diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17857 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 2545,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019646365422396855,
+      "grad_norm": 2.869349943562663,
+      "learning_rate": 3.1372549019607843e-07,
+      "loss": 0.9772,
+      "step": 1
+    },
+    {
+      "epoch": 0.003929273084479371,
+      "grad_norm": 2.8858084462639275,
+      "learning_rate": 6.274509803921569e-07,
+      "loss": 0.9718,
+      "step": 2
+    },
+    {
+      "epoch": 0.005893909626719057,
+      "grad_norm": 2.9653247684664272,
+      "learning_rate": 9.411764705882353e-07,
+      "loss": 0.9949,
+      "step": 3
+    },
+    {
+      "epoch": 0.007858546168958742,
+      "grad_norm": 2.9184493250912475,
+      "learning_rate": 1.2549019607843137e-06,
+      "loss": 0.9814,
+      "step": 4
+    },
+    {
+      "epoch": 0.009823182711198428,
+      "grad_norm": 2.7593729823255377,
+      "learning_rate": 1.5686274509803923e-06,
+      "loss": 0.9623,
+      "step": 5
+    },
+    {
+      "epoch": 0.011787819253438114,
+      "grad_norm": 2.6934821485631963,
+      "learning_rate": 1.8823529411764707e-06,
+      "loss": 0.9624,
+      "step": 6
+    },
+    {
+      "epoch": 0.0137524557956778,
+      "grad_norm": 2.252060241335818,
+      "learning_rate": 2.196078431372549e-06,
+      "loss": 0.9522,
+      "step": 7
+    },
+    {
+      "epoch": 0.015717092337917484,
+      "grad_norm": 2.1117855149575404,
+      "learning_rate": 2.5098039215686274e-06,
+      "loss": 0.9314,
+      "step": 8
+    },
+    {
+      "epoch": 0.01768172888015717,
+      "grad_norm": 1.479461804674372,
+      "learning_rate": 2.8235294117647062e-06,
+      "loss": 0.9061,
+      "step": 9
+    },
+    {
+      "epoch": 0.019646365422396856,
+      "grad_norm": 1.4474468533385711,
+      "learning_rate": 3.1372549019607846e-06,
+      "loss": 0.9031,
+      "step": 10
+    },
+    {
+      "epoch": 0.021611001964636542,
+      "grad_norm": 1.3397031584820531,
+      "learning_rate": 3.450980392156863e-06,
+      "loss": 0.8933,
+      "step": 11
+    },
+    {
+      "epoch": 0.023575638506876228,
+      "grad_norm": 1.182411324723917,
+      "learning_rate": 3.7647058823529414e-06,
+      "loss": 0.8628,
+      "step": 12
+    },
+    {
+      "epoch": 0.025540275049115914,
+      "grad_norm": 1.1160347777000645,
+      "learning_rate": 4.07843137254902e-06,
+      "loss": 0.8545,
+      "step": 13
+    },
+    {
+      "epoch": 0.0275049115913556,
+      "grad_norm": 1.0276542919401468,
+      "learning_rate": 4.392156862745098e-06,
+      "loss": 0.8355,
+      "step": 14
+    },
+    {
+      "epoch": 0.029469548133595286,
+      "grad_norm": 0.9465429581884693,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.831,
+      "step": 15
+    },
+    {
+      "epoch": 0.03143418467583497,
+      "grad_norm": 0.8436630866510695,
+      "learning_rate": 5.019607843137255e-06,
+      "loss": 0.7929,
+      "step": 16
+    },
+    {
+      "epoch": 0.03339882121807466,
+      "grad_norm": 0.8761715738287036,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.7865,
+      "step": 17
+    },
+    {
+      "epoch": 0.03536345776031434,
+      "grad_norm": 0.7965792415340047,
+      "learning_rate": 5.6470588235294125e-06,
+      "loss": 0.7758,
+      "step": 18
+    },
+    {
+      "epoch": 0.03732809430255403,
+      "grad_norm": 0.6842217892843347,
+      "learning_rate": 5.96078431372549e-06,
+      "loss": 0.7556,
+      "step": 19
+    },
+    {
+      "epoch": 0.03929273084479371,
+      "grad_norm": 0.6268997243594878,
+      "learning_rate": 6.274509803921569e-06,
+      "loss": 0.7459,
+      "step": 20
+    },
+    {
+      "epoch": 0.0412573673870334,
+      "grad_norm": 0.5586952540569388,
+      "learning_rate": 6.588235294117647e-06,
+      "loss": 0.7395,
+      "step": 21
+    },
+    {
+      "epoch": 0.043222003929273084,
+      "grad_norm": 0.7177195642985513,
+      "learning_rate": 6.901960784313726e-06,
+      "loss": 0.7296,
+      "step": 22
+    },
+    {
+      "epoch": 0.04518664047151277,
+      "grad_norm": 0.7513631652437482,
+      "learning_rate": 7.215686274509805e-06,
+      "loss": 0.7226,
+      "step": 23
+    },
+    {
+      "epoch": 0.047151277013752456,
+      "grad_norm": 0.6191635784394427,
+      "learning_rate": 7.529411764705883e-06,
+      "loss": 0.7227,
+      "step": 24
+    },
+    {
+      "epoch": 0.04911591355599214,
+      "grad_norm": 0.5110376284732959,
+      "learning_rate": 7.84313725490196e-06,
+      "loss": 0.7063,
+      "step": 25
+    },
+    {
+      "epoch": 0.05108055009823183,
+      "grad_norm": 0.5384303624853751,
+      "learning_rate": 8.15686274509804e-06,
+      "loss": 0.7139,
+      "step": 26
+    },
+    {
+      "epoch": 0.05304518664047151,
+      "grad_norm": 0.5250855662306244,
+      "learning_rate": 8.470588235294118e-06,
+      "loss": 0.6938,
+      "step": 27
+    },
+    {
+      "epoch": 0.0550098231827112,
+      "grad_norm": 0.46643819775205486,
+      "learning_rate": 8.784313725490196e-06,
+      "loss": 0.6887,
+      "step": 28
+    },
+    {
+      "epoch": 0.05697445972495088,
+      "grad_norm": 0.4086100645648967,
+      "learning_rate": 9.098039215686276e-06,
+      "loss": 0.7011,
+      "step": 29
+    },
+    {
+      "epoch": 0.05893909626719057,
+      "grad_norm": 0.34417937464113085,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.6883,
+      "step": 30
+    },
+    {
+      "epoch": 0.060903732809430254,
+      "grad_norm": 0.32829822440264084,
+      "learning_rate": 9.725490196078432e-06,
+      "loss": 0.6904,
+      "step": 31
+    },
+    {
+      "epoch": 0.06286836935166994,
+      "grad_norm": 0.343227898778333,
+      "learning_rate": 1.003921568627451e-05,
+      "loss": 0.6767,
+      "step": 32
+    },
+    {
+      "epoch": 0.06483300589390963,
+      "grad_norm": 0.38346177713094465,
+      "learning_rate": 1.035294117647059e-05,
+      "loss": 0.6796,
+      "step": 33
+    },
+    {
+      "epoch": 0.06679764243614932,
+      "grad_norm": 0.3475636790918237,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 0.6724,
+      "step": 34
+    },
+    {
+      "epoch": 0.068762278978389,
+      "grad_norm": 0.268886503827006,
+      "learning_rate": 1.0980392156862747e-05,
+      "loss": 0.671,
+      "step": 35
+    },
+    {
+      "epoch": 0.07072691552062868,
+      "grad_norm": 0.24493979118941558,
+      "learning_rate": 1.1294117647058825e-05,
+      "loss": 0.6623,
+      "step": 36
+    },
+    {
+      "epoch": 0.07269155206286837,
+      "grad_norm": 0.2820306716391014,
+      "learning_rate": 1.1607843137254905e-05,
+      "loss": 0.6548,
+      "step": 37
+    },
+    {
+      "epoch": 0.07465618860510806,
+      "grad_norm": 0.27566630422867244,
+      "learning_rate": 1.192156862745098e-05,
+      "loss": 0.6586,
+      "step": 38
+    },
+    {
+      "epoch": 0.07662082514734773,
+      "grad_norm": 0.24401555517026216,
+      "learning_rate": 1.223529411764706e-05,
+      "loss": 0.6512,
+      "step": 39
+    },
+    {
+      "epoch": 0.07858546168958742,
+      "grad_norm": 0.226232900709781,
+      "learning_rate": 1.2549019607843138e-05,
+      "loss": 0.6464,
+      "step": 40
+    },
+    {
+      "epoch": 0.08055009823182711,
+      "grad_norm": 0.23065071641812576,
+      "learning_rate": 1.2862745098039218e-05,
+      "loss": 0.6607,
+      "step": 41
+    },
+    {
+      "epoch": 0.0825147347740668,
+      "grad_norm": 0.22438748215837995,
+      "learning_rate": 1.3176470588235294e-05,
+      "loss": 0.6434,
+      "step": 42
+    },
+    {
+      "epoch": 0.08447937131630648,
+      "grad_norm": 0.22545995895982532,
+      "learning_rate": 1.3490196078431374e-05,
+      "loss": 0.6492,
+      "step": 43
+    },
+    {
+      "epoch": 0.08644400785854617,
+      "grad_norm": 0.18874070152783246,
+      "learning_rate": 1.3803921568627452e-05,
+      "loss": 0.6325,
+      "step": 44
+    },
+    {
+      "epoch": 0.08840864440078586,
+      "grad_norm": 0.1782563263533947,
+      "learning_rate": 1.4117647058823532e-05,
+      "loss": 0.6445,
+      "step": 45
+    },
+    {
+      "epoch": 0.09037328094302555,
+      "grad_norm": 0.18346791290687817,
+      "learning_rate": 1.443137254901961e-05,
+      "loss": 0.6397,
+      "step": 46
+    },
+    {
+      "epoch": 0.09233791748526522,
+      "grad_norm": 0.2128712872996269,
+      "learning_rate": 1.474509803921569e-05,
+      "loss": 0.6295,
+      "step": 47
+    },
+    {
+      "epoch": 0.09430255402750491,
+      "grad_norm": 0.22388663040158974,
+      "learning_rate": 1.5058823529411765e-05,
+      "loss": 0.6243,
+      "step": 48
+    },
+    {
+      "epoch": 0.0962671905697446,
+      "grad_norm": 0.16495797553731292,
+      "learning_rate": 1.5372549019607847e-05,
+      "loss": 0.6292,
+      "step": 49
+    },
+    {
+      "epoch": 0.09823182711198428,
+      "grad_norm": 0.1765537548031483,
+      "learning_rate": 1.568627450980392e-05,
+      "loss": 0.6399,
+      "step": 50
+    },
+    {
+      "epoch": 0.10019646365422397,
+      "grad_norm": 0.19368461003233384,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6359,
+      "step": 51
+    },
+    {
+      "epoch": 0.10216110019646366,
+      "grad_norm": 0.14899603368188405,
+      "learning_rate": 1.631372549019608e-05,
+      "loss": 0.6238,
+      "step": 52
+    },
+    {
+      "epoch": 0.10412573673870335,
+      "grad_norm": 0.17339020436965724,
+      "learning_rate": 1.662745098039216e-05,
+      "loss": 0.6206,
+      "step": 53
+    },
+    {
+      "epoch": 0.10609037328094302,
+      "grad_norm": 0.16192355095677685,
+      "learning_rate": 1.6941176470588237e-05,
+      "loss": 0.6295,
+      "step": 54
+    },
+    {
+      "epoch": 0.10805500982318271,
+      "grad_norm": 0.13636696971257764,
+      "learning_rate": 1.7254901960784314e-05,
+      "loss": 0.6376,
+      "step": 55
+    },
+    {
+      "epoch": 0.1100196463654224,
+      "grad_norm": 0.12992519749165568,
+      "learning_rate": 1.7568627450980392e-05,
+      "loss": 0.6233,
+      "step": 56
+    },
+    {
+      "epoch": 0.11198428290766209,
+      "grad_norm": 0.16834550898642528,
+      "learning_rate": 1.7882352941176474e-05,
+      "loss": 0.6169,
+      "step": 57
+    },
+    {
+      "epoch": 0.11394891944990176,
+      "grad_norm": 0.14248743350328516,
+      "learning_rate": 1.8196078431372552e-05,
+      "loss": 0.6205,
+      "step": 58
+    },
+    {
+      "epoch": 0.11591355599214145,
+      "grad_norm": 0.16634353596662652,
+      "learning_rate": 1.850980392156863e-05,
+      "loss": 0.6294,
+      "step": 59
+    },
+    {
+      "epoch": 0.11787819253438114,
+      "grad_norm": 0.14932347463361945,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.6204,
+      "step": 60
+    },
+    {
+      "epoch": 0.11984282907662082,
+      "grad_norm": 0.11990395489226027,
+      "learning_rate": 1.9137254901960786e-05,
+      "loss": 0.6147,
+      "step": 61
+    },
+    {
+      "epoch": 0.12180746561886051,
+      "grad_norm": 0.1397150153462449,
+      "learning_rate": 1.9450980392156864e-05,
+      "loss": 0.6115,
+      "step": 62
+    },
+    {
+      "epoch": 0.1237721021611002,
+      "grad_norm": 0.17410375396670338,
+      "learning_rate": 1.9764705882352945e-05,
+      "loss": 0.6161,
+      "step": 63
+    },
+    {
+      "epoch": 0.12573673870333987,
+      "grad_norm": 0.13563438919796725,
+      "learning_rate": 2.007843137254902e-05,
+      "loss": 0.6112,
+      "step": 64
+    },
+    {
+      "epoch": 0.12770137524557956,
+      "grad_norm": 0.1394781911000326,
+      "learning_rate": 2.0392156862745097e-05,
+      "loss": 0.6117,
+      "step": 65
+    },
+    {
+      "epoch": 0.12966601178781925,
+      "grad_norm": 0.13959863704945039,
+      "learning_rate": 2.070588235294118e-05,
+      "loss": 0.6108,
+      "step": 66
+    },
+    {
+      "epoch": 0.13163064833005894,
+      "grad_norm": 0.13152013400392104,
+      "learning_rate": 2.1019607843137257e-05,
+      "loss": 0.607,
+      "step": 67
+    },
+    {
+      "epoch": 0.13359528487229863,
+      "grad_norm": 0.16968446992608743,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 0.6081,
+      "step": 68
+    },
+    {
+      "epoch": 0.13555992141453832,
+      "grad_norm": 0.2060911726933915,
+      "learning_rate": 2.1647058823529413e-05,
+      "loss": 0.6165,
+      "step": 69
+    },
+    {
+      "epoch": 0.137524557956778,
+      "grad_norm": 0.30815384285852987,
+      "learning_rate": 2.1960784313725494e-05,
+      "loss": 0.5979,
+      "step": 70
+    },
+    {
+      "epoch": 0.13948919449901767,
+      "grad_norm": 0.4203038764367161,
+      "learning_rate": 2.2274509803921572e-05,
+      "loss": 0.6075,
+      "step": 71
+    },
+    {
+      "epoch": 0.14145383104125736,
+      "grad_norm": 0.353471448879735,
