0.4123, "step": 2274 }, { "epoch": 1.3383721357902092, "grad_norm": 0.4185636840116669, "learning_rate": 2.9850919700643422e-06, "loss": 0.5215, "step": 2275 }, { "epoch": 1.3389606090698445, "grad_norm": 0.3380470056141121, "learning_rate": 2.980394975640526e-06, "loss": 0.4413, "step": 2276 }, { "epoch": 1.3395490823494796, "grad_norm": 0.36073036408806985, "learning_rate": 2.975700109960023e-06, "loss": 0.4417, "step": 2277 }, { "epoch": 1.3401375556291146, "grad_norm": 0.3760003158259704, "learning_rate": 2.9710073779714053e-06, "loss": 0.4665, "step": 2278 }, { "epoch": 1.34072602890875, "grad_norm": 0.36009926115099555, "learning_rate": 2.966316784621e-06, "loss": 0.4333, "step": 2279 }, { "epoch": 1.341314502188385, "grad_norm": 0.3462612084658827, "learning_rate": 2.9616283348528756e-06, "loss": 0.3983, "step": 2280 }, { "epoch": 1.3419029754680203, "grad_norm": 0.3875196459262894, "learning_rate": 2.956942033608843e-06, "loss": 0.4866, "step": 2281 }, { "epoch": 1.3424914487476554, "grad_norm": 0.40137674761732434, "learning_rate": 2.9522578858284496e-06, "loss": 0.4994, "step": 2282 }, { "epoch": 1.3430799220272904, "grad_norm": 0.3875834263297586, "learning_rate": 2.947575896448971e-06, "loss": 0.5017, "step": 2283 }, { "epoch": 1.3436683953069255, "grad_norm": 0.38680356207818506, "learning_rate": 2.9428960704054075e-06, "loss": 0.4973, "step": 2284 }, { "epoch": 1.3442568685865608, "grad_norm": 0.3470231794641354, "learning_rate": 2.9382184126304834e-06, "loss": 0.4588, "step": 2285 }, { "epoch": 1.3448453418661959, "grad_norm": 0.3582588258601038, "learning_rate": 2.933542928054632e-06, "loss": 0.4485, "step": 2286 }, { "epoch": 1.3454338151458312, "grad_norm": 0.36195890067289427, "learning_rate": 2.9288696216059977e-06, "loss": 0.467, "step": 2287 }, { "epoch": 1.3460222884254662, "grad_norm": 0.3760632896434293, "learning_rate": 2.924198498210432e-06, "loss": 0.4533, "step": 2288 }, { "epoch": 1.3466107617051013, "grad_norm": 0.34836794051505454, "learning_rate": 2.919529562791482e-06, "loss": 0.4264, "step": 2289 }, { "epoch": 1.3471992349847364, "grad_norm": 0.3402696363912907, "learning_rate": 2.9148628202703882e-06, "loss": 0.398, "step": 2290 }, { "epoch": 1.3477877082643717, "grad_norm": 0.42185782867753696, "learning_rate": 2.910198275566085e-06, "loss": 0.483, "step": 2291 }, { "epoch": 1.3483761815440067, "grad_norm": 0.424338091346998, "learning_rate": 2.9055359335951814e-06, "loss": 0.4547, "step": 2292 }, { "epoch": 1.348964654823642, "grad_norm": 0.38239214972955515, "learning_rate": 2.9008757992719734e-06, "loss": 0.5262, "step": 2293 }, { "epoch": 1.349553128103277, "grad_norm": 0.3595469464166676, "learning_rate": 2.8962178775084267e-06, "loss": 0.4493, "step": 2294 }, { "epoch": 1.3501416013829122, "grad_norm": 0.38773176482622895, "learning_rate": 2.891562173214171e-06, "loss": 0.5185, "step": 2295 }, { "epoch": 1.3507300746625472, "grad_norm": 0.3978819982617313, "learning_rate": 2.886908691296504e-06, "loss": 0.4492, "step": 2296 }, { "epoch": 1.3513185479421825, "grad_norm": 0.37184466476961014, "learning_rate": 2.8822574366603804e-06, "loss": 0.4369, "step": 2297 }, { "epoch": 1.3519070212218176, "grad_norm": 0.3302561475892476, "learning_rate": 2.8776084142084036e-06, "loss": 0.3814, "step": 2298 }, { "epoch": 1.352495494501453, "grad_norm": 0.36209891716409265, "learning_rate": 2.872961628840828e-06, "loss": 0.4361, "step": 2299 }, { "epoch": 1.353083967781088, "grad_norm": 0.39391940610833764, "learning_rate": 2.8683170854555497e-06, "loss": 0.4322, "step": 2300 }, { "epoch": 1.353672441060723, "grad_norm": 0.36960846568776995, "learning_rate": 2.863674788948097e-06, "loss": 0.4462, "step": 2301 }, { "epoch": 1.3542609143403581, "grad_norm": 0.3651455638282179, "learning_rate": 2.859034744211633e-06, "loss": 0.4058, "step": 2302 }, { "epoch": 1.3548493876199934, "grad_norm": 0.3602590922577809, "learning_rate": 2.8543969561369556e-06, "loss": 0.481, "step": 2303 }, { "epoch": 1.3554378608996285, "grad_norm": 0.3812182475431378, "learning_rate": 2.8497614296124683e-06, "loss": 0.4334, "step": 2304 }, { "epoch": 1.3560263341792638, "grad_norm": 0.372029360388829, "learning_rate": 2.8451281695242013e-06, "loss": 0.4594, "step": 2305 }, { "epoch": 1.3566148074588988, "grad_norm": 0.36409713118162, "learning_rate": 2.8404971807557957e-06, "loss": 0.3922, "step": 2306 }, { "epoch": 1.357203280738534, "grad_norm": 0.38300606827691275, "learning_rate": 2.8358684681884917e-06, "loss": 0.4689, "step": 2307 }, { "epoch": 1.357791754018169, "grad_norm": 0.4112787532902843, "learning_rate": 2.831242036701137e-06, "loss": 0.4667, "step": 2308 }, { "epoch": 1.3583802272978043, "grad_norm": 0.34207934132199463, "learning_rate": 2.8266178911701757e-06, "loss": 0.4978, "step": 2309 }, { "epoch": 1.3589687005774393, "grad_norm": 0.3770292851622675, "learning_rate": 2.8219960364696343e-06, "loss": 0.4506, "step": 2310 }, { "epoch": 1.3595571738570746, "grad_norm": 0.36445901715688306, "learning_rate": 2.817376477471132e-06, "loss": 0.4023, "step": 2311 }, { "epoch": 1.3601456471367097, "grad_norm": 0.354784031814639, "learning_rate": 2.812759219043869e-06, "loss": 0.3846, "step": 2312 }, { "epoch": 1.3607341204163448, "grad_norm": 0.43047092949402244, "learning_rate": 2.8081442660546126e-06, "loss": 0.4691, "step": 2313 }, { "epoch": 1.3613225936959799, "grad_norm": 0.3213956171458106, "learning_rate": 2.803531623367708e-06, "loss": 0.4078, "step": 2314 }, { "epoch": 1.3619110669756151, "grad_norm": 0.3805136248698277, "learning_rate": 2.798921295845064e-06, "loss": 0.4033, "step": 2315 }, { "epoch": 1.3624995402552502, "grad_norm": 0.38453071391424676, "learning_rate": 2.7943132883461434e-06, "loss": 0.4506, "step": 2316 }, { "epoch": 1.3630880135348855, "grad_norm": 0.35784292514808036, "learning_rate": 2.7897076057279703e-06, "loss": 0.4501, "step": 2317 }, { "epoch": 1.3636764868145206, "grad_norm": 0.3615561086567372, "learning_rate": 2.7851042528451174e-06, "loss": 0.4727, "step": 2318 }, { "epoch": 1.3642649600941557, "grad_norm": 0.47126323774861867, "learning_rate": 2.7805032345496956e-06, "loss": 0.4376, "step": 2319 }, { "epoch": 1.364853433373791, "grad_norm": 0.357082414925822, "learning_rate": 2.775904555691363e-06, "loss": 0.4068, "step": 2320 }, { "epoch": 1.365441906653426, "grad_norm": 0.325842889700605, "learning_rate": 2.771308221117309e-06, "loss": 0.3594, "step": 2321 }, { "epoch": 1.366030379933061, "grad_norm": 0.3475905599986919, "learning_rate": 2.7667142356722487e-06, "loss": 0.4292, "step": 2322 }, { "epoch": 1.3666188532126964, "grad_norm": 0.36225751675615625, "learning_rate": 2.762122604198424e-06, "loss": 0.5018, "step": 2323 }, { "epoch": 1.3672073264923315, "grad_norm": 0.3476856978767723, "learning_rate": 2.7575333315355994e-06, "loss": 0.3901, "step": 2324 }, { "epoch": 1.3677957997719665, "grad_norm": 0.34774638657737267, "learning_rate": 2.7529464225210447e-06, "loss": 0.3699, "step": 2325 }, { "epoch": 1.3683842730516018, "grad_norm": 0.3629110580171729, "learning_rate": 2.748361881989546e-06, "loss": 0.4454, "step": 2326 }, { "epoch": 1.368972746331237, "grad_norm": 0.38724384864580513, "learning_rate": 2.743779714773386e-06, "loss": 0.4931, "step": 2327 }, { "epoch": 1.3695612196108722, "grad_norm": 0.3987334522457393, "learning_rate": 2.7391999257023525e-06, "loss": 0.4415, "step": 2328 }, { "epoch": 1.3701496928905073, "grad_norm": 0.3842625840838843, "learning_rate": 2.734622519603726e-06, "loss": 0.5016, "step": 2329 }, { "epoch": 1.3707381661701423, "grad_norm": 0.3690772368255247, "learning_rate": 2.7300475013022666e-06, "loss": 0.4366, "step": 2330 }, { "epoch": 1.3713266394497774, "grad_norm": 0.36464344448123237, "learning_rate": 2.725474875620228e-06, "loss": 0.4262, "step": 2331 }, { "epoch": 1.3719151127294127, "grad_norm": 0.3714169311097765, "learning_rate": 2.7209046473773393e-06, "loss": 0.4484, "step": 2332 }, { "epoch": 1.3725035860090478, "grad_norm": 0.3236254746491754, "learning_rate": 2.7163368213907975e-06, "loss": 0.4011, "step": 2333 }, { "epoch": 1.373092059288683, "grad_norm": 0.38896783345179925, "learning_rate": 2.711771402475273e-06, "loss": 0.4234, "step": 2334 }, { "epoch": 1.3736805325683181, "grad_norm": 0.4206875637398667, "learning_rate": 2.7072083954428997e-06, "loss": 0.4385, "step": 2335 }, { "epoch": 1.3742690058479532, "grad_norm": 0.39187264921284626, "learning_rate": 2.7026478051032625e-06, "loss": 0.493, "step": 2336 }, { "epoch": 1.3748574791275883, "grad_norm": 0.3366852620331259, "learning_rate": 2.698089636263405e-06, "loss": 0.3868, "step": 2337 }, { "epoch": 1.3754459524072236, "grad_norm": 0.38562001372689236, "learning_rate": 2.69353389372782e-06, "loss": 0.4371, "step": 2338 }, { "epoch": 1.3760344256868586, "grad_norm": 0.3623678467868721, "learning_rate": 2.6889805822984348e-06, "loss": 0.4693, "step": 2339 }, { "epoch": 1.376622898966494, "grad_norm": 0.4255946671514023, "learning_rate": 2.684429706774622e-06, "loss": 0.4608, "step": 2340 }, { "epoch": 1.377211372246129, "grad_norm": 0.3781958954362802, "learning_rate": 2.6798812719531843e-06, "loss": 0.4051, "step": 2341 }, { "epoch": 1.377799845525764, "grad_norm": 0.35312530580616147, "learning_rate": 2.6753352826283484e-06, "loss": 0.4464, "step": 2342 }, { "epoch": 1.3783883188053991, "grad_norm": 0.3313386283971185, "learning_rate": 2.6707917435917687e-06, "loss": 0.4225, "step": 2343 }, { "epoch": 1.3789767920850344, "grad_norm": 0.33435748257270326, "learning_rate": 2.6662506596325156e-06, "loss": 0.3879, "step": 2344 }, { "epoch": 1.3795652653646695, "grad_norm": 0.4067906032121379, "learning_rate": 2.6617120355370667e-06, "loss": 0.5296, "step": 2345 }, { "epoch": 1.3801537386443048, "grad_norm": 0.3548123278258963, "learning_rate": 2.6571758760893127e-06, "loss": 0.4607, "step": 2346 }, { "epoch": 1.3807422119239399, "grad_norm": 0.3325996769087891, "learning_rate": 2.6526421860705474e-06, "loss": 0.4477, "step": 2347 }, { "epoch": 1.381330685203575, "grad_norm": 0.3670541336632051, "learning_rate": 2.648110970259454e-06, "loss": 0.4098, "step": 2348 }, { "epoch": 1.38191915848321, "grad_norm": 0.34927844186560364, "learning_rate": 2.643582233432115e-06, "loss": 0.3962, "step": 2349 }, { "epoch": 1.3825076317628453, "grad_norm": 0.35529725738531853, "learning_rate": 2.6390559803619984e-06, "loss": 0.436, "step": 2350 }, { "epoch": 1.3830961050424804, "grad_norm": 0.3385709688222969, "learning_rate": 2.6345322158199503e-06, "loss": 0.3588, "step": 2351 }, { "epoch": 1.3836845783221157, "grad_norm": 0.35192377016323384, "learning_rate": 2.6300109445741995e-06, "loss": 0.4947, "step": 2352 }, { "epoch": 1.3842730516017507, "grad_norm": 0.37016455025822276, "learning_rate": 2.6254921713903447e-06, "loss": 0.4009, "step": 2353 }, { "epoch": 1.3848615248813858, "grad_norm": 0.35625798535369263, "learning_rate": 2.620975901031348e-06, "loss": 0.3813, "step": 2354 }, { "epoch": 1.3854499981610209, "grad_norm": 0.383513286242068, "learning_rate": 2.616462138257538e-06, "loss": 0.4785, "step": 2355 }, { "epoch": 1.3860384714406562, "grad_norm": 0.3533533792684791, "learning_rate": 2.611950887826601e-06, "loss": 0.3858, "step": 2356 }, { "epoch": 1.3866269447202912, "grad_norm": 0.438953055248515, "learning_rate": 2.607442154493568e-06, "loss": 0.4399, "step": 2357 }, { "epoch": 1.3872154179999265, "grad_norm": 0.42136786005841287, "learning_rate": 2.602935943010825e-06, "loss": 0.4833, "step": 2358 }, { "epoch": 1.3878038912795616, "grad_norm": 0.3867400330021937, "learning_rate": 2.5984322581280967e-06, "loss": 0.4519, "step": 2359 }, { "epoch": 1.3883923645591967, "grad_norm": 0.36850786329054624, "learning_rate": 2.593931104592442e-06, "loss": 0.5157, "step": 2360 }, { "epoch": 1.3889808378388318, "grad_norm": 0.5371419093149321, "learning_rate": 2.5894324871482557e-06, "loss": 0.4256, "step": 2361 }, { "epoch": 1.389569311118467, "grad_norm": 0.40547698266896665, "learning_rate": 2.58493641053726e-06, "loss": 0.4178, "step": 2362 }, { "epoch": 1.3901577843981021, "grad_norm": 0.31767914656707735, "learning_rate": 2.5804428794984926e-06, "loss": 0.3814, "step": 2363 }, { "epoch": 1.3907462576777374, "grad_norm": 0.42447744208871013, "learning_rate": 2.5759518987683154e-06, "loss": 0.5055, "step": 2364 }, { "epoch": 1.3913347309573725, "grad_norm": 0.3659726339110198, "learning_rate": 2.5714634730803993e-06, "loss": 0.4647, "step": 2365 }, { "epoch": 1.3919232042370075, "grad_norm": 0.334033603471114, "learning_rate": 2.5669776071657194e-06, "loss": 0.3992, "step": 2366 }, { "epoch": 1.3925116775166428, "grad_norm": 0.43422537689117213, "learning_rate": 2.562494305752557e-06, "loss": 0.3892, "step": 2367 }, { "epoch": 1.393100150796278, "grad_norm": 0.3885273038442574, "learning_rate": 2.55801357356649e-06, "loss": 0.4504, "step": 2368 }, { "epoch": 1.393688624075913, "grad_norm": 0.3753387443287163, "learning_rate": 2.5535354153303827e-06, "loss": 0.4286, "step": 2369 }, { "epoch": 1.3942770973555483, "grad_norm": 0.44755425227016804, "learning_rate": 2.549059835764394e-06, "loss": 0.4506, "step": 2370 }, { "epoch": 1.3948655706351833, "grad_norm": 0.3324412106194817, "learning_rate": 2.544586839585961e-06, "loss": 0.4163, "step": 2371 }, { "epoch": 1.3954540439148184, "grad_norm": 0.3438655924723392, "learning_rate": 2.540116431509795e-06, "loss": 0.3607, "step": 2372 }, { "epoch": 1.3960425171944537, "grad_norm": 0.3099107474008321, "learning_rate": 2.5356486162478843e-06, "loss": 0.3894, "step": 2373 }, { "epoch": 1.3966309904740888, "grad_norm": 0.3564262421594166, "learning_rate": 2.5311833985094845e-06, "loss": 0.4237, "step": 2374 }, { "epoch": 1.3972194637537239, "grad_norm": 0.3711872077191632, "learning_rate": 2.526720783001107e-06, "loss": 0.4504, "step": 2375 }, { "epoch": 1.3978079370333591, "grad_norm": 0.3456598458917385, "learning_rate": 2.5222607744265266e-06, "loss": 0.4336, "step": 2376 }, { "epoch": 1.3983964103129942, "grad_norm": 0.33907165510372295, "learning_rate": 2.5178033774867692e-06, "loss": 0.4322, "step": 2377 }, { "epoch": 1.3989848835926293, "grad_norm": 0.3255787029096176, "learning_rate": 2.513348596880105e-06, "loss": 0.3764, "step": 2378 }, { "epoch": 1.3995733568722646, "grad_norm": 0.394153606452258, "learning_rate": 2.5088964373020485e-06, "loss": 0.4239, "step": 2379 }, { "epoch": 1.4001618301518997, "grad_norm": 0.4168265474021788, "learning_rate": 2.5044469034453535e-06, "loss": 0.4268, "step": 2380 }, { "epoch": 1.400750303431535, "grad_norm": 0.3344749250698253, "learning_rate": 2.5000000000000015e-06, "loss": 0.3914, "step": 2381 }, { "epoch": 1.40133877671117, "grad_norm": 0.3320804932274252, "learning_rate": 2.495555731653204e-06, "loss": 0.3882, "step": 2382 }, { "epoch": 1.401927249990805, "grad_norm": 0.384799448549601, "learning_rate": 2.4911141030893994e-06, "loss": 0.4903, "step": 2383 }, { "epoch": 1.4025157232704402, "grad_norm": 0.35455091041879433, "learning_rate": 2.486675118990233e-06, "loss": 0.449, "step": 2384 }, { "epoch": 1.4031041965500755, "grad_norm": 0.3491160036568239, "learning_rate": 2.482238784034573e-06, "loss": 0.4162, "step": 2385 }, { "epoch": 1.4036926698297105, "grad_norm": 0.34173850708760206, "learning_rate": 2.4778051028984918e-06, "loss": 0.3985, "step": 2386 }, { "epoch": 1.4042811431093458, "grad_norm": 0.3987829784103688, "learning_rate": 2.473374080255261e-06, "loss": 0.5338, "step": 2387 }, { "epoch": 1.404869616388981, "grad_norm": 0.37900166255408446, "learning_rate": 2.468945720775355e-06, "loss": 0.3921, "step": 2388 }, { "epoch": 1.405458089668616, "grad_norm": 0.34885189158226226, "learning_rate": 2.464520029126441e-06, "loss": 0.4267, "step": 2389 }, { "epoch": 1.406046562948251, "grad_norm": 0.41007372976585466, "learning_rate": 2.4600970099733684e-06, "loss": 0.4976, "step": 2390 }, { "epoch": 1.4066350362278863, "grad_norm": 0.3360199681473403, "learning_rate": 2.4556766679781763e-06, "loss": 0.4229, "step": 2391 }, { "epoch": 1.4072235095075214, "grad_norm": 0.41476354274469085, "learning_rate": 2.451259007800081e-06, "loss": 0.5126, "step": 2392 }, { "epoch": 1.4078119827871567, "grad_norm": 0.41461924572832715, "learning_rate": 2.4468440340954664e-06, "loss": 0.4495, "step": 2393 }, { "epoch": 1.4084004560667918, "grad_norm": 0.33698521594965597, "learning_rate": 2.4424317515178914e-06, "loss": 0.3792, "step": 2394 }, { "epoch": 1.4089889293464268, "grad_norm": 0.3870404506072342, "learning_rate": 2.438022164718077e-06, "loss": 0.4251, "step": 2395 }, { "epoch": 1.409577402626062, "grad_norm": 0.455520170727391, "learning_rate": 2.4336152783438984e-06, "loss": 0.4452, "step": 2396 }, { "epoch": 1.4101658759056972, "grad_norm": 0.35150060492885504, "learning_rate": 2.4292110970403885e-06, "loss": 0.4103, "step": 2397 }, { "epoch": 1.4107543491853323, "grad_norm": 0.31930002948139063, "learning_rate": 2.424809625449729e-06, "loss": 0.3418, "step": 2398 }, { "epoch": 1.4113428224649676, "grad_norm": 0.36311039684320195, "learning_rate": 2.4204108682112443e-06, "loss": 0.4178, "step": 2399 }, { "epoch": 1.4119312957446026, "grad_norm": 0.3526312406909505, "learning_rate": 2.416014829961399e-06, "loss": 0.4529, "step": 2400 }, { "epoch": 1.4125197690242377, "grad_norm": 0.3404711504055876, "learning_rate": 2.411621515333788e-06, "loss": 0.3799, "step": 2401 }, { "epoch": 1.4131082423038728, "grad_norm": 0.3457444887489907, "learning_rate": 2.4072309289591394e-06, "loss": 0.4148, "step": 2402 }, { "epoch": 1.413696715583508, "grad_norm": 0.4562214163549887, "learning_rate": 2.4028430754653063e-06, "loss": 0.498, "step": 2403 }, { "epoch": 1.4142851888631431, "grad_norm": 0.3929055074489524, "learning_rate": 2.3984579594772557e-06, "loss": 0.4639, "step": 2404 }, { "epoch": 1.4148736621427784, "grad_norm": 0.3446790119301245, "learning_rate": 2.3940755856170744e-06, "loss": 0.3868, "step": 2405 }, { "epoch": 1.4154621354224135, "grad_norm": 0.3863251097107733, "learning_rate": 2.3896959585039585e-06, "loss": 0.5007, "step": 2406 }, { "epoch": 1.4160506087020486, "grad_norm": 0.3844655641244538, "learning_rate": 2.3853190827542056e-06, "loss": 0.4515, "step": 2407 }, { "epoch": 1.4166390819816836, "grad_norm": 0.3362768603456256, "learning_rate": 2.3809449629812154e-06, "loss": 0.4143, "step": 2408 }, { "epoch": 1.417227555261319, "grad_norm": 0.3434769275039902, "learning_rate": 2.376573603795486e-06, "loss": 0.3871, "step": 2409 }, { "epoch": 1.417816028540954, "grad_norm": 0.31958689294266585, "learning_rate": 2.3722050098045964e-06, "loss": 0.3847, "step": 2410 }, { "epoch": 1.4184045018205893, "grad_norm": 0.35803829691866823, "learning_rate": 2.3678391856132203e-06, "loss": 0.445, "step": 2411 }, { "epoch": 1.4189929751002244, "grad_norm": 0.38213962512680416, "learning_rate": 2.3634761358231105e-06, "loss": 0.4124, "step": 2412 }, { "epoch": 1.4195814483798594, "grad_norm": 0.3529753210249102, "learning_rate": 2.3591158650330896e-06, "loss": 0.4188, "step": 2413 }, { "epoch": 1.4201699216594945, "grad_norm": 0.37378398360196596, "learning_rate": 2.3547583778390564e-06, "loss": 0.4279, "step": 2414 }, { "epoch": 1.4207583949391298, "grad_norm": 0.4059528259152301, "learning_rate": 2.3504036788339763e-06, "loss": 0.5125, "step": 2415 }, { "epoch": 1.4213468682187649, "grad_norm": 0.48243035996989375, "learning_rate": 2.3460517726078696e-06, "loss": 0.4997, "step": 2416 }, { "epoch": 1.4219353414984002, "grad_norm": 0.4062621917379012, "learning_rate": 2.341702663747819e-06, "loss": 0.4626, "step": 2417 }, { "epoch": 1.4225238147780352, "grad_norm": 0.3476719070284656, "learning_rate": 2.3373563568379595e-06, "loss": 0.4534, "step": 2418 }, { "epoch": 1.4231122880576703, "grad_norm": 0.35455312108147546, "learning_rate": 2.3330128564594645e-06, "loss": 0.446, "step": 2419 }, { "epoch": 1.4237007613373056, "grad_norm": 0.4297025078110375, "learning_rate": 2.328672167190558e-06, "loss": 0.4018, "step": 2420 }, { "epoch": 1.4242892346169407, "grad_norm": 0.36290240905212195, "learning_rate": 2.324334293606499e-06, "loss": 0.4192, "step": 2421 }, { "epoch": 1.4248777078965758, "grad_norm": 0.3498629944403821, "learning_rate": 2.3199992402795734e-06, "loss": 0.409, "step": 2422 }, { "epoch": 1.425466181176211, "grad_norm": 0.3551533822048732, "learning_rate": 2.3156670117790996e-06, "loss": 0.464, "step": 2423 }, { "epoch": 1.4260546544558461, "grad_norm": 0.40437354998078495, "learning_rate": 2.3113376126714204e-06, "loss": 0.4463, "step": 2424 }, { "epoch": 1.4266431277354812, "grad_norm": 0.34897931485962586, "learning_rate": 2.307011047519888e-06, "loss": 0.4328, "step": 2425 }, { "epoch": 1.4272316010151165, "grad_norm": 0.38921264998094407, "learning_rate": 2.302687320884876e-06, "loss": 0.4833, "step": 2426 }, { "epoch": 1.4278200742947516, "grad_norm": 0.3373374954382049, "learning_rate": 2.2983664373237635e-06, "loss": 0.3762, "step": 2427 }, { "epoch": 1.4284085475743866, "grad_norm": 0.3919117954414549, "learning_rate": 2.2940484013909287e-06, "loss": 0.4548, "step": 2428 }, { "epoch": 1.428997020854022, "grad_norm": 0.36528579910632153, "learning_rate": 2.289733217637753e-06, "loss": 0.4014, "step": 2429 }, { "epoch": 1.429585494133657, "grad_norm": 0.3563377004628945, "learning_rate": 2.285420890612613e-06, "loss": 0.3942, "step": 2430 }, { "epoch": 1.430173967413292, "grad_norm": 0.37029371104649256, "learning_rate": 2.2811114248608675e-06, "loss": 0.4969, "step": 2431 }, { "epoch": 1.4307624406929274, "grad_norm": 0.34558327702546976, "learning_rate": 2.2768048249248648e-06, "loss": 0.4021, "step": 2432 }, { "epoch": 1.4313509139725624, "grad_norm": 0.36822153315132883, "learning_rate": 2.2725010953439323e-06, "loss": 0.4949, "step": 2433 }, { "epoch": 1.4319393872521977, "grad_norm": 0.3480060093630333, "learning_rate": 2.2682002406543687e-06, "loss": 0.4236, "step": 2434 }, { "epoch": 1.4325278605318328, "grad_norm": 0.325831854318672, "learning_rate": 2.2639022653894443e-06, "loss": 0.3693, "step": 2435 }, { "epoch": 1.4331163338114679, "grad_norm": 0.362935159398456, "learning_rate": 2.2596071740793974e-06, "loss": 0.5167, "step": 2436 }, { "epoch": 1.433704807091103, "grad_norm": 0.40569018237324134, "learning_rate": 2.2553149712514193e-06, "loss": 0.4381, "step": 2437 }, { "epoch": 1.4342932803707382, "grad_norm": 0.3609759026153452, "learning_rate": 2.2510256614296638e-06, "loss": 0.4961, "step": 2438 }, { "epoch": 1.4348817536503733, "grad_norm": 0.3607160267465604, "learning_rate": 2.2467392491352335e-06, "loss": 0.4191, "step": 2439 }, { "epoch": 1.4354702269300086, "grad_norm": 0.4324824729788767, "learning_rate": 2.242455738886172e-06, "loss": 0.5054, "step": 2440 }, { "epoch": 1.4360587002096437, "grad_norm": 0.37757600017095155, "learning_rate": 2.238175135197471e-06, "loss": 0.4694, "step": 2441 }, { "epoch": 1.4366471734892787, "grad_norm": 0.3545454359252546, "learning_rate": 2.2338974425810567e-06, "loss": 0.3987, "step": 2442 }, { "epoch": 1.4372356467689138, "grad_norm": 0.38707426664461286, "learning_rate": 2.2296226655457816e-06, "loss": 0.4497, "step": 2443 }, { "epoch": 1.437824120048549, "grad_norm": 0.37536351675103724, "learning_rate": 2.225350808597432e-06, "loss": 0.3617, "step": 2444 }, { "epoch": 1.4384125933281842, "grad_norm": 0.3305913723507797, "learning_rate": 2.2210818762387143e-06, "loss": 0.4359, "step": 2445 }, { "epoch": 1.4390010666078195, "grad_norm": 0.4091920382476855, "learning_rate": 2.2168158729692484e-06, "loss": 0.4234, "step": 2446 }, { "epoch": 1.4395895398874545, "grad_norm": 0.3578588899954817, "learning_rate": 2.2125528032855727e-06, "loss": 0.3906, "step": 2447 }, { "epoch": 1.4401780131670896, "grad_norm": 0.355086442359979, "learning_rate": 2.208292671681131e-06, "loss": 0.4748, "step": 2448 }, { "epoch": 1.4407664864467247, "grad_norm": 0.376296404618454, "learning_rate": 2.204035482646267e-06, "loss": 0.4694, "step": 2449 }, { "epoch": 1.44135495972636, "grad_norm": 0.361856787290203, "learning_rate": 2.1997812406682266e-06, "loss": 0.5082, "step": 2450 }, { "epoch": 1.441943433005995, "grad_norm": 0.3254994235673778, "learning_rate": 2.1955299502311523e-06, "loss": 0.4322, "step": 2451 }, { "epoch": 1.4425319062856303, "grad_norm": 0.3688837208519056, "learning_rate": 2.191281615816067e-06, "loss": 0.4382, "step": 2452 }, { "epoch": 1.4431203795652654, "grad_norm": 0.3513302356373729, "learning_rate": 2.1870362419008844e-06, "loss": 0.4356, "step": 2453 }, { "epoch": 1.4437088528449005, "grad_norm": 0.33469553868752244, "learning_rate": 2.1827938329603987e-06, "loss": 0.4058, "step": 2454 }, { "epoch": 1.4442973261245355, "grad_norm": 0.4390620682206549, "learning_rate": 2.178554393466271e-06, "loss": 0.5367, "step": 2455 }, { "epoch": 1.4448857994041708, "grad_norm": 0.3712269289034155, "learning_rate": 2.174317927887041e-06, "loss": 0.4545, "step": 2456 }, { "epoch": 1.445474272683806, "grad_norm": 0.3870727296199058, "learning_rate": 2.170084440688111e-06, "loss": 0.4822, "step": 2457 }, { "epoch": 1.4460627459634412, "grad_norm": 0.34061539296509463, "learning_rate": 2.1658539363317405e-06, "loss": 0.4269, "step": 2458 }, { "epoch": 1.4466512192430763, "grad_norm": 0.4228179599654926, "learning_rate": 2.1616264192770496e-06, "loss": 0.496, "step": 2459 }, { "epoch": 1.4472396925227113, "grad_norm": 0.35072373047620553, "learning_rate": 2.1574018939800094e-06, "loss": 0.4538, "step": 2460 }, { "epoch": 1.4478281658023464, "grad_norm": 0.3668561483340891, "learning_rate": 2.1531803648934333e-06, "loss": 0.4219, "step": 2461 }, { "epoch": 1.4484166390819817, "grad_norm": 0.3535750167672447, "learning_rate": 2.14896183646698e-06, "loss": 0.4333, "step": 2462 }, { "epoch": 1.4490051123616168, "grad_norm": 0.36079116910930975, "learning_rate": 2.144746313147148e-06, "loss": 0.4497, "step": 2463 }, { "epoch": 1.449593585641252, "grad_norm": 0.3438995765752842, "learning_rate": 2.14053379937726e-06, "loss": 0.4147, "step": 2464 }, { "epoch": 1.4501820589208871, "grad_norm": 0.35368683002758533, "learning_rate": 2.136324299597474e-06, "loss": 0.4584, "step": 2465 }, { "epoch": 1.4507705322005222, "grad_norm": 0.3639283965276241, "learning_rate": 2.132117818244771e-06, "loss": 0.428, "step": 2466 }, { "epoch": 1.4513590054801573, "grad_norm": 0.3564459604727192, "learning_rate": 2.1279143597529438e-06, "loss": 0.4392, "step": 2467 }, { "epoch": 1.4519474787597926, "grad_norm": 0.4088898640147898, "learning_rate": 2.1237139285526058e-06, "loss": 0.4733, "step": 2468 }, { "epoch": 1.4525359520394276, "grad_norm": 0.34692316397211637, "learning_rate": 2.1195165290711782e-06, "loss": 0.457, "step": 2469 }, { "epoch": 1.453124425319063, "grad_norm": 0.38877715872021046, "learning_rate": 2.1153221657328827e-06, "loss": 0.4205, "step": 2470 }, { "epoch": 1.453712898598698, "grad_norm": 0.3896133238034156, "learning_rate": 2.1111308429587446e-06, "loss": 0.5218, "step": 2471 }, { "epoch": 1.454301371878333, "grad_norm": 0.3824089896445475, "learning_rate": 2.106942565166586e-06, "loss": 0.4979, "step": 2472 }, { "epoch": 1.4548898451579684, "grad_norm": 0.3500399989384875, "learning_rate": 2.1027573367710124e-06, "loss": 0.4394, "step": 2473 }, { "epoch": 1.4554783184376034, "grad_norm": 0.3578014959791625, "learning_rate": 2.098575162183422e-06, "loss": 0.3795, "step": 2474 }, { "epoch": 1.4560667917172385, "grad_norm": 0.36849780246958364, "learning_rate": 2.0943960458119926e-06, "loss": 0.4346, "step": 2475 }, { "epoch": 1.4566552649968738, "grad_norm": 0.38102608691714973, "learning_rate": 2.090219992061675e-06, "loss": 0.436, "step": 2476 }, { "epoch": 1.4572437382765089, "grad_norm": 0.3502083500729829, "learning_rate": 2.0860470053341957e-06, "loss": 0.4383, "step": 2477 }, { "epoch": 1.457832211556144, "grad_norm": 0.4016387826102091, "learning_rate": 2.0818770900280495e-06, "loss": 0.4827, "step": 2478 }, { "epoch": 1.4584206848357792, "grad_norm": 0.37055759075817774, "learning_rate": 2.0777102505384874e-06, "loss": 0.489, "step": 2479 }, { "epoch": 1.4590091581154143, "grad_norm": 0.394648520051973, "learning_rate": 2.073546491257524e-06, "loss": 0.4464, "step": 2480 }, { "epoch": 1.4595976313950494, "grad_norm": 0.36987944217642166, "learning_rate": 2.069385816573928e-06, "loss": 0.4306, "step": 2481 }, { "epoch": 1.4601861046746847, "grad_norm": 0.4074078450645316, "learning_rate": 2.065228230873212e-06, "loss": 0.5663, "step": 2482 }, { "epoch": 1.4607745779543198, "grad_norm": 0.38128922860873143, "learning_rate": 2.061073738537635e-06, "loss": 0.415, "step": 2483 }, { "epoch": 1.4613630512339548, "grad_norm": 0.40419126473752515, "learning_rate": 2.056922343946199e-06, "loss": 0.3992, "step": 2484 }, { "epoch": 1.4619515245135901, "grad_norm": 0.3620713009427078, "learning_rate": 2.0527740514746324e-06, "loss": 0.4625, "step": 2485 }, { "epoch": 1.4625399977932252, "grad_norm": 0.4156510343060423, "learning_rate": 2.048628865495403e-06, "loss": 0.5345, "step": 2486 }, { "epoch": 1.4631284710728605, "grad_norm": 0.3467220643508277, "learning_rate": 2.0444867903777007e-06, "loss": 0.4391, "step": 2487 }, { "epoch": 1.4637169443524956, "grad_norm": 0.35112733292771686, "learning_rate": 2.0403478304874317e-06, "loss": 0.4128, "step": 2488 }, { "epoch": 1.4643054176321306, "grad_norm": 0.3679124520765004, "learning_rate": 2.0362119901872262e-06, "loss": 0.4174, "step": 2489 }, { "epoch": 1.4648938909117657, "grad_norm": 0.3571290400785807, "learning_rate": 2.0320792738364235e-06, "loss": 0.4857, "step": 2490 }, { "epoch": 1.465482364191401, "grad_norm": 0.39034261127604236, "learning_rate": 2.0279496857910667e-06, "loss": 0.4583, "step": 2491 }, { "epoch": 1.466070837471036, "grad_norm": 0.3347978222979776, "learning_rate": 2.023823230403907e-06, "loss": 0.4007, "step": 2492 }, { "epoch": 1.4666593107506714, "grad_norm": 0.35472376050159926, "learning_rate": 2.0196999120243886e-06, "loss": 0.4262, "step": 2493 }, { "epoch": 1.4672477840303064, "grad_norm": 0.36815113217648926, "learning_rate": 2.0155797349986537e-06, "loss": 0.4138, "step": 2494 }, { "epoch": 1.4678362573099415, "grad_norm": 0.37810158110438025, "learning_rate": 2.011462703669532e-06, "loss": 0.4088, "step": 2495 }, { "epoch": 1.4684247305895766, "grad_norm": 0.3700236853404871, "learning_rate": 2.0073488223765346e-06, "loss": 0.4964, "step": 2496 }, { "epoch": 1.4690132038692119, "grad_norm": 0.35578317437704227, "learning_rate": 2.003238095455855e-06, "loss": 0.3786, "step": 2497 }, { "epoch": 1.469601677148847, "grad_norm": 0.6983851937728272, "learning_rate": 1.9991305272403637e-06, "loss": 0.3996, "step": 2498 }, { "epoch": 1.4701901504284822, "grad_norm": 0.3524828140244812, "learning_rate": 1.9950261220595956e-06, "loss": 0.4549, "step": 2499 }, { "epoch": 1.4707786237081173, "grad_norm": 0.33363402779353807, "learning_rate": 1.990924884239758e-06, "loss": 0.3796, "step": 2500 }, { "epoch": 1.4713670969877524, "grad_norm": 0.3641225425404599, "learning_rate": 1.9868268181037186e-06, "loss": 0.4359, "step": 2501 }, { "epoch": 1.4719555702673874, "grad_norm": 0.39442620321859345, "learning_rate": 1.982731927970997e-06, "loss": 0.4286, "step": 2502 }, { "epoch": 1.4725440435470227, "grad_norm": 0.357121273144524, "learning_rate": 1.9786402181577713e-06, "loss": 0.4653, "step": 2503 }, { "epoch": 1.4731325168266578, "grad_norm": 0.34546709732974296, "learning_rate": 1.974551692976866e-06, "loss": 0.4703, "step": 2504 }, { "epoch": 1.473720990106293, "grad_norm": 0.41404312162512724, "learning_rate": 1.9704663567377445e-06, "loss": 0.4238, "step": 2505 }, { "epoch": 1.4743094633859282, "grad_norm": 0.39996411719073366, "learning_rate": 1.9663842137465144e-06, "loss": 0.4392, "step": 2506 }, { "epoch": 1.4748979366655632, "grad_norm": 0.355831195544288, "learning_rate": 1.9623052683059164e-06, "loss": 0.4483, "step": 2507 }, { "epoch": 1.4754864099451983, "grad_norm": 0.3725759279454592, "learning_rate": 1.9582295247153165e-06, "loss": 0.5126, "step": 2508 }, { "epoch": 1.4760748832248336, "grad_norm": 0.3892676428463921, "learning_rate": 1.954156987270711e-06, "loss": 0.4316, "step": 2509 }, { "epoch": 1.4766633565044687, "grad_norm": 0.38794733607824133, "learning_rate": 1.9500876602647167e-06, "loss": 0.4667, "step": 2510 }, { "epoch": 1.477251829784104, "grad_norm": 0.37146607334064874, "learning_rate": 1.9460215479865613e-06, "loss": 0.4438, "step": 2511 }, { "epoch": 1.477840303063739, "grad_norm": 0.34465580846062427, "learning_rate": 1.9419586547220888e-06, "loss": 0.4052, "step": 2512 }, { "epoch": 1.478428776343374, "grad_norm": 0.41558967508599715, "learning_rate": 1.937898984753751e-06, "loss": 0.4832, "step": 2513 }, { "epoch": 1.4790172496230092, "grad_norm": 0.34680898074004185, "learning_rate": 1.9338425423605963e-06, "loss": 0.4474, "step": 2514 }, { "epoch": 1.4796057229026445, "grad_norm": 0.3632083520590157, "learning_rate": 1.929789331818277e-06, "loss": 0.4138, "step": 2515 }, { "epoch": 1.4801941961822795, "grad_norm": 0.3626049356931775, "learning_rate": 1.925739357399038e-06, "loss": 0.3747, "step": 2516 }, { "epoch": 1.4807826694619148, "grad_norm": 0.36004170476938524, "learning_rate": 1.9216926233717087e-06, "loss": 0.4138, "step": 2517 }, { "epoch": 1.48137114274155, "grad_norm": 0.36722679654682394, "learning_rate": 1.9176491340017083e-06, "loss": 0.4429, "step": 2518 }, { "epoch": 1.481959616021185, "grad_norm": 0.3680628479296944, "learning_rate": 1.913608893551036e-06, "loss": 0.4288, "step": 2519 }, { "epoch": 1.48254808930082, "grad_norm": 0.34308474599595534, "learning_rate": 1.9095719062782603e-06, "loss": 0.4114, "step": 2520 }, { "epoch": 1.4831365625804553, "grad_norm": 0.37464801489172656, "learning_rate": 1.9055381764385272e-06, "loss": 0.4339, "step": 2521 }, { "epoch": 1.4837250358600904, "grad_norm": 0.3841349147261042, "learning_rate": 1.9015077082835487e-06, "loss": 0.4125, "step": 2522 }, { "epoch": 1.4843135091397257, "grad_norm": 0.36828247611436865, "learning_rate": 1.8974805060615931e-06, "loss": 0.4494, "step": 2523 }, { "epoch": 1.4849019824193608, "grad_norm": 0.4341537116671341, "learning_rate": 1.8934565740174927e-06, "loss": 0.4491, "step": 2524 }, { "epoch": 1.4854904556989958, "grad_norm": 0.4084034423302373, "learning_rate": 1.8894359163926312e-06, "loss": 0.4784, "step": 2525 }, { "epoch": 1.4860789289786311, "grad_norm": 0.38818762619766034, "learning_rate": 1.8854185374249384e-06, "loss": 0.4666, "step": 2526 }, { "epoch": 1.4866674022582662, "grad_norm": 0.3460895109075847, "learning_rate": 1.8814044413488908e-06, "loss": 0.4498, "step": 2527 }, { "epoch": 1.4872558755379013, "grad_norm": 0.4277232243857831, "learning_rate": 1.8773936323955055e-06, "loss": 0.4559, "step": 2528 }, { "epoch": 1.4878443488175366, "grad_norm": 0.36725315139138703, "learning_rate": 1.8733861147923298e-06, "loss": 0.4376, "step": 2529 }, { "epoch": 1.4884328220971716, "grad_norm": 0.39218177566559465, "learning_rate": 1.8693818927634484e-06, "loss": 0.5387, "step": 2530 }, { "epoch": 1.4890212953768067, "grad_norm": 0.3703187524887761, "learning_rate": 1.865380970529469e-06, "loss": 0.4069, "step": 2531 }, { "epoch": 1.489609768656442, "grad_norm": 0.3523734686825101, "learning_rate": 1.8613833523075187e-06, "loss": 0.4107, "step": 2532 }, { "epoch": 1.490198241936077, "grad_norm": 0.3555244611311032, "learning_rate": 1.8573890423112461e-06, "loss": 0.4545, "step": 2533 }, { "epoch": 1.4907867152157122, "grad_norm": 0.35935346702840576, "learning_rate": 1.8533980447508138e-06, "loss": 0.4653, "step": 2534 }, { "epoch": 1.4913751884953474, "grad_norm": 0.37299880900105603, "learning_rate": 1.8494103638328865e-06, "loss": 0.4777, "step": 2535 }, { "epoch": 1.4919636617749825, "grad_norm": 0.3798736230624653, "learning_rate": 1.8454260037606391e-06, "loss": 0.4047, "step": 2536 }, { "epoch": 1.4925521350546176, "grad_norm": 0.4084110197483552, "learning_rate": 1.8414449687337467e-06, "loss": 0.5597, "step": 2537 }, { "epoch": 1.4931406083342529, "grad_norm": 0.3630940303248563, "learning_rate": 1.837467262948372e-06, "loss": 0.4411, "step": 2538 }, { "epoch": 1.493729081613888, "grad_norm": 0.3438754360264229, "learning_rate": 1.833492890597176e-06, "loss": 0.4405, "step": 2539 }, { "epoch": 1.4943175548935232, "grad_norm": 0.34154326341364616, "learning_rate": 1.8295218558693067e-06, "loss": 0.3924, "step": 2540 }, { "epoch": 1.4949060281731583, "grad_norm": 0.376747104421291, "learning_rate": 1.8255541629503865e-06, "loss": 0.5058, "step": 2541 }, { "epoch": 1.4954945014527934, "grad_norm": 0.3267946878020161, "learning_rate": 1.8215898160225226e-06, "loss": 0.3717, "step": 2542 }, { "epoch": 1.4960829747324285, "grad_norm": 0.35393268247548043, "learning_rate": 1.8176288192642944e-06, "loss": 0.3864, "step": 2543 }, { "epoch": 1.4966714480120638, "grad_norm": 0.34915200520933015, "learning_rate": 1.8136711768507448e-06, "loss": 0.3712, "step": 2544 }, { "epoch": 1.4972599212916988, "grad_norm": 0.352584108762193, "learning_rate": 1.8097168929533876e-06, "loss": 0.4519, "step": 2545 }, { "epoch": 1.4978483945713341, "grad_norm": 0.3525095118952767, "learning_rate": 1.8057659717401948e-06, "loss": 0.4254, "step": 2546 }, { "epoch": 1.4984368678509692, "grad_norm": 0.36109474082261167, "learning_rate": 1.80181841737559e-06, "loss": 0.4665, "step": 2547 }, { "epoch": 1.4990253411306043, "grad_norm": 0.40531804539507715, "learning_rate": 1.7978742340204526e-06, "loss": 0.4979, "step": 2548 }, { "epoch": 1.4996138144102393, "grad_norm": 0.3541013185693249, "learning_rate": 1.7939334258321094e-06, "loss": 0.4469, "step": 2549 }, { "epoch": 1.5002022876898746, "grad_norm": 0.3608362421545421, "learning_rate": 1.7899959969643238e-06, "loss": 0.4431, "step": 2550 }, { "epoch": 1.5007907609695097, "grad_norm": 0.31980477268910984, "learning_rate": 1.7860619515673034e-06, "loss": 0.4164, "step": 2551 }, { "epoch": 1.501379234249145, "grad_norm": 0.35238095010625636, "learning_rate": 1.7821312937876883e-06, "loss": 0.4142, "step": 2552 }, { "epoch": 1.50196770752878, "grad_norm": 0.39612827471528694, "learning_rate": 1.7782040277685436e-06, "loss": 0.4617, "step": 2553 }, { "epoch": 1.5025561808084151, "grad_norm": 0.4861527508252182, "learning_rate": 1.774280157649364e-06, "loss": 0.5068, "step": 2554 }, { "epoch": 1.5031446540880502, "grad_norm": 0.48799565991185895, "learning_rate": 1.7703596875660645e-06, "loss": 0.4742, "step": 2555 }, { "epoch": 1.5037331273676855, "grad_norm": 0.41611646658241097, "learning_rate": 1.7664426216509728e-06, "loss": 