+      "learning_rate": 2.258823529411765e-05,
+      "loss": 0.6059,
+      "step": 72
+    },
+    {
+      "epoch": 0.14341846758349705,
+      "grad_norm": 0.30318315594265116,
+      "learning_rate": 2.2901960784313724e-05,
+      "loss": 0.6049,
+      "step": 73
+    },
+    {
+      "epoch": 0.14538310412573674,
+      "grad_norm": 0.30537092257982645,
+      "learning_rate": 2.321568627450981e-05,
+      "loss": 0.6111,
+      "step": 74
+    },
+    {
+      "epoch": 0.14734774066797643,
+      "grad_norm": 0.2317117119260459,
+      "learning_rate": 2.3529411764705884e-05,
+      "loss": 0.601,
+      "step": 75
+    },
+    {
+      "epoch": 0.14931237721021612,
+      "grad_norm": 0.2116602536913387,
+      "learning_rate": 2.384313725490196e-05,
+      "loss": 0.6061,
+      "step": 76
+    },
+    {
+      "epoch": 0.1512770137524558,
+      "grad_norm": 0.27402775201531815,
+      "learning_rate": 2.415686274509804e-05,
+      "loss": 0.6034,
+      "step": 77
+    },
+    {
+      "epoch": 0.15324165029469547,
+      "grad_norm": 0.2977498026827592,
+      "learning_rate": 2.447058823529412e-05,
+      "loss": 0.6008,
+      "step": 78
+    },
+    {
+      "epoch": 0.15520628683693516,
+      "grad_norm": 0.27801529550991316,
+      "learning_rate": 2.47843137254902e-05,
+      "loss": 0.5982,
+      "step": 79
+    },
+    {
+      "epoch": 0.15717092337917485,
+      "grad_norm": 0.25256172602345045,
+      "learning_rate": 2.5098039215686277e-05,
+      "loss": 0.5991,
+      "step": 80
+    },
+    {
+      "epoch": 0.15913555992141454,
+      "grad_norm": 0.23771392228667046,
+      "learning_rate": 2.5411764705882355e-05,
+      "loss": 0.5867,
+      "step": 81
+    },
+    {
+      "epoch": 0.16110019646365423,
+      "grad_norm": 0.18863808327524728,
+      "learning_rate": 2.5725490196078436e-05,
+      "loss": 0.5923,
+      "step": 82
+    },
+    {
+      "epoch": 0.16306483300589392,
+      "grad_norm": 0.19804839112634723,
+      "learning_rate": 2.6039215686274514e-05,
+      "loss": 0.6086,
+      "step": 83
+    },
+    {
+      "epoch": 0.1650294695481336,
+      "grad_norm": 0.2316749425214643,
+      "learning_rate": 2.635294117647059e-05,
+      "loss": 0.5998,
+      "step": 84
+    },
+    {
+      "epoch": 0.16699410609037327,
+      "grad_norm": 0.25394279539661124,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5906,
+      "step": 85
+    },
+    {
+      "epoch": 0.16895874263261296,
+      "grad_norm": 0.3227137035570147,
+      "learning_rate": 2.6980392156862748e-05,
+      "loss": 0.5873,
+      "step": 86
+    },
+    {
+      "epoch": 0.17092337917485265,
+      "grad_norm": 0.3472145869159207,
+      "learning_rate": 2.7294117647058826e-05,
+      "loss": 0.596,
+      "step": 87
+    },
+    {
+      "epoch": 0.17288801571709234,
+      "grad_norm": 0.30459848801000117,
+      "learning_rate": 2.7607843137254904e-05,
+      "loss": 0.5904,
+      "step": 88
+    },
+    {
+      "epoch": 0.17485265225933203,
+      "grad_norm": 0.3207461968369133,
+      "learning_rate": 2.7921568627450982e-05,
+      "loss": 0.5853,
+      "step": 89
+    },
+    {
+      "epoch": 0.17681728880157171,
+      "grad_norm": 0.27057170636674666,
+      "learning_rate": 2.8235294117647063e-05,
+      "loss": 0.59,
+      "step": 90
+    },
+    {
+      "epoch": 0.1787819253438114,
+      "grad_norm": 0.2756953873050925,
+      "learning_rate": 2.854901960784314e-05,
+      "loss": 0.5883,
+      "step": 91
+    },
+    {
+      "epoch": 0.1807465618860511,
+      "grad_norm": 0.2872416937226475,
+      "learning_rate": 2.886274509803922e-05,
+      "loss": 0.602,
+      "step": 92
+    },
+    {
+      "epoch": 0.18271119842829076,
+      "grad_norm": 0.30800836560469785,
+      "learning_rate": 2.9176470588235294e-05,
+      "loss": 0.5993,
+      "step": 93
+    },
+    {
+      "epoch": 0.18467583497053044,
+      "grad_norm": 0.33804690423398226,
+      "learning_rate": 2.949019607843138e-05,
+      "loss": 0.587,
+      "step": 94
+    },
+    {
+      "epoch": 0.18664047151277013,
+      "grad_norm": 0.38147052243516966,
+      "learning_rate": 2.9803921568627453e-05,
+      "loss": 0.5942,
+      "step": 95
+    },
+    {
+      "epoch": 0.18860510805500982,
+      "grad_norm": 0.38610713989422984,
+      "learning_rate": 3.011764705882353e-05,
+      "loss": 0.5875,
+      "step": 96
+    },
+    {
+      "epoch": 0.1905697445972495,
+      "grad_norm": 0.36123368433216513,
+      "learning_rate": 3.043137254901961e-05,
+      "loss": 0.5838,
+      "step": 97
+    },
+    {
+      "epoch": 0.1925343811394892,
+      "grad_norm": 0.4136102931466299,
+      "learning_rate": 3.0745098039215694e-05,
+      "loss": 0.5918,
+      "step": 98
+    },
+    {
+      "epoch": 0.1944990176817289,
+      "grad_norm": 0.47561373425600084,
+      "learning_rate": 3.1058823529411765e-05,
+      "loss": 0.5897,
+      "step": 99
+    },
+    {
+      "epoch": 0.19646365422396855,
+      "grad_norm": 0.44450244018126395,
+      "learning_rate": 3.137254901960784e-05,
+      "loss": 0.5919,
+      "step": 100
+    },
+    {
+      "epoch": 0.19842829076620824,
+      "grad_norm": 0.4574733294813543,
+      "learning_rate": 3.168627450980392e-05,
+      "loss": 0.5872,
+      "step": 101
+    },
+    {
+      "epoch": 0.20039292730844793,
+      "grad_norm": 0.42758643254664114,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.5874,
+      "step": 102
+    },
+    {
+      "epoch": 0.20235756385068762,
+      "grad_norm": 0.3457844917510654,
+      "learning_rate": 3.231372549019608e-05,
+      "loss": 0.5932,
+      "step": 103
+    },
+    {
+      "epoch": 0.2043222003929273,
+      "grad_norm": 0.2997207165056099,
+      "learning_rate": 3.262745098039216e-05,
+      "loss": 0.5869,
+      "step": 104
+    },
+    {
+      "epoch": 0.206286836935167,
+      "grad_norm": 0.3250161298040834,
+      "learning_rate": 3.294117647058824e-05,
+      "loss": 0.5857,
+      "step": 105
+    },
+    {
+      "epoch": 0.2082514734774067,
+      "grad_norm": 0.35709303268280485,
+      "learning_rate": 3.325490196078432e-05,
+      "loss": 0.586,
+      "step": 106
+    },
+    {
+      "epoch": 0.21021611001964635,
+      "grad_norm": 0.3504384861576393,
+      "learning_rate": 3.3568627450980395e-05,
+      "loss": 0.5928,
+      "step": 107
+    },
+    {
+      "epoch": 0.21218074656188604,
+      "grad_norm": 0.2850506109210298,
+      "learning_rate": 3.388235294117647e-05,
+      "loss": 0.5914,
+      "step": 108
+    },
+    {
+      "epoch": 0.21414538310412573,
+      "grad_norm": 0.23775553707570063,
+      "learning_rate": 3.419607843137255e-05,
+      "loss": 0.582,
+      "step": 109
+    },
+    {
+      "epoch": 0.21611001964636542,
+      "grad_norm": 0.26710197855885015,
+      "learning_rate": 3.450980392156863e-05,
+      "loss": 0.5838,
+      "step": 110
+    },
+    {
+      "epoch": 0.2180746561886051,
+      "grad_norm": 0.3014314981087465,
+      "learning_rate": 3.482352941176471e-05,
+      "loss": 0.5786,
+      "step": 111
+    },
+    {
+      "epoch": 0.2200392927308448,
+      "grad_norm": 0.3558874198496929,
+      "learning_rate": 3.5137254901960785e-05,
+      "loss": 0.586,
+      "step": 112
+    },
+    {
+      "epoch": 0.2220039292730845,
+      "grad_norm": 0.40226538364358416,
+      "learning_rate": 3.545098039215686e-05,
+      "loss": 0.5878,
+      "step": 113
+    },
+    {
+      "epoch": 0.22396856581532418,
+      "grad_norm": 0.3942559937206116,
+      "learning_rate": 3.576470588235295e-05,
+      "loss": 0.586,
+      "step": 114
+    },
+    {
+      "epoch": 0.22593320235756384,
+      "grad_norm": 0.418531956523001,
+      "learning_rate": 3.6078431372549025e-05,
+      "loss": 0.5841,
+      "step": 115
+    },
+    {
+      "epoch": 0.22789783889980353,
+      "grad_norm": 0.4471611314486502,
+      "learning_rate": 3.6392156862745103e-05,
+      "loss": 0.5871,
+      "step": 116
+    },
+    {
+      "epoch": 0.22986247544204322,
+      "grad_norm": 0.519536898916548,
+      "learning_rate": 3.6705882352941175e-05,
+      "loss": 0.5779,
+      "step": 117
+    },
+    {
+      "epoch": 0.2318271119842829,
+      "grad_norm": 0.5128072557499016,
+      "learning_rate": 3.701960784313726e-05,
+      "loss": 0.5762,
+      "step": 118
+    },
+    {
+      "epoch": 0.2337917485265226,
+      "grad_norm": 0.4455860972278374,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 0.5778,
+      "step": 119
+    },
+    {
+      "epoch": 0.2357563850687623,
+      "grad_norm": 0.4755882708542053,
+      "learning_rate": 3.7647058823529415e-05,
+      "loss": 0.584,
+      "step": 120
+    },
+    {
+      "epoch": 0.23772102161100198,
+      "grad_norm": 0.49885300644794933,
+      "learning_rate": 3.796078431372549e-05,
+      "loss": 0.5757,
+      "step": 121
+    },
+    {
+      "epoch": 0.23968565815324164,
+      "grad_norm": 0.5049104034056423,
+      "learning_rate": 3.827450980392157e-05,
+      "loss": 0.5811,
+      "step": 122
+    },
+    {
+      "epoch": 0.24165029469548133,
+      "grad_norm": 0.5153911457057996,
+      "learning_rate": 3.858823529411765e-05,
+      "loss": 0.5782,
+      "step": 123
+    },
+    {
+      "epoch": 0.24361493123772102,
+      "grad_norm": 0.5946852956252856,
+      "learning_rate": 3.890196078431373e-05,
+      "loss": 0.5862,
+      "step": 124
+    },
+    {
+      "epoch": 0.2455795677799607,
+      "grad_norm": 0.6914028211029504,
+      "learning_rate": 3.9215686274509805e-05,
+      "loss": 0.5704,
+      "step": 125
+    },
+    {
+      "epoch": 0.2475442043222004,
+      "grad_norm": 0.7663543204331588,
+      "learning_rate": 3.952941176470589e-05,
+      "loss": 0.5816,
+      "step": 126
+    },
+    {
+      "epoch": 0.24950884086444008,
+      "grad_norm": 0.7431466031758991,
+      "learning_rate": 3.984313725490197e-05,
+      "loss": 0.5895,
+      "step": 127
+    },
+    {
+      "epoch": 0.25147347740667975,
+      "grad_norm": 0.7806618522501877,
+      "learning_rate": 4.015686274509804e-05,
+      "loss": 0.5792,
+      "step": 128
+    },
+    {
+      "epoch": 0.25343811394891946,
+      "grad_norm": 0.8275071289011663,
+      "learning_rate": 4.047058823529412e-05,
+      "loss": 0.6002,
+      "step": 129
+    },
+    {
+      "epoch": 0.2554027504911591,
+      "grad_norm": 0.6050894722579638,
+      "learning_rate": 4.0784313725490195e-05,
+      "loss": 0.5778,
+      "step": 130
+    },
+    {
+      "epoch": 0.25736738703339884,
+      "grad_norm": 0.4962560275368978,
+      "learning_rate": 4.109803921568627e-05,
+      "loss": 0.5777,
+      "step": 131
+    },
+    {
+      "epoch": 0.2593320235756385,
+      "grad_norm": 0.7196156873281503,
+      "learning_rate": 4.141176470588236e-05,
+      "loss": 0.584,
+      "step": 132
+    },
+    {
+      "epoch": 0.26129666011787817,
+      "grad_norm": 0.5639095738838849,
+      "learning_rate": 4.1725490196078435e-05,
+      "loss": 0.581,
+      "step": 133
+    },
+    {
+      "epoch": 0.2632612966601179,
+      "grad_norm": 0.5895022934585601,
+      "learning_rate": 4.203921568627451e-05,
+      "loss": 0.5797,
+      "step": 134
+    },
+    {
+      "epoch": 0.26522593320235754,
+      "grad_norm": 0.6197426772338325,
+      "learning_rate": 4.235294117647059e-05,
+      "loss": 0.5759,
+      "step": 135
+    },
+    {
+      "epoch": 0.26719056974459726,
+      "grad_norm": 0.42802386749742066,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 0.5798,
+      "step": 136
+    },
+    {
+      "epoch": 0.2691552062868369,
+      "grad_norm": 0.5334701835528505,
+      "learning_rate": 4.298039215686275e-05,
+      "loss": 0.5819,
+      "step": 137
+    },
+    {
+      "epoch": 0.27111984282907664,
+      "grad_norm": 0.35217469379020394,
+      "learning_rate": 4.3294117647058825e-05,
+      "loss": 0.5691,
+      "step": 138
+    },
+    {
+      "epoch": 0.2730844793713163,
+      "grad_norm": 0.46589992141673775,
+      "learning_rate": 4.36078431372549e-05,
+      "loss": 0.5779,
+      "step": 139
+    },
+    {
+      "epoch": 0.275049115913556,
+      "grad_norm": 0.4032429080390111,
+      "learning_rate": 4.392156862745099e-05,
+      "loss": 0.5608,
+      "step": 140
+    },
+    {
+      "epoch": 0.2770137524557957,
+      "grad_norm": 0.34083102871350945,
+      "learning_rate": 4.4235294117647066e-05,
+      "loss": 0.5673,
+      "step": 141
+    },
+    {
+      "epoch": 0.27897838899803534,
+      "grad_norm": 0.4423126076850681,
+      "learning_rate": 4.4549019607843144e-05,
+      "loss": 0.5728,
+      "step": 142
+    },
+    {
+      "epoch": 0.28094302554027506,
+      "grad_norm": 0.3333848575121668,
+      "learning_rate": 4.486274509803922e-05,
+      "loss": 0.5784,
+      "step": 143
+    },
+    {
+      "epoch": 0.2829076620825147,
+      "grad_norm": 0.41568025443087114,
+      "learning_rate": 4.51764705882353e-05,