0.4134, "step": 2556 }, { "epoch": 1.5043216006473206, "grad_norm": 0.39624308868996694, "learning_rate": 1.762528964032832e-06, "loss": 0.4168, "step": 2557 }, { "epoch": 1.5049100739269559, "grad_norm": 0.3325326367935813, "learning_rate": 1.758618718836793e-06, "loss": 0.405, "step": 2558 }, { "epoch": 1.505498547206591, "grad_norm": 0.35440713376960364, "learning_rate": 1.7547118901844057e-06, "loss": 0.4266, "step": 2559 }, { "epoch": 1.506087020486226, "grad_norm": 0.3656108572316197, "learning_rate": 1.7508084821936223e-06, "loss": 0.4448, "step": 2560 }, { "epoch": 1.506675493765861, "grad_norm": 0.3626883979015528, "learning_rate": 1.746908498978791e-06, "loss": 0.4264, "step": 2561 }, { "epoch": 1.5072639670454964, "grad_norm": 0.3719026559779689, "learning_rate": 1.743011944650644e-06, "loss": 0.4344, "step": 2562 }, { "epoch": 1.5078524403251314, "grad_norm": 0.3769129753304877, "learning_rate": 1.7391188233163054e-06, "loss": 0.4648, "step": 2563 }, { "epoch": 1.5084409136047667, "grad_norm": 0.3542480369241033, "learning_rate": 1.7352291390792798e-06, "loss": 0.4114, "step": 2564 }, { "epoch": 1.5090293868844018, "grad_norm": 0.36768134195262, "learning_rate": 1.731342896039444e-06, "loss": 0.421, "step": 2565 }, { "epoch": 1.5096178601640369, "grad_norm": 0.37045557643328403, "learning_rate": 1.7274600982930544e-06, "loss": 0.4131, "step": 2566 }, { "epoch": 1.510206333443672, "grad_norm": 0.3443734131327738, "learning_rate": 1.7235807499327335e-06, "loss": 0.3933, "step": 2567 }, { "epoch": 1.5107948067233072, "grad_norm": 0.3499089768411563, "learning_rate": 1.7197048550474643e-06, "loss": 0.4004, "step": 2568 }, { "epoch": 1.5113832800029425, "grad_norm": 0.3686946977585929, "learning_rate": 1.7158324177225948e-06, "loss": 0.3417, "step": 2569 }, { "epoch": 1.5119717532825776, "grad_norm": 0.36755420523638327, "learning_rate": 1.7119634420398273e-06, "loss": 0.453, "step": 2570 }, { "epoch": 1.5125602265622127, "grad_norm": 0.3615696868893597, "learning_rate": 1.708097932077213e-06, "loss": 0.4587, "step": 2571 }, { "epoch": 1.5131486998418477, "grad_norm": 0.34238277272151346, "learning_rate": 1.7042358919091518e-06, "loss": 0.3687, "step": 2572 }, { "epoch": 1.5137371731214828, "grad_norm": 0.3317115137117115, "learning_rate": 1.7003773256063882e-06, "loss": 0.3667, "step": 2573 }, { "epoch": 1.514325646401118, "grad_norm": 0.36183055332731934, "learning_rate": 1.6965222372360012e-06, "loss": 0.4453, "step": 2574 }, { "epoch": 1.5149141196807534, "grad_norm": 0.3490571140825648, "learning_rate": 1.6926706308614061e-06, "loss": 0.3839, "step": 2575 }, { "epoch": 1.5155025929603885, "grad_norm": 0.37145945012107334, "learning_rate": 1.6888225105423505e-06, "loss": 0.4693, "step": 2576 }, { "epoch": 1.5160910662400235, "grad_norm": 0.3460496162959425, "learning_rate": 1.684977880334901e-06, "loss": 0.3757, "step": 2577 }, { "epoch": 1.5166795395196586, "grad_norm": 0.3351315188206505, "learning_rate": 1.681136744291451e-06, "loss": 0.3799, "step": 2578 }, { "epoch": 1.5172680127992937, "grad_norm": 0.34797675630585834, "learning_rate": 1.6772991064607113e-06, "loss": 0.4236, "step": 2579 }, { "epoch": 1.517856486078929, "grad_norm": 0.34424668232551686, "learning_rate": 1.6734649708877004e-06, "loss": 0.3854, "step": 2580 }, { "epoch": 1.5184449593585643, "grad_norm": 0.34909452569914773, "learning_rate": 1.6696343416137495e-06, "loss": 0.405, "step": 2581 }, { "epoch": 1.5190334326381993, "grad_norm": 0.3808751441110743, "learning_rate": 1.6658072226764949e-06, "loss": 0.4842, "step": 2582 }, { "epoch": 1.5196219059178344, "grad_norm": 0.3819674221060536, "learning_rate": 1.661983618109867e-06, "loss": 0.466, "step": 2583 }, { "epoch": 1.5202103791974695, "grad_norm": 0.37013254839404425, "learning_rate": 1.658163531944098e-06, "loss": 0.4803, "step": 2584 }, { "epoch": 1.5207988524771046, "grad_norm": 0.3450578800437742, "learning_rate": 1.6543469682057105e-06, "loss": 0.4342, "step": 2585 }, { "epoch": 1.5213873257567399, "grad_norm": 0.364072833602955, "learning_rate": 1.65053393091751e-06, "loss": 0.438, "step": 2586 }, { "epoch": 1.5219757990363751, "grad_norm": 0.32419446248475786, "learning_rate": 1.6467244240985903e-06, "loss": 0.4034, "step": 2587 }, { "epoch": 1.5225642723160102, "grad_norm": 0.436349658639181, "learning_rate": 1.6429184517643222e-06, "loss": 0.4555, "step": 2588 }, { "epoch": 1.5231527455956453, "grad_norm": 0.3761314087700032, "learning_rate": 1.6391160179263467e-06, "loss": 0.4657, "step": 2589 }, { "epoch": 1.5237412188752804, "grad_norm": 0.3696591441299406, "learning_rate": 1.635317126592585e-06, "loss": 0.4237, "step": 2590 }, { "epoch": 1.5243296921549154, "grad_norm": 0.37896143890563094, "learning_rate": 1.6315217817672142e-06, "loss": 0.4479, "step": 2591 }, { "epoch": 1.5249181654345507, "grad_norm": 0.3614941383845958, "learning_rate": 1.6277299874506774e-06, "loss": 0.4361, "step": 2592 }, { "epoch": 1.525506638714186, "grad_norm": 0.3617060980175792, "learning_rate": 1.623941747639679e-06, "loss": 0.4337, "step": 2593 }, { "epoch": 1.526095111993821, "grad_norm": 0.4272377768609459, "learning_rate": 1.6201570663271682e-06, "loss": 0.4542, "step": 2594 }, { "epoch": 1.5266835852734562, "grad_norm": 0.3834573709308226, "learning_rate": 1.6163759475023505e-06, "loss": 0.448, "step": 2595 }, { "epoch": 1.5272720585530912, "grad_norm": 0.33890283850443603, "learning_rate": 1.6125983951506758e-06, "loss": 0.3866, "step": 2596 }, { "epoch": 1.5278605318327265, "grad_norm": 0.3534452633707597, "learning_rate": 1.60882441325383e-06, "loss": 0.3735, "step": 2597 }, { "epoch": 1.5284490051123616, "grad_norm": 0.3547657755880191, "learning_rate": 1.6050540057897407e-06, "loss": 0.4625, "step": 2598 }, { "epoch": 1.5290374783919969, "grad_norm": 0.37046875314477196, "learning_rate": 1.6012871767325671e-06, "loss": 0.4574, "step": 2599 }, { "epoch": 1.529625951671632, "grad_norm": 0.42554933891868074, "learning_rate": 1.5975239300526924e-06, "loss": 0.3879, "step": 2600 }, { "epoch": 1.530214424951267, "grad_norm": 0.363966723022285, "learning_rate": 1.5937642697167288e-06, "loss": 0.4119, "step": 2601 }, { "epoch": 1.530802898230902, "grad_norm": 0.3742826922045975, "learning_rate": 1.5900081996875083e-06, "loss": 0.4396, "step": 2602 }, { "epoch": 1.5313913715105374, "grad_norm": 0.49621156519977233, "learning_rate": 1.5862557239240729e-06, "loss": 0.4218, "step": 2603 }, { "epoch": 1.5319798447901725, "grad_norm": 0.3501856217007502, "learning_rate": 1.5825068463816817e-06, "loss": 0.3851, "step": 2604 }, { "epoch": 1.5325683180698078, "grad_norm": 0.37299746820479956, "learning_rate": 1.578761571011802e-06, "loss": 0.4211, "step": 2605 }, { "epoch": 1.5331567913494428, "grad_norm": 0.4278154359868545, "learning_rate": 1.575019901762097e-06, "loss": 0.5333, "step": 2606 }, { "epoch": 1.533745264629078, "grad_norm": 0.32949644453060367, "learning_rate": 1.5712818425764365e-06, "loss": 0.4395, "step": 2607 }, { "epoch": 1.534333737908713, "grad_norm": 0.4217847785857017, "learning_rate": 1.5675473973948835e-06, "loss": 0.5407, "step": 2608 }, { "epoch": 1.5349222111883483, "grad_norm": 0.3831329283212342, "learning_rate": 1.5638165701536866e-06, "loss": 0.4458, "step": 2609 }, { "epoch": 1.5355106844679833, "grad_norm": 0.3558625232136852, "learning_rate": 1.560089364785287e-06, "loss": 0.3928, "step": 2610 }, { "epoch": 1.5360991577476186, "grad_norm": 0.33997548262301136, "learning_rate": 1.5563657852183072e-06, "loss": 0.4335, "step": 2611 }, { "epoch": 1.5366876310272537, "grad_norm": 0.35704864692657656, "learning_rate": 1.5526458353775447e-06, "loss": 0.4015, "step": 2612 }, { "epoch": 1.5372761043068888, "grad_norm": 0.3386346597172168, "learning_rate": 1.5489295191839738e-06, "loss": 0.452, "step": 2613 }, { "epoch": 1.5378645775865238, "grad_norm": 0.35578218408327356, "learning_rate": 1.5452168405547402e-06, "loss": 0.4193, "step": 2614 }, { "epoch": 1.5384530508661591, "grad_norm": 0.37408410205222536, "learning_rate": 1.54150780340315e-06, "loss": 0.4084, "step": 2615 }, { "epoch": 1.5390415241457942, "grad_norm": 0.3788979545265764, "learning_rate": 1.537802411638677e-06, "loss": 0.4288, "step": 2616 }, { "epoch": 1.5396299974254295, "grad_norm": 0.36333687022110833, "learning_rate": 1.5341006691669513e-06, "loss": 0.4431, "step": 2617 }, { "epoch": 1.5402184707050646, "grad_norm": 0.35687874019612265, "learning_rate": 1.5304025798897521e-06, "loss": 0.3871, "step": 2618 }, { "epoch": 1.5408069439846996, "grad_norm": 0.35472588701185864, "learning_rate": 1.5267081477050132e-06, "loss": 0.4571, "step": 2619 }, { "epoch": 1.5413954172643347, "grad_norm": 0.37152737608002967, "learning_rate": 1.5230173765068135e-06, "loss": 0.4456, "step": 2620 }, { "epoch": 1.54198389054397, "grad_norm": 0.38777090895222416, "learning_rate": 1.5193302701853674e-06, "loss": 0.4339, "step": 2621 }, { "epoch": 1.5425723638236053, "grad_norm": 0.4264367402724305, "learning_rate": 1.5156468326270324e-06, "loss": 0.4093, "step": 2622 }, { "epoch": 1.5431608371032404, "grad_norm": 0.39301644414471415, "learning_rate": 1.5119670677142984e-06, "loss": 0.4938, "step": 2623 }, { "epoch": 1.5437493103828754, "grad_norm": 0.4027585849992698, "learning_rate": 1.5082909793257795e-06, "loss": 0.4774, "step": 2624 }, { "epoch": 1.5443377836625105, "grad_norm": 0.3240919478151283, "learning_rate": 1.5046185713362199e-06, "loss": 0.374, "step": 2625 }, { "epoch": 1.5449262569421456, "grad_norm": 0.37772976413989223, "learning_rate": 1.500949847616483e-06, "loss": 0.4356, "step": 2626 }, { "epoch": 1.5455147302217809, "grad_norm": 0.34491689811445403, "learning_rate": 1.4972848120335453e-06, "loss": 0.4149, "step": 2627 }, { "epoch": 1.5461032035014162, "grad_norm": 0.37019484923070245, "learning_rate": 1.4936234684505012e-06, "loss": 0.4605, "step": 2628 }, { "epoch": 1.5466916767810512, "grad_norm": 0.38771587325247076, "learning_rate": 1.489965820726552e-06, "loss": 0.4702, "step": 2629 }, { "epoch": 1.5472801500606863, "grad_norm": 0.40379185539418894, "learning_rate": 1.486311872717e-06, "loss": 0.5178, "step": 2630 }, { "epoch": 1.5478686233403214, "grad_norm": 0.3684651426748907, "learning_rate": 1.4826616282732509e-06, "loss": 0.4227, "step": 2631 }, { "epoch": 1.5484570966199565, "grad_norm": 0.3299020857705006, "learning_rate": 1.4790150912428091e-06, "loss": 0.4602, "step": 2632 }, { "epoch": 1.5490455698995917, "grad_norm": 0.3530115738368142, "learning_rate": 1.475372265469265e-06, "loss": 0.434, "step": 2633 }, { "epoch": 1.549634043179227, "grad_norm": 0.4648081248421497, "learning_rate": 1.4717331547923009e-06, "loss": 0.4337, "step": 2634 }, { "epoch": 1.550222516458862, "grad_norm": 0.3387744248768024, "learning_rate": 1.468097763047686e-06, "loss": 0.3924, "step": 2635 }, { "epoch": 1.5508109897384972, "grad_norm": 0.4116849960333714, "learning_rate": 1.4644660940672628e-06, "loss": 0.3966, "step": 2636 }, { "epoch": 1.5513994630181323, "grad_norm": 0.36068613712894043, "learning_rate": 1.460838151678955e-06, "loss": 0.4621, "step": 2637 }, { "epoch": 1.5519879362977673, "grad_norm": 0.35471873272655924, "learning_rate": 1.4572139397067586e-06, "loss": 0.4207, "step": 2638 }, { "epoch": 1.5525764095774026, "grad_norm": 0.35031353237630103, "learning_rate": 1.453593461970733e-06, "loss": 0.4182, "step": 2639 }, { "epoch": 1.553164882857038, "grad_norm": 0.36091814168755315, "learning_rate": 1.4499767222870058e-06, "loss": 0.4283, "step": 2640 }, { "epoch": 1.553753356136673, "grad_norm": 0.3376215751472809, "learning_rate": 1.4463637244677648e-06, "loss": 0.361, "step": 2641 }, { "epoch": 1.554341829416308, "grad_norm": 0.35188060832486706, "learning_rate": 1.4427544723212489e-06, "loss": 0.448, "step": 2642 }, { "epoch": 1.5549303026959431, "grad_norm": 0.37259567428016094, "learning_rate": 1.4391489696517535e-06, "loss": 0.421, "step": 2643 }, { "epoch": 1.5555187759755782, "grad_norm": 0.36778366784597344, "learning_rate": 1.4355472202596226e-06, "loss": 0.4521, "step": 2644 }, { "epoch": 1.5561072492552135, "grad_norm": 0.35585983947555894, "learning_rate": 1.4319492279412388e-06, "loss": 0.4105, "step": 2645 }, { "epoch": 1.5566957225348488, "grad_norm": 0.35707909349882405, "learning_rate": 1.4283549964890298e-06, "loss": 0.4518, "step": 2646 }, { "epoch": 1.5572841958144839, "grad_norm": 0.3406574023017485, "learning_rate": 1.4247645296914574e-06, "loss": 0.3719, "step": 2647 }, { "epoch": 1.557872669094119, "grad_norm": 0.32601798126458653, "learning_rate": 1.4211778313330137e-06, "loss": 0.4066, "step": 2648 }, { "epoch": 1.558461142373754, "grad_norm": 0.36185433829204994, "learning_rate": 1.4175949051942207e-06, "loss": 0.3893, "step": 2649 }, { "epoch": 1.5590496156533893, "grad_norm": 0.35447048288274136, "learning_rate": 1.4140157550516258e-06, "loss": 0.4808, "step": 2650 }, { "epoch": 1.5596380889330244, "grad_norm": 0.40577005047236925, "learning_rate": 1.410440384677791e-06, "loss": 0.4821, "step": 2651 }, { "epoch": 1.5602265622126597, "grad_norm": 0.3466633515767074, "learning_rate": 1.4068687978412992e-06, "loss": 0.3618, "step": 2652 }, { "epoch": 1.5608150354922947, "grad_norm": 0.38152499629264314, "learning_rate": 1.4033009983067454e-06, "loss": 0.4341, "step": 2653 }, { "epoch": 1.5614035087719298, "grad_norm": 0.39176745592206724, "learning_rate": 1.399736989834728e-06, "loss": 0.5173, "step": 2654 }, { "epoch": 1.5619919820515649, "grad_norm": 0.3526013215986296, "learning_rate": 1.3961767761818534e-06, "loss": 0.398, "step": 2655 }, { "epoch": 1.5625804553312002, "grad_norm": 0.3797909097289101, "learning_rate": 1.3926203611007294e-06, "loss": 0.4581, "step": 2656 }, { "epoch": 1.5631689286108352, "grad_norm": 0.37595209747233604, "learning_rate": 1.389067748339954e-06, "loss": 0.4117, "step": 2657 }, { "epoch": 1.5637574018904705, "grad_norm": 0.3624735994789221, "learning_rate": 1.3855189416441227e-06, "loss": 0.3998, "step": 2658 }, { "epoch": 1.5643458751701056, "grad_norm": 0.33056296525128576, "learning_rate": 1.38197394475382e-06, "loss": 0.398, "step": 2659 }, { "epoch": 1.5649343484497407, "grad_norm": 0.3332272907986276, "learning_rate": 1.3784327614056087e-06, "loss": 0.3902, "step": 2660 }, { "epoch": 1.5655228217293757, "grad_norm": 0.3558275119420867, "learning_rate": 1.374895395332037e-06, "loss": 0.4283, "step": 2661 }, { "epoch": 1.566111295009011, "grad_norm": 0.39281833118532306, "learning_rate": 1.3713618502616298e-06, "loss": 0.489, "step": 2662 }, { "epoch": 1.566699768288646, "grad_norm": 0.3479526102381695, "learning_rate": 1.3678321299188802e-06, "loss": 0.4404, "step": 2663 }, { "epoch": 1.5672882415682814, "grad_norm": 0.38737467534450026, "learning_rate": 1.364306238024254e-06, "loss": 0.4167, "step": 2664 }, { "epoch": 1.5678767148479165, "grad_norm": 0.3756736956188446, "learning_rate": 1.360784178294181e-06, "loss": 0.4081, "step": 2665 }, { "epoch": 1.5684651881275515, "grad_norm": 0.3595554187172362, "learning_rate": 1.3572659544410493e-06, "loss": 0.4159, "step": 2666 }, { "epoch": 1.5690536614071866, "grad_norm": 0.3119424094658967, "learning_rate": 1.3537515701732063e-06, "loss": 0.3667, "step": 2667 }, { "epoch": 1.569642134686822, "grad_norm": 0.34827902410002626, "learning_rate": 1.350241029194953e-06, "loss": 0.4428, "step": 2668 }, { "epoch": 1.570230607966457, "grad_norm": 0.37303879955246033, "learning_rate": 1.3467343352065349e-06, "loss": 0.4384, "step": 2669 }, { "epoch": 1.5708190812460923, "grad_norm": 0.3527293003519052, "learning_rate": 1.3432314919041478e-06, "loss": 0.4068, "step": 2670 }, { "epoch": 1.5714075545257273, "grad_norm": 0.3592794698456468, "learning_rate": 1.339732502979928e-06, "loss": 0.4308, "step": 2671 }, { "epoch": 1.5719960278053624, "grad_norm": 0.3474875645151671, "learning_rate": 1.336237372121944e-06, "loss": 0.4338, "step": 2672 }, { "epoch": 1.5725845010849975, "grad_norm": 0.37338814015303357, "learning_rate": 1.3327461030142037e-06, "loss": 0.4373, "step": 2673 }, { "epoch": 1.5731729743646328, "grad_norm": 0.3677312159909903, "learning_rate": 1.329258699336644e-06, "loss": 0.4675, "step": 2674 }, { "epoch": 1.573761447644268, "grad_norm": 0.4111574015494885, "learning_rate": 1.3257751647651223e-06, "loss": 0.4379, "step": 2675 }, { "epoch": 1.5743499209239031, "grad_norm": 0.35853474091550513, "learning_rate": 1.322295502971423e-06, "loss": 0.4112, "step": 2676 }, { "epoch": 1.5749383942035382, "grad_norm": 0.36516589887494216, "learning_rate": 1.3188197176232486e-06, "loss": 0.4459, "step": 2677 }, { "epoch": 1.5755268674831733, "grad_norm": 0.367307948689375, "learning_rate": 1.3153478123842111e-06, "loss": 0.456, "step": 2678 }, { "epoch": 1.5761153407628083, "grad_norm": 0.3594356579042598, "learning_rate": 1.3118797909138364e-06, "loss": 0.4549, "step": 2679 }, { "epoch": 1.5767038140424436, "grad_norm": 0.36000266401219133, "learning_rate": 1.3084156568675577e-06, "loss": 0.4885, "step": 2680 }, { "epoch": 1.577292287322079, "grad_norm": 0.37318110707422547, "learning_rate": 1.3049554138967052e-06, "loss": 0.4573, "step": 2681 }, { "epoch": 1.577880760601714, "grad_norm": 0.3769497817085647, "learning_rate": 1.3014990656485139e-06, "loss": 0.4346, "step": 2682 }, { "epoch": 1.578469233881349, "grad_norm": 0.3887041848742039, "learning_rate": 1.298046615766112e-06, "loss": 0.4921, "step": 2683 }, { "epoch": 1.5790577071609841, "grad_norm": 0.38873696778101474, "learning_rate": 1.2945980678885134e-06, "loss": 0.431, "step": 2684 }, { "epoch": 1.5796461804406192, "grad_norm": 0.34468169631268436, "learning_rate": 1.2911534256506297e-06, "loss": 0.3819, "step": 2685 }, { "epoch": 1.5802346537202545, "grad_norm": 0.3616312186025826, "learning_rate": 1.2877126926832446e-06, "loss": 0.423, "step": 2686 }, { "epoch": 1.5808231269998898, "grad_norm": 0.3934871872980845, "learning_rate": 1.2842758726130283e-06, "loss": 0.4923, "step": 2687 }, { "epoch": 1.5814116002795249, "grad_norm": 0.36138609071263095, "learning_rate": 1.2808429690625256e-06, "loss": 0.3764, "step": 2688 }, { "epoch": 1.58200007355916, "grad_norm": 0.39511594252055315, "learning_rate": 1.27741398565015e-06, "loss": 0.4373, "step": 2689 }, { "epoch": 1.582588546838795, "grad_norm": 0.3699517469665755, "learning_rate": 1.2739889259901866e-06, "loss": 0.428, "step": 2690 }, { "epoch": 1.58317702011843, "grad_norm": 0.40812010395100123, "learning_rate": 1.2705677936927841e-06, "loss": 0.4482, "step": 2691 }, { "epoch": 1.5837654933980654, "grad_norm": 0.37342450077716544, "learning_rate": 1.2671505923639477e-06, "loss": 0.4339, "step": 2692 }, { "epoch": 1.5843539666777007, "grad_norm": 0.35924726195149803, "learning_rate": 1.2637373256055445e-06, "loss": 0.4505, "step": 2693 }, { "epoch": 1.5849424399573357, "grad_norm": 0.35191022292951424, "learning_rate": 1.2603279970152937e-06, "loss": 0.4265, "step": 2694 }, { "epoch": 1.5855309132369708, "grad_norm": 0.3399204232689882, "learning_rate": 1.2569226101867587e-06, "loss": 0.4145, "step": 2695 }, { "epoch": 1.586119386516606, "grad_norm": 0.3290475594470617, "learning_rate": 1.2535211687093535e-06, "loss": 0.4015, "step": 2696 }, { "epoch": 1.5867078597962412, "grad_norm": 0.3562782249511113, "learning_rate": 1.2501236761683321e-06, "loss": 0.4103, "step": 2697 }, { "epoch": 1.5872963330758763, "grad_norm": 0.3425967753442431, "learning_rate": 1.246730136144783e-06, "loss": 0.3896, "step": 2698 }, { "epoch": 1.5878848063555115, "grad_norm": 0.36657259596436104, "learning_rate": 1.2433405522156334e-06, "loss": 0.4781, "step": 2699 }, { "epoch": 1.5884732796351466, "grad_norm": 0.3736279265341975, "learning_rate": 1.2399549279536383e-06, "loss": 0.4022, "step": 2700 }, { "epoch": 1.5890617529147817, "grad_norm": 0.39277571343640827, "learning_rate": 1.2365732669273778e-06, "loss": 0.4727, "step": 2701 }, { "epoch": 1.5896502261944168, "grad_norm": 0.35080304596279493, "learning_rate": 1.2331955727012562e-06, "loss": 0.4016, "step": 2702 }, { "epoch": 1.590238699474052, "grad_norm": 0.3093022586449962, "learning_rate": 1.229821848835499e-06, "loss": 0.3764, "step": 2703 }, { "epoch": 1.5908271727536871, "grad_norm": 0.34566871544582345, "learning_rate": 1.22645209888614e-06, "loss": 0.3938, "step": 2704 }, { "epoch": 1.5914156460333224, "grad_norm": 0.383798906246259, "learning_rate": 1.2230863264050308e-06, "loss": 0.459, "step": 2705 }, { "epoch": 1.5920041193129575, "grad_norm": 0.3918675804549293, "learning_rate": 1.2197245349398296e-06, "loss": 0.4986, "step": 2706 }, { "epoch": 1.5925925925925926, "grad_norm": 0.3462202970065113, "learning_rate": 1.2163667280339937e-06, "loss": 0.4323, "step": 2707 }, { "epoch": 1.5931810658722276, "grad_norm": 0.3486908610559686, "learning_rate": 1.213012909226786e-06, "loss": 0.3711, "step": 2708 }, { "epoch": 1.593769539151863, "grad_norm": 0.3307183335047943, "learning_rate": 1.2096630820532652e-06, "loss": 0.3879, "step": 2709 }, { "epoch": 1.594358012431498, "grad_norm": 0.3719572530535317, "learning_rate": 1.2063172500442777e-06, "loss": 0.5049, "step": 2710 }, { "epoch": 1.5949464857111333, "grad_norm": 0.3915497364701786, "learning_rate": 1.202975416726464e-06, "loss": 0.3809, "step": 2711 }, { "epoch": 1.5955349589907684, "grad_norm": 0.3659659504102223, "learning_rate": 1.1996375856222503e-06, "loss": 0.4171, "step": 2712 }, { "epoch": 1.5961234322704034, "grad_norm": 0.40262423492350746, "learning_rate": 1.1963037602498385e-06, "loss": 0.4773, "step": 2713 }, { "epoch": 1.5967119055500385, "grad_norm": 0.3414411824668187, "learning_rate": 1.1929739441232136e-06, "loss": 0.42, "step": 2714 }, { "epoch": 1.5973003788296738, "grad_norm": 0.42841517522346495, "learning_rate": 1.1896481407521343e-06, "loss": 0.5104, "step": 2715 }, { "epoch": 1.5978888521093089, "grad_norm": 0.344088824442517, "learning_rate": 1.1863263536421261e-06, "loss": 0.4502, "step": 2716 }, { "epoch": 1.5984773253889442, "grad_norm": 0.40159105525674926, "learning_rate": 1.1830085862944851e-06, "loss": 0.4086, "step": 2717 }, { "epoch": 1.5990657986685792, "grad_norm": 0.3558905300634563, "learning_rate": 1.1796948422062693e-06, "loss": 0.3831, "step": 2718 }, { "epoch": 1.5996542719482143, "grad_norm": 0.37180126220136533, "learning_rate": 1.176385124870294e-06, "loss": 0.3911, "step": 2719 }, { "epoch": 1.6002427452278494, "grad_norm": 0.37622597664371743, "learning_rate": 1.1730794377751327e-06, "loss": 0.4351, "step": 2720 }, { "epoch": 1.6008312185074847, "grad_norm": 0.365824832886149, "learning_rate": 1.1697777844051105e-06, "loss": 0.4265, "step": 2721 }, { "epoch": 1.6014196917871197, "grad_norm": 0.3558880177828918, "learning_rate": 1.1664801682402988e-06, "loss": 0.3659, "step": 2722 }, { "epoch": 1.602008165066755, "grad_norm": 0.3840088844383575, "learning_rate": 1.163186592756515e-06, "loss": 0.4818, "step": 2723 }, { "epoch": 1.60259663834639, "grad_norm": 0.3572556218036145, "learning_rate": 1.1598970614253202e-06, "loss": 0.4343, "step": 2724 }, { "epoch": 1.6031851116260252, "grad_norm": 0.3966914627522023, "learning_rate": 1.1566115777140069e-06, "loss": 0.4763, "step": 2725 }, { "epoch": 1.6037735849056602, "grad_norm": 0.3671188669237545, "learning_rate": 1.1533301450856054e-06, "loss": 0.4201, "step": 2726 }, { "epoch": 1.6043620581852955, "grad_norm": 0.36031440884363825, "learning_rate": 1.1500527669988765e-06, "loss": 0.4143, "step": 2727 }, { "epoch": 1.6049505314649308, "grad_norm": 0.35360583873631407, "learning_rate": 1.1467794469083039e-06, "loss": 0.4296, "step": 2728 }, { "epoch": 1.605539004744566, "grad_norm": 0.3581877960896892, "learning_rate": 1.1435101882640964e-06, "loss": 0.3989, "step": 2729 }, { "epoch": 1.606127478024201, "grad_norm": 0.4793477811612881, "learning_rate": 1.1402449945121842e-06, "loss": 0.486, "step": 2730 }, { "epoch": 1.606715951303836, "grad_norm": 0.35191575134072595, "learning_rate": 1.1369838690942059e-06, "loss": 0.438, "step": 2731 }, { "epoch": 1.6073044245834711, "grad_norm": 0.37535922655446763, "learning_rate": 1.1337268154475184e-06, "loss": 0.3737, "step": 2732 }, { "epoch": 1.6078928978631064, "grad_norm": 0.4164816290977961, "learning_rate": 1.130473837005186e-06, "loss": 0.4793, "step": 2733 }, { "epoch": 1.6084813711427417, "grad_norm": 0.3577256747873454, "learning_rate": 1.1272249371959737e-06, "loss": 0.4321, "step": 2734 }, { "epoch": 1.6090698444223768, "grad_norm": 0.37150934311484624, "learning_rate": 1.1239801194443507e-06, "loss": 0.4035, "step": 2735 }, { "epoch": 1.6096583177020118, "grad_norm": 0.33241119271515396, "learning_rate": 1.1207393871704852e-06, "loss": 0.3715, "step": 2736 }, { "epoch": 1.610246790981647, "grad_norm": 0.37620755891869845, "learning_rate": 1.117502743790233e-06, "loss": 0.3997, "step": 2737 }, { "epoch": 1.610835264261282, "grad_norm": 0.3483977831811481, "learning_rate": 1.1142701927151456e-06, "loss": 0.3825, "step": 2738 }, { "epoch": 1.6114237375409173, "grad_norm": 0.3614065190853514, "learning_rate": 1.1110417373524607e-06, "loss": 0.4412, "step": 2739 }, { "epoch": 1.6120122108205526, "grad_norm": 0.3371730450787164, "learning_rate": 1.1078173811050946e-06, "loss": 0.4041, "step": 2740 }, { "epoch": 1.6126006841001876, "grad_norm": 0.3414571645475162, "learning_rate": 1.1045971273716476e-06, "loss": 0.3797, "step": 2741 }, { "epoch": 1.6131891573798227, "grad_norm": 0.39640147409017046, "learning_rate": 1.1013809795463954e-06, "loss": 0.4715, "step": 2742 }, { "epoch": 1.6137776306594578, "grad_norm": 0.38373502382072305, "learning_rate": 1.0981689410192815e-06, "loss": 0.4707, "step": 2743 }, { "epoch": 1.6143661039390929, "grad_norm": 0.36270023430622056, "learning_rate": 1.0949610151759233e-06, "loss": 0.4084, "step": 2744 }, { "epoch": 1.6149545772187281, "grad_norm": 0.3932206767314402, "learning_rate": 1.091757205397601e-06, "loss": 0.4755, "step": 2745 }, { "epoch": 1.6155430504983634, "grad_norm": 0.36865849062664513, "learning_rate": 1.0885575150612549e-06, "loss": 0.413, "step": 2746 }, { "epoch": 1.6161315237779985, "grad_norm": 0.38509971126844084, "learning_rate": 1.085361947539486e-06, "loss": 0.4518, "step": 2747 }, { "epoch": 1.6167199970576336, "grad_norm": 0.3765467326364313, "learning_rate": 1.0821705062005494e-06, "loss": 0.5161, "step": 2748 }, { "epoch": 1.6173084703372687, "grad_norm": 0.35590397867280515, "learning_rate": 1.0789831944083478e-06, "loss": 0.4195, "step": 2749 }, { "epoch": 1.617896943616904, "grad_norm": 0.35977629836396935, "learning_rate": 1.0758000155224341e-06, "loss": 0.389, "step": 2750 }, { "epoch": 1.618485416896539, "grad_norm": 0.36907220516113276, "learning_rate": 1.072620972898007e-06, "loss": 0.4166, "step": 2751 }, { "epoch": 1.6190738901761743, "grad_norm": 0.35858884626320037, "learning_rate": 1.0694460698858982e-06, "loss": 0.4364, "step": 2752 }, { "epoch": 1.6196623634558094, "grad_norm": 0.38765443107099445, "learning_rate": 1.066275309832584e-06, "loss": 0.5128, "step": 2753 }, { "epoch": 1.6202508367354445, "grad_norm": 0.3811249937575916, "learning_rate": 1.06310869608017e-06, "loss": 0.4029, "step": 2754 }, { "epoch": 1.6208393100150795, "grad_norm": 0.39437399915541066, "learning_rate": 1.0599462319663906e-06, "loss": 0.4731, "step": 2755 }, { "epoch": 1.6214277832947148, "grad_norm": 0.3508760874485875, "learning_rate": 1.0567879208246084e-06, "loss": 0.3618, "step": 2756 }, { "epoch": 1.62201625657435, "grad_norm": 0.3833129216589638, "learning_rate": 1.0536337659838092e-06, "loss": 0.3908, "step": 2757 }, { "epoch": 1.6226047298539852, "grad_norm": 0.3454321469189728, "learning_rate": 1.0504837707685945e-06, "loss": 0.3956, "step": 2758 }, { "epoch": 1.6231932031336203, "grad_norm": 0.3244577686030128, "learning_rate": 1.0473379384991833e-06, "loss": 0.3851, "step": 2759 }, { "epoch": 1.6237816764132553, "grad_norm": 0.34853834617282736, "learning_rate": 1.0441962724914095e-06, "loss": 0.4396, "step": 2760 }, { "epoch": 1.6243701496928904, "grad_norm": 0.3411247813244153, "learning_rate": 1.0410587760567104e-06, "loss": 0.3691, "step": 2761 }, { "epoch": 1.6249586229725257, "grad_norm": 0.37692070939742317, "learning_rate": 1.037925452502131e-06, "loss": 0.466, "step": 2762 }, { "epoch": 1.6255470962521608, "grad_norm": 0.3381178008589521, "learning_rate": 1.0347963051303201e-06, "loss": 0.4251, "step": 2763 }, { "epoch": 1.626135569531796, "grad_norm": 0.42768781858925825, "learning_rate": 1.0316713372395192e-06, "loss": 0.4461, "step": 2764 }, { "epoch": 1.6267240428114311, "grad_norm": 0.40357636079888837, "learning_rate": 1.02855055212357e-06, "loss": 0.4655, "step": 2765 }, { "epoch": 1.6273125160910662, "grad_norm": 0.38187391039488244, "learning_rate": 1.0254339530719031e-06, "loss": 0.4068, "step": 2766 }, { "epoch": 1.6279009893707013, "grad_norm": 0.3758291004870841, "learning_rate": 1.022321543369535e-06, "loss": 0.5019, "step": 2767 }, { "epoch": 1.6284894626503366, "grad_norm": 0.35567235638323424, "learning_rate": 1.0192133262970688e-06, "loss": 0.4077, "step": 2768 }, { "epoch": 1.6290779359299716, "grad_norm": 0.3995581234620785, "learning_rate": 1.0161093051306896e-06, "loss": 0.4474, "step": 2769 }, { "epoch": 1.629666409209607, "grad_norm": 0.3742300855347833, "learning_rate": 1.0130094831421555e-06, "loss": 0.4113, "step": 2770 }, { "epoch": 1.630254882489242, "grad_norm": 0.3745573201494477, "learning_rate": 1.0099138635988026e-06, "loss": 0.4097, "step": 2771 }, { "epoch": 1.630843355768877, "grad_norm": 0.37907717307020095, "learning_rate": 1.006822449763537e-06, "loss": 0.4721, "step": 2772 }, { "epoch": 1.6314318290485121, "grad_norm": 0.37372621571008985, "learning_rate": 1.0037352448948278e-06, "loss": 0.4299, "step": 2773 }, { "epoch": 1.6320203023281474, "grad_norm": 0.36773043328803734, "learning_rate": 1.0006522522467121e-06, "loss": 0.3929, "step": 2774 }, { "epoch": 1.6326087756077827, "grad_norm": 0.34604034926555005, "learning_rate": 9.975734750687876e-07, "loss": 0.42, "step": 2775 }, { "epoch": 1.6331972488874178, "grad_norm": 0.37596956949292004, "learning_rate": 9.944989166062031e-07, "loss": 0.4468, "step": 2776 }, { "epoch": 1.6337857221670529, "grad_norm": 0.3465398252250362, "learning_rate": 9.91428580099667e-07, "loss": 0.4019, "step": 2777 }, { "epoch": 1.634374195446688, "grad_norm": 0.3670067132554242, "learning_rate": 9.883624687854354e-07, "loss": 0.4187, "step": 2778 }, { "epoch": 1.634962668726323, "grad_norm": 0.3422949951232779, "learning_rate": 9.85300585895308e-07, "loss": 0.4103, "step": 2779 }, { "epoch": 1.6355511420059583, "grad_norm": 0.35616823797955666, "learning_rate": 9.822429346566314e-07, "loss": 0.4473, "step": 2780 }, { "epoch": 1.6361396152855936, "grad_norm": 0.31933702142702514, "learning_rate": 9.791895182922911e-07, "loss": 0.4016, "step": 2781 }, { "epoch": 1.6367280885652287, "grad_norm": 0.3476605051961055, "learning_rate": 9.761403400207081e-07, "loss": 0.41, "step": 2782 }, { "epoch": 1.6373165618448637, "grad_norm": 0.39220698032247886, "learning_rate": 9.73095403055837e-07, "loss": 0.4258, "step": 2783 }, { "epoch": 1.6379050351244988, "grad_norm": 0.36235795763756845, "learning_rate": 9.7005471060716e-07, "loss": 0.3964, "step": 2784 }, { "epoch": 1.6384935084041339, "grad_norm": 0.33809331068294307, "learning_rate": 9.670182658796872e-07, "loss": 0.4104, "step": 2785 }, { "epoch": 1.6390819816837692, "grad_norm": 0.36295672606219337, "learning_rate": 9.639860720739524e-07, "loss": 0.3877, "step": 2786 }, { "epoch": 1.6396704549634045, "grad_norm": 0.38426970941264044, "learning_rate": 9.609581323860056e-07, "loss": 0.4151, "step": 2787 }, { "epoch": 1.6402589282430395, "grad_norm": 0.34256170777308237, "learning_rate": 9.579344500074156e-07, "loss": 0.4527, "step": 2788 }, { "epoch": 1.6408474015226746, "grad_norm": 0.34495598382535086, "learning_rate": 9.549150281252633e-07, "loss": 0.3749, "step": 2789 }, { "epoch": 1.6414358748023097, "grad_norm": 0.42943507881384957, "learning_rate": 9.518998699221366e-07, "loss": 0.5067, "step": 2790 }, { "epoch": 1.6420243480819448, "grad_norm": 0.3821401906033587, "learning_rate": 9.488889785761324e-07, "loss": 0.4784, "step": 2791 }, { "epoch": 1.64261282136158, "grad_norm": 0.36128733748345065, "learning_rate": 9.458823572608494e-07, "loss": 0.3968, "step": 2792 }, { "epoch": 1.6432012946412153, "grad_norm": 0.3387529453939102, "learning_rate": 9.428800091453838e-07, "loss": 0.3873, "step": 2793 }, { "epoch": 1.6437897679208504, "grad_norm": 0.34722974163899306, "learning_rate": 9.398819373943291e-07, "loss": 0.3652, "step": 2794 }, { "epoch": 1.6443782412004855, "grad_norm": 0.3714747220166461, "learning_rate": 9.368881451677725e-07, "loss": 0.429, "step": 2795 }, { "epoch": 1.6449667144801206, "grad_norm": 0.3993741878976646, "learning_rate": 9.338986356212865e-07, "loss": 0.528, "step": 2796 }, { "epoch": 1.6455551877597556, "grad_norm": 0.3341821676612746, "learning_rate": 9.309134119059337e-07, "loss": 0.4056, "step": 2797 }, { "epoch": 1.646143661039391, "grad_norm": 0.40872068567922165, "learning_rate": 9.279324771682586e-07, "loss": 0.4282, "step": 2798 }, { "epoch": 1.6467321343190262, "grad_norm": 0.38700399526663937, "learning_rate": 9.249558345502807e-07, "loss": 0.424, "step": 2799 }, { "epoch": 1.6473206075986613, "grad_norm": 0.35391785513314844, "learning_rate": 9.219834871895011e-07, "loss": 0.4233, "step": 2800 }, { "epoch": 1.6479090808782964, "grad_norm": 0.3449348947675704, "learning_rate": 9.190154382188921e-07, "loss": 0.39, "step": 2801 }, { "epoch": 1.6484975541579314, "grad_norm": 0.3514266782642334, "learning_rate": 9.160516907668909e-07, "loss": 0.4492, "step": 2802 }, { "epoch": 1.6490860274375667, "grad_norm": 0.396605817906325, "learning_rate": 9.130922479574067e-07, "loss": 0.4063, "step": 2803 }, { "epoch": 1.6496745007172018, "grad_norm": 0.42432826407487473, "learning_rate": 9.101371129098091e-07, "loss": 0.4702, "step": 2804 }, { "epoch": 1.650262973996837, "grad_norm": 0.35858234079231344, "learning_rate": 9.071862887389255e-07, "loss": 0.4245, "step": 2805 }, { "epoch": 1.6508514472764722, "grad_norm": 0.36261128971807655, "learning_rate": 9.042397785550405e-07, "loss": 0.3762, "step": 2806 }, { "epoch": 1.6514399205561072, "grad_norm": 0.38468931232145925, "learning_rate": 9.01297585463895e-07, "loss": 0.4433, "step": 2807 }, { "epoch": 1.6520283938357423, "grad_norm": 0.3809808384302746, "learning_rate": 8.983597125666727e-07, "loss": 0.4251, "step": 2808 }, { "epoch": 1.6526168671153776, "grad_norm": 0.40124482046568116, "learning_rate": 8.954261629600097e-07, "loss": 0.4131, "step": 2809 }, { "epoch": 1.6532053403950127, "grad_norm": 0.3718061811454687, "learning_rate": 8.924969397359834e-07, "loss": 0.4358, "step": 2810 }, { "epoch": 1.653793813674648, "grad_norm": 0.3980175906292977, "learning_rate": 8.895720459821089e-07, "loss": 0.4133, "step": 2811 }, { "epoch": 1.654382286954283, "grad_norm": 0.3578803466686018, "learning_rate": 8.866514847813395e-07, "loss": 0.4833, "step": 2812 }, { "epoch": 1.654970760233918, "grad_norm": 0.4175252821982685, "learning_rate": 8.837352592120646e-07, "loss": 0.4559, "step": 2813 }, { "epoch": 1.6555592335135532, "grad_norm": 0.3757755865094569, "learning_rate": 8.808233723480968e-07, "loss": 0.4684, "step": 2814 }, { "epoch": 1.6561477067931885, "grad_norm": 0.3581521776355447, "learning_rate": 8.779158272586824e-07, "loss": 0.4123, "step": 2815 }, { "epoch": 1.6567361800728235, "grad_norm": 0.3414518872124365, "learning_rate": 8.750126270084891e-07, "loss": 0.3745, "step": 2816 }, { "epoch": 1.6573246533524588, "grad_norm": 0.3682930867162556, "learning_rate": 8.721137746576019e-07, "loss": 0.4202, "step": 2817 }, { "epoch": 1.657913126632094, "grad_norm": 0.3589145769331684, "learning_rate": 8.692192732615267e-07, "loss": 0.4343, "step": 2818 }, { "epoch": 1.658501599911729, "grad_norm": 0.3660969522188726, "learning_rate": 8.663291258711831e-07, "loss": 0.4217, "step": 2819 }, { "epoch": 1.659090073191364, "grad_norm": 0.3703731531298175, "learning_rate": 8.634433355328986e-07, "loss": 0.4071, "step": 2820 }, { "epoch": 1.6596785464709993, "grad_norm": 0.37421162486517046, "learning_rate": 8.605619052884106e-07, "loss": 0.4867, "step": 2821 }, { "epoch": 1.6602670197506344, "grad_norm": 0.39486780436650504, "learning_rate": 8.57684838174862e-07, "loss": 0.5041, "step": 2822 }, { "epoch": 1.6608554930302697, "grad_norm": 0.3817247602738236, "learning_rate": 8.54812137224792e-07, "loss": 0.4746, "step": 2823 }, { "epoch": 1.6614439663099048, "grad_norm": 0.3877031623765114, "learning_rate": 8.519438054661416e-07, "loss": 0.4345, "step": 2824 }, { "epoch": 1.6620324395895398, "grad_norm": 0.3443369308775897, "learning_rate": 8.490798459222477e-07, "loss": 0.3786, "step": 2825 }, { "epoch": 1.662620912869175, "grad_norm": 0.39493338455493743, "learning_rate": 8.462202616118331e-07, "loss": 0.5239, "step": 2826 }, { "epoch": 1.6632093861488102, "grad_norm": 0.3675987955670444, "learning_rate": 8.433650555490142e-07, "loss": 0.47, "step": 2827 }, { "epoch": 1.6637978594284455, "grad_norm": 0.33671573576794817, "learning_rate": 8.405142307432917e-07, "loss": 0.4171, "step": 2828 }, { "epoch": 1.6643863327080806, "grad_norm": 0.35573307167012735, "learning_rate": 8.376677901995456e-07, "loss": 0.4024, "step": 2829 }, { "epoch": 1.6649748059877156, "grad_norm": 0.37048984061313534, "learning_rate": 8.348257369180368e-07, "loss": 0.4464, "step": 2830 }, { "epoch": 1.6655632792673507, "grad_norm": 0.42245012501157764, "learning_rate": 8.31988073894403e-07, "loss": 0.4858, "step": 2831 }, { "epoch": 1.6661517525469858, "grad_norm": 0.36935779680649805, "learning_rate": 8.291548041196507e-07, "loss": 0.4711, "step": 2832 }, { "epoch": 1.666740225826621, "grad_norm": 0.3756624009640286, "learning_rate": 8.263259305801591e-07, "loss": 0.5037, "step": 2833 }, { "epoch": 1.6673286991062564, "grad_norm": 0.38682984772547435, "learning_rate": 8.235014562576732e-07, "loss": 0.4391, "step": 2834 }, { "epoch": 1.6679171723858914, "grad_norm": 0.36681421768332106, "learning_rate": 8.206813841292983e-07, "loss": 0.4403, "step": 2835 }, { "epoch": 1.6685056456655265, "grad_norm": 0.36790707399806827, "learning_rate": 8.178657171675025e-07, "loss": 0.4549, "step": 2836 }, { "epoch": 1.6690941189451616, "grad_norm": 0.36859523337483474, "learning_rate": 8.150544583401116e-07, "loss": 0.4813, "step": 2837 }, { "epoch": 1.6696825922247966, "grad_norm": 0.3916800455099263, "learning_rate": 8.122476106103e-07, "loss": 