+      "loss": 0.5804,
+      "step": 144
+    },
+    {
+      "epoch": 0.28487229862475444,
+      "grad_norm": 0.4138460383825594,
+      "learning_rate": 4.549019607843138e-05,
+      "loss": 0.5649,
+      "step": 145
+    },
+    {
+      "epoch": 0.2868369351669941,
+      "grad_norm": 0.4065129888016761,
+      "learning_rate": 4.580392156862745e-05,
+      "loss": 0.5705,
+      "step": 146
+    },
+    {
+      "epoch": 0.2888015717092338,
+      "grad_norm": 0.376793741234179,
+      "learning_rate": 4.611764705882353e-05,
+      "loss": 0.5797,
+      "step": 147
+    },
+    {
+      "epoch": 0.2907662082514735,
+      "grad_norm": 0.4112410016392894,
+      "learning_rate": 4.643137254901962e-05,
+      "loss": 0.5815,
+      "step": 148
+    },
+    {
+      "epoch": 0.29273084479371314,
+      "grad_norm": 0.5007734324234949,
+      "learning_rate": 4.674509803921569e-05,
+      "loss": 0.5669,
+      "step": 149
+    },
+    {
+      "epoch": 0.29469548133595286,
+      "grad_norm": 0.6551242179060432,
+      "learning_rate": 4.705882352941177e-05,
+      "loss": 0.5842,
+      "step": 150
+    },
+    {
+      "epoch": 0.2966601178781925,
+      "grad_norm": 0.8244930256848831,
+      "learning_rate": 4.7372549019607845e-05,
+      "loss": 0.5756,
+      "step": 151
+    },
+    {
+      "epoch": 0.29862475442043224,
+      "grad_norm": 0.8057685957620361,
+      "learning_rate": 4.768627450980392e-05,
+      "loss": 0.5684,
+      "step": 152
+    },
+    {
+      "epoch": 0.3005893909626719,
+      "grad_norm": 0.5317869581740617,
+      "learning_rate": 4.8e-05,
+      "loss": 0.5687,
+      "step": 153
+    },
+    {
+      "epoch": 0.3025540275049116,
+      "grad_norm": 0.5048199944428914,
+      "learning_rate": 4.831372549019608e-05,
+      "loss": 0.5698,
+      "step": 154
+    },
+    {
+      "epoch": 0.3045186640471513,
+      "grad_norm": 0.5233410544293879,
+      "learning_rate": 4.862745098039216e-05,
+      "loss": 0.5697,
+      "step": 155
+    },
+    {
+      "epoch": 0.30648330058939094,
+      "grad_norm": 0.6902078867673104,
+      "learning_rate": 4.894117647058824e-05,
+      "loss": 0.5634,
+      "step": 156
+    },
+    {
+      "epoch": 0.30844793713163066,
+      "grad_norm": 0.6331596758134145,
+      "learning_rate": 4.925490196078432e-05,
+      "loss": 0.5728,
+      "step": 157
+    },
+    {
+      "epoch": 0.3104125736738703,
+      "grad_norm": 0.45926091471292046,
+      "learning_rate": 4.95686274509804e-05,
+      "loss": 0.5747,
+      "step": 158
+    },
+    {
+      "epoch": 0.31237721021611004,
+      "grad_norm": 0.7842006949184721,
+      "learning_rate": 4.9882352941176476e-05,
+      "loss": 0.5682,
+      "step": 159
+    },
+    {
+      "epoch": 0.3143418467583497,
+      "grad_norm": 0.5383765257031949,
+      "learning_rate": 5.0196078431372554e-05,
+      "loss": 0.5664,
+      "step": 160
+    },
+    {
+      "epoch": 0.3163064833005894,
+      "grad_norm": 0.6142508334815271,
+      "learning_rate": 5.050980392156863e-05,
+      "loss": 0.5789,
+      "step": 161
+    },
+    {
+      "epoch": 0.3182711198428291,
+      "grad_norm": 0.5151627988533322,
+      "learning_rate": 5.082352941176471e-05,
+      "loss": 0.5753,
+      "step": 162
+    },
+    {
+      "epoch": 0.32023575638506874,
+      "grad_norm": 0.5422137523676791,
+      "learning_rate": 5.113725490196079e-05,
+      "loss": 0.5686,
+      "step": 163
+    },
+    {
+      "epoch": 0.32220039292730845,
+      "grad_norm": 0.5969299478043332,
+      "learning_rate": 5.145098039215687e-05,
+      "loss": 0.5786,
+      "step": 164
+    },
+    {
+      "epoch": 0.3241650294695481,
+      "grad_norm": 0.46776882698750627,
+      "learning_rate": 5.176470588235295e-05,
+      "loss": 0.5691,
+      "step": 165
+    },
+    {
+      "epoch": 0.32612966601178783,
+      "grad_norm": 0.5647896528493653,
+      "learning_rate": 5.207843137254903e-05,
+      "loss": 0.5681,
+      "step": 166
+    },
+    {
+      "epoch": 0.3280943025540275,
+      "grad_norm": 0.46259845802981747,
+      "learning_rate": 5.23921568627451e-05,
+      "loss": 0.5676,
+      "step": 167
+    },
+    {
+      "epoch": 0.3300589390962672,
+      "grad_norm": 0.5037869499583382,
+      "learning_rate": 5.270588235294118e-05,
+      "loss": 0.5572,
+      "step": 168
+    },
+    {
+      "epoch": 0.3320235756385069,
+      "grad_norm": 0.5182477227868953,
+      "learning_rate": 5.3019607843137255e-05,
+      "loss": 0.5673,
+      "step": 169
+    },
+    {
+      "epoch": 0.33398821218074654,
+      "grad_norm": 0.3525527212584046,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.5534,
+      "step": 170
+    },
+    {
+      "epoch": 0.33595284872298625,
+      "grad_norm": 0.5468318713914483,
+      "learning_rate": 5.364705882352941e-05,
+      "loss": 0.5707,
+      "step": 171
+    },
+    {
+      "epoch": 0.3379174852652259,
+      "grad_norm": 0.34703884370142285,
+      "learning_rate": 5.3960784313725496e-05,
+      "loss": 0.5608,
+      "step": 172
+    },
+    {
+      "epoch": 0.33988212180746563,
+      "grad_norm": 0.49009719439924027,
+      "learning_rate": 5.4274509803921574e-05,
+      "loss": 0.5562,
+      "step": 173
+    },
+    {
+      "epoch": 0.3418467583497053,
+      "grad_norm": 0.40696324441241855,
+      "learning_rate": 5.458823529411765e-05,
+      "loss": 0.5685,
+      "step": 174
+    },
+    {
+      "epoch": 0.343811394891945,
+      "grad_norm": 0.3201099831542352,
+      "learning_rate": 5.490196078431373e-05,
+      "loss": 0.566,
+      "step": 175
+    },
+    {
+      "epoch": 0.34577603143418467,
+      "grad_norm": 0.3983721406550538,
+      "learning_rate": 5.521568627450981e-05,
+      "loss": 0.5668,
+      "step": 176
+    },
+    {
+      "epoch": 0.3477406679764244,
+      "grad_norm": 0.3722762755747041,
+      "learning_rate": 5.5529411764705886e-05,
+      "loss": 0.5786,
+      "step": 177
+    },
+    {
+      "epoch": 0.34970530451866405,
+      "grad_norm": 0.5044908300254344,
+      "learning_rate": 5.5843137254901964e-05,
+      "loss": 0.5633,
+      "step": 178
+    },
+    {
+      "epoch": 0.3516699410609037,
+      "grad_norm": 0.5319396845037461,
+      "learning_rate": 5.615686274509804e-05,
+      "loss": 0.5594,
+      "step": 179
+    },
+    {
+      "epoch": 0.35363457760314343,
+      "grad_norm": 0.6785576019964707,
+      "learning_rate": 5.6470588235294126e-05,
+      "loss": 0.5597,
+      "step": 180
+    },
+    {
+      "epoch": 0.3555992141453831,
+      "grad_norm": 0.8103386994176834,
+      "learning_rate": 5.6784313725490204e-05,
+      "loss": 0.5818,
+      "step": 181
+    },
+    {
+      "epoch": 0.3575638506876228,
+      "grad_norm": 0.7548468304684407,
+      "learning_rate": 5.709803921568628e-05,
+      "loss": 0.5598,
+      "step": 182
+    },
+    {
+      "epoch": 0.35952848722986247,
+      "grad_norm": 0.4872834495898871,
+      "learning_rate": 5.741176470588236e-05,
+      "loss": 0.5556,
+      "step": 183
+    },
+    {
+      "epoch": 0.3614931237721022,
+      "grad_norm": 0.4093411958457091,
+      "learning_rate": 5.772549019607844e-05,
+      "loss": 0.554,
+      "step": 184
+    },
+    {
+      "epoch": 0.36345776031434185,
+      "grad_norm": 0.5322076079319503,
+      "learning_rate": 5.8039215686274516e-05,
+      "loss": 0.5687,
+      "step": 185
+    },
+    {
+      "epoch": 0.3654223968565815,
+      "grad_norm": 0.43172101806687746,
+      "learning_rate": 5.835294117647059e-05,
+      "loss": 0.5617,
+      "step": 186
+    },
+    {
+      "epoch": 0.36738703339882123,
+      "grad_norm": 0.49115107703621674,
+      "learning_rate": 5.8666666666666665e-05,
+      "loss": 0.5682,
+      "step": 187
+    },
+    {
+      "epoch": 0.3693516699410609,
+      "grad_norm": 0.598686545400832,
+      "learning_rate": 5.898039215686276e-05,
+      "loss": 0.5716,
+      "step": 188
+    },
+    {
+      "epoch": 0.3713163064833006,
+      "grad_norm": 0.6159554677370218,
+      "learning_rate": 5.929411764705883e-05,
+      "loss": 0.5551,
+      "step": 189
+    },
+    {
+      "epoch": 0.37328094302554027,
+      "grad_norm": 0.5942338295716559,
+      "learning_rate": 5.9607843137254906e-05,
+      "loss": 0.562,
+      "step": 190
+    },
+    {
+      "epoch": 0.37524557956778,
+      "grad_norm": 0.539945526100254,
+      "learning_rate": 5.9921568627450984e-05,
+      "loss": 0.5599,
+      "step": 191
+    },
+    {
+      "epoch": 0.37721021611001965,
+      "grad_norm": 0.44605784780723545,
+      "learning_rate": 6.023529411764706e-05,
+      "loss": 0.5631,
+      "step": 192
+    },
+    {
+      "epoch": 0.3791748526522593,
+      "grad_norm": 0.4197322777909507,
+      "learning_rate": 6.054901960784314e-05,
+      "loss": 0.5644,
+      "step": 193
+    },
+    {
+      "epoch": 0.381139489194499,
+      "grad_norm": 0.3872094920862164,
+      "learning_rate": 6.086274509803922e-05,
+      "loss": 0.5605,
+      "step": 194
+    },
+    {
+      "epoch": 0.3831041257367387,
+      "grad_norm": 0.4474671493267813,
+      "learning_rate": 6.117647058823529e-05,
+      "loss": 0.5583,
+      "step": 195
+    },
+    {
+      "epoch": 0.3850687622789784,
+      "grad_norm": 0.5494866068109173,
+      "learning_rate": 6.149019607843139e-05,
+      "loss": 0.5584,
+      "step": 196
+    },
+    {
+      "epoch": 0.38703339882121807,
+      "grad_norm": 0.3807522242158805,
+      "learning_rate": 6.180392156862746e-05,
+      "loss": 0.5571,
+      "step": 197
+    },
+    {
+      "epoch": 0.3889980353634578,
+      "grad_norm": 0.313773115742069,
+      "learning_rate": 6.211764705882353e-05,
+      "loss": 0.5502,
+      "step": 198
+    },
+    {
+      "epoch": 0.39096267190569745,
+      "grad_norm": 0.38046415055185523,
+      "learning_rate": 6.243137254901961e-05,
+      "loss": 0.5615,
+      "step": 199
+    },
+    {
+      "epoch": 0.3929273084479371,
+      "grad_norm": 0.32096778392174274,
+      "learning_rate": 6.274509803921569e-05,
+      "loss": 0.5577,
+      "step": 200
+    },
+    {
+      "epoch": 0.3948919449901768,
+      "grad_norm": 0.25051066379310505,
+      "learning_rate": 6.305882352941177e-05,
+      "loss": 0.5603,
+      "step": 201
+    },
+    {
+      "epoch": 0.3968565815324165,
+      "grad_norm": 0.27959950767566505,
+      "learning_rate": 6.337254901960784e-05,
+      "loss": 0.5577,
+      "step": 202
+    },
+    {
+      "epoch": 0.3988212180746562,
+      "grad_norm": 0.28207916549231027,
+      "learning_rate": 6.368627450980393e-05,
+      "loss": 0.5377,
+      "step": 203
+    },
+    {
+      "epoch": 0.40078585461689586,
+      "grad_norm": 0.30558508915968063,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.5597,
+      "step": 204
+    },
+    {
+      "epoch": 0.4027504911591356,
+      "grad_norm": 0.3404426664222926,
+      "learning_rate": 6.431372549019608e-05,
+      "loss": 0.5627,
+      "step": 205
+    },
+    {
+      "epoch": 0.40471512770137524,
+      "grad_norm": 0.32994814631972563,
+      "learning_rate": 6.462745098039217e-05,
+      "loss": 0.5587,
+      "step": 206
+    },
+    {
+      "epoch": 0.4066797642436149,
+      "grad_norm": 0.5080956748849085,
+      "learning_rate": 6.494117647058824e-05,
+      "loss": 0.5582,
+      "step": 207
+    },
+    {
+      "epoch": 0.4086444007858546,
+      "grad_norm": 0.6852559520494648,
+      "learning_rate": 6.525490196078432e-05,
+      "loss": 0.5598,
+      "step": 208
+    },
+    {
+      "epoch": 0.4106090373280943,
+      "grad_norm": 0.822055329991665,
+      "learning_rate": 6.55686274509804e-05,
+      "loss": 0.5688,
+      "step": 209
+    },
+    {
+      "epoch": 0.412573673870334,
+      "grad_norm": 0.6599002099862327,
+      "learning_rate": 6.588235294117648e-05,
+      "loss": 0.5638,
+      "step": 210
+    },
+    {
+      "epoch": 0.41453831041257366,
+      "grad_norm": 0.4809338255958477,
+      "learning_rate": 6.619607843137255e-05,
+      "loss": 0.555,
+      "step": 211
+    },
+    {
+      "epoch": 0.4165029469548134,
+      "grad_norm": 0.5846649111693321,
+      "learning_rate": 6.650980392156863e-05,
+      "loss": 0.5574,
+      "step": 212
+    },
+    {
+      "epoch": 0.41846758349705304,
+      "grad_norm": 0.4888013933561947,
+      "learning_rate": 6.682352941176472e-05,
+      "loss": 0.5638,
+      "step": 213
+    },
+    {
+      "epoch": 0.4204322200392927,
+      "grad_norm": 0.5182119669802906,
+      "learning_rate": 6.713725490196079e-05,
+      "loss": 0.5566,
+      "step": 214
+    },
+    {
+      "epoch": 0.4223968565815324,
+      "grad_norm": 0.6715876760875897,
+      "learning_rate": 6.745098039215687e-05,
+      "loss": 0.566,
+      "step": 215
+    },
+    {
+      "epoch": 0.4243614931237721,
+      "grad_norm": 0.48467709150137156,
+      "learning_rate": 6.776470588235295e-05,
+      "loss": 0.5567,
+      "step": 216