0.48, "step": 2838 }, { "epoch": 1.670271065504432, "grad_norm": 0.39430333275673374, "learning_rate": 8.094451769365974e-07, "loss": 0.453, "step": 2839 }, { "epoch": 1.6708595387840672, "grad_norm": 0.40532981590180717, "learning_rate": 8.066471602728804e-07, "loss": 0.4803, "step": 2840 }, { "epoch": 1.6714480120637023, "grad_norm": 0.3660964604778748, "learning_rate": 8.03853563568367e-07, "loss": 0.4168, "step": 2841 }, { "epoch": 1.6720364853433374, "grad_norm": 0.37980527268495295, "learning_rate": 8.01064389767619e-07, "loss": 0.4753, "step": 2842 }, { "epoch": 1.6726249586229724, "grad_norm": 0.3464803031158565, "learning_rate": 7.98279641810537e-07, "loss": 0.3791, "step": 2843 }, { "epoch": 1.6732134319026075, "grad_norm": 0.38128815286161066, "learning_rate": 7.954993226323527e-07, "loss": 0.4259, "step": 2844 }, { "epoch": 1.6738019051822428, "grad_norm": 0.3020848051088086, "learning_rate": 7.927234351636331e-07, "loss": 0.3679, "step": 2845 }, { "epoch": 1.674390378461878, "grad_norm": 0.3849848641710649, "learning_rate": 7.899519823302743e-07, "loss": 0.4809, "step": 2846 }, { "epoch": 1.6749788517415132, "grad_norm": 0.3826827609632322, "learning_rate": 7.871849670534947e-07, "loss": 0.445, "step": 2847 }, { "epoch": 1.6755673250211482, "grad_norm": 0.37290588614438125, "learning_rate": 7.844223922498384e-07, "loss": 0.4376, "step": 2848 }, { "epoch": 1.6761557983007833, "grad_norm": 0.36041227738104314, "learning_rate": 7.816642608311692e-07, "loss": 0.4085, "step": 2849 }, { "epoch": 1.6767442715804184, "grad_norm": 0.35603891831620677, "learning_rate": 7.789105757046644e-07, "loss": 0.4614, "step": 2850 }, { "epoch": 1.6773327448600537, "grad_norm": 0.3673587057361143, "learning_rate": 7.761613397728174e-07, "loss": 0.4597, "step": 2851 }, { "epoch": 1.677921218139689, "grad_norm": 0.3575475890091003, "learning_rate": 7.734165559334327e-07, "loss": 0.3755, "step": 2852 }, { "epoch": 1.678509691419324, "grad_norm": 0.3249318403054078, "learning_rate": 7.706762270796186e-07, "loss": 0.413, "step": 2853 }, { "epoch": 1.6790981646989591, "grad_norm": 0.4082776821175431, "learning_rate": 7.679403560997906e-07, "loss": 0.471, "step": 2854 }, { "epoch": 1.6796866379785942, "grad_norm": 0.42234280596157275, "learning_rate": 7.652089458776651e-07, "loss": 0.4524, "step": 2855 }, { "epoch": 1.6802751112582295, "grad_norm": 0.3802076229773006, "learning_rate": 7.62481999292255e-07, "loss": 0.4071, "step": 2856 }, { "epoch": 1.6808635845378646, "grad_norm": 0.40284563211569885, "learning_rate": 7.597595192178702e-07, "loss": 0.4646, "step": 2857 }, { "epoch": 1.6814520578174998, "grad_norm": 0.36249385355498553, "learning_rate": 7.570415085241129e-07, "loss": 0.3933, "step": 2858 }, { "epoch": 1.682040531097135, "grad_norm": 0.46304102386733276, "learning_rate": 7.543279700758715e-07, "loss": 0.5175, "step": 2859 }, { "epoch": 1.68262900437677, "grad_norm": 0.3864940820523421, "learning_rate": 7.516189067333241e-07, "loss": 0.4898, "step": 2860 }, { "epoch": 1.683217477656405, "grad_norm": 0.3542777758319617, "learning_rate": 7.489143213519301e-07, "loss": 0.4328, "step": 2861 }, { "epoch": 1.6838059509360404, "grad_norm": 0.384178415652854, "learning_rate": 7.462142167824277e-07, "loss": 0.3957, "step": 2862 }, { "epoch": 1.6843944242156754, "grad_norm": 0.36014785381675757, "learning_rate": 7.435185958708352e-07, "loss": 0.4443, "step": 2863 }, { "epoch": 1.6849828974953107, "grad_norm": 0.3811232000217508, "learning_rate": 7.408274614584432e-07, "loss": 0.4678, "step": 2864 }, { "epoch": 1.6855713707749458, "grad_norm": 0.37354083915485214, "learning_rate": 7.381408163818121e-07, "loss": 0.3729, "step": 2865 }, { "epoch": 1.6861598440545809, "grad_norm": 0.31173515824921455, "learning_rate": 7.354586634727729e-07, "loss": 0.3646, "step": 2866 }, { "epoch": 1.686748317334216, "grad_norm": 0.37955595827126365, "learning_rate": 7.327810055584211e-07, "loss": 0.3843, "step": 2867 }, { "epoch": 1.6873367906138512, "grad_norm": 0.3831203643888604, "learning_rate": 7.301078454611116e-07, "loss": 0.5355, "step": 2868 }, { "epoch": 1.6879252638934863, "grad_norm": 0.3590966977264916, "learning_rate": 7.27439185998462e-07, "loss": 0.4112, "step": 2869 }, { "epoch": 1.6885137371731216, "grad_norm": 0.39086099853109385, "learning_rate": 7.24775029983345e-07, "loss": 0.4521, "step": 2870 }, { "epoch": 1.6891022104527567, "grad_norm": 0.3578664937944716, "learning_rate": 7.221153802238845e-07, "loss": 0.4181, "step": 2871 }, { "epoch": 1.6896906837323917, "grad_norm": 0.3551238629969985, "learning_rate": 7.194602395234573e-07, "loss": 0.4242, "step": 2872 }, { "epoch": 1.6902791570120268, "grad_norm": 0.4111398482533741, "learning_rate": 7.168096106806871e-07, "loss": 0.4675, "step": 2873 }, { "epoch": 1.690867630291662, "grad_norm": 0.37149200364342555, "learning_rate": 7.141634964894389e-07, "loss": 0.4833, "step": 2874 }, { "epoch": 1.6914561035712972, "grad_norm": 0.3354641977737812, "learning_rate": 7.115218997388229e-07, "loss": 0.412, "step": 2875 }, { "epoch": 1.6920445768509325, "grad_norm": 0.3263551173706522, "learning_rate": 7.088848232131862e-07, "loss": 0.4032, "step": 2876 }, { "epoch": 1.6926330501305675, "grad_norm": 0.3784525434644735, "learning_rate": 7.062522696921109e-07, "loss": 0.4299, "step": 2877 }, { "epoch": 1.6932215234102026, "grad_norm": 0.39989057092522945, "learning_rate": 7.036242419504141e-07, "loss": 0.4673, "step": 2878 }, { "epoch": 1.6938099966898377, "grad_norm": 0.33777244273015217, "learning_rate": 7.010007427581378e-07, "loss": 0.4006, "step": 2879 }, { "epoch": 1.694398469969473, "grad_norm": 0.37623012345335294, "learning_rate": 6.983817748805549e-07, "loss": 0.423, "step": 2880 }, { "epoch": 1.6949869432491083, "grad_norm": 0.3676472000588866, "learning_rate": 6.957673410781617e-07, "loss": 0.4045, "step": 2881 }, { "epoch": 1.6955754165287433, "grad_norm": 0.38056925176235096, "learning_rate": 6.931574441066725e-07, "loss": 0.4399, "step": 2882 }, { "epoch": 1.6961638898083784, "grad_norm": 0.38248366642137366, "learning_rate": 6.905520867170223e-07, "loss": 0.5129, "step": 2883 }, { "epoch": 1.6967523630880135, "grad_norm": 0.3334691071400812, "learning_rate": 6.879512716553622e-07, "loss": 0.4556, "step": 2884 }, { "epoch": 1.6973408363676485, "grad_norm": 0.36661612151432793, "learning_rate": 6.853550016630517e-07, "loss": 0.4215, "step": 2885 }, { "epoch": 1.6979293096472838, "grad_norm": 0.3645120463817611, "learning_rate": 6.82763279476662e-07, "loss": 0.4362, "step": 2886 }, { "epoch": 1.6985177829269191, "grad_norm": 0.3614475211862146, "learning_rate": 6.801761078279723e-07, "loss": 0.3975, "step": 2887 }, { "epoch": 1.6991062562065542, "grad_norm": 0.360034026348207, "learning_rate": 6.775934894439606e-07, "loss": 0.4326, "step": 2888 }, { "epoch": 1.6996947294861893, "grad_norm": 0.3488147947097181, "learning_rate": 6.750154270468102e-07, "loss": 0.4511, "step": 2889 }, { "epoch": 1.7002832027658243, "grad_norm": 0.36945048768936883, "learning_rate": 6.724419233539014e-07, "loss": 0.4719, "step": 2890 }, { "epoch": 1.7008716760454594, "grad_norm": 0.37200255197720095, "learning_rate": 6.698729810778065e-07, "loss": 0.4657, "step": 2891 }, { "epoch": 1.7014601493250947, "grad_norm": 0.37944926107323357, "learning_rate": 6.673086029262932e-07, "loss": 0.4409, "step": 2892 }, { "epoch": 1.70204862260473, "grad_norm": 0.375367895435164, "learning_rate": 6.647487916023176e-07, "loss": 0.4708, "step": 2893 }, { "epoch": 1.702637095884365, "grad_norm": 0.4477513731628918, "learning_rate": 6.6219354980402e-07, "loss": 0.529, "step": 2894 }, { "epoch": 1.7032255691640001, "grad_norm": 0.35483552441804317, "learning_rate": 6.596428802247267e-07, "loss": 0.4241, "step": 2895 }, { "epoch": 1.7038140424436352, "grad_norm": 0.36427726504347635, "learning_rate": 6.570967855529453e-07, "loss": 0.4804, "step": 2896 }, { "epoch": 1.7044025157232703, "grad_norm": 0.39670394063962955, "learning_rate": 6.545552684723583e-07, "loss": 0.4235, "step": 2897 }, { "epoch": 1.7049909890029056, "grad_norm": 0.3751711825628715, "learning_rate": 6.520183316618256e-07, "loss": 0.4505, "step": 2898 }, { "epoch": 1.7055794622825409, "grad_norm": 0.3501003759901407, "learning_rate": 6.494859777953793e-07, "loss": 0.4136, "step": 2899 }, { "epoch": 1.706167935562176, "grad_norm": 0.3700680650507097, "learning_rate": 6.469582095422183e-07, "loss": 0.3896, "step": 2900 }, { "epoch": 1.706756408841811, "grad_norm": 0.3828716890476839, "learning_rate": 6.444350295667112e-07, "loss": 0.474, "step": 2901 }, { "epoch": 1.707344882121446, "grad_norm": 0.4063717592215896, "learning_rate": 6.419164405283907e-07, "loss": 0.4758, "step": 2902 }, { "epoch": 1.7079333554010812, "grad_norm": 0.36236856951128826, "learning_rate": 6.394024450819458e-07, "loss": 0.3917, "step": 2903 }, { "epoch": 1.7085218286807164, "grad_norm": 0.36981240250435793, "learning_rate": 6.368930458772283e-07, "loss": 0.4597, "step": 2904 }, { "epoch": 1.7091103019603517, "grad_norm": 0.46074312811648505, "learning_rate": 6.343882455592454e-07, "loss": 0.4857, "step": 2905 }, { "epoch": 1.7096987752399868, "grad_norm": 0.37805514836885284, "learning_rate": 6.318880467681527e-07, "loss": 0.4245, "step": 2906 }, { "epoch": 1.7102872485196219, "grad_norm": 0.3761187952337355, "learning_rate": 6.293924521392591e-07, "loss": 0.4298, "step": 2907 }, { "epoch": 1.710875721799257, "grad_norm": 0.34448977297864153, "learning_rate": 6.269014643030214e-07, "loss": 0.4237, "step": 2908 }, { "epoch": 1.7114641950788922, "grad_norm": 0.3628570946443965, "learning_rate": 6.244150858850368e-07, "loss": 0.4494, "step": 2909 }, { "epoch": 1.7120526683585273, "grad_norm": 0.3470104020847903, "learning_rate": 6.219333195060468e-07, "loss": 0.3786, "step": 2910 }, { "epoch": 1.7126411416381626, "grad_norm": 0.32888146564100046, "learning_rate": 6.194561677819327e-07, "loss": 0.3632, "step": 2911 }, { "epoch": 1.7132296149177977, "grad_norm": 0.37541667587867017, "learning_rate": 6.169836333237073e-07, "loss": 0.4007, "step": 2912 }, { "epoch": 1.7138180881974328, "grad_norm": 0.40084453817339394, "learning_rate": 6.145157187375212e-07, "loss": 0.4595, "step": 2913 }, { "epoch": 1.7144065614770678, "grad_norm": 0.3603719554866741, "learning_rate": 6.120524266246536e-07, "loss": 0.3956, "step": 2914 }, { "epoch": 1.7149950347567031, "grad_norm": 0.34532199961983695, "learning_rate": 6.095937595815104e-07, "loss": 0.3696, "step": 2915 }, { "epoch": 1.7155835080363382, "grad_norm": 0.3511222735967298, "learning_rate": 6.071397201996243e-07, "loss": 0.4638, "step": 2916 }, { "epoch": 1.7161719813159735, "grad_norm": 0.44154308979183493, "learning_rate": 6.046903110656505e-07, "loss": 0.4422, "step": 2917 }, { "epoch": 1.7167604545956086, "grad_norm": 0.3639513805299209, "learning_rate": 6.022455347613609e-07, "loss": 0.43, "step": 2918 }, { "epoch": 1.7173489278752436, "grad_norm": 0.4811355859854881, "learning_rate": 5.998053938636472e-07, "loss": 0.5255, "step": 2919 }, { "epoch": 1.7179374011548787, "grad_norm": 0.35483082102334684, "learning_rate": 5.973698909445147e-07, "loss": 0.411, "step": 2920 }, { "epoch": 1.718525874434514, "grad_norm": 0.3705259078708883, "learning_rate": 5.949390285710777e-07, "loss": 0.4835, "step": 2921 }, { "epoch": 1.719114347714149, "grad_norm": 0.3511598375542283, "learning_rate": 5.925128093055621e-07, "loss": 0.4442, "step": 2922 }, { "epoch": 1.7197028209937844, "grad_norm": 0.40750471385256487, "learning_rate": 5.900912357052996e-07, "loss": 0.4381, "step": 2923 }, { "epoch": 1.7202912942734194, "grad_norm": 0.34377319891578095, "learning_rate": 5.876743103227217e-07, "loss": 0.4093, "step": 2924 }, { "epoch": 1.7208797675530545, "grad_norm": 0.3758803796791573, "learning_rate": 5.852620357053651e-07, "loss": 0.4214, "step": 2925 }, { "epoch": 1.7214682408326896, "grad_norm": 0.36154770630092414, "learning_rate": 5.828544143958625e-07, "loss": 0.4343, "step": 2926 }, { "epoch": 1.7220567141123249, "grad_norm": 0.3272056005011584, "learning_rate": 5.804514489319402e-07, "loss": 0.3665, "step": 2927 }, { "epoch": 1.72264518739196, "grad_norm": 0.38914445236919865, "learning_rate": 5.780531418464191e-07, "loss": 0.513, "step": 2928 }, { "epoch": 1.7232336606715952, "grad_norm": 0.338273194636521, "learning_rate": 5.756594956672107e-07, "loss": 0.3887, "step": 2929 }, { "epoch": 1.7238221339512303, "grad_norm": 0.345950165618757, "learning_rate": 5.7327051291731e-07, "loss": 0.4021, "step": 2930 }, { "epoch": 1.7244106072308654, "grad_norm": 0.37232371833767863, "learning_rate": 5.708861961148004e-07, "loss": 0.4639, "step": 2931 }, { "epoch": 1.7249990805105004, "grad_norm": 0.37755500945064785, "learning_rate": 5.685065477728469e-07, "loss": 0.4418, "step": 2932 }, { "epoch": 1.7255875537901357, "grad_norm": 0.4118787000084046, "learning_rate": 5.661315703996905e-07, "loss": 0.4384, "step": 2933 }, { "epoch": 1.726176027069771, "grad_norm": 0.3217672162811514, "learning_rate": 5.637612664986519e-07, "loss": 0.3538, "step": 2934 }, { "epoch": 1.726764500349406, "grad_norm": 0.37291382764117353, "learning_rate": 5.613956385681263e-07, "loss": 0.4134, "step": 2935 }, { "epoch": 1.7273529736290412, "grad_norm": 0.3348679851344902, "learning_rate": 5.590346891015758e-07, "loss": 0.4284, "step": 2936 }, { "epoch": 1.7279414469086762, "grad_norm": 0.38265545110567767, "learning_rate": 5.566784205875364e-07, "loss": 0.4533, "step": 2937 }, { "epoch": 1.7285299201883113, "grad_norm": 0.34908593315833275, "learning_rate": 5.543268355096082e-07, "loss": 0.4402, "step": 2938 }, { "epoch": 1.7291183934679466, "grad_norm": 0.3350070097747319, "learning_rate": 5.519799363464523e-07, "loss": 0.401, "step": 2939 }, { "epoch": 1.729706866747582, "grad_norm": 0.37070140017275105, "learning_rate": 5.496377255717949e-07, "loss": 0.4061, "step": 2940 }, { "epoch": 1.730295340027217, "grad_norm": 0.38756875223152654, "learning_rate": 5.473002056544191e-07, "loss": 0.4566, "step": 2941 }, { "epoch": 1.730883813306852, "grad_norm": 0.3355439205887286, "learning_rate": 5.449673790581611e-07, "loss": 0.3823, "step": 2942 }, { "epoch": 1.731472286586487, "grad_norm": 0.3950840272221328, "learning_rate": 5.426392482419124e-07, "loss": 0.4224, "step": 2943 }, { "epoch": 1.7320607598661222, "grad_norm": 0.35015256318353316, "learning_rate": 5.403158156596167e-07, "loss": 0.3773, "step": 2944 }, { "epoch": 1.7326492331457575, "grad_norm": 0.38030499584395805, "learning_rate": 5.379970837602611e-07, "loss": 0.4541, "step": 2945 }, { "epoch": 1.7332377064253928, "grad_norm": 0.4089577846635698, "learning_rate": 5.356830549878817e-07, "loss": 0.4762, "step": 2946 }, { "epoch": 1.7338261797050278, "grad_norm": 0.39883217494400824, "learning_rate": 5.333737317815563e-07, "loss": 0.4759, "step": 2947 }, { "epoch": 1.734414652984663, "grad_norm": 0.377580630401089, "learning_rate": 5.310691165754012e-07, "loss": 0.5038, "step": 2948 }, { "epoch": 1.735003126264298, "grad_norm": 0.3593124607714127, "learning_rate": 5.287692117985732e-07, "loss": 0.4362, "step": 2949 }, { "epoch": 1.735591599543933, "grad_norm": 0.3815663978557555, "learning_rate": 5.26474019875261e-07, "loss": 0.4375, "step": 2950 }, { "epoch": 1.7361800728235683, "grad_norm": 0.35133632048650687, "learning_rate": 5.241835432246888e-07, "loss": 0.4171, "step": 2951 }, { "epoch": 1.7367685461032036, "grad_norm": 0.40275516498333874, "learning_rate": 5.218977842611095e-07, "loss": 0.4243, "step": 2952 }, { "epoch": 1.7373570193828387, "grad_norm": 0.3306128312867849, "learning_rate": 5.196167453938011e-07, "loss": 0.3747, "step": 2953 }, { "epoch": 1.7379454926624738, "grad_norm": 0.3913049081420622, "learning_rate": 5.173404290270706e-07, "loss": 0.4655, "step": 2954 }, { "epoch": 1.7385339659421088, "grad_norm": 0.378869923854262, "learning_rate": 5.150688375602447e-07, "loss": 0.4131, "step": 2955 }, { "epoch": 1.7391224392217441, "grad_norm": 0.42219494504877936, "learning_rate": 5.128019733876699e-07, "loss": 0.4562, "step": 2956 }, { "epoch": 1.7397109125013792, "grad_norm": 0.3588778982229446, "learning_rate": 5.105398388987098e-07, "loss": 0.4281, "step": 2957 }, { "epoch": 1.7402993857810145, "grad_norm": 0.3553078385322833, "learning_rate": 5.082824364777455e-07, "loss": 0.4199, "step": 2958 }, { "epoch": 1.7408878590606496, "grad_norm": 0.37082519576378825, "learning_rate": 5.06029768504166e-07, "loss": 0.4961, "step": 2959 }, { "epoch": 1.7414763323402846, "grad_norm": 0.39519559819590944, "learning_rate": 5.037818373523723e-07, "loss": 0.4716, "step": 2960 }, { "epoch": 1.7420648056199197, "grad_norm": 0.35031020308054844, "learning_rate": 5.015386453917742e-07, "loss": 0.3745, "step": 2961 }, { "epoch": 1.742653278899555, "grad_norm": 0.32312832053490875, "learning_rate": 4.993001949867826e-07, "loss": 0.3729, "step": 2962 }, { "epoch": 1.74324175217919, "grad_norm": 0.36508364984808245, "learning_rate": 4.970664884968135e-07, "loss": 0.4584, "step": 2963 }, { "epoch": 1.7438302254588254, "grad_norm": 0.37111964712170015, "learning_rate": 4.948375282762818e-07, "loss": 0.409, "step": 2964 }, { "epoch": 1.7444186987384604, "grad_norm": 0.3144695107781569, "learning_rate": 4.926133166745978e-07, "loss": 0.3748, "step": 2965 }, { "epoch": 1.7450071720180955, "grad_norm": 0.37741364557645546, "learning_rate": 4.903938560361698e-07, "loss": 0.4815, "step": 2966 }, { "epoch": 1.7455956452977306, "grad_norm": 0.37955591006226796, "learning_rate": 4.881791487003967e-07, "loss": 0.4251, "step": 2967 }, { "epoch": 1.7461841185773659, "grad_norm": 0.3603654446516699, "learning_rate": 4.859691970016667e-07, "loss": 0.44, "step": 2968 }, { "epoch": 1.746772591857001, "grad_norm": 0.33256921322158656, "learning_rate": 4.837640032693558e-07, "loss": 0.4032, "step": 2969 }, { "epoch": 1.7473610651366362, "grad_norm": 0.3270068571773785, "learning_rate": 4.815635698278271e-07, "loss": 0.3728, "step": 2970 }, { "epoch": 1.7479495384162713, "grad_norm": 0.3849385651285699, "learning_rate": 4.793678989964207e-07, "loss": 0.4366, "step": 2971 }, { "epoch": 1.7485380116959064, "grad_norm": 0.5775997946732617, "learning_rate": 4.771769930894632e-07, "loss": 0.4571, "step": 2972 }, { "epoch": 1.7491264849755415, "grad_norm": 0.36348523894243584, "learning_rate": 4.7499085441625515e-07, "loss": 0.4077, "step": 2973 }, { "epoch": 1.7497149582551768, "grad_norm": 0.38733479879764904, "learning_rate": 4.7280948528107216e-07, "loss": 0.4534, "step": 2974 }, { "epoch": 1.7503034315348118, "grad_norm": 0.35347505370039495, "learning_rate": 4.7063288798316397e-07, "loss": 0.4123, "step": 2975 }, { "epoch": 1.7508919048144471, "grad_norm": 0.3641238653331403, "learning_rate": 4.6846106481675035e-07, "loss": 0.3729, "step": 2976 }, { "epoch": 1.7514803780940822, "grad_norm": 0.3834523684606994, "learning_rate": 4.662940180710168e-07, "loss": 0.3588, "step": 2977 }, { "epoch": 1.7520688513737173, "grad_norm": 0.3948672664618167, "learning_rate": 4.641317500301173e-07, "loss": 0.4336, "step": 2978 }, { "epoch": 1.7526573246533523, "grad_norm": 0.37371204030030986, "learning_rate": 4.619742629731683e-07, "loss": 0.4446, "step": 2979 }, { "epoch": 1.7532457979329876, "grad_norm": 0.578588824352222, "learning_rate": 4.598215591742444e-07, "loss": 0.5016, "step": 2980 }, { "epoch": 1.7538342712126227, "grad_norm": 0.39486813485832023, "learning_rate": 4.576736409023813e-07, "loss": 0.4307, "step": 2981 }, { "epoch": 1.754422744492258, "grad_norm": 0.366238805677391, "learning_rate": 4.5553051042157015e-07, "loss": 0.4585, "step": 2982 }, { "epoch": 1.755011217771893, "grad_norm": 0.3898685314528434, "learning_rate": 4.533921699907534e-07, "loss": 0.4376, "step": 2983 }, { "epoch": 1.7555996910515281, "grad_norm": 0.3640033019545571, "learning_rate": 4.512586218638265e-07, "loss": 0.3841, "step": 2984 }, { "epoch": 1.7561881643311632, "grad_norm": 0.35022786991150096, "learning_rate": 4.4912986828963555e-07, "loss": 0.4042, "step": 2985 }, { "epoch": 1.7567766376107985, "grad_norm": 0.3579313658071663, "learning_rate": 4.470059115119674e-07, "loss": 0.3978, "step": 2986 }, { "epoch": 1.7573651108904338, "grad_norm": 0.3585460491588702, "learning_rate": 4.448867537695578e-07, "loss": 0.465, "step": 2987 }, { "epoch": 1.7579535841700689, "grad_norm": 0.3998236721693461, "learning_rate": 4.42772397296084e-07, "loss": 0.4355, "step": 2988 }, { "epoch": 1.758542057449704, "grad_norm": 0.3482349441785635, "learning_rate": 4.4066284432015825e-07, "loss": 0.3731, "step": 2989 }, { "epoch": 1.759130530729339, "grad_norm": 0.39052893753148654, "learning_rate": 4.3855809706533357e-07, "loss": 0.4321, "step": 2990 }, { "epoch": 1.759719004008974, "grad_norm": 0.3880389351982452, "learning_rate": 4.364581577500987e-07, "loss": 0.4952, "step": 2991 }, { "epoch": 1.7603074772886094, "grad_norm": 0.35224988909916977, "learning_rate": 4.3436302858786914e-07, "loss": 0.4119, "step": 2992 }, { "epoch": 1.7608959505682447, "grad_norm": 0.3468123016643849, "learning_rate": 4.322727117869951e-07, "loss": 0.4259, "step": 2993 }, { "epoch": 1.7614844238478797, "grad_norm": 0.3547242757530285, "learning_rate": 4.301872095507542e-07, "loss": 0.4466, "step": 2994 }, { "epoch": 1.7620728971275148, "grad_norm": 0.3007633041102438, "learning_rate": 4.281065240773452e-07, "loss": 0.3441, "step": 2995 }, { "epoch": 1.7626613704071499, "grad_norm": 0.3507769608291761, "learning_rate": 4.2603065755989493e-07, "loss": 0.4509, "step": 2996 }, { "epoch": 1.763249843686785, "grad_norm": 0.3545871877161918, "learning_rate": 4.239596121864481e-07, "loss": 0.4518, "step": 2997 }, { "epoch": 1.7638383169664202, "grad_norm": 0.35504030693683797, "learning_rate": 4.218933901399674e-07, "loss": 0.4278, "step": 2998 }, { "epoch": 1.7644267902460555, "grad_norm": 0.45068530133863566, "learning_rate": 4.198319935983325e-07, "loss": 0.477, "step": 2999 }, { "epoch": 1.7650152635256906, "grad_norm": 0.37975551297021914, "learning_rate": 4.1777542473433743e-07, "loss": 0.4306, "step": 3000 }, { "epoch": 1.7656037368053257, "grad_norm": 0.379778031534214, "learning_rate": 4.15723685715686e-07, "loss": 0.4623, "step": 3001 }, { "epoch": 1.7661922100849607, "grad_norm": 0.39416392799503264, "learning_rate": 4.1367677870499166e-07, "loss": 0.4152, "step": 3002 }, { "epoch": 1.7667806833645958, "grad_norm": 0.38833163843581553, "learning_rate": 4.1163470585977694e-07, "loss": 0.4222, "step": 3003 }, { "epoch": 1.767369156644231, "grad_norm": 0.3570519914318645, "learning_rate": 4.0959746933246503e-07, "loss": 0.4335, "step": 3004 }, { "epoch": 1.7679576299238664, "grad_norm": 0.38183708877261363, "learning_rate": 4.0756507127038494e-07, "loss": 0.3616, "step": 3005 }, { "epoch": 1.7685461032035015, "grad_norm": 0.37580712678042627, "learning_rate": 4.0553751381576414e-07, "loss": 0.4875, "step": 3006 }, { "epoch": 1.7691345764831365, "grad_norm": 0.36022698746823073, "learning_rate": 4.03514799105727e-07, "loss": 0.4436, "step": 3007 }, { "epoch": 1.7697230497627716, "grad_norm": 0.36831693949210664, "learning_rate": 4.0149692927229533e-07, "loss": 0.4194, "step": 3008 }, { "epoch": 1.770311523042407, "grad_norm": 0.3526493556144966, "learning_rate": 3.99483906442385e-07, "loss": 0.4286, "step": 3009 }, { "epoch": 1.770899996322042, "grad_norm": 0.46047505315639053, "learning_rate": 3.9747573273779816e-07, "loss": 0.5179, "step": 3010 }, { "epoch": 1.7714884696016773, "grad_norm": 0.3523809618247567, "learning_rate": 3.9547241027523164e-07, "loss": 0.3876, "step": 3011 }, { "epoch": 1.7720769428813123, "grad_norm": 0.3819224899803335, "learning_rate": 3.9347394116626636e-07, "loss": 0.4819, "step": 3012 }, { "epoch": 1.7726654161609474, "grad_norm": 0.3568101165843601, "learning_rate": 3.9148032751736564e-07, "loss": 0.4245, "step": 3013 }, { "epoch": 1.7732538894405825, "grad_norm": 0.3588898356765622, "learning_rate": 3.894915714298775e-07, "loss": 0.4215, "step": 3014 }, { "epoch": 1.7738423627202178, "grad_norm": 0.4017073610130721, "learning_rate": 3.8750767500003117e-07, "loss": 0.4833, "step": 3015 }, { "epoch": 1.7744308359998529, "grad_norm": 0.36026793822930253, "learning_rate": 3.85528640318929e-07, "loss": 0.4412, "step": 3016 }, { "epoch": 1.7750193092794881, "grad_norm": 0.3172497081414483, "learning_rate": 3.8355446947255293e-07, "loss": 0.3679, "step": 3017 }, { "epoch": 1.7756077825591232, "grad_norm": 0.3872030379929806, "learning_rate": 3.815851645417579e-07, "loss": 0.4248, "step": 3018 }, { "epoch": 1.7761962558387583, "grad_norm": 0.3569845493232082, "learning_rate": 3.796207276022662e-07, "loss": 0.4338, "step": 3019 }, { "epoch": 1.7767847291183934, "grad_norm": 0.38596490265798233, "learning_rate": 3.776611607246744e-07, "loss": 0.4193, "step": 3020 }, { "epoch": 1.7773732023980287, "grad_norm": 0.35673409215784024, "learning_rate": 3.7570646597444196e-07, "loss": 0.4606, "step": 3021 }, { "epoch": 1.7779616756776637, "grad_norm": 0.3663393582017516, "learning_rate": 3.7375664541189347e-07, "loss": 0.4392, "step": 3022 }, { "epoch": 1.778550148957299, "grad_norm": 0.3485414474177656, "learning_rate": 3.71811701092219e-07, "loss": 0.4069, "step": 3023 }, { "epoch": 1.779138622236934, "grad_norm": 0.3936498030135818, "learning_rate": 3.6987163506546465e-07, "loss": 0.419, "step": 3024 }, { "epoch": 1.7797270955165692, "grad_norm": 0.3720467232412673, "learning_rate": 3.679364493765364e-07, "loss": 0.4651, "step": 3025 }, { "epoch": 1.7803155687962042, "grad_norm": 0.3727897887555289, "learning_rate": 3.660061460651981e-07, "loss": 0.4535, "step": 3026 }, { "epoch": 1.7809040420758395, "grad_norm": 0.36408844206006347, "learning_rate": 3.6408072716606346e-07, "loss": 0.4728, "step": 3027 }, { "epoch": 1.7814925153554746, "grad_norm": 0.34456066164593324, "learning_rate": 3.6216019470860063e-07, "loss": 0.4122, "step": 3028 }, { "epoch": 1.7820809886351099, "grad_norm": 0.3832883592421097, "learning_rate": 3.602445507171276e-07, "loss": 0.4667, "step": 3029 }, { "epoch": 1.782669461914745, "grad_norm": 0.37710833236180846, "learning_rate": 3.583337972108064e-07, "loss": 0.4675, "step": 3030 }, { "epoch": 1.78325793519438, "grad_norm": 0.34052800206603673, "learning_rate": 3.564279362036488e-07, "loss": 0.4084, "step": 3031 }, { "epoch": 1.783846408474015, "grad_norm": 0.37170194593057404, "learning_rate": 3.5452696970450674e-07, "loss": 0.4366, "step": 3032 }, { "epoch": 1.7844348817536504, "grad_norm": 0.39506626423574354, "learning_rate": 3.5263089971707245e-07, "loss": 0.4234, "step": 3033 }, { "epoch": 1.7850233550332857, "grad_norm": 0.3643793233166181, "learning_rate": 3.5073972823988e-07, "loss": 0.4146, "step": 3034 }, { "epoch": 1.7856118283129208, "grad_norm": 0.38991006208101003, "learning_rate": 3.488534572662994e-07, "loss": 0.4168, "step": 3035 }, { "epoch": 1.7862003015925558, "grad_norm": 0.37244613225904866, "learning_rate": 3.469720887845324e-07, "loss": 0.47, "step": 3036 }, { "epoch": 1.786788774872191, "grad_norm": 0.420799263754605, "learning_rate": 3.450956247776166e-07, "loss": 0.496, "step": 3037 }, { "epoch": 1.787377248151826, "grad_norm": 0.36800566681282854, "learning_rate": 3.4322406722341993e-07, "loss": 0.4825, "step": 3038 }, { "epoch": 1.7879657214314613, "grad_norm": 0.374933307489788, "learning_rate": 3.4135741809463606e-07, "loss": 0.4394, "step": 3039 }, { "epoch": 1.7885541947110966, "grad_norm": 0.3628214678318104, "learning_rate": 3.394956793587878e-07, "loss": 0.3989, "step": 3040 }, { "epoch": 1.7891426679907316, "grad_norm": 0.34592336025886444, "learning_rate": 3.3763885297822153e-07, "loss": 0.4191, "step": 3041 }, { "epoch": 1.7897311412703667, "grad_norm": 0.3239401193711494, "learning_rate": 3.3578694091010443e-07, "loss": 0.3829, "step": 3042 }, { "epoch": 1.7903196145500018, "grad_norm": 0.37449617217844866, "learning_rate": 3.3393994510642566e-07, "loss": 0.4938, "step": 3043 }, { "epoch": 1.7909080878296368, "grad_norm": 0.34809272331188096, "learning_rate": 3.320978675139919e-07, "loss": 0.3905, "step": 3044 }, { "epoch": 1.7914965611092721, "grad_norm": 0.33851247857370076, "learning_rate": 3.3026071007442426e-07, "loss": 0.379, "step": 3045 }, { "epoch": 1.7920850343889074, "grad_norm": 0.3500567506611919, "learning_rate": 3.284284747241601e-07, "loss": 0.4529, "step": 3046 }, { "epoch": 1.7926735076685425, "grad_norm": 0.387882205417678, "learning_rate": 3.266011633944477e-07, "loss": 0.4514, "step": 3047 }, { "epoch": 1.7932619809481776, "grad_norm": 0.3545106482660138, "learning_rate": 3.247787780113443e-07, "loss": 0.4172, "step": 3048 }, { "epoch": 1.7938504542278126, "grad_norm": 0.3453047461867485, "learning_rate": 3.229613204957166e-07, "loss": 0.3835, "step": 3049 }, { "epoch": 1.7944389275074477, "grad_norm": 0.3831999995690117, "learning_rate": 3.2114879276323783e-07, "loss": 0.4289, "step": 3050 }, { "epoch": 1.795027400787083, "grad_norm": 0.41025365958616145, "learning_rate": 3.1934119672438093e-07, "loss": 0.5072, "step": 3051 }, { "epoch": 1.7956158740667183, "grad_norm": 0.3636109601930009, "learning_rate": 3.1753853428442503e-07, "loss": 0.462, "step": 3052 }, { "epoch": 1.7962043473463534, "grad_norm": 0.3595657737971085, "learning_rate": 3.1574080734344757e-07, "loss": 0.3805, "step": 3053 }, { "epoch": 1.7967928206259884, "grad_norm": 0.4113619766780667, "learning_rate": 3.1394801779632276e-07, "loss": 0.4872, "step": 3054 }, { "epoch": 1.7973812939056235, "grad_norm": 0.3495864222998627, "learning_rate": 3.121601675327218e-07, "loss": 0.3959, "step": 3055 }, { "epoch": 1.7979697671852586, "grad_norm": 0.3543163142283438, "learning_rate": 3.103772584371106e-07, "loss": 0.415, "step": 3056 }, { "epoch": 1.7985582404648939, "grad_norm": 0.4101448315356161, "learning_rate": 3.085992923887432e-07, "loss": 0.5026, "step": 3057 }, { "epoch": 1.7991467137445292, "grad_norm": 0.39169171367110645, "learning_rate": 3.068262712616676e-07, "loss": 0.4059, "step": 3058 }, { "epoch": 1.7997351870241642, "grad_norm": 0.37307486869304535, "learning_rate": 3.0505819692471797e-07, "loss": 0.4716, "step": 3059 }, { "epoch": 1.8003236603037993, "grad_norm": 0.3613313551524517, "learning_rate": 3.032950712415134e-07, "loss": 0.4466, "step": 3060 }, { "epoch": 1.8009121335834344, "grad_norm": 0.35075425237804847, "learning_rate": 3.015368960704584e-07, "loss": 0.4637, "step": 3061 }, { "epoch": 1.8015006068630697, "grad_norm": 0.3529924763155372, "learning_rate": 2.9978367326473966e-07, "loss": 0.4229, "step": 3062 }, { "epoch": 1.8020890801427047, "grad_norm": 0.376109535705695, "learning_rate": 2.980354046723216e-07, "loss": 0.3678, "step": 3063 }, { "epoch": 1.80267755342234, "grad_norm": 0.3288729182682341, "learning_rate": 2.9629209213594914e-07, "loss": 0.3837, "step": 3064 }, { "epoch": 1.8032660267019751, "grad_norm": 0.36944861057031814, "learning_rate": 2.9455373749314285e-07, "loss": 0.444, "step": 3065 }, { "epoch": 1.8038544999816102, "grad_norm": 0.37907660299804175, "learning_rate": 2.928203425761961e-07, "loss": 0.4319, "step": 3066 }, { "epoch": 1.8044429732612453, "grad_norm": 0.34732204111911696, "learning_rate": 2.91091909212175e-07, "loss": 0.4558, "step": 3067 }, { "epoch": 1.8050314465408805, "grad_norm": 0.38674088884440666, "learning_rate": 2.893684392229185e-07, "loss": 0.471, "step": 3068 }, { "epoch": 1.8056199198205156, "grad_norm": 0.3559113702054375, "learning_rate": 2.8764993442503053e-07, "loss": 0.4463, "step": 3069 }, { "epoch": 1.806208393100151, "grad_norm": 0.38185464809946684, "learning_rate": 2.859363966298828e-07, "loss": 0.4803, "step": 3070 }, { "epoch": 1.806796866379786, "grad_norm": 0.39982824486185164, "learning_rate": 2.842278276436128e-07, "loss": 0.4417, "step": 3071 }, { "epoch": 1.807385339659421, "grad_norm": 0.3303424053220174, "learning_rate": 2.825242292671182e-07, "loss": 0.3862, "step": 3072 }, { "epoch": 1.8079738129390561, "grad_norm": 0.3802181188987624, "learning_rate": 2.8082560329605934e-07, "loss": 0.3999, "step": 3073 }, { "epoch": 1.8085622862186914, "grad_norm": 0.369607609906856, "learning_rate": 2.791319515208557e-07, "loss": 0.4052, "step": 3074 }, { "epoch": 1.8091507594983265, "grad_norm": 0.3677959947651576, "learning_rate": 2.7744327572668206e-07, "loss": 0.4407, "step": 3075 }, { "epoch": 1.8097392327779618, "grad_norm": 0.35952246911599717, "learning_rate": 2.757595776934691e-07, "loss": 0.3959, "step": 3076 }, { "epoch": 1.8103277060575969, "grad_norm": 0.3573614287481565, "learning_rate": 2.7408085919590265e-07, "loss": 0.4613, "step": 3077 }, { "epoch": 1.810916179337232, "grad_norm": 0.3851728692777491, "learning_rate": 2.724071220034158e-07, "loss": 0.4378, "step": 3078 }, { "epoch": 1.811504652616867, "grad_norm": 0.37196186568715, "learning_rate": 2.7073836788019503e-07, "loss": 0.3971, "step": 3079 }, { "epoch": 1.8120931258965023, "grad_norm": 0.34879350774428064, "learning_rate": 2.6907459858517336e-07, "loss": 0.426, "step": 3080 }, { "epoch": 1.8126815991761374, "grad_norm": 0.3870350054138436, "learning_rate": 2.6741581587202747e-07, "loss": 0.4774, "step": 3081 }, { "epoch": 1.8132700724557727, "grad_norm": 0.3587016855513123, "learning_rate": 2.6576202148918095e-07, "loss": 0.4318, "step": 3082 }, { "epoch": 1.8138585457354077, "grad_norm": 0.3455498221745496, "learning_rate": 2.6411321717979886e-07, "loss": 0.4045, "step": 3083 }, { "epoch": 1.8144470190150428, "grad_norm": 0.38018123675341786, "learning_rate": 2.624694046817844e-07, "loss": 0.3995, "step": 3084 }, { "epoch": 1.8150354922946779, "grad_norm": 0.42591605854697395, "learning_rate": 2.6083058572778155e-07, "loss": 0.5203, "step": 3085 }, { "epoch": 1.8156239655743132, "grad_norm": 0.3402011032830822, "learning_rate": 2.5919676204517073e-07, "loss": 0.3979, "step": 3086 }, { "epoch": 1.8162124388539485, "grad_norm": 0.3559086257332086, "learning_rate": 2.5756793535606504e-07, "loss": 0.4141, "step": 3087 }, { "epoch": 1.8168009121335835, "grad_norm": 0.35151690722878987, "learning_rate": 2.5594410737731334e-07, "loss": 0.3865, "step": 3088 }, { "epoch": 1.8173893854132186, "grad_norm": 0.3665820429884153, "learning_rate": 2.5432527982049424e-07, "loss": 0.4276, "step": 3089 }, { "epoch": 1.8179778586928537, "grad_norm": 0.3702767526005032, "learning_rate": 2.527114543919146e-07, "loss": 0.3959, "step": 3090 }, { "epoch": 1.8185663319724887, "grad_norm": 0.39508140412031334, "learning_rate": 2.511026327926114e-07, "loss": 0.4417, "step": 3091 }, { "epoch": 1.819154805252124, "grad_norm": 0.39333052092005477, "learning_rate": 2.4949881671834487e-07, "loss": 0.4406, "step": 3092 }, { "epoch": 1.8197432785317593, "grad_norm": 0.3555298491774012, "learning_rate": 2.4790000785960113e-07, "loss": 0.4316, "step": 3093 }, { "epoch": 1.8203317518113944, "grad_norm": 0.37637845250714186, "learning_rate": 2.4630620790158763e-07, "loss": 0.4416, "step": 3094 }, { "epoch": 1.8209202250910295, "grad_norm": 0.6316424021103049, "learning_rate": 2.447174185242324e-07, "loss": 0.4878, "step": 3095 }, { "epoch": 1.8215086983706645, "grad_norm": 0.3365864268496033, "learning_rate": 2.431336414021818e-07, "loss": 0.3703, "step": 3096 }, { "epoch": 1.8220971716502996, "grad_norm": 0.39257838341121015, "learning_rate": 2.4155487820479997e-07, "loss": 0.3775, "step": 3097 }, { "epoch": 1.822685644929935, "grad_norm": 0.30612539228427843, "learning_rate": 2.399811305961647e-07, "loss": 0.3761, "step": 3098 }, { "epoch": 1.8232741182095702, "grad_norm": 0.3763250159911193, "learning_rate": 2.3841240023506795e-07, "loss": 0.507, "step": 3099 }, { "epoch": 1.8238625914892053, "grad_norm": 0.36511186385741584, "learning_rate": 2.3684868877501488e-07, "loss": 0.4284, "step": 3100 }, { "epoch": 1.8244510647688403, "grad_norm": 0.3257113656775898, "learning_rate": 2.3528999786421758e-07, "loss": 0.3619, "step": 3101 }, { "epoch": 1.8250395380484754, "grad_norm": 0.3439653485635246, "learning_rate": 2.3373632914559797e-07, "loss": 0.4157, "step": 3102 }, { "epoch": 1.8256280113281105, "grad_norm": 0.3648874138869474, "learning_rate": 2.321876842567844e-07, "loss": 0.4343, "step": 3103 }, { "epoch": 1.8262164846077458, "grad_norm": 0.37366029932283507, "learning_rate": 2.3064406483010947e-07, "loss": 0.4053, "step": 3104 }, { "epoch": 1.826804957887381, "grad_norm": 0.3866820334277549, "learning_rate": 2.291054724926084e-07, "loss": 0.4582, "step": 3105 }, { "epoch": 1.8273934311670161, "grad_norm": 0.3122995459366454, "learning_rate": 2.2757190886602e-07, "loss": 0.3769, "step": 3106 }, { "epoch": 1.8279819044466512, "grad_norm": 0.3826917778811463, "learning_rate": 2.2604337556677846e-07, "loss": 0.4104, "step": 3107 }, { "epoch": 1.8285703777262863, "grad_norm": 0.3556680262325681, "learning_rate": 2.245198742060195e-07, "loss": 0.4394, "step": 3108 }, { "epoch": 1.8291588510059213, "grad_norm": 0.35348950100501925, "learning_rate": 2.2300140638957358e-07, "loss": 0.4674, "step": 3109 }, { "epoch": 1.8297473242855566, "grad_norm": 0.40140174647785626, "learning_rate": 2.2148797371796483e-07, "loss": 0.4918, "step": 3110 }, { "epoch": 1.830335797565192, "grad_norm": 0.40196452684492534, "learning_rate": 2.1997957778641166e-07, "loss": 0.4593, "step": 3111 }, { "epoch": 1.830924270844827, "grad_norm": 0.3775443465832813, "learning_rate": 2.1847622018482283e-07, "loss": 0.4678, "step": 3112 }, { "epoch": 1.831512744124462, "grad_norm": 0.39693172282848854, "learning_rate": 2.1697790249779638e-07, "loss": 0.5023, "step": 3113 }, { "epoch": 1.8321012174040971, "grad_norm": 0.3913381268859045, "learning_rate": 2.1548462630461787e-07, "loss": 0.4466, "step": 3114 }, { "epoch": 1.8326896906837324, "grad_norm": 0.34459740748163303, "learning_rate": 2.1399639317925991e-07, "loss": 0.4377, "step": 3115 }, { "epoch": 1.8332781639633675, "grad_norm": 0.3315275552630866, "learning_rate": 2.1251320469037827e-07, "loss": 0.3828, "step": 3116 }, { "epoch": 1.8338666372430028, "grad_norm": 0.3889860162201658, "learning_rate": 2.1103506240131243e-07, "loss": 0.4547, "step": 3117 }, { "epoch": 1.8344551105226379, "grad_norm": 0.37832770183824677, "learning_rate": 2.0956196787008333e-07, "loss": 0.4518, "step": 3118 }, { "epoch": 1.835043583802273, "grad_norm": 0.44466916289319874, "learning_rate": 2.080939226493889e-07, "loss": 0.5369, "step": 3119 }, { "epoch": 1.835632057081908, "grad_norm": 0.37800287098256413, "learning_rate": 2.0663092828660802e-07, "loss": 0.5049, "step": 3120 }, { "epoch": 1.8362205303615433, "grad_norm": 0.4233912001029953, "learning_rate": 