+    },
+    {
+      "epoch": 0.4263261296660118,
+      "grad_norm": 0.48137304474394016,
+      "learning_rate": 6.807843137254902e-05,
+      "loss": 0.5637,
+      "step": 217
+    },
+    {
+      "epoch": 0.42829076620825146,
+      "grad_norm": 0.48497814698892505,
+      "learning_rate": 6.83921568627451e-05,
+      "loss": 0.564,
+      "step": 218
+    },
+    {
+      "epoch": 0.4302554027504912,
+      "grad_norm": 0.402641798925953,
+      "learning_rate": 6.870588235294117e-05,
+      "loss": 0.5578,
+      "step": 219
+    },
+    {
+      "epoch": 0.43222003929273084,
+      "grad_norm": 0.4908603167792517,
+      "learning_rate": 6.901960784313726e-05,
+      "loss": 0.5636,
+      "step": 220
+    },
+    {
+      "epoch": 0.43418467583497056,
+      "grad_norm": 0.3220611027661207,
+      "learning_rate": 6.933333333333334e-05,
+      "loss": 0.5594,
+      "step": 221
+    },
+    {
+      "epoch": 0.4361493123772102,
+      "grad_norm": 0.5006692647799165,
+      "learning_rate": 6.964705882352941e-05,
+      "loss": 0.559,
+      "step": 222
+    },
+    {
+      "epoch": 0.4381139489194499,
+      "grad_norm": 0.5483745447456514,
+      "learning_rate": 6.99607843137255e-05,
+      "loss": 0.5471,
+      "step": 223
+    },
+    {
+      "epoch": 0.4400785854616896,
+      "grad_norm": 0.46459622931667643,
+      "learning_rate": 7.027450980392157e-05,
+      "loss": 0.5587,
+      "step": 224
+    },
+    {
+      "epoch": 0.44204322200392926,
+      "grad_norm": 0.4467959518076016,
+      "learning_rate": 7.058823529411765e-05,
+      "loss": 0.5555,
+      "step": 225
+    },
+    {
+      "epoch": 0.444007858546169,
+      "grad_norm": 0.3464956908811292,
+      "learning_rate": 7.090196078431373e-05,
+      "loss": 0.5621,
+      "step": 226
+    },
+    {
+      "epoch": 0.44597249508840864,
+      "grad_norm": 0.43965783111822543,
+      "learning_rate": 7.121568627450981e-05,
+      "loss": 0.5514,
+      "step": 227
+    },
+    {
+      "epoch": 0.44793713163064836,
+      "grad_norm": 0.5909625469259163,
+      "learning_rate": 7.15294117647059e-05,
+      "loss": 0.5484,
+      "step": 228
+    },
+    {
+      "epoch": 0.449901768172888,
+      "grad_norm": 0.7073611169097613,
+      "learning_rate": 7.184313725490197e-05,
+      "loss": 0.5577,
+      "step": 229
+    },
+    {
+      "epoch": 0.4518664047151277,
+      "grad_norm": 0.761958390879125,
+      "learning_rate": 7.215686274509805e-05,
+      "loss": 0.5573,
+      "step": 230
+    },
+    {
+      "epoch": 0.4538310412573674,
+      "grad_norm": 0.6187090313784889,
+      "learning_rate": 7.247058823529412e-05,
+      "loss": 0.5523,
+      "step": 231
+    },
+    {
+      "epoch": 0.45579567779960706,
+      "grad_norm": 0.3907538119086157,
+      "learning_rate": 7.278431372549021e-05,
+      "loss": 0.5605,
+      "step": 232
+    },
+    {
+      "epoch": 0.4577603143418468,
+      "grad_norm": 0.3742471522609025,
+      "learning_rate": 7.309803921568628e-05,
+      "loss": 0.5549,
+      "step": 233
+    },
+    {
+      "epoch": 0.45972495088408644,
+      "grad_norm": 0.4694972568518388,
+      "learning_rate": 7.341176470588235e-05,
+      "loss": 0.5546,
+      "step": 234
+    },
+    {
+      "epoch": 0.46168958742632615,
+      "grad_norm": 0.445186577655185,
+      "learning_rate": 7.372549019607843e-05,
+      "loss": 0.5595,
+      "step": 235
+    },
+    {
+      "epoch": 0.4636542239685658,
+      "grad_norm": 0.4644240068315893,
+      "learning_rate": 7.403921568627452e-05,
+      "loss": 0.5531,
+      "step": 236
+    },
+    {
+      "epoch": 0.4656188605108055,
+      "grad_norm": 0.49838029385047106,
+      "learning_rate": 7.435294117647059e-05,
+      "loss": 0.5442,
+      "step": 237
+    },
+    {
+      "epoch": 0.4675834970530452,
+      "grad_norm": 0.5465017444606399,
+      "learning_rate": 7.466666666666667e-05,
+      "loss": 0.5558,
+      "step": 238
+    },
+    {
+      "epoch": 0.46954813359528486,
+      "grad_norm": 0.5840369858326687,
+      "learning_rate": 7.498039215686275e-05,
+      "loss": 0.562,
+      "step": 239
+    },
+    {
+      "epoch": 0.4715127701375246,
+      "grad_norm": 0.6002001686842443,
+      "learning_rate": 7.529411764705883e-05,
+      "loss": 0.5636,
+      "step": 240
+    },
+    {
+      "epoch": 0.47347740667976423,
+      "grad_norm": 0.6237357819404895,
+      "learning_rate": 7.56078431372549e-05,
+      "loss": 0.5527,
+      "step": 241
+    },
+    {
+      "epoch": 0.47544204322200395,
+      "grad_norm": 0.6810690653739004,
+      "learning_rate": 7.592156862745099e-05,
+      "loss": 0.5595,
+      "step": 242
+    },
+    {
+      "epoch": 0.4774066797642436,
+      "grad_norm": 0.5229908242814304,
+      "learning_rate": 7.623529411764706e-05,
+      "loss": 0.5558,
+      "step": 243
+    },
+    {
+      "epoch": 0.4793713163064833,
+      "grad_norm": 0.5093188065291652,
+      "learning_rate": 7.654901960784314e-05,
+      "loss": 0.5636,
+      "step": 244
+    },
+    {
+      "epoch": 0.481335952848723,
+      "grad_norm": 0.6173276618382788,
+      "learning_rate": 7.686274509803923e-05,
+      "loss": 0.554,
+      "step": 245
+    },
+    {
+      "epoch": 0.48330058939096265,
+      "grad_norm": 0.5110018520637413,
+      "learning_rate": 7.71764705882353e-05,
+      "loss": 0.5546,
+      "step": 246
+    },
+    {
+      "epoch": 0.48526522593320237,
+      "grad_norm": 0.5784265144762075,
+      "learning_rate": 7.749019607843138e-05,
+      "loss": 0.5577,
+      "step": 247
+    },
+    {
+      "epoch": 0.48722986247544203,
+      "grad_norm": 0.5371267626960764,
+      "learning_rate": 7.780392156862745e-05,
+      "loss": 0.5489,
+      "step": 248
+    },
+    {
+      "epoch": 0.48919449901768175,
+      "grad_norm": 0.5079124491845124,
+      "learning_rate": 7.811764705882354e-05,
+      "loss": 0.5672,
+      "step": 249
+    },
+    {
+      "epoch": 0.4911591355599214,
+      "grad_norm": 0.575928562954962,
+      "learning_rate": 7.843137254901961e-05,
+      "loss": 0.5639,
+      "step": 250
+    },
+    {
+      "epoch": 0.4931237721021611,
+      "grad_norm": 0.3855544031347316,
+      "learning_rate": 7.87450980392157e-05,
+      "loss": 0.5525,
+      "step": 251
+    },
+    {
+      "epoch": 0.4950884086444008,
+      "grad_norm": 0.42894020347220135,
+      "learning_rate": 7.905882352941178e-05,
+      "loss": 0.5597,
+      "step": 252
+    },
+    {
+      "epoch": 0.49705304518664045,
+      "grad_norm": 0.46454249020133503,
+      "learning_rate": 7.937254901960785e-05,
+      "loss": 0.56,
+      "step": 253
+    },
+    {
+      "epoch": 0.49901768172888017,
+      "grad_norm": 0.42628564469267377,
+      "learning_rate": 7.968627450980394e-05,
+      "loss": 0.5529,
+      "step": 254
+    },
+    {
+      "epoch": 0.5009823182711198,
+      "grad_norm": 0.40345380023114147,
+      "learning_rate": 8e-05,
+      "loss": 0.5481,
+      "step": 255
+    },
+    {
+      "epoch": 0.5029469548133595,
+      "grad_norm": 0.494049611808842,
+      "learning_rate": 7.999996235921187e-05,
+      "loss": 0.5632,
+      "step": 256
+    },
+    {
+      "epoch": 0.5049115913555993,
+      "grad_norm": 0.43859315092594225,
+      "learning_rate": 7.999984943691831e-05,
+      "loss": 0.5436,
+      "step": 257
+    },
+    {
+      "epoch": 0.5068762278978389,
+      "grad_norm": 0.35434156992045557,
+      "learning_rate": 7.999966123333183e-05,
+      "loss": 0.5485,
+      "step": 258
+    },
+    {
+      "epoch": 0.5088408644400786,
+      "grad_norm": 0.36175301878930605,
+      "learning_rate": 7.999939774880666e-05,
+      "loss": 0.5457,
+      "step": 259
+    },
+    {
+      "epoch": 0.5108055009823183,
+      "grad_norm": 0.4628458654111991,
+      "learning_rate": 7.999905898383868e-05,
+      "loss": 0.5535,
+      "step": 260
+    },
+    {
+      "epoch": 0.5127701375245579,
+      "grad_norm": 0.5236005635491273,
+      "learning_rate": 7.999864493906545e-05,
+      "loss": 0.5456,
+      "step": 261
+    },
+    {
+      "epoch": 0.5147347740667977,
+      "grad_norm": 0.4307501713518782,
+      "learning_rate": 7.999815561526623e-05,
+      "loss": 0.5565,
+      "step": 262
+    },
+    {
+      "epoch": 0.5166994106090373,
+      "grad_norm": 0.41155415335474865,
+      "learning_rate": 7.999759101336195e-05,
+      "loss": 0.5639,
+      "step": 263
+    },
+    {
+      "epoch": 0.518664047151277,
+      "grad_norm": 0.4363890852785939,
+      "learning_rate": 7.999695113441519e-05,
+      "loss": 0.549,
+      "step": 264
+    },
+    {
+      "epoch": 0.5206286836935167,
+      "grad_norm": 0.5490776466892815,
+      "learning_rate": 7.999623597963026e-05,
+      "loss": 0.5564,
+      "step": 265
+    },
+    {
+      "epoch": 0.5225933202357563,
+      "grad_norm": 0.4971985944233417,
+      "learning_rate": 7.999544555035308e-05,
+      "loss": 0.5455,
+      "step": 266
+    },
+    {
+      "epoch": 0.5245579567779961,
+      "grad_norm": 0.35156499543471664,
+      "learning_rate": 7.999457984807128e-05,
+      "loss": 0.5523,
+      "step": 267
+    },
+    {
+      "epoch": 0.5265225933202358,
+      "grad_norm": 0.38699945925405027,
+      "learning_rate": 7.999363887441416e-05,
+      "loss": 0.5628,
+      "step": 268
+    },
+    {
+      "epoch": 0.5284872298624754,
+      "grad_norm": 0.3948566129120866,
+      "learning_rate": 7.999262263115267e-05,
+      "loss": 0.5567,
+      "step": 269
+    },
+    {
+      "epoch": 0.5304518664047151,
+      "grad_norm": 0.31256596587711855,
+      "learning_rate": 7.999153112019939e-05,
+      "loss": 0.5555,
+      "step": 270
+    },
+    {
+      "epoch": 0.5324165029469549,
+      "grad_norm": 0.21962406185356131,
+      "learning_rate": 7.99903643436086e-05,
+      "loss": 0.5524,
+      "step": 271
+    },
+    {
+      "epoch": 0.5343811394891945,
+      "grad_norm": 0.25968543311333897,
+      "learning_rate": 7.998912230357623e-05,
+      "loss": 0.5483,
+      "step": 272
+    },
+    {
+      "epoch": 0.5363457760314342,
+      "grad_norm": 0.3046545384146074,
+      "learning_rate": 7.998780500243983e-05,
+      "loss": 0.549,
+      "step": 273
+    },
+    {
+      "epoch": 0.5383104125736738,
+      "grad_norm": 0.3503696045286193,
+      "learning_rate": 7.998641244267863e-05,
+      "loss": 0.5536,
+      "step": 274
+    },
+    {
+      "epoch": 0.5402750491159135,
+      "grad_norm": 0.3898446604690689,
+      "learning_rate": 7.998494462691349e-05,
+      "loss": 0.5569,
+      "step": 275
+    },
+    {
+      "epoch": 0.5422396856581533,
+      "grad_norm": 0.42404689988985134,
+      "learning_rate": 7.998340155790686e-05,
+      "loss": 0.5486,
+      "step": 276
+    },
+    {
+      "epoch": 0.5442043222003929,
+      "grad_norm": 0.407641218918114,
+      "learning_rate": 7.99817832385629e-05,
+      "loss": 0.5411,
+      "step": 277
+    },
+    {
+      "epoch": 0.5461689587426326,
+      "grad_norm": 0.3336024388679025,
+      "learning_rate": 7.998008967192732e-05,
+      "loss": 0.5545,
+      "step": 278
+    },
+    {
+      "epoch": 0.5481335952848723,
+      "grad_norm": 0.34055434440996035,
+      "learning_rate": 7.997832086118749e-05,
+      "loss": 0.5422,
+      "step": 279
+    },
+    {
+      "epoch": 0.550098231827112,
+      "grad_norm": 0.36880809906233075,
+      "learning_rate": 7.99764768096724e-05,
+      "loss": 0.553,
+      "step": 280
+    },
+    {
+      "epoch": 0.5520628683693517,
+      "grad_norm": 0.37508057839073095,
+      "learning_rate": 7.997455752085258e-05,
+      "loss": 0.5539,
+      "step": 281
+    },
+    {
+      "epoch": 0.5540275049115914,
+      "grad_norm": 0.42147999201374325,
+      "learning_rate": 7.997256299834025e-05,
+      "loss": 0.5598,
+      "step": 282
+    },
+    {
+      "epoch": 0.555992141453831,
+      "grad_norm": 0.48881719520672323,
+      "learning_rate": 7.997049324588917e-05,
+      "loss": 0.5515,
+      "step": 283
+    },
+    {
+      "epoch": 0.5579567779960707,
+      "grad_norm": 0.5843522138612633,
+      "learning_rate": 7.996834826739469e-05,
+      "loss": 0.5519,
+      "step": 284
+    },
+    {
+      "epoch": 0.5599214145383105,
+      "grad_norm": 0.6119803404319576,
+      "learning_rate": 7.996612806689376e-05,
+      "loss": 0.5526,
+      "step": 285
+    },
+    {
+      "epoch": 0.5618860510805501,
+      "grad_norm": 0.5929922039643432,
+      "learning_rate": 7.996383264856485e-05,
+      "loss": 0.5444,
+      "step": 286
+    },
+    {
+      "epoch": 0.5638506876227898,
+      "grad_norm": 0.4636431245916106,
+      "learning_rate": 7.996146201672806e-05,
+      "loss": 0.5541,
+      "step": 287
+    },
+    {
+      "epoch": 0.5658153241650294,
+      "grad_norm": 0.3021401367564679,
+      "learning_rate": 7.995901617584499e-05,
+      "loss": 0.5487,
+      "step": 288
+    },
+    {
+      "epoch": 0.5677799607072691,
+      "grad_norm": 0.40289663735677794,
+      "learning_rate": 7.995649513051882e-05,
+      "loss": 0.5534,
+      "step": 289
+    },
+    {
+      "epoch": 0.5697445972495089,
+      "grad_norm": 0.47575335540937996,
+      "learning_rate": 7.995389888549428e-05,
+      "loss": 0.5506,
+      "step": 290
+    },
+    {
+      "epoch": 0.5717092337917485,
+      "grad_norm": 0.34122856493929127,
+      "learning_rate": 7.995122744565757e-05,
+      "loss": 0.5488,
+      "step": 291
+    },
+    {
+      "epoch": 0.5736738703339882,
+      "grad_norm": 0.3317989031857925,
+      "learning_rate": 7.994848081603644e-05,
+      "loss": 0.5518,
+      "step": 292
+    },
+    {
+      "epoch": 0.5756385068762279,
+      "grad_norm": 0.43178971823867923,
+      "learning_rate": 7.99456590018002e-05,
+      "loss": 0.5439,
+      "step": 293
+    },
+    {
+      "epoch": 0.5776031434184676,
+      "grad_norm": 0.35900973946011006,
+      "learning_rate": 7.994276200825958e-05,
+      "loss": 0.5467,
+      "step": 294
+    },
+    {
+      "epoch": 0.5795677799607073,
+      "grad_norm": 0.26678477153968366,
+      "learning_rate": 7.993978984086683e-05,
+      "loss": 0.5385,
+      "step": 295
+    },
+    {
+      "epoch": 0.581532416502947,
+      "grad_norm": 0.32670037931691337,
+      "learning_rate": 7.993674250521572e-05,
+      "loss": 0.5508,
+      "step": 296
+    },
+    {
+      "epoch": 0.5834970530451866,
+      "grad_norm": 0.35897322333388537,
+      "learning_rate": 7.993362000704143e-05,
+      "loss": 0.5451,
+      "step": 297
+    },
+    {
+      "epoch": 0.5854616895874263,
+      "grad_norm": 0.27287664271908985,
+      "learning_rate": 7.993042235222062e-05,
+      "loss": 0.5441,
+      "step": 298
+    },
+    {
+      "epoch": 0.587426326129666,
+      "grad_norm": 0.27513487678293275,
+      "learning_rate": 7.992714954677143e-05,
+      "loss": 0.5423,
+      "step": 299
+    },
+    {
+      "epoch": 0.5893909626719057,
+      "grad_norm": 0.3165613698667714,
+      "learning_rate": 7.992380159685338e-05,
+      "loss": 0.5388,
+      "step": 300
+    },
+    {
+      "epoch": 0.5913555992141454,
+      "grad_norm": 0.33063182873383273,
+      "learning_rate": 7.992037850876745e-05,
+      "loss": 0.5425,
+      "step": 301
+    },
+    {
+      "epoch": 0.593320235756385,
+      "grad_norm": 0.32378900026428226,
+      "learning_rate": 7.991688028895605e-05,
+      "loss": 0.5448,
+      "step": 302
+    },
+    {
+      "epoch": 0.5952848722986247,
+      "grad_norm": 0.3073320949171006,
+      "learning_rate": 7.991330694400293e-05,
+      "loss": 0.5578,
+      "step": 303
+    },
+    {
+      "epoch": 0.5972495088408645,
+      "grad_norm": 0.3669524370860771,
+      "learning_rate": 7.99096584806333e-05,
+      "loss": 0.5447,
+      "step": 304
+    },
+    {
+      "epoch": 0.5992141453831041,
+      "grad_norm": 0.28861052791844194,
+      "learning_rate": 7.990593490571369e-05,
+      "loss": 0.551,
+      "step": 305
+    },
+    {
+      "epoch": 0.6011787819253438,
+      "grad_norm": 0.26524157251457975,
+      "learning_rate": 7.990213622625202e-05,
+      "loss": 0.5422,
+      "step": 306
+    },
+    {
+      "epoch": 0.6031434184675835,
+      "grad_norm": 0.4109382006860738,
+      "learning_rate": 7.989826244939757e-05,
+      "loss": 0.542,
+      "step": 307
+    },
+    {
+      "epoch": 0.6051080550098232,
+      "grad_norm": 0.3827568551058756,
+      "learning_rate": 7.989431358244093e-05,
+      "loss": 0.5451,
+      "step": 308
+    },
+    {
+      "epoch": 0.6070726915520629,
+      "grad_norm": 0.3296018316450615,
+      "learning_rate": 7.989028963281401e-05,
+      "loss": 0.5465,
+      "step": 309
+    },
+    {
+      "epoch": 0.6090373280943026,
+      "grad_norm": 0.4631502800785598,
+      "learning_rate": 7.988619060809005e-05,
+      "loss": 0.5414,
+      "step": 310
+    },
+    {
+      "epoch": 0.6110019646365422,
+      "grad_norm": 0.5442524444958319,
+      "learning_rate": 7.988201651598359e-05,
+      "loss": 0.5576,
+      "step": 311
+    },
+    {
+      "epoch": 0.6129666011787819,
+      "grad_norm": 0.5472854342648696,
+      "learning_rate": 7.987776736435042e-05,
+      "loss": 0.5447,
+      "step": 312
+    },
+    {
+      "epoch": 0.6149312377210217,
+      "grad_norm": 0.5408196886880645,
+      "learning_rate": 7.987344316118761e-05,
+      "loss": 0.5386,
+      "step": 313
+    },
+    {
+      "epoch": 0.6168958742632613,
+      "grad_norm": 0.5185764572016152,
+      "learning_rate": 7.986904391463349e-05,
+      "loss": 0.542,
+      "step": 314
+    },
+    {
+      "epoch": 0.618860510805501,
+      "grad_norm": 0.5195497241692139,
+      "learning_rate": 7.98645696329676e-05,
+      "loss": 0.5537,
+      "step": 315
+    },
+    {
+      "epoch": 0.6208251473477406,
+      "grad_norm": 0.48966336143949907,
+      "learning_rate": 7.986002032461074e-05,
+      "loss": 0.5395,
+      "step": 316
+    },
+    {
+      "epoch": 0.6227897838899804,
+      "grad_norm": 0.43646881924921466,
+      "learning_rate": 7.985539599812488e-05,
+      "loss": 0.5461,
+      "step": 317
+    },
+    {
+      "epoch": 0.6247544204322201,
+      "grad_norm": 0.3315139983697266,
+      "learning_rate": 7.985069666221316e-05,
+      "loss": 0.5397,
+      "step": 318
+    },
+    {
+      "epoch": 0.6267190569744597,
+      "grad_norm": 0.3026579355529067,
+      "learning_rate": 7.984592232571994e-05,
+      "loss": 0.5427,
+      "step": 319
+    },
+    {
+      "epoch": 0.6286836935166994,
+      "grad_norm": 0.2796280560506545,
+      "learning_rate": 7.98410729976307e-05,
+      "loss": 0.5449,
+      "step": 320
+    },
+    {
+      "epoch": 0.630648330058939,
+      "grad_norm": 0.3174701570455909,
+      "learning_rate": 7.983614868707208e-05,
+      "loss": 0.5462,
+      "step": 321
+    },
+    {
+      "epoch": 0.6326129666011788,
+      "grad_norm": 0.3916479195826402,
+      "learning_rate": 7.983114940331181e-05,
+      "loss": 0.5525,
+      "step": 322
+    },
+    {
+      "epoch": 0.6345776031434185,
+      "grad_norm": 0.41193413052313926,
+      "learning_rate": 7.982607515575874e-05,
+      "loss": 0.5515,
+      "step": 323
+    },
+    {
+      "epoch": 0.6365422396856582,
+      "grad_norm": 0.3256002295454151,
+      "learning_rate": 7.982092595396282e-05,
+      "loss": 0.545,
+      "step": 324
+    },
+    {
+      "epoch": 0.6385068762278978,
+      "grad_norm": 0.2729061678156102,
+      "learning_rate": 7.981570180761503e-05,
+      "loss": 0.5465,
+      "step": 325
+    },
+    {
+      "epoch": 0.6404715127701375,
+      "grad_norm": 0.24563854138299804,
+      "learning_rate": 7.981040272654743e-05,
+      "loss": 0.5452,
+      "step": 326
+    },
+    {
+      "epoch": 0.6424361493123772,
+      "grad_norm": 0.24181716985544116,
+      "learning_rate": 7.98050287207331e-05,
+      "loss": 0.553,
+      "step": 327
+    },
+    {
+      "epoch": 0.6444007858546169,
+      "grad_norm": 0.2843727487384882,
+      "learning_rate": 7.979957980028612e-05,
+      "loss": 0.5479,
+      "step": 328
+    },
+    {
+      "epoch": 0.6463654223968566,
+      "grad_norm": 0.3350302061172736,
+      "learning_rate": 7.97940559754616e-05,
+      "loss": 0.5435,
+      "step": 329
+    },
+    {
+      "epoch": 0.6483300589390962,
+      "grad_norm": 0.3382536829532006,
+      "learning_rate": 7.978845725665557e-05,
+      "loss": 0.5408,
+      "step": 330
+    },
+    {
+      "epoch": 0.650294695481336,
+      "grad_norm": 0.29230242279409513,
+      "learning_rate": 7.978278365440504e-05,
+      "loss": 0.5474,
+      "step": 331
+    },
+    {
+      "epoch": 0.6522593320235757,
+      "grad_norm": 0.3378408792023042,
+      "learning_rate": 7.977703517938797e-05,
+      "loss": 0.5424,
+      "step": 332
+    },
+    {
+      "epoch": 0.6542239685658153,
+      "grad_norm": 0.3194231418253994,
+      "learning_rate": 7.977121184242321e-05,
+      "loss": 0.5455,
+      "step": 333
+    },
+    {
+      "epoch": 0.656188605108055,
+      "grad_norm": 0.28606293016318435,
+      "learning_rate": 7.97653136544705e-05,
+      "loss": 0.55,
+      "step": 334
+    },
+    {
+      "epoch": 0.6581532416502947,
+      "grad_norm": 0.3037536657401747,
+      "learning_rate": 7.975934062663046e-05,
+      "loss": 0.5509,
+      "step": 335
+    },
+    {
+      "epoch": 0.6601178781925344,
+      "grad_norm": 0.31303338662808516,
+      "learning_rate": 7.975329277014459e-05,
+      "loss": 0.5487,
+      "step": 336
+    },
+    {
+      "epoch": 0.6620825147347741,
+      "grad_norm": 0.3050481926074019,
+      "learning_rate": 7.974717009639516e-05,
+      "loss": 0.5451,
+      "step": 337
+    },
+    {
+      "epoch": 0.6640471512770137,
+      "grad_norm": 0.29835600102755555,
+      "learning_rate": 7.974097261690531e-05,
+      "loss": 0.5494,
+      "step": 338
+    },
+    {
+      "epoch": 0.6660117878192534,
+      "grad_norm": 0.3265789184762612,
+      "learning_rate": 7.973470034333894e-05,
+      "loss": 0.5488,
+      "step": 339
+    },
+    {
+      "epoch": 0.6679764243614931,
+      "grad_norm": 0.33635070022463237,
+      "learning_rate": 7.97283532875007e-05,
+      "loss": 0.54,
+      "step": 340
+    },
+    {
+      "epoch": 0.6699410609037328,
+      "grad_norm": 0.2616837787015939,
+      "learning_rate": 7.972193146133599e-05,
+      "loss": 0.5401,
+      "step": 341
+    },
+    {
+      "epoch": 0.6719056974459725,
+      "grad_norm": 0.20363469029162648,
+      "learning_rate": 7.971543487693097e-05,
+      "loss": 0.5455,
+      "step": 342
+    },
+    {
+      "epoch": 0.6738703339882122,
+      "grad_norm": 0.18040490616813468,
+      "learning_rate": 7.970886354651244e-05,
+      "loss": 0.5338,
+      "step": 343
+    },
+    {
+      "epoch": 0.6758349705304518,
+      "grad_norm": 0.1957362846066182,
+      "learning_rate": 7.970221748244793e-05,
+      "loss": 0.5443,
+      "step": 344
+    },
+    {
+      "epoch": 0.6777996070726916,
+      "grad_norm": 0.243584965365042,
+      "learning_rate": 7.969549669724558e-05,
+      "loss": 0.5473,
+      "step": 345
+    },
+    {
+      "epoch": 0.6797642436149313,
+      "grad_norm": 0.30523707171411807,
+      "learning_rate": 7.968870120355418e-05,
+      "loss": 0.5456,
+      "step": 346
+    },
+    {
+      "epoch": 0.6817288801571709,
+      "grad_norm": 0.3723251081181164,
+      "learning_rate": 7.96818310141631e-05,
+      "loss": 0.5435,
+      "step": 347
+    },
+    {
+      "epoch": 0.6836935166994106,
+      "grad_norm": 0.420855920298727,
+      "learning_rate": 7.967488614200232e-05,
+      "loss": 0.5379,
+      "step": 348
+    },
+    {
+      "epoch": 0.6856581532416502,
+      "grad_norm": 0.44945534304593987,
+      "learning_rate": 7.966786660014236e-05,
+      "loss": 0.5465,
+      "step": 349
+    },
+    {
+      "epoch": 0.68762278978389,
+      "grad_norm": 0.4738869865388266,
+      "learning_rate": 7.966077240179428e-05,
+      "loss": 0.5446,
+      "step": 350
+    },
+    {
+      "epoch": 0.6895874263261297,
+      "grad_norm": 0.49399832740102345,
+      "learning_rate": 7.965360356030962e-05,
+      "loss": 0.5474,
+      "step": 351
+    },
+    {
+      "epoch": 0.6915520628683693,
+      "grad_norm": 0.5245149011524607,
+      "learning_rate": 7.964636008918045e-05,
+      "loss": 0.5541,
+      "step": 352
+    },
+    {
+      "epoch": 0.693516699410609,
+      "grad_norm": 0.46917369921619284,
+      "learning_rate": 7.963904200203926e-05,
+      "loss": 0.5381,
+      "step": 353
+    },
+    {
+      "epoch": 0.6954813359528488,
+      "grad_norm": 0.4034469678893911,
+      "learning_rate": 7.963164931265898e-05,
+      "loss": 0.5374,
+      "step": 354
+    },
+    {
+      "epoch": 0.6974459724950884,
+      "grad_norm": 0.3450692418877584,
+      "learning_rate": 7.962418203495292e-05,
+      "loss": 0.5348,
+      "step": 355
+    },
+    {
+      "epoch": 0.6994106090373281,
+      "grad_norm": 0.3017445456854127,
+      "learning_rate": 7.96166401829748e-05,
+      "loss": 0.5444,
+      "step": 356
+    },
+    {
+      "epoch": 0.7013752455795678,
+      "grad_norm": 0.3489487509556447,
+      "learning_rate": 7.960902377091872e-05,
+      "loss": 0.5349,
+      "step": 357
+    },
+    {
+      "epoch": 0.7033398821218074,
+      "grad_norm": 0.4112001580899121,
+      "learning_rate": 7.960133281311902e-05,
+      "loss": 0.5397,
+      "step": 358
+    },
+    {
+      "epoch": 0.7053045186640472,
+      "grad_norm": 0.38633785944944243,
+      "learning_rate": 7.95935673240504e-05,
+      "loss": 0.5464,
+      "step": 359
+    },
+    {
+      "epoch": 0.7072691552062869,
+      "grad_norm": 0.40833523174206937,
+      "learning_rate": 7.958572731832784e-05,
+      "loss": 0.5347,
+      "step": 360
+    },
+    {
+      "epoch": 0.7092337917485265,
+      "grad_norm": 0.4889732397207154,
+      "learning_rate": 7.95778128107065e-05,
+      "loss": 0.5372,
+      "step": 361
+    },
+    {
+      "epoch": 0.7111984282907662,
+      "grad_norm": 0.5069817461691449,
+      "learning_rate": 7.956982381608184e-05,
+      "loss": 0.5512,
+      "step": 362
+    },
+    {
+      "epoch": 0.7131630648330058,
+      "grad_norm": 0.48794762899255706,
+      "learning_rate": 7.956176034948942e-05,
+      "loss": 0.5464,
+      "step": 363
+    },
+    {
+      "epoch": 0.7151277013752456,
+      "grad_norm": 0.42009880200008526,
+      "learning_rate": 7.955362242610502e-05,
+      "loss": 0.5494,
+      "step": 364
+    },
+    {
+      "epoch": 0.7170923379174853,
+      "grad_norm": 0.3178283395727616,
+      "learning_rate": 7.954541006124453e-05,
+      "loss": 0.5353,
+      "step": 365
+    },
+    {
+      "epoch": 0.7190569744597249,
+      "grad_norm": 0.3361843800411505,
+      "learning_rate": 7.953712327036396e-05,
+      "loss": 0.5355,
+      "step": 366
+    },
+    {
+      "epoch": 0.7210216110019646,
+      "grad_norm": 0.3833110968417026,
+      "learning_rate": 7.952876206905936e-05,
+      "loss": 0.5423,
+      "step": 367
+    },
+    {
+      "epoch": 0.7229862475442044,
+      "grad_norm": 0.33336613461039183,
+      "learning_rate": 7.952032647306684e-05,
+      "loss": 0.5361,
+      "step": 368
+    },
+    {
+      "epoch": 0.724950884086444,
+      "grad_norm": 0.2479685810755696,
+      "learning_rate": 7.951181649826252e-05,
+      "loss": 0.5385,
+      "step": 369
+    },
+    {
+      "epoch": 0.7269155206286837,
+      "grad_norm": 0.3381048646387299,
+      "learning_rate": 7.950323216066252e-05,
+      "loss": 0.5463,
+      "step": 370
+    },
+    {
+      "epoch": 0.7288801571709234,
+      "grad_norm": 0.34579602627181333,
+      "learning_rate": 7.949457347642292e-05,
+      "loss": 0.5346,
+      "step": 371
+    },
+    {
+      "epoch": 0.730844793713163,
+      "grad_norm": 0.2969811146410946,
+      "learning_rate": 7.948584046183967e-05,
+      "loss": 0.5422,
+      "step": 372
+    },
+    {
+      "epoch": 0.7328094302554028,
+      "grad_norm": 0.324163596152794,
+      "learning_rate": 7.947703313334866e-05,
+      "loss": 0.5352,
+      "step": 373
+    },
+    {
+      "epoch": 0.7347740667976425,
+      "grad_norm": 0.28582696439142324,
+      "learning_rate": 7.94681515075256e-05,
+      "loss": 0.5477,
+      "step": 374
+    },
+    {
+      "epoch": 0.7367387033398821,
+      "grad_norm": 0.30566283165408414,
+      "learning_rate": 7.945919560108612e-05,
+      "loss": 0.5367,
+      "step": 375
+    },
+    {
+      "epoch": 0.7387033398821218,
+      "grad_norm": 0.3636336981918442,
+      "learning_rate": 7.945016543088556e-05,
+      "loss": 0.5382,
+      "step": 376
+    },
+    {
+      "epoch": 0.7406679764243614,
+      "grad_norm": 0.35048808319162317,
+      "learning_rate": 7.944106101391906e-05,
+      "loss": 0.5382,
+      "step": 377
+    },
+    {
+      "epoch": 0.7426326129666012,
+      "grad_norm": 0.2633494377661315,
+      "learning_rate": 7.943188236732147e-05,
+      "loss": 0.5344,
+      "step": 378
+    },
+    {
+      "epoch": 0.7445972495088409,
+      "grad_norm": 0.2476617392102118,
+      "learning_rate": 7.942262950836738e-05,
+      "loss": 0.5434,
+      "step": 379
+    },
+    {
+      "epoch": 0.7465618860510805,
+      "grad_norm": 0.28098912120295927,
+      "learning_rate": 7.941330245447103e-05,
+      "loss": 0.5372,
+      "step": 380
+    },
+    {
+      "epoch": 0.7485265225933202,
+      "grad_norm": 0.2803139840331155,
+      "learning_rate": 7.940390122318633e-05,
+      "loss": 0.5351,
+      "step": 381
+    },
+    {
+      "epoch": 0.75049115913556,
+      "grad_norm": 0.2381284603421722,
+      "learning_rate": 7.939442583220674e-05,
+      "loss": 0.5403,
+      "step": 382
+    },
+    {
+      "epoch": 0.7524557956777996,
+      "grad_norm": 0.25735779447447044,
+      "learning_rate": 7.938487629936532e-05,
+      "loss": 0.5461,
+      "step": 383
+    },
+    {
+      "epoch": 0.7544204322200393,
+      "grad_norm": 0.2477724085525229,
+      "learning_rate": 7.937525264263468e-05,
+      "loss": 0.5478,
+      "step": 384
+    },
+    {
+      "epoch": 0.756385068762279,
+      "grad_norm": 0.24095606736445105,
+      "learning_rate": 7.936555488012691e-05,
+      "loss": 0.5428,
+      "step": 385
+    },
+    {
+      "epoch": 0.7583497053045186,
+      "grad_norm": 0.2646760905425485,
+      "learning_rate": 7.935578303009358e-05,
+      "loss": 0.5307,
+      "step": 386
+    },
+    {
+      "epoch": 0.7603143418467584,
+      "grad_norm": 0.3144208006778055,
+      "learning_rate": 7.93459371109257e-05,
+      "loss": 0.5487,
+      "step": 387
+    },
+    {
+      "epoch": 0.762278978388998,
+      "grad_norm": 0.32881002504409884,
+      "learning_rate": 7.933601714115369e-05,
+      "loss": 0.5344,
+      "step": 388
+    },
+    {
+      "epoch": 0.7642436149312377,
+      "grad_norm": 0.3550915610091036,
+      "learning_rate": 7.932602313944731e-05,
+      "loss": 0.543,
+      "step": 389
+    },
+    {
+      "epoch": 0.7662082514734774,
+      "grad_norm": 0.39028849733166887,
+      "learning_rate": 7.931595512461566e-05,
+      "loss": 0.538,
+      "step": 390
+    },
+    {
+      "epoch": 0.768172888015717,
+      "grad_norm": 0.3791533893673213,
+      "learning_rate": 7.930581311560717e-05,
+      "loss": 0.5527,
+      "step": 391
+    },
+    {
+      "epoch": 0.7701375245579568,
+      "grad_norm": 0.38707683109307855,
+      "learning_rate": 7.929559713150946e-05,
+      "loss": 0.5329,
+      "step": 392
+    },
+    {
+      "epoch": 0.7721021611001965,
+      "grad_norm": 0.40376208430028165,
+      "learning_rate": 7.928530719154944e-05,
+      "loss": 0.5453,
+      "step": 393
+    },
+    {
+      "epoch": 0.7740667976424361,
+      "grad_norm": 0.3652906836386957,
+      "learning_rate": 7.92749433150932e-05,
+      "loss": 0.5444,
+      "step": 394
+    },
+    {
+      "epoch": 0.7760314341846758,
+      "grad_norm": 0.34378544340380746,
+      "learning_rate": 7.926450552164591e-05,
+      "loss": 0.5216,
+      "step": 395
+    },
+    {
+      "epoch": 0.7779960707269156,
+      "grad_norm": 0.36708771367149,
+      "learning_rate": 7.925399383085195e-05,
+      "loss": 0.5324,
+      "step": 396
+    },
+    {
+      "epoch": 0.7799607072691552,
+      "grad_norm": 0.26918941956701975,
+      "learning_rate": 7.924340826249473e-05,
+      "loss": 0.5319,
+      "step": 397
+    },
+    {
+      "epoch": 0.7819253438113949,
+      "grad_norm": 0.36855500449173456,
+      "learning_rate": 7.923274883649671e-05,
+      "loss": 0.545,
+      "step": 398
+    },
+    {
+      "epoch": 0.7838899803536346,
+      "grad_norm": 0.4629018012986128,
+      "learning_rate": 7.922201557291933e-05,
+      "loss": 0.5341,
+      "step": 399
+    },
+    {
+      "epoch": 0.7858546168958742,
+      "grad_norm": 0.37513808261414827,
+      "learning_rate": 7.921120849196303e-05,
+      "loss": 0.5324,
+      "step": 400
+    },
+    {
+      "epoch": 0.787819253438114,
+      "grad_norm": 0.28985091528274187,
+      "learning_rate": 7.920032761396716e-05,
+      "loss": 0.5286,
+      "step": 401
+    },
+    {
+      "epoch": 0.7897838899803536,
+      "grad_norm": 0.29682885641576273,
+      "learning_rate": 7.918937295940997e-05,
+      "loss": 0.5325,
+      "step": 402
+    },
+    {
+      "epoch": 0.7917485265225933,
+      "grad_norm": 0.3412705259598063,
+      "learning_rate": 7.917834454890854e-05,
+      "loss": 0.5347,
+      "step": 403
+    },
+    {
+      "epoch": 0.793713163064833,
+      "grad_norm": 0.38660086602459337,
+      "learning_rate": 7.916724240321878e-05,
+      "loss": 0.544,
+      "step": 404
+    },
+    {
+      "epoch": 0.7956777996070727,
+      "grad_norm": 0.37175900069405626,
+      "learning_rate": 7.915606654323534e-05,
+      "loss": 0.5305,
+      "step": 405
+    },
+    {
+      "epoch": 0.7976424361493124,
+      "grad_norm": 0.354777701021265,
+      "learning_rate": 7.914481698999168e-05,
+      "loss": 0.5237,
+      "step": 406
+    },
+    {
+      "epoch": 0.7996070726915521,
+      "grad_norm": 0.34871065414617325,
+      "learning_rate": 7.913349376465984e-05,
+      "loss": 0.535,
+      "step": 407
+    },
+    {
+      "epoch": 0.8015717092337917,
+      "grad_norm": 0.36144424921795226,
+      "learning_rate": 7.912209688855063e-05,
+      "loss": 0.5452,
+      "step": 408
+    },
+    {
+      "epoch": 0.8035363457760314,
+      "grad_norm": 0.28866233925937934,
+      "learning_rate": 7.91106263831134e-05,
+      "loss": 0.5231,
+      "step": 409
+    },
+    {
+      "epoch": 0.8055009823182712,
+      "grad_norm": 0.2388237233330627,
+      "learning_rate": 7.909908226993609e-05,
+      "loss": 0.5491,
+      "step": 410
+    },
+    {
+      "epoch": 0.8074656188605108,
+      "grad_norm": 0.25411095698852826,
+      "learning_rate": 7.908746457074518e-05,
+      "loss": 0.5304,
+      "step": 411
+    },
+    {
+      "epoch": 0.8094302554027505,
+      "grad_norm": 0.22664123676817086,
+      "learning_rate": 7.907577330740563e-05,
+      "loss": 0.5417,
+      "step": 412
+    },
+    {
+      "epoch": 0.8113948919449901,
+      "grad_norm": 0.24169484656636106,
+      "learning_rate": 7.906400850192087e-05,
+      "loss": 0.5418,
+      "step": 413
+    },
+    {
+      "epoch": 0.8133595284872298,
+      "grad_norm": 0.2597383324417893,
+      "learning_rate": 7.905217017643272e-05,
+      "loss": 0.5392,
+      "step": 414
+    },
+    {
+      "epoch": 0.8153241650294696,
+      "grad_norm": 0.2519442870976418,
+      "learning_rate": 7.904025835322139e-05,
+      "loss": 0.5336,
+      "step": 415
+    },
+    {
+      "epoch": 0.8172888015717092,
+      "grad_norm": 0.21198706694507977,
+      "learning_rate": 7.902827305470535e-05,
+      "loss": 0.5287,
+      "step": 416
+    },
+    {
+      "epoch": 0.8192534381139489,
+      "grad_norm": 0.19980499179205108,
+      "learning_rate": 7.901621430344148e-05,
+      "loss": 0.5244,
+      "step": 417
+    },
+    {
+      "epoch": 0.8212180746561886,
+      "grad_norm": 0.2564327537455601,
+      "learning_rate": 7.900408212212478e-05,
+      "loss": 0.5364,
+      "step": 418
+    },
+    {
+      "epoch": 0.8231827111984283,
+      "grad_norm": 0.2585215005140894,
+      "learning_rate": 7.899187653358849e-05,
+      "loss": 0.5407,
+      "step": 419
+    },
+    {
+      "epoch": 0.825147347740668,
+      "grad_norm": 0.23789669125085935,
+      "learning_rate": 7.897959756080402e-05,
+      "loss": 0.5409,
+      "step": 420
+    },
+    {
+      "epoch": 0.8271119842829077,
+      "grad_norm": 0.2386595506297681,
+      "learning_rate": 7.896724522688088e-05,
+      "loss": 0.5351,
+      "step": 421
+    },
+    {
+      "epoch": 0.8290766208251473,
+      "grad_norm": 0.2502975887898376,
+      "learning_rate": 7.895481955506665e-05,
+      "loss": 0.5262,
+      "step": 422
+    },
+    {
+      "epoch": 0.831041257367387,
+      "grad_norm": 0.2891950526516659,
+      "learning_rate": 7.894232056874694e-05,
+      "loss": 0.5329,
+      "step": 423
+    },
+    {
+      "epoch": 0.8330058939096268,
+      "grad_norm": 0.2826894968634538,
+      "learning_rate": 7.892974829144532e-05,
+      "loss": 0.5386,
+      "step": 424
+    },
+    {
+      "epoch": 0.8349705304518664,
+      "grad_norm": 0.3130100776882286,
+      "learning_rate": 7.891710274682333e-05,
+      "loss": 0.5506,
+      "step": 425
+    },
+    {
+      "epoch": 0.8369351669941061,
+      "grad_norm": 0.3874012841065107,
+      "learning_rate": 7.890438395868035e-05,
+      "loss": 0.5341,
+      "step": 426
+    },
+    {
+      "epoch": 0.8388998035363457,
+      "grad_norm": 0.43244769046170695,
+      "learning_rate": 7.889159195095367e-05,
+      "loss": 0.5402,
+      "step": 427
+    },
+    {
+      "epoch": 0.8408644400785854,
+      "grad_norm": 0.4126277544859472,
+      "learning_rate": 7.887872674771836e-05,
+      "loss": 0.5382,
+      "step": 428
+    },
+    {
+      "epoch": 0.8428290766208252,
+      "grad_norm": 0.3784128024246673,
+      "learning_rate": 7.886578837318722e-05,
+      "loss": 0.5348,
+      "step": 429
+    },
+    {
+      "epoch": 0.8447937131630648,
+      "grad_norm": 0.33578073995267765,
+      "learning_rate": 7.885277685171078e-05,
+      "loss": 0.5339,
+      "step": 430
+    },
+    {
+      "epoch": 0.8467583497053045,
+      "grad_norm": 0.2856979263294091,
+      "learning_rate": 7.883969220777724e-05,
+      "loss": 0.5343,
+      "step": 431
+    },
+    {
+      "epoch": 0.8487229862475442,
+      "grad_norm": 0.35169222355695645,
+      "learning_rate": 7.882653446601242e-05,
+      "loss": 0.5416,
+      "step": 432
+    },
+    {
+      "epoch": 0.8506876227897839,
+      "grad_norm": 0.4396162498426928,
+      "learning_rate": 7.88133036511797e-05,
+      "loss": 0.5376,
+      "step": 433
+    },
+    {
+      "epoch": 0.8526522593320236,
+      "grad_norm": 0.4621230231088642,
+      "learning_rate": 7.879999978818001e-05,
+      "loss": 0.5382,
+      "step": 434
+    },
+    {
+      "epoch": 0.8546168958742633,
+      "grad_norm": 0.38791159875746223,
+      "learning_rate": 7.878662290205173e-05,
+      "loss": 0.5417,
+      "step": 435
+    },
+    {
+      "epoch": 0.8565815324165029,
+      "grad_norm": 0.27326186571237004,
+      "learning_rate": 7.877317301797069e-05,
+      "loss": 0.5438,
+      "step": 436
+    },
+    {
+      "epoch": 0.8585461689587426,
+      "grad_norm": 0.31786643061392283,
+      "learning_rate": 7.875965016125011e-05,
+      "loss": 0.5369,
+      "step": 437
+    },
+    {
+      "epoch": 0.8605108055009824,
+      "grad_norm": 0.3023833736668177,
+      "learning_rate": 7.874605435734052e-05,
+      "loss": 0.5355,
+      "step": 438
+    },
+    {
+      "epoch": 0.862475442043222,
+      "grad_norm": 0.2677546893379095,
+      "learning_rate": 7.873238563182979e-05,
+      "loss": 0.5316,
+      "step": 439
+    },
+    {
+      "epoch": 0.8644400785854617,
+      "grad_norm": 0.28031574717473506,
+      "learning_rate": 7.871864401044297e-05,
+      "loss": 0.5378,
+      "step": 440
+    },
+    {
+      "epoch": 0.8664047151277013,
+      "grad_norm": 0.32613207216405454,
+      "learning_rate": 7.870482951904235e-05,
+      "loss": 0.5351,
+      "step": 441
+    },
+    {
+      "epoch": 0.8683693516699411,
+      "grad_norm": 0.35691568664920914,
+      "learning_rate": 7.869094218362735e-05,
+      "loss": 0.5288,
+      "step": 442
+    },
+    {
+      "epoch": 0.8703339882121808,
+      "grad_norm": 0.28792306023819103,
+      "learning_rate": 7.867698203033446e-05,
+      "loss": 0.5324,
+      "step": 443
+    },
+    {
+      "epoch": 0.8722986247544204,
+      "grad_norm": 0.22752753256898703,
+      "learning_rate": 7.866294908543726e-05,
+      "loss": 0.5396,
+      "step": 444
+    },
+    {
+      "epoch": 0.8742632612966601,
+      "grad_norm": 0.21960804138331325,
+      "learning_rate": 7.86488433753463e-05,
+      "loss": 0.5267,
+      "step": 445
+    },
+    {
+      "epoch": 0.8762278978388998,
+      "grad_norm": 0.3249248684231161,
+      "learning_rate": 7.863466492660907e-05,
+      "loss": 0.55,
+      "step": 446
+    },
+    {
+      "epoch": 0.8781925343811395,
+      "grad_norm": 0.331088213109672,
+      "learning_rate": 7.862041376590999e-05,
+      "loss": 0.5262,
+      "step": 447
+    },
+    {
+      "epoch": 0.8801571709233792,
+      "grad_norm": 0.322215193797039,
+      "learning_rate": 7.860608992007028e-05,
+      "loss": 0.5246,
+      "step": 448
+    },
+    {
+      "epoch": 0.8821218074656189,
+      "grad_norm": 0.357008827193408,
+      "learning_rate": 7.859169341604802e-05,
+      "loss": 0.5269,
+      "step": 449
+    },
+    {
+      "epoch": 0.8840864440078585,
+      "grad_norm": 0.29298366877337845,
+      "learning_rate": 7.857722428093797e-05,
+      "loss": 0.5401,
+      "step": 450
+    },
+    {
+      "epoch": 0.8860510805500982,
+      "grad_norm": 0.21767477336647134,
+      "learning_rate": 7.85626825419716e-05,
+      "loss": 0.5415,
+      "step": 451
+    },
+    {
+      "epoch": 0.888015717092338,
+      "grad_norm": 0.2562211403629511,
+      "learning_rate": 7.854806822651707e-05,
+      "loss": 0.5367,
+      "step": 452
+    },
+    {
+      "epoch": 0.8899803536345776,
+      "grad_norm": 0.2746457985762298,
+      "learning_rate": 7.853338136207908e-05,
+      "loss": 0.5293,
+      "step": 453
+    },
+    {
+      "epoch": 0.8919449901768173,
+      "grad_norm": 0.30040407972624467,
+      "learning_rate": 7.851862197629888e-05,
+      "loss": 0.5373,
+      "step": 454
+    },
+    {
+      "epoch": 0.8939096267190569,
+      "grad_norm": 0.2987097799633544,
+      "learning_rate": 7.850379009695426e-05,
+      "loss": 0.5346,
+      "step": 455
+    },
+    {
+      "epoch": 0.8958742632612967,
+      "grad_norm": 0.27880826554266713,
+      "learning_rate": 7.848888575195933e-05,
+      "loss": 0.5474,
+      "step": 456
+    },
+    {
+      "epoch": 0.8978388998035364,
+      "grad_norm": 0.28292813751057355,
+      "learning_rate": 7.84739089693647e-05,
+      "loss": 0.5318,
+      "step": 457
+    },
+    {
+      "epoch": 0.899803536345776,
+      "grad_norm": 0.2904010290667057,
+      "learning_rate": 7.845885977735726e-05,
+      "loss": 0.5271,
+      "step": 458
+    },
+    {
+      "epoch": 0.9017681728880157,
+      "grad_norm": 0.26132113504340065,
+      "learning_rate": 7.844373820426018e-05,
+      "loss": 0.5342,
+      "step": 459
+    },
+    {
+      "epoch": 0.9037328094302554,
+      "grad_norm": 0.21475199033307743,
+      "learning_rate": 7.842854427853285e-05,
+      "loss": 0.5357,
+      "step": 460
+    },
+    {
+      "epoch": 0.9056974459724951,
+      "grad_norm": 0.24591731160110863,
+      "learning_rate": 7.841327802877085e-05,
+      "loss": 0.5308,
+      "step": 461
+    },
+    {
+      "epoch": 0.9076620825147348,
+      "grad_norm": 0.214971886770936,
+      "learning_rate": 7.839793948370585e-05,
+      "loss": 0.5339,
+      "step": 462
+    },
+    {
+      "epoch": 0.9096267190569745,
+      "grad_norm": 0.22708325947140506,
+      "learning_rate": 7.838252867220559e-05,
+      "loss": 0.5363,
+      "step": 463
+    },
+    {
+      "epoch": 0.9115913555992141,
+      "grad_norm": 0.2811869565965541,
+      "learning_rate": 7.836704562327387e-05,
+      "loss": 0.5393,
+      "step": 464
+    },
+    {
+      "epoch": 0.9135559921414538,
+      "grad_norm": 0.305839427960017,
+      "learning_rate": 7.835149036605033e-05,
+      "loss": 0.5429,
+      "step": 465
+    },
+    {
+      "epoch": 0.9155206286836935,
+      "grad_norm": 0.26074368992439134,
+      "learning_rate": 7.833586292981063e-05,
+      "loss": 0.5344,
+      "step": 466
+    },
+    {
+      "epoch": 0.9174852652259332,
+      "grad_norm": 0.21548029936799717,
+      "learning_rate": 7.832016334396618e-05,
+      "loss": 0.5338,
+      "step": 467
+    },
+    {
+      "epoch": 0.9194499017681729,
+      "grad_norm": 0.28310760906614774,
+      "learning_rate": 7.830439163806427e-05,
+      "loss": 0.5349,
+      "step": 468
+    },
+    {
+      "epoch": 0.9214145383104125,
+      "grad_norm": 0.3601995696438711,
+      "learning_rate": 7.828854784178784e-05,
+      "loss": 0.5315,
+      "step": 469
+    },
+    {
+      "epoch": 0.9233791748526523,
+      "grad_norm": 0.3455779640061523,
+      "learning_rate": 7.827263198495554e-05,
+      "loss": 0.5344,
+      "step": 470
+    },
+    {
+      "epoch": 0.925343811394892,
+      "grad_norm": 0.28891107362130247,
+      "learning_rate": 7.825664409752163e-05,
+      "loss": 0.5378,
+      "step": 471
+    },
+    {
+      "epoch": 0.9273084479371316,
+      "grad_norm": 0.2915661621427824,
+      "learning_rate": 7.824058420957598e-05,
+      "loss": 0.5354,
+      "step": 472
+    },
+    {
+      "epoch": 0.9292730844793713,
+      "grad_norm": 0.298727733596555,
+      "learning_rate": 7.822445235134388e-05,
+      "loss": 0.5406,
+      "step": 473
+    },
+    {
+      "epoch": 0.931237721021611,
+      "grad_norm": 0.3003693963343444,
+      "learning_rate": 7.820824855318616e-05,
+      "loss": 0.5313,
+      "step": 474
+    },
+    {
+      "epoch": 0.9332023575638507,
+      "grad_norm": 0.3332900039370254,
+      "learning_rate": 7.819197284559901e-05,
+      "loss": 0.5371,
+      "step": 475
+    },
+    {
+      "epoch": 0.9351669941060904,
+      "grad_norm": 0.3619706966296112,
+      "learning_rate": 7.817562525921394e-05,
+      "loss": 0.5369,
+      "step": 476
+    },
+    {
+      "epoch": 0.93713163064833,
+      "grad_norm": 0.3801132677009124,
+      "learning_rate": 7.815920582479775e-05,
+      "loss": 0.5287,
+      "step": 477
+    },
+    {
+      "epoch": 0.9390962671905697,
+      "grad_norm": 0.3850854396295641,
+      "learning_rate": 7.814271457325247e-05,
+      "loss": 0.5446,
+      "step": 478
+    },
+    {
+      "epoch": 0.9410609037328095,
+      "grad_norm": 0.31431774256477024,
+      "learning_rate": 7.812615153561526e-05,
+      "loss": 0.5258,
+      "step": 479
+    },
+    {
+      "epoch": 0.9430255402750491,
+      "grad_norm": 0.32071034122623515,
+      "learning_rate": 7.810951674305845e-05,
+      "loss": 0.5305,
+      "step": 480
+    },
+    {
+      "epoch": 0.9449901768172888,
+      "grad_norm": 0.3370970289088498,
+      "learning_rate": 7.809281022688935e-05,
+      "loss": 0.5256,
+      "step": 481
+    },
+    {
+      "epoch": 0.9469548133595285,
+      "grad_norm": 0.25175603469391544,
+      "learning_rate": 7.80760320185503e-05,
+      "loss": 0.5484,
+      "step": 482
+    },
+    {
+      "epoch": 0.9489194499017681,
+      "grad_norm": 0.1756626334431688,
+      "learning_rate": 7.805918214961851e-05,
+      "loss": 0.5186,
+      "step": 483
+    },
+    {
+      "epoch": 0.9508840864440079,
+      "grad_norm": 0.2498209859288056,
+      "learning_rate": 7.804226065180615e-05,
+      "loss": 0.5333,
+      "step": 484
+    },
+    {
+      "epoch": 0.9528487229862476,
+      "grad_norm": 0.24854698796356253,
+      "learning_rate": 7.802526755696011e-05,
+      "loss": 0.5346,
+      "step": 485
+    },
+    {
+      "epoch": 0.9548133595284872,
+      "grad_norm": 0.17980083982927972,
+      "learning_rate": 7.800820289706207e-05,
+      "loss": 0.5306,
+      "step": 486
+    },
+    {
+      "epoch": 0.9567779960707269,
+      "grad_norm": 0.17083863676461128,
+      "learning_rate": 7.79910667042284e-05,
+      "loss": 0.5385,
+      "step": 487
+    },
+    {
+      "epoch": 0.9587426326129665,
+      "grad_norm": 0.2341994820849114,
+      "learning_rate": 7.797385901071009e-05,
+      "loss": 0.5348,
+      "step": 488
+    },
+    {
+      "epoch": 0.9607072691552063,
+      "grad_norm": 0.271240072211406,
+      "learning_rate": 7.79565798488927e-05,
+      "loss": 0.5264,
+      "step": 489
+    },
+    {
+      "epoch": 0.962671905697446,
+      "grad_norm": 0.28990250044837557,
+      "learning_rate": 7.793922925129626e-05,
+      "loss": 0.5307,
+      "step": 490
+    },
+    {
+      "epoch": 0.9646365422396856,
+      "grad_norm": 0.28345644217229937,
+      "learning_rate": 7.792180725057533e-05,
+      "loss": 0.5263,
+      "step": 491
+    },
+    {
+      "epoch": 0.9666011787819253,
+      "grad_norm": 0.24866463828879862,
+      "learning_rate": 7.790431387951878e-05,
+      "loss": 0.5308,
+      "step": 492
+    },
+    {
+      "epoch": 0.9685658153241651,
+      "grad_norm": 0.197484371529423,
+      "learning_rate": 7.78867491710498e-05,
+      "loss": 0.5296,
+      "step": 493
+    },
+    {
+      "epoch": 0.9705304518664047,
+      "grad_norm": 0.1505157965673145,
+      "learning_rate": 7.78691131582259e-05,
+      "loss": 0.5138,
+      "step": 494
+    },
+    {
+      "epoch": 0.9724950884086444,
+      "grad_norm": 0.18461441730183856,
+      "learning_rate": 7.785140587423872e-05,
+      "loss": 0.5284,
+      "step": 495
+    },
+    {
+      "epoch": 0.9744597249508841,
+      "grad_norm": 0.2247481862129197,
+      "learning_rate": 7.78336273524141e-05,
+      "loss": 0.5342,
+      "step": 496
+    },
+    {
+      "epoch": 0.9764243614931237,
+      "grad_norm": 0.22548270454534866,
+      "learning_rate": 7.78157776262119e-05,
+      "loss": 0.5288,
+      "step": 497
+    },
+    {
+      "epoch": 0.9783889980353635,
+      "grad_norm": 0.20850432156565557,
+      "learning_rate": 7.7797856729226e-05,
+      "loss": 0.5333,
+      "step": 498
+    },
+    {
+      "epoch": 0.9803536345776032,
+      "grad_norm": 0.21850997165067135,
+      "learning_rate": 7.777986469518424e-05,
+      "loss": 0.5285,
+      "step": 499
+    },
+    {
+      "epoch": 0.9823182711198428,
+      "grad_norm": 0.1987007817790852,
+      "learning_rate": 7.776180155794836e-05,
+      "loss": 0.5339,
+      "step": 500
+    },
+    {
+      "epoch": 0.9842829076620825,
+      "grad_norm": 0.1784179250594391,
+      "learning_rate": 7.774366735151386e-05,
+      "loss": 0.5207,
+      "step": 501
+    },
+    {
+      "epoch": 0.9862475442043221,
+      "grad_norm": 0.1613881565016949,
+      "learning_rate": 7.772546211001005e-05,
+      "loss": 0.5337,
+      "step": 502
+    },
+    {
+      "epoch": 0.9882121807465619,
+      "grad_norm": 0.20343759862720978,
+      "learning_rate": 7.770718586769991e-05,
+      "loss": 0.5411,
+      "step": 503
+    },
+    {
+      "epoch": 0.9901768172888016,
+      "grad_norm": 0.3048124906733787,
+      "learning_rate": 7.768883865898005e-05,
+      "loss": 0.5333,
+      "step": 504
+    },
+    {
+      "epoch": 0.9921414538310412,
+      "grad_norm": 0.3927625928792787,
+      "learning_rate": 7.767042051838064e-05,
+      "loss": 0.5347,
+      "step": 505
+    },
+    {
+      "epoch": 0.9941060903732809,
+      "grad_norm": 0.46285293188842946,
+      "learning_rate": 7.765193148056533e-05,
+      "loss": 0.5387,
+      "step": 506
+    },
+    {
+      "epoch": 0.9960707269155207,
+      "grad_norm": 0.5444948846542476,
+      "learning_rate": 7.763337158033125e-05,
+      "loss": 0.5383,
+      "step": 507
+    },
+    {
+      "epoch": 0.9980353634577603,
+      "grad_norm": 0.5560459250947489,
+      "learning_rate": 7.761474085260883e-05,
+      "loss": 0.532,
+      "step": 508
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4205108457653066,
+      "learning_rate": 7.759603933246186e-05,
+      "loss": 0.5388,
+      "step": 509
+    },
+    {
+      "epoch": 1.0019646365422397,
+      "grad_norm": 0.31766584959906796,
+      "learning_rate": 7.757726705508732e-05,
+      "loss": 0.5105,
+      "step": 510
+    },
+    {
+      "epoch": 1.0039292730844793,
+      "grad_norm": 0.32704620413161517,
+      "learning_rate": 7.755842405581538e-05,
+      "loss": 0.5083,
+      "step": 511
+    },
+    {
+      "epoch": 1.005893909626719,
+      "grad_norm": 0.5447132250953503,
+      "learning_rate": 7.753951037010931e-05,
+      "loss": 0.5073,
+      "step": 512
+    },
+    {
+      "epoch": 1.0078585461689586,
+      "grad_norm": 0.7199342248075017,
+      "learning_rate": 7.752052603356542e-05,
+      "loss": 0.5072,
+      "step": 513
+    },
+    {
+      "epoch": 1.0098231827111985,
+      "grad_norm": 0.7219954910048539,
+      "learning_rate": 7.750147108191297e-05,
+      "loss": 0.5074,
+      "step": 514
+    },
+    {
+      "epoch": 1.0117878192534382,
+      "grad_norm": 0.4959520033073526,
+      "learning_rate": 7.748234555101413e-05,
+      "loss": 0.5105,
+      "step": 515
+    },
+    {
+      "epoch": 1.0137524557956779,
+      "grad_norm": 0.3371791263551942,
+      "learning_rate": 7.74631494768639e-05,
+      "loss": 0.5085,
+      "step": 516
+    },
+    {
+      "epoch": 1.0157170923379175,
+      "grad_norm": 0.5099499436751466,
+      "learning_rate": 7.744388289559006e-05,
+      "loss": 0.5002,
+      "step": 517
+    },
+    {
+      "epoch": 1.0176817288801572,
+      "grad_norm": 0.4755384716027486,
+      "learning_rate": 7.742454584345306e-05,
+      "loss": 0.5063,
+      "step": 518
+    },
+    {
+      "epoch": 1.0196463654223968,
+      "grad_norm": 0.35312791982457026,
+      "learning_rate": 7.740513835684601e-05,
+      "loss": 0.5077,
+      "step": 519
+    },
+    {
+      "epoch": 1.0216110019646365,
+      "grad_norm": 0.3873193404721078,
+      "learning_rate": 7.738566047229454e-05,
+      "loss": 0.5031,
+      "step": 520
+    },
+    {
+      "epoch": 1.0235756385068762,
+      "grad_norm": 0.33415264133200623,
+      "learning_rate": 7.736611222645682e-05,
+      "loss": 0.5073,
+      "step": 521
+    },
+    {
+      "epoch": 1.0255402750491158,
+      "grad_norm": 0.3139014298742204,
+      "learning_rate": 7.734649365612341e-05,
+      "loss": 0.5038,
+      "step": 522
+    },
+    {
+      "epoch": 1.0275049115913557,
+      "grad_norm": 0.26131530555213617,
+      "learning_rate": 7.732680479821724e-05,
+      "loss": 0.5051,
+      "step": 523
+    },
+    {
+      "epoch": 1.0294695481335954,
+      "grad_norm": 0.28163895062277255,
+      "learning_rate": 7.730704568979351e-05,
+      "loss": 0.4962,
+      "step": 524
+    },
+    {
+      "epoch": 1.031434184675835,
+      "grad_norm": 0.2949964673471062,
+      "learning_rate": 7.728721636803964e-05,
+      "loss": 0.5101,
+      "step": 525
+    },
+    {
+      "epoch": 1.0333988212180747,
+      "grad_norm": 0.24596617695820613,
+      "learning_rate": 7.726731687027518e-05,
+      "loss": 0.4954,
+      "step": 526
+    },
+    {
+      "epoch": 1.0353634577603144,
+      "grad_norm": 0.24524141845752986,
+      "learning_rate": 7.72473472339518e-05,
+      "loss": 0.4979,
+      "step": 527
+    },
+    {
+      "epoch": 1.037328094302554,
+      "grad_norm": 0.2932215610604043,
+      "learning_rate": 7.722730749665312e-05,
+      "loss": 0.5058,
+      "step": 528
+    },
+    {
+      "epoch": 1.0392927308447937,
+      "grad_norm": 0.2389072819251357,
+      "learning_rate": 7.720719769609473e-05,
+      "loss": 0.5037,
+      "step": 529
+    },
+    {
+      "epoch": 1.0412573673870333,
+      "grad_norm": 0.22411941454140152,
+      "learning_rate": 7.718701787012406e-05,
+      "loss": 0.5123,
+      "step": 530
+    },
+    {
+      "epoch": 1.043222003929273,
+      "grad_norm": 0.2308985475397428,
+      "learning_rate": 7.716676805672033e-05,
+      "loss": 0.4984,
+      "step": 531
+    },
+    {
+      "epoch": 1.0451866404715129,
+      "grad_norm": 0.22820683077183254,
+      "learning_rate": 7.714644829399449e-05,
+      "loss": 0.5016,
+      "step": 532
+    },
+    {
+      "epoch": 1.0471512770137525,
+      "grad_norm": 0.21374519324766586,
+      "learning_rate": 7.712605862018914e-05,
+      "loss": 0.4995,
+      "step": 533
+    },
+    {
+      "epoch": 1.0491159135559922,
+      "grad_norm": 0.18508481332509374,
+      "learning_rate": 7.710559907367846e-05,
+      "loss": 0.5098,
+      "step": 534
+    },
+    {
+      "epoch": 1.0510805500982319,
+      "grad_norm": 0.21610795396411828,
+      "learning_rate": 7.708506969296811e-05,
+      "loss": 0.4989,
+      "step": 535
+    },
+    {
+      "epoch": 1.0530451866404715,
+      "grad_norm": 0.22502089766833389,
+      "learning_rate": 7.706447051669518e-05,
+      "loss": 0.5004,
+      "step": 536
+    },
+    {
+      "epoch": 1.0550098231827112,
+      "grad_norm": 0.2118170724076432,
+      "learning_rate": 7.704380158362816e-05,
+      "loss": 0.5172,
+      "step": 537
+    },
+    {
+      "epoch": 1.0569744597249509,
+      "grad_norm": 0.2067955735196444,
+      "learning_rate": 7.702306293266677e-05,
+      "loss": 0.5054,
+      "step": 538
+    },
+    {
+      "epoch": 1.0589390962671905,
+      "grad_norm": 0.26158772784903417,
+      "learning_rate": 7.700225460284199e-05,
+      "loss": 0.5067,
+      "step": 539
+    },
+    {
+      "epoch": 1.0609037328094302,
+      "grad_norm": 0.22752185668588695,
+      "learning_rate": 7.69813766333159e-05,
+      "loss": 0.5062,
+      "step": 540
+    },
+    {
+      "epoch": 1.0628683693516698,
+      "grad_norm": 0.20888538070608897,
+      "learning_rate": 7.696042906338166e-05,
+      "loss": 0.5029,
+      "step": 541
+    },
+    {
+      "epoch": 1.0648330058939097,
+      "grad_norm": 0.2603501134337154,
+      "learning_rate": 7.693941193246344e-05,
+      "loss": 0.5004,
+      "step": 542
+    },
+    {
+      "epoch": 1.0667976424361494,
+      "grad_norm": 0.22058158548502163,
+      "learning_rate": 7.691832528011632e-05,
+      "loss": 0.4928,
+      "step": 543
+    },
+    {
+      "epoch": 1.068762278978389,
+      "grad_norm": 0.19417493102027514,
+      "learning_rate": 7.689716914602616e-05,
+      "loss": 0.4989,
+      "step": 544
+    },
+    {
+      "epoch": 1.0707269155206287,
+      "grad_norm": 0.2832574730031631,
+      "learning_rate": 7.687594357000968e-05,
+      "loss": 0.5046,
+      "step": 545
+    },
+    {
+      "epoch": 1.0726915520628684,
+      "grad_norm": 0.3417062272542928,
+      "learning_rate": 7.685464859201424e-05,
+      "loss": 0.5038,
+      "step": 546
+    },
+    {
+      "epoch": 1.074656188605108,
+      "grad_norm": 0.3479630889195829,
+      "learning_rate": 7.683328425211783e-05,
+      "loss": 0.4897,
+      "step": 547
+    },
+    {
+      "epoch": 1.0766208251473477,
+      "grad_norm": 0.3276440128073062,
+      "learning_rate": 7.681185059052895e-05,
+      "loss": 0.5019,
+      "step": 548
+    },
+    {
+      "epoch": 1.0785854616895874,
+      "grad_norm": 0.25701763685638207,
+      "learning_rate": 7.679034764758665e-05,
+      "loss": 0.503,
+      "step": 549
+    },
+    {
+      "epoch": 1.080550098231827,
+      "grad_norm": 0.29090633627160983,
+      "learning_rate": 7.676877546376028e-05,
+      "loss": 0.5097,
+      "step": 550
+    },
+    {
+      "epoch": 1.082514734774067,
+      "grad_norm": 0.3233623297225609,
+      "learning_rate": 7.674713407964954e-05,
+      "loss": 0.4999,
+      "step": 551
+    },
+    {
+      "epoch": 1.0844793713163066,
+      "grad_norm": 0.3589736224258201,
+      "learning_rate": 7.672542353598438e-05,
+      "loss": 0.5055,
+      "step": 552
+    },
+    {
+      "epoch": 1.0864440078585462,
+      "grad_norm": 0.3335816034314487,
+      "learning_rate": 7.670364387362491e-05,
+      "loss": 0.5041,
+      "step": 553
+    },
+    {
+      "epoch": 1.0884086444007859,
+      "grad_norm": 0.3371430110639185,
+      "learning_rate": 7.668179513356128e-05,
+      "loss": 0.4928,
+      "step": 554
+    },
+    {
+      "epoch": 1.0903732809430255,
+      "grad_norm": 0.3418161821560029,
+      "learning_rate": 7.66598773569137e-05,
+      "loss": 0.4977,
+      "step": 555
+    },
+    {
+      "epoch": 1.0923379174852652,
+      "grad_norm": 0.32142783521327656,
+      "learning_rate": 7.663789058493228e-05,
+      "loss": 0.498,
+      "step": 556
+    },
+    {
+      "epoch": 1.0943025540275049,
+      "grad_norm": 0.2919420630849687,
+      "learning_rate": 7.6615834858997e-05,
+      "loss": 0.5004,
+      "step": 557
+    },
+    {
+      "epoch": 1.0962671905697445,
+      "grad_norm": 0.25934305607585056,
+      "learning_rate": 7.659371022061762e-05,
+      "loss": 0.5038,
+      "step": 558
+    },
+    {
+      "epoch": 1.0982318271119842,
+      "grad_norm": 0.257842441826305,
+      "learning_rate": 7.657151671143354e-05,
+      "loss": 0.5043,
+      "step": 559
+    },
+    {
+      "epoch": 1.1001964636542239,
+      "grad_norm": 0.28705799843163254,
+      "learning_rate": 7.654925437321386e-05,
+      "loss": 0.4987,
+      "step": 560
+    },
+    {
+      "epoch": 1.1021611001964637,
+      "grad_norm": 0.27757519296240407,
+      "learning_rate": 7.652692324785713e-05,
+      "loss": 0.5078,
+      "step": 561
+    },
+    {
+      "epoch": 1.1041257367387034,
+      "grad_norm": 0.28527053648678463,
+      "learning_rate": 7.650452337739147e-05,
+      "loss": 0.4926,
+      "step": 562
+    },
+    {
+      "epoch": 1.106090373280943,
+      "grad_norm": 0.2711697219404111,
+      "learning_rate": 7.648205480397428e-05,
+      "loss": 0.4988,
+      "step": 563
+    },
+    {
+      "epoch": 1.1080550098231827,
+      "grad_norm": 0.2769895421049087,
+      "learning_rate": 7.645951756989231e-05,
+      "loss": 0.5044,
+      "step": 564
+    },
+    {
+      "epoch": 1.1100196463654224,
+      "grad_norm": 0.2321943656124972,
+      "learning_rate": 7.643691171756151e-05,
+      "loss": 0.5037,
+      "step": 565
+    },
+    {
+      "epoch": 1.111984282907662,
+      "grad_norm": 0.26648334615499536,
+      "learning_rate": 7.641423728952701e-05,
+      "loss": 0.503,
+      "step": 566
+    },
+    {
+      "epoch": 1.1139489194499017,
+      "grad_norm": 0.28091819057637496,
+      "learning_rate": 7.639149432846296e-05,
+      "loss": 0.5031,
+      "step": 567
+    },
+    {
+      "epoch": 1.1159135559921414,
+      "grad_norm": 0.25205088001713144,
+      "learning_rate": 7.63686828771725e-05,
+      "loss": 0.5026,
+      "step": 568
+    },
+    {
+      "epoch": 1.117878192534381,
+      "grad_norm": 0.23652567770743932,
+      "learning_rate": 7.63458029785877e-05,
+      "loss": 0.4954,
+      "step": 569
+    },
+    {
+      "epoch": 1.119842829076621,
+      "grad_norm": 0.2288225593242025,
+      "learning_rate": 7.632285467576943e-05,
+      "loss": 0.4946,
+      "step": 570
+    },
+    {
+      "epoch": 1.1218074656188606,
+      "grad_norm": 0.2318951089683919,
+      "learning_rate": 7.629983801190729e-05,
+      "loss": 0.5001,
+      "step": 571
+    },
+    {
+      "epoch": 1.1237721021611002,
+      "grad_norm": 0.26630122604903145,
+      "learning_rate": 7.627675303031954e-05,
+      "loss": 0.5057,
+      "step": 572
+    },
+    {
+      "epoch": 1.12573673870334,
+      "grad_norm": 0.27820268321667246,
+      "learning_rate": 7.625359977445303e-05,
+      "loss": 0.506,
+      "step": 573
+    },
+    {
+      "epoch": 1.1277013752455796,
+      "grad_norm": 0.25402051432414857,
+      "learning_rate": 7.623037828788314e-05,
+      "loss": 0.4987,
+      "step": 574
+    },
+    {
+      "epoch": 1.1296660117878192,
+      "grad_norm": 0.3015349727174611,
+      "learning_rate": 7.620708861431356e-05,
+      "loss": 0.5025,
+      "step": 575
+    },
+    {
+      "epoch": 1.1316306483300589,
+      "grad_norm": 0.3094348739979955,
+      "learning_rate": 7.618373079757641e-05,
+      "loss": 0.5054,
+      "step": 576
+    },
+    {
+      "epoch": 1.1335952848722985,
+      "grad_norm": 0.2533474663039733,
+      "learning_rate": 7.616030488163201e-05,
+      "loss": 0.5047,
+      "step": 577
+    },
+    {
+