2.0517298632379445e-07, "loss": 0.5081, "step": 3121 }, { "epoch": 1.8368090036411784, "grad_norm": 0.4737211340685146, "learning_rate": 2.0372009829767558e-07, "loss": 0.4537, "step": 3122 }, { "epoch": 1.8373974769208137, "grad_norm": 0.3675240476562199, "learning_rate": 2.0227226573965364e-07, "loss": 0.4433, "step": 3123 }, { "epoch": 1.8379859502004487, "grad_norm": 0.37153156503311213, "learning_rate": 2.0082949017580122e-07, "loss": 0.4531, "step": 3124 }, { "epoch": 1.8385744234800838, "grad_norm": 0.38574709047823874, "learning_rate": 1.9939177312685963e-07, "loss": 0.434, "step": 3125 }, { "epoch": 1.839162896759719, "grad_norm": 0.3667875187858281, "learning_rate": 1.9795911610824004e-07, "loss": 0.4592, "step": 3126 }, { "epoch": 1.8397513700393542, "grad_norm": 0.312917468834572, "learning_rate": 1.9653152063002002e-07, "loss": 0.3673, "step": 3127 }, { "epoch": 1.8403398433189893, "grad_norm": 0.3874260157724371, "learning_rate": 1.951089881969398e-07, "loss": 0.3869, "step": 3128 }, { "epoch": 1.8409283165986245, "grad_norm": 0.36215964636725945, "learning_rate": 1.9369152030840553e-07, "loss": 0.3937, "step": 3129 }, { "epoch": 1.8415167898782596, "grad_norm": 0.3929954968951567, "learning_rate": 1.9227911845848546e-07, "loss": 0.394, "step": 3130 }, { "epoch": 1.8421052631578947, "grad_norm": 0.3546176095082159, "learning_rate": 1.908717841359048e-07, "loss": 0.3662, "step": 3131 }, { "epoch": 1.8426937364375298, "grad_norm": 0.3928803992714585, "learning_rate": 1.8946951882405084e-07, "loss": 0.4516, "step": 3132 }, { "epoch": 1.843282209717165, "grad_norm": 0.31388959757081936, "learning_rate": 1.8807232400096575e-07, "loss": 0.4239, "step": 3133 }, { "epoch": 1.8438706829968001, "grad_norm": 0.34197956913898386, "learning_rate": 1.8668020113934816e-07, "loss": 0.3948, "step": 3134 }, { "epoch": 1.8444591562764354, "grad_norm": 0.35969112287454974, "learning_rate": 1.8529315170655037e-07, "loss": 0.4466, "step": 3135 }, { "epoch": 1.8450476295560705, "grad_norm": 0.39715303383250833, "learning_rate": 1.839111771645774e-07, "loss": 0.4812, "step": 3136 }, { "epoch": 1.8456361028357056, "grad_norm": 0.3531313565413089, "learning_rate": 1.825342789700846e-07, "loss": 0.4017, "step": 3137 }, { "epoch": 1.8462245761153406, "grad_norm": 0.31629228841317464, "learning_rate": 1.8116245857437708e-07, "loss": 0.3759, "step": 3138 }, { "epoch": 1.846813049394976, "grad_norm": 0.3636915576276675, "learning_rate": 1.797957174234083e-07, "loss": 0.4234, "step": 3139 }, { "epoch": 1.8474015226746112, "grad_norm": 0.3287947443726984, "learning_rate": 1.7843405695777582e-07, "loss": 0.3592, "step": 3140 }, { "epoch": 1.8479899959542463, "grad_norm": 0.35072889399328133, "learning_rate": 1.770774786127244e-07, "loss": 0.4193, "step": 3141 }, { "epoch": 1.8485784692338814, "grad_norm": 0.3810439184876691, "learning_rate": 1.7572598381814188e-07, "loss": 0.4241, "step": 3142 }, { "epoch": 1.8491669425135164, "grad_norm": 0.39616215591788856, "learning_rate": 1.7437957399855488e-07, "loss": 0.4625, "step": 3143 }, { "epoch": 1.8497554157931515, "grad_norm": 0.40415528492294184, "learning_rate": 1.7303825057313428e-07, "loss": 0.4453, "step": 3144 }, { "epoch": 1.8503438890727868, "grad_norm": 0.40472951638615967, "learning_rate": 1.717020149556875e-07, "loss": 0.4408, "step": 3145 }, { "epoch": 1.850932362352422, "grad_norm": 0.46929902665382695, "learning_rate": 1.7037086855465902e-07, "loss": 0.4635, "step": 3146 }, { "epoch": 1.8515208356320572, "grad_norm": 0.3532945370986443, "learning_rate": 1.690448127731298e-07, "loss": 0.4393, "step": 3147 }, { "epoch": 1.8521093089116922, "grad_norm": 0.43440807239181717, "learning_rate": 1.6772384900881566e-07, "loss": 0.4233, "step": 3148 }, { "epoch": 1.8526977821913273, "grad_norm": 0.3781415046210508, "learning_rate": 1.664079786540629e-07, "loss": 0.427, "step": 3149 }, { "epoch": 1.8532862554709624, "grad_norm": 0.376677284343264, "learning_rate": 1.6509720309585198e-07, "loss": 0.4283, "step": 3150 }, { "epoch": 1.8538747287505977, "grad_norm": 0.3504489764352753, "learning_rate": 1.6379152371579277e-07, "loss": 0.4164, "step": 3151 }, { "epoch": 1.854463202030233, "grad_norm": 0.3778963185220897, "learning_rate": 1.624909418901205e-07, "loss": 0.45, "step": 3152 }, { "epoch": 1.855051675309868, "grad_norm": 0.3791355565883499, "learning_rate": 1.6119545898970136e-07, "loss": 0.4461, "step": 3153 }, { "epoch": 1.855640148589503, "grad_norm": 0.3502483878002617, "learning_rate": 1.599050763800253e-07, "loss": 0.4164, "step": 3154 }, { "epoch": 1.8562286218691382, "grad_norm": 0.36039631081266776, "learning_rate": 1.5861979542120598e-07, "loss": 0.4394, "step": 3155 }, { "epoch": 1.8568170951487732, "grad_norm": 0.3807342033002436, "learning_rate": 1.5733961746798032e-07, "loss": 0.4254, "step": 3156 }, { "epoch": 1.8574055684284085, "grad_norm": 0.3433419432890105, "learning_rate": 1.5606454386970672e-07, "loss": 0.3952, "step": 3157 }, { "epoch": 1.8579940417080438, "grad_norm": 0.38617663902193833, "learning_rate": 1.547945759703623e-07, "loss": 0.4249, "step": 3158 }, { "epoch": 1.858582514987679, "grad_norm": 0.3499599568148045, "learning_rate": 1.5352971510854354e-07, "loss": 0.4002, "step": 3159 }, { "epoch": 1.859170988267314, "grad_norm": 0.3367214015335166, "learning_rate": 1.5226996261746451e-07, "loss": 0.4555, "step": 3160 }, { "epoch": 1.859759461546949, "grad_norm": 0.3744624920212778, "learning_rate": 1.510153198249531e-07, "loss": 0.4381, "step": 3161 }, { "epoch": 1.8603479348265841, "grad_norm": 0.3528159648353572, "learning_rate": 1.4976578805345198e-07, "loss": 0.4465, "step": 3162 }, { "epoch": 1.8609364081062194, "grad_norm": 0.3568203662467883, "learning_rate": 1.4852136862001766e-07, "loss": 0.4608, "step": 3163 }, { "epoch": 1.8615248813858547, "grad_norm": 0.3685837008710632, "learning_rate": 1.4728206283631697e-07, "loss": 0.4336, "step": 3164 }, { "epoch": 1.8621133546654898, "grad_norm": 0.3526309970116767, "learning_rate": 1.460478720086278e-07, "loss": 0.4297, "step": 3165 }, { "epoch": 1.8627018279451248, "grad_norm": 0.4159181245112371, "learning_rate": 1.44818797437834e-07, "loss": 0.5091, "step": 3166 }, { "epoch": 1.86329030122476, "grad_norm": 0.3635331363878162, "learning_rate": 1.435948404194304e-07, "loss": 0.4709, "step": 3167 }, { "epoch": 1.8638787745043952, "grad_norm": 0.40240214298653554, "learning_rate": 1.423760022435161e-07, "loss": 0.4911, "step": 3168 }, { "epoch": 1.8644672477840303, "grad_norm": 0.3396580697096603, "learning_rate": 1.411622841947935e-07, "loss": 0.4457, "step": 3169 }, { "epoch": 1.8650557210636656, "grad_norm": 0.3788403593495051, "learning_rate": 1.399536875525692e-07, "loss": 0.4684, "step": 3170 }, { "epoch": 1.8656441943433006, "grad_norm": 0.2916250946701604, "learning_rate": 1.3875021359075257e-07, "loss": 0.3477, "step": 3171 }, { "epoch": 1.8662326676229357, "grad_norm": 0.33042394173126, "learning_rate": 1.3755186357785222e-07, "loss": 0.4296, "step": 3172 }, { "epoch": 1.8668211409025708, "grad_norm": 0.39747735771494125, "learning_rate": 1.363586387769761e-07, "loss": 0.4689, "step": 3173 }, { "epoch": 1.867409614182206, "grad_norm": 0.36610632901401635, "learning_rate": 1.3517054044583088e-07, "loss": 0.455, "step": 3174 }, { "epoch": 1.8679980874618412, "grad_norm": 0.3447877759740957, "learning_rate": 1.3398756983671822e-07, "loss": 0.4439, "step": 3175 }, { "epoch": 1.8685865607414764, "grad_norm": 0.34257871797960704, "learning_rate": 1.328097281965357e-07, "loss": 0.4022, "step": 3176 }, { "epoch": 1.8691750340211115, "grad_norm": 0.38425365158765173, "learning_rate": 1.3163701676677632e-07, "loss": 0.41, "step": 3177 }, { "epoch": 1.8697635073007466, "grad_norm": 0.33553737347489226, "learning_rate": 1.3046943678352186e-07, "loss": 0.4041, "step": 3178 }, { "epoch": 1.8703519805803817, "grad_norm": 0.36666793220723437, "learning_rate": 1.2930698947744957e-07, "loss": 0.4191, "step": 3179 }, { "epoch": 1.870940453860017, "grad_norm": 0.3559666488408952, "learning_rate": 1.2814967607382433e-07, "loss": 0.4462, "step": 3180 }, { "epoch": 1.871528927139652, "grad_norm": 0.36676755771897374, "learning_rate": 1.2699749779249926e-07, "loss": 0.423, "step": 3181 }, { "epoch": 1.8721174004192873, "grad_norm": 0.35460672605789884, "learning_rate": 1.2585045584791623e-07, "loss": 0.4573, "step": 3182 }, { "epoch": 1.8727058736989224, "grad_norm": 0.3617566237787531, "learning_rate": 1.247085514491031e-07, "loss": 0.4159, "step": 3183 }, { "epoch": 1.8732943469785575, "grad_norm": 0.4188126085885247, "learning_rate": 1.23571785799671e-07, "loss": 0.4812, "step": 3184 }, { "epoch": 1.8738828202581925, "grad_norm": 0.3629267383848308, "learning_rate": 1.22440160097817e-07, "loss": 0.417, "step": 3185 }, { "epoch": 1.8744712935378278, "grad_norm": 0.3526529612446012, "learning_rate": 1.2131367553631867e-07, "loss": 0.3703, "step": 3186 }, { "epoch": 1.875059766817463, "grad_norm": 0.38504804072809, "learning_rate": 1.2019233330253456e-07, "loss": 0.4381, "step": 3187 }, { "epoch": 1.8756482400970982, "grad_norm": 0.3415217162272222, "learning_rate": 1.1907613457840472e-07, "loss": 0.429, "step": 3188 }, { "epoch": 1.8762367133767333, "grad_norm": 0.3924822737311609, "learning_rate": 1.1796508054044586e-07, "loss": 0.4166, "step": 3189 }, { "epoch": 1.8768251866563683, "grad_norm": 0.3663283645210549, "learning_rate": 1.1685917235975286e-07, "loss": 0.4339, "step": 3190 }, { "epoch": 1.8774136599360034, "grad_norm": 0.40100275041894945, "learning_rate": 1.157584112019966e-07, "loss": 0.4888, "step": 3191 }, { "epoch": 1.8780021332156387, "grad_norm": 0.39771947592948426, "learning_rate": 1.1466279822742287e-07, "loss": 0.4389, "step": 3192 }, { "epoch": 1.878590606495274, "grad_norm": 0.37043165558257635, "learning_rate": 1.135723345908507e-07, "loss": 0.4355, "step": 3193 }, { "epoch": 1.879179079774909, "grad_norm": 0.3771837765170204, "learning_rate": 1.1248702144167123e-07, "loss": 0.4422, "step": 3194 }, { "epoch": 1.8797675530545441, "grad_norm": 0.36939284759509006, "learning_rate": 1.1140685992384881e-07, "loss": 0.5017, "step": 3195 }, { "epoch": 1.8803560263341792, "grad_norm": 0.37388968595068667, "learning_rate": 1.1033185117591494e-07, "loss": 0.474, "step": 3196 }, { "epoch": 1.8809444996138143, "grad_norm": 0.3448939833213339, "learning_rate": 1.0926199633097156e-07, "loss": 0.4449, "step": 3197 }, { "epoch": 1.8815329728934496, "grad_norm": 0.34049084576036537, "learning_rate": 1.0819729651668887e-07, "loss": 0.4383, "step": 3198 }, { "epoch": 1.8821214461730849, "grad_norm": 0.34437590304377796, "learning_rate": 1.0713775285530026e-07, "loss": 0.4554, "step": 3199 }, { "epoch": 1.88270991945272, "grad_norm": 0.35110403154004827, "learning_rate": 1.060833664636085e-07, "loss": 0.4434, "step": 3200 }, { "epoch": 1.883298392732355, "grad_norm": 0.41779543613685094, "learning_rate": 1.0503413845297739e-07, "loss": 0.4247, "step": 3201 }, { "epoch": 1.88388686601199, "grad_norm": 0.34906858869119817, "learning_rate": 1.0399006992933503e-07, "loss": 0.4268, "step": 3202 }, { "epoch": 1.8844753392916251, "grad_norm": 0.35081023121682964, "learning_rate": 1.0295116199317057e-07, "loss": 0.4594, "step": 3203 }, { "epoch": 1.8850638125712604, "grad_norm": 0.4248539809245603, "learning_rate": 1.0191741573953473e-07, "loss": 0.4999, "step": 3204 }, { "epoch": 1.8856522858508957, "grad_norm": 0.34516781366761773, "learning_rate": 1.0088883225803536e-07, "loss": 0.4101, "step": 3205 }, { "epoch": 1.8862407591305308, "grad_norm": 0.3820740870947898, "learning_rate": 9.986541263284077e-08, "loss": 0.4458, "step": 3206 }, { "epoch": 1.8868292324101659, "grad_norm": 0.38703501753864983, "learning_rate": 9.884715794267586e-08, "loss": 0.4572, "step": 3207 }, { "epoch": 1.887417705689801, "grad_norm": 0.38089826056947573, "learning_rate": 9.783406926082095e-08, "loss": 0.4512, "step": 3208 }, { "epoch": 1.888006178969436, "grad_norm": 0.37945375822684413, "learning_rate": 9.682614765511134e-08, "loss": 0.455, "step": 3209 }, { "epoch": 1.8885946522490713, "grad_norm": 0.3936988571169145, "learning_rate": 9.582339418793662e-08, "loss": 0.4724, "step": 3210 }, { "epoch": 1.8891831255287066, "grad_norm": 0.3814828520480913, "learning_rate": 9.482580991623747e-08, "loss": 0.4059, "step": 3211 }, { "epoch": 1.8897715988083417, "grad_norm": 0.3785257158646565, "learning_rate": 9.383339589150776e-08, "loss": 0.4527, "step": 3212 }, { "epoch": 1.8903600720879767, "grad_norm": 0.3835535281521533, "learning_rate": 9.284615315979129e-08, "loss": 0.4082, "step": 3213 }, { "epoch": 1.8909485453676118, "grad_norm": 0.389163366141509, "learning_rate": 9.186408276168012e-08, "loss": 0.4607, "step": 3214 }, { "epoch": 1.891537018647247, "grad_norm": 0.4297105229137311, "learning_rate": 9.08871857323157e-08, "loss": 0.5029, "step": 3215 }, { "epoch": 1.8921254919268822, "grad_norm": 0.3897671304368152, "learning_rate": 8.991546310138599e-08, "loss": 0.4564, "step": 3216 }, { "epoch": 1.8927139652065175, "grad_norm": 0.3519757606965559, "learning_rate": 8.894891589312449e-08, "loss": 0.4151, "step": 3217 }, { "epoch": 1.8933024384861525, "grad_norm": 0.33585917937577314, "learning_rate": 8.798754512631013e-08, "loss": 0.393, "step": 3218 }, { "epoch": 1.8938909117657876, "grad_norm": 0.40316036632357116, "learning_rate": 8.703135181426625e-08, "loss": 0.4335, "step": 3219 }, { "epoch": 1.8944793850454227, "grad_norm": 0.34771761737557044, "learning_rate": 8.608033696485718e-08, "loss": 0.4138, "step": 3220 }, { "epoch": 1.895067858325058, "grad_norm": 0.43078283879791635, "learning_rate": 8.513450158049109e-08, "loss": 0.4683, "step": 3221 }, { "epoch": 1.895656331604693, "grad_norm": 0.3845007688986671, "learning_rate": 8.419384665811547e-08, "loss": 0.4339, "step": 3222 }, { "epoch": 1.8962448048843283, "grad_norm": 0.3904736708771806, "learning_rate": 8.325837318921781e-08, "loss": 0.4328, "step": 3223 }, { "epoch": 1.8968332781639634, "grad_norm": 0.38457061976138884, "learning_rate": 8.232808215982325e-08, "loss": 0.4925, "step": 3224 }, { "epoch": 1.8974217514435985, "grad_norm": 0.3263435168255387, "learning_rate": 8.140297455049628e-08, "loss": 0.4325, "step": 3225 }, { "epoch": 1.8980102247232336, "grad_norm": 0.34852937628521174, "learning_rate": 8.048305133633583e-08, "loss": 0.3973, "step": 3226 }, { "epoch": 1.8985986980028688, "grad_norm": 0.3563606210870138, "learning_rate": 7.956831348697791e-08, "loss": 0.4841, "step": 3227 }, { "epoch": 1.899187171282504, "grad_norm": 0.38294261795223344, "learning_rate": 7.865876196659239e-08, "loss": 0.4289, "step": 3228 }, { "epoch": 1.8997756445621392, "grad_norm": 0.41562683359274355, "learning_rate": 7.775439773388182e-08, "loss": 0.5107, "step": 3229 }, { "epoch": 1.9003641178417743, "grad_norm": 0.4433995708520549, "learning_rate": 7.685522174208205e-08, "loss": 0.4344, "step": 3230 }, { "epoch": 1.9009525911214094, "grad_norm": 0.3689772414594358, "learning_rate": 7.59612349389599e-08, "loss": 0.436, "step": 3231 }, { "epoch": 1.9015410644010444, "grad_norm": 0.3430616940867066, "learning_rate": 7.507243826681276e-08, "loss": 0.4325, "step": 3232 }, { "epoch": 1.9021295376806797, "grad_norm": 0.3874877145813225, "learning_rate": 7.418883266246734e-08, "loss": 0.5036, "step": 3233 }, { "epoch": 1.9027180109603148, "grad_norm": 0.36546582712513626, "learning_rate": 7.331041905727975e-08, "loss": 0.3889, "step": 3234 }, { "epoch": 1.90330648423995, "grad_norm": 0.369167335950123, "learning_rate": 7.243719837713048e-08, "loss": 0.4378, "step": 3235 }, { "epoch": 1.9038949575195852, "grad_norm": 0.427987179319835, "learning_rate": 7.156917154243048e-08, "loss": 0.4419, "step": 3236 }, { "epoch": 1.9044834307992202, "grad_norm": 0.4168269777283528, "learning_rate": 7.070633946811344e-08, "loss": 0.4066, "step": 3237 }, { "epoch": 1.9050719040788553, "grad_norm": 0.3628913890511648, "learning_rate": 6.9848703063638e-08, "loss": 0.4147, "step": 3238 }, { "epoch": 1.9056603773584906, "grad_norm": 0.3894980813092583, "learning_rate": 6.899626323298714e-08, "loss": 0.4244, "step": 3239 }, { "epoch": 1.9062488506381257, "grad_norm": 0.36753703384978104, "learning_rate": 6.814902087466657e-08, "loss": 0.4238, "step": 3240 }, { "epoch": 1.906837323917761, "grad_norm": 0.3565074063390282, "learning_rate": 6.730697688170251e-08, "loss": 0.4511, "step": 3241 }, { "epoch": 1.907425797197396, "grad_norm": 0.38159623503539347, "learning_rate": 6.647013214164277e-08, "loss": 0.4525, "step": 3242 }, { "epoch": 1.908014270477031, "grad_norm": 0.4063495104612175, "learning_rate": 6.563848753655456e-08, "loss": 0.4529, "step": 3243 }, { "epoch": 1.9086027437566662, "grad_norm": 0.37799902368262395, "learning_rate": 6.481204394302388e-08, "loss": 0.4758, "step": 3244 }, { "epoch": 1.9091912170363015, "grad_norm": 0.3622250859975022, "learning_rate": 6.399080223215503e-08, "loss": 0.4318, "step": 3245 }, { "epoch": 1.9097796903159368, "grad_norm": 0.35215207411882854, "learning_rate": 6.31747632695695e-08, "loss": 0.3984, "step": 3246 }, { "epoch": 1.9103681635955718, "grad_norm": 0.353667943272911, "learning_rate": 6.236392791540424e-08, "loss": 0.3954, "step": 3247 }, { "epoch": 1.910956636875207, "grad_norm": 0.3637009054393379, "learning_rate": 6.15582970243117e-08, "loss": 0.4092, "step": 3248 }, { "epoch": 1.911545110154842, "grad_norm": 0.3820371186466404, "learning_rate": 6.075787144545875e-08, "loss": 0.4386, "step": 3249 }, { "epoch": 1.912133583434477, "grad_norm": 0.34986220035746385, "learning_rate": 5.996265202252494e-08, "loss": 0.4228, "step": 3250 }, { "epoch": 1.9127220567141123, "grad_norm": 0.3904805116611563, "learning_rate": 5.917263959370312e-08, "loss": 0.4416, "step": 3251 }, { "epoch": 1.9133105299937476, "grad_norm": 0.3665110108622097, "learning_rate": 5.838783499169831e-08, "loss": 0.4352, "step": 3252 }, { "epoch": 1.9138990032733827, "grad_norm": 0.36178688980760376, "learning_rate": 5.7608239043724366e-08, "loss": 0.4295, "step": 3253 }, { "epoch": 1.9144874765530178, "grad_norm": 0.36874692227852185, "learning_rate": 5.683385257150675e-08, "loss": 0.4276, "step": 3254 }, { "epoch": 1.9150759498326528, "grad_norm": 0.3585379269784683, "learning_rate": 5.606467639127977e-08, "loss": 0.4458, "step": 3255 }, { "epoch": 1.915664423112288, "grad_norm": 0.3740499023996214, "learning_rate": 5.530071131378434e-08, "loss": 0.4412, "step": 3256 }, { "epoch": 1.9162528963919232, "grad_norm": 0.3444149719010313, "learning_rate": 5.454195814427021e-08, "loss": 0.4189, "step": 3257 }, { "epoch": 1.9168413696715585, "grad_norm": 0.34931872567819894, "learning_rate": 5.378841768249377e-08, "loss": 0.3858, "step": 3258 }, { "epoch": 1.9174298429511936, "grad_norm": 0.3931508268367171, "learning_rate": 5.304009072271521e-08, "loss": 0.4183, "step": 3259 }, { "epoch": 1.9180183162308286, "grad_norm": 0.33623762875691504, "learning_rate": 5.2296978053702486e-08, "loss": 0.4109, "step": 3260 }, { "epoch": 1.9186067895104637, "grad_norm": 0.3442389693587118, "learning_rate": 5.155908045872349e-08, "loss": 0.4153, "step": 3261 }, { "epoch": 1.9191952627900988, "grad_norm": 0.3641166124680275, "learning_rate": 5.082639871555273e-08, "loss": 0.46, "step": 3262 }, { "epoch": 1.919783736069734, "grad_norm": 0.4116702767627094, "learning_rate": 5.009893359646523e-08, "loss": 0.403, "step": 3263 }, { "epoch": 1.9203722093493694, "grad_norm": 0.3295457314764197, "learning_rate": 4.937668586823763e-08, "loss": 0.3587, "step": 3264 }, { "epoch": 1.9209606826290044, "grad_norm": 0.4029655576698524, "learning_rate": 4.865965629214819e-08, "loss": 0.4414, "step": 3265 }, { "epoch": 1.9215491559086395, "grad_norm": 0.34910048836485735, "learning_rate": 4.794784562397459e-08, "loss": 0.4473, "step": 3266 }, { "epoch": 1.9221376291882746, "grad_norm": 0.3602687003055281, "learning_rate": 4.724125461399165e-08, "loss": 0.4512, "step": 3267 }, { "epoch": 1.9227261024679099, "grad_norm": 0.3823490188207246, "learning_rate": 4.653988400697585e-08, "loss": 0.4716, "step": 3268 }, { "epoch": 1.923314575747545, "grad_norm": 0.3714690645974321, "learning_rate": 4.584373454219859e-08, "loss": 0.4933, "step": 3269 }, { "epoch": 1.9239030490271802, "grad_norm": 0.3448511222588735, "learning_rate": 4.515280695342905e-08, "loss": 0.4035, "step": 3270 }, { "epoch": 1.9244915223068153, "grad_norm": 0.37695552589349296, "learning_rate": 4.446710196893245e-08, "loss": 0.4356, "step": 3271 }, { "epoch": 1.9250799955864504, "grad_norm": 0.37241633214298037, "learning_rate": 4.378662031146896e-08, "loss": 0.4426, "step": 3272 }, { "epoch": 1.9256684688660854, "grad_norm": 0.34120396584960233, "learning_rate": 4.311136269829208e-08, "loss": 0.3914, "step": 3273 }, { "epoch": 1.9262569421457207, "grad_norm": 0.3659575992762857, "learning_rate": 4.244132984115079e-08, "loss": 0.379, "step": 3274 }, { "epoch": 1.9268454154253558, "grad_norm": 0.33630486634463425, "learning_rate": 4.177652244628627e-08, "loss": 0.4268, "step": 3275 }, { "epoch": 1.927433888704991, "grad_norm": 0.3587611422008153, "learning_rate": 4.111694121443133e-08, "loss": 0.3638, "step": 3276 }, { "epoch": 1.9280223619846262, "grad_norm": 0.3629157136695993, "learning_rate": 4.046258684081039e-08, "loss": 0.4592, "step": 3277 }, { "epoch": 1.9286108352642612, "grad_norm": 0.379211009977427, "learning_rate": 3.9813460015139524e-08, "loss": 0.4136, "step": 3278 }, { "epoch": 1.9291993085438963, "grad_norm": 0.34275957155814113, "learning_rate": 3.916956142162309e-08, "loss": 0.3967, "step": 3279 }, { "epoch": 1.9297877818235316, "grad_norm": 0.3698901757478187, "learning_rate": 3.853089173895597e-08, "loss": 0.4723, "step": 3280 }, { "epoch": 1.9303762551031667, "grad_norm": 0.35020477754541174, "learning_rate": 3.7897451640321326e-08, "loss": 0.3846, "step": 3281 }, { "epoch": 1.930964728382802, "grad_norm": 0.3349742212992856, "learning_rate": 3.726924179339009e-08, "loss": 0.4041, "step": 3282 }, { "epoch": 1.931553201662437, "grad_norm": 0.3323803130585093, "learning_rate": 3.6646262860319804e-08, "loss": 0.4232, "step": 3283 }, { "epoch": 1.9321416749420721, "grad_norm": 0.38559266903161343, "learning_rate": 3.602851549775521e-08, "loss": 0.4581, "step": 3284 }, { "epoch": 1.9327301482217072, "grad_norm": 0.3656186647304697, "learning_rate": 3.541600035682602e-08, "loss": 0.4338, "step": 3285 }, { "epoch": 1.9333186215013425, "grad_norm": 0.36369502019105193, "learning_rate": 3.4808718083147454e-08, "loss": 0.4209, "step": 3286 }, { "epoch": 1.9339070947809776, "grad_norm": 0.34825244693822976, "learning_rate": 3.4206669316819155e-08, "loss": 0.4123, "step": 3287 }, { "epoch": 1.9344955680606128, "grad_norm": 0.38461304446842337, "learning_rate": 3.360985469242406e-08, "loss": 0.4811, "step": 3288 }, { "epoch": 1.935084041340248, "grad_norm": 0.40001317082464605, "learning_rate": 3.301827483902842e-08, "loss": 0.4469, "step": 3289 }, { "epoch": 1.935672514619883, "grad_norm": 0.3773025434110401, "learning_rate": 3.243193038018122e-08, "loss": 0.4734, "step": 3290 }, { "epoch": 1.936260987899518, "grad_norm": 0.38625058736347884, "learning_rate": 3.185082193391143e-08, "loss": 0.4697, "step": 3291 }, { "epoch": 1.9368494611791534, "grad_norm": 0.340139632084032, "learning_rate": 3.127495011273074e-08, "loss": 0.355, "step": 3292 }, { "epoch": 1.9374379344587886, "grad_norm": 0.3604319677641891, "learning_rate": 3.0704315523631956e-08, "loss": 0.4188, "step": 3293 }, { "epoch": 1.9380264077384237, "grad_norm": 0.3648290516051878, "learning_rate": 3.0138918768084503e-08, "loss": 0.4669, "step": 3294 }, { "epoch": 1.9386148810180588, "grad_norm": 0.35451652083244334, "learning_rate": 2.9578760442039998e-08, "loss": 0.4097, "step": 3295 }, { "epoch": 1.9392033542976939, "grad_norm": 0.3709370983138792, "learning_rate": 2.9023841135927822e-08, "loss": 0.4489, "step": 3296 }, { "epoch": 1.939791827577329, "grad_norm": 0.3426288891258348, "learning_rate": 2.8474161434653424e-08, "loss": 0.4124, "step": 3297 }, { "epoch": 1.9403803008569642, "grad_norm": 0.35345603140176973, "learning_rate": 2.792972191760168e-08, "loss": 0.4072, "step": 3298 }, { "epoch": 1.9409687741365995, "grad_norm": 0.360045092073745, "learning_rate": 2.7390523158633552e-08, "loss": 0.4315, "step": 3299 }, { "epoch": 1.9415572474162346, "grad_norm": 0.3533192999440985, "learning_rate": 2.685656572608497e-08, "loss": 0.3906, "step": 3300 }, { "epoch": 1.9421457206958697, "grad_norm": 0.36659626858595673, "learning_rate": 2.6327850182769065e-08, "loss": 0.4556, "step": 3301 }, { "epoch": 1.9427341939755047, "grad_norm": 0.3573763373318939, "learning_rate": 2.5804377085972278e-08, "loss": 0.4553, "step": 3302 }, { "epoch": 1.9433226672551398, "grad_norm": 0.35117193943088015, "learning_rate": 2.5286146987456017e-08, "loss": 0.4395, "step": 3303 }, { "epoch": 1.943911140534775, "grad_norm": 0.36329339088788404, "learning_rate": 2.4773160433455013e-08, "loss": 0.3789, "step": 3304 }, { "epoch": 1.9444996138144104, "grad_norm": 0.3835216150023388, "learning_rate": 2.426541796467785e-08, "loss": 0.465, "step": 3305 }, { "epoch": 1.9450880870940455, "grad_norm": 0.340901026231439, "learning_rate": 2.3762920116304767e-08, "loss": 0.3866, "step": 3306 }, { "epoch": 1.9456765603736805, "grad_norm": 0.34899550428823634, "learning_rate": 2.3265667417988192e-08, "loss": 0.441, "step": 3307 }, { "epoch": 1.9462650336533156, "grad_norm": 0.39240723054704946, "learning_rate": 2.2773660393852758e-08, "loss": 0.3969, "step": 3308 }, { "epoch": 1.9468535069329507, "grad_norm": 0.35278352038774763, "learning_rate": 2.228689956249197e-08, "loss": 0.4156, "step": 3309 }, { "epoch": 1.947441980212586, "grad_norm": 0.36861092637807313, "learning_rate": 2.180538543697208e-08, "loss": 0.4097, "step": 3310 }, { "epoch": 1.9480304534922213, "grad_norm": 0.38175047808575985, "learning_rate": 2.1329118524827662e-08, "loss": 0.4081, "step": 3311 }, { "epoch": 1.9486189267718563, "grad_norm": 0.4042004767167224, "learning_rate": 2.0858099328062153e-08, "loss": 0.4081, "step": 3312 }, { "epoch": 1.9492074000514914, "grad_norm": 0.3277541328004718, "learning_rate": 2.0392328343149525e-08, "loss": 0.3588, "step": 3313 }, { "epoch": 1.9497958733311265, "grad_norm": 0.38188887986452574, "learning_rate": 1.9931806061029846e-08, "loss": 0.4281, "step": 3314 }, { "epoch": 1.9503843466107615, "grad_norm": 0.3548458647992218, "learning_rate": 1.9476532967112605e-08, "loss": 0.3723, "step": 3315 }, { "epoch": 1.9509728198903968, "grad_norm": 0.381409420327931, "learning_rate": 1.9026509541272276e-08, "loss": 0.4753, "step": 3316 }, { "epoch": 1.9515612931700321, "grad_norm": 0.44087159809649973, "learning_rate": 1.8581736257852756e-08, "loss": 0.4807, "step": 3317 }, { "epoch": 1.9521497664496672, "grad_norm": 0.3685749766400254, "learning_rate": 1.814221358566126e-08, "loss": 0.4361, "step": 3318 }, { "epoch": 1.9527382397293023, "grad_norm": 0.3620465428944035, "learning_rate": 1.770794198797332e-08, "loss": 0.455, "step": 3319 }, { "epoch": 1.9533267130089373, "grad_norm": 0.3759150758755786, "learning_rate": 1.7278921922527224e-08, "loss": 0.4186, "step": 3320 }, { "epoch": 1.9539151862885726, "grad_norm": 0.32690197514903707, "learning_rate": 1.6855153841527915e-08, "loss": 0.3761, "step": 3321 }, { "epoch": 1.9545036595682077, "grad_norm": 0.34900970886456434, "learning_rate": 1.6436638191643094e-08, "loss": 0.414, "step": 3322 }, { "epoch": 1.955092132847843, "grad_norm": 0.3265634049360124, "learning_rate": 1.6023375414004894e-08, "loss": 0.3712, "step": 3323 }, { "epoch": 1.955680606127478, "grad_norm": 0.4475164712802862, "learning_rate": 1.5615365944208205e-08, "loss": 0.4743, "step": 3324 }, { "epoch": 1.9562690794071131, "grad_norm": 0.35480721063180587, "learning_rate": 1.5212610212311796e-08, "loss": 0.3957, "step": 3325 }, { "epoch": 1.9568575526867482, "grad_norm": 0.3504201266195561, "learning_rate": 1.481510864283553e-08, "loss": 0.4446, "step": 3326 }, { "epoch": 1.9574460259663835, "grad_norm": 0.34701656602997705, "learning_rate": 1.4422861654762588e-08, "loss": 0.4651, "step": 3327 }, { "epoch": 1.9580344992460186, "grad_norm": 0.4180428829827467, "learning_rate": 1.4035869661535584e-08, "loss": 0.5008, "step": 3328 }, { "epoch": 1.9586229725256539, "grad_norm": 0.3727301747007, "learning_rate": 1.3654133071059894e-08, "loss": 0.4305, "step": 3329 }, { "epoch": 1.959211445805289, "grad_norm": 0.35109429998563674, "learning_rate": 1.3277652285700881e-08, "loss": 0.4332, "step": 3330 }, { "epoch": 1.959799919084924, "grad_norm": 0.4045756433871753, "learning_rate": 1.2906427702284452e-08, "loss": 0.4464, "step": 3331 }, { "epoch": 1.960388392364559, "grad_norm": 0.3805493153114484, "learning_rate": 1.2540459712095387e-08, "loss": 0.4202, "step": 3332 }, { "epoch": 1.9609768656441944, "grad_norm": 0.3515726821355353, "learning_rate": 1.2179748700879013e-08, "loss": 0.4151, "step": 3333 }, { "epoch": 1.9615653389238294, "grad_norm": 0.3509826713841071, "learning_rate": 1.182429504883842e-08, "loss": 0.3814, "step": 3334 }, { "epoch": 1.9621538122034647, "grad_norm": 0.355681217868795, "learning_rate": 1.1474099130635575e-08, "loss": 0.4046, "step": 3335 }, { "epoch": 1.9627422854830998, "grad_norm": 0.33445066500061066, "learning_rate": 1.1129161315391879e-08, "loss": 0.4125, "step": 3336 }, { "epoch": 1.9633307587627349, "grad_norm": 0.3703393276012962, "learning_rate": 1.0789481966683724e-08, "loss": 0.4258, "step": 3337 }, { "epoch": 1.96391923204237, "grad_norm": 0.33750526397026687, "learning_rate": 1.0455061442548597e-08, "loss": 0.437, "step": 3338 }, { "epoch": 1.9645077053220052, "grad_norm": 0.36732003731450513, "learning_rate": 1.012590009547676e-08, "loss": 0.5169, "step": 3339 }, { "epoch": 1.9650961786016403, "grad_norm": 0.35512202201802623, "learning_rate": 9.801998272419011e-09, "loss": 0.3977, "step": 3340 }, { "epoch": 1.9656846518812756, "grad_norm": 0.357014858561202, "learning_rate": 9.48335631477948e-09, "loss": 0.4159, "step": 3341 }, { "epoch": 1.9662731251609107, "grad_norm": 0.4118799961935701, "learning_rate": 9.169974558420058e-09, "loss": 0.4381, "step": 3342 }, { "epoch": 1.9668615984405458, "grad_norm": 0.3539970126398682, "learning_rate": 8.861853333657077e-09, "loss": 0.4093, "step": 3343 }, { "epoch": 1.9674500717201808, "grad_norm": 0.39992926291950676, "learning_rate": 8.558992965262969e-09, "loss": 0.4719, "step": 3344 }, { "epoch": 1.9680385449998161, "grad_norm": 0.3575148718344588, "learning_rate": 8.261393772463488e-09, "loss": 0.4203, "step": 3345 }, { "epoch": 1.9686270182794514, "grad_norm": 0.36652851144683174, "learning_rate": 7.96905606894105e-09, "loss": 0.4061, "step": 3346 }, { "epoch": 1.9692154915590865, "grad_norm": 0.3850357417710372, "learning_rate": 7.681980162830283e-09, "loss": 0.4692, "step": 3347 }, { "epoch": 1.9698039648387216, "grad_norm": 0.3350029872871535, "learning_rate": 7.400166356720806e-09, "loss": 0.416, "step": 3348 }, { "epoch": 1.9703924381183566, "grad_norm": 0.3700539624871692, "learning_rate": 7.123614947655566e-09, "loss": 0.4484, "step": 3349 }, { "epoch": 1.9709809113979917, "grad_norm": 0.3955049606510261, "learning_rate": 6.852326227130835e-09, "loss": 0.4191, "step": 3350 }, { "epoch": 1.971569384677627, "grad_norm": 0.3746312020978786, "learning_rate": 6.586300481095098e-09, "loss": 0.4135, "step": 3351 }, { "epoch": 1.9721578579572623, "grad_norm": 0.37676363716234645, "learning_rate": 6.32553798995017e-09, "loss": 0.4176, "step": 3352 }, { "epoch": 1.9727463312368974, "grad_norm": 0.36278969181689086, "learning_rate": 6.070039028550634e-09, "loss": 0.3671, "step": 3353 }, { "epoch": 1.9733348045165324, "grad_norm": 0.3654932346121125, "learning_rate": 5.819803866201068e-09, "loss": 0.4628, "step": 3354 }, { "epoch": 1.9739232777961675, "grad_norm": 0.37610147682765754, "learning_rate": 5.574832766660487e-09, "loss": 0.4431, "step": 3355 }, { "epoch": 1.9745117510758026, "grad_norm": 0.3838411317360214, "learning_rate": 5.3351259881379016e-09, "loss": 0.4248, "step": 3356 }, { "epoch": 1.9751002243554379, "grad_norm": 0.40777082348497695, "learning_rate": 5.1006837832928705e-09, "loss": 0.5021, "step": 3357 }, { "epoch": 1.9756886976350732, "grad_norm": 0.44789036988436204, "learning_rate": 4.8715063992377245e-09, "loss": 0.4092, "step": 3358 }, { "epoch": 1.9762771709147082, "grad_norm": 0.426979304828034, "learning_rate": 4.647594077534235e-09, "loss": 0.4809, "step": 3359 }, { "epoch": 1.9768656441943433, "grad_norm": 0.36048670775532415, "learning_rate": 4.428947054194166e-09, "loss": 0.4132, "step": 3360 }, { "epoch": 1.9774541174739784, "grad_norm": 0.4161258395443248, "learning_rate": 4.2155655596809455e-09, "loss": 0.4831, "step": 3361 }, { "epoch": 1.9780425907536134, "grad_norm": 0.3995746423436599, "learning_rate": 4.007449818906328e-09, "loss": 0.4726, "step": 3362 }, { "epoch": 1.9786310640332487, "grad_norm": 0.44308984030373544, "learning_rate": 3.804600051232621e-09, "loss": 0.4842, "step": 3363 }, { "epoch": 1.979219537312884, "grad_norm": 0.34786435522493536, "learning_rate": 3.6070164704726796e-09, "loss": 0.3992, "step": 3364 }, { "epoch": 1.979808010592519, "grad_norm": 0.3626464850896519, "learning_rate": 3.41469928488547e-09, "loss": 0.4359, "step": 3365 }, { "epoch": 1.9803964838721542, "grad_norm": 0.3944270514055894, "learning_rate": 3.227648697182173e-09, "loss": 0.4227, "step": 3366 }, { "epoch": 1.9809849571517892, "grad_norm": 0.35743717458145935, "learning_rate": 3.0458649045211897e-09, "loss": 0.4782, "step": 3367 }, { "epoch": 1.9815734304314243, "grad_norm": 0.3606690916261432, "learning_rate": 2.8693480985098056e-09, "loss": 0.401, "step": 3368 }, { "epoch": 1.9821619037110596, "grad_norm": 0.3753146807063232, "learning_rate": 2.698098465203636e-09, "loss": 0.424, "step": 3369 }, { "epoch": 1.982750376990695, "grad_norm": 0.3691861241400832, "learning_rate": 2.5321161851066257e-09, "loss": 0.4546, "step": 3370 }, { "epoch": 1.98333885027033, "grad_norm": 0.3414805157379496, "learning_rate": 2.371401433170495e-09, "loss": 0.4018, "step": 3371 }, { "epoch": 1.983927323549965, "grad_norm": 0.34995647466862123, "learning_rate": 2.215954378795293e-09, "loss": 0.4526, "step": 3372 }, { "epoch": 1.9845157968296, "grad_norm": 0.3739565677003356, "learning_rate": 2.0657751858277342e-09, "loss": 0.4583, "step": 3373 }, { "epoch": 1.9851042701092354, "grad_norm": 0.3619568650636465, "learning_rate": 1.9208640125628618e-09, "loss": 0.4218, "step": 3374 }, { "epoch": 1.9856927433888705, "grad_norm": 0.33776617897015526, "learning_rate": 1.7812210117429396e-09, "loss": 0.4228, "step": 3375 }, { "epoch": 1.9862812166685058, "grad_norm": 0.3402129406528821, "learning_rate": 1.64684633055745e-09, "loss": 0.4137, "step": 3376 }, { "epoch": 1.9868696899481408, "grad_norm": 0.3707451308515246, "learning_rate": 1.5177401106419853e-09, "loss": 0.4588, "step": 3377 }, { "epoch": 1.987458163227776, "grad_norm": 0.3994666270208093, "learning_rate": 1.3939024880799124e-09, "loss": 0.4817, "step": 3378 }, { "epoch": 1.988046636507411, "grad_norm": 0.33925349511692987, "learning_rate": 1.2753335934007073e-09, "loss": 0.4489, "step": 3379 }, { "epoch": 1.9886351097870463, "grad_norm": 0.3483741634592876, "learning_rate": 1.1620335515810654e-09, "loss": 0.4055, "step": 3380 }, { "epoch": 1.9892235830666813, "grad_norm": 0.31690319164270464, "learning_rate": 1.054002482043237e-09, "loss": 0.3802, "step": 3381 }, { "epoch": 1.9898120563463166, "grad_norm": 0.38327754616392384, "learning_rate": 9.512404986566914e-10, "loss": 0.4189, "step": 3382 }, { "epoch": 1.9904005296259517, "grad_norm": 0.3602852874444872, "learning_rate": 8.537477097364522e-10, "loss": 0.4527, "step": 3383 }, { "epoch": 1.9909890029055868, "grad_norm": 0.3601216126497891, "learning_rate": 7.615242180436521e-10, "loss": 0.4388, "step": 3384 }, { "epoch": 1.9915774761852219, "grad_norm": 0.3860114122286655, "learning_rate": 6.745701207855337e-10, "loss": 0.5484, "step": 3385 }, { "epoch": 1.9921659494648571, "grad_norm": 0.35294274681389687, "learning_rate": 5.928855096154485e-10, "loss": 0.4316, "step": 3386 }, { "epoch": 1.9927544227444922, "grad_norm": 0.374456342884089, "learning_rate": 5.164704706317469e-10, "loss": 0.502, "step": 3387 }, { "epoch": 1.9933428960241275, "grad_norm": 0.4142130329227272, "learning_rate": 4.4532508437888923e-10, "loss": 0.4772, "step": 3388 }, { "epoch": 1.9939313693037626, "grad_norm": 0.36873532407473225, "learning_rate": 3.7944942584688947e-10, "loss": 0.3993, "step": 3389 }, { "epoch": 1.9945198425833977, "grad_norm": 0.33300579161998695, "learning_rate": 3.1884356447187125e-10, "loss": 0.3917, "step": 3390 }, { "epoch": 1.9951083158630327, "grad_norm": 0.35702460125498287, "learning_rate": 2.6350756413440203e-10, "loss": 0.4269, "step": 3391 }, { "epoch": 1.995696789142668, "grad_norm": 0.3797901292513818, "learning_rate": 2.1344148316060352e-10, "loss": 0.4992, "step": 3392 }, { "epoch": 1.996285262422303, "grad_norm": 0.32462367642767903, "learning_rate": 1.6864537432270677e-10, "loss": 0.3852, "step": 3393 }, { "epoch": 1.9968737357019384, "grad_norm": 0.4024401724341136, "learning_rate": 1.2911928483683168e-10, "loss": 0.4726, "step": 3394 }, { "epoch": 1.9974622089815735, "grad_norm": 0.3699735020485836, "learning_rate": 9.486325636576254e-11, "loss": 0.4584, "step": 3395 }, { "epoch": 1.9980506822612085, "grad_norm": 0.386548696184757, "learning_rate": 6.58773250156175e-11, "loss": 0.4539, "step": 3396 }, { "epoch": 1.9986391555408436, "grad_norm": 0.39723457043875005, "learning_rate": 4.216152134028928e-11, "loss": 0.3775, "step": 3397 }, { "epoch": 1.9992276288204789, "grad_norm": 0.34312597411085316, "learning_rate": 2.3715870335894175e-11, "loss": 0.4388, "step": 3398 }, { "epoch": 1.9998161021001142, "grad_norm": 0.3747425963843166, "learning_rate": 1.0540391445212905e-11, "loss": 0.3805, "step": 3399 }, { "epoch": 2.0, "grad_norm": 0.6298526102920563, "learning_rate": 2.6350985554701724e-12, "loss": 0.3638, "step": 3400 }, { "epoch": 2.0, "step": 3400, "total_flos": 918889926033408.0, "train_loss": 0.502564754082876, "train_runtime": 78581.2699, "train_samples_per_second": 11.072, "train_steps_per_second": 0.043 } ], "logging_steps": 1, "max_steps": 3400, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 250, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 918889926033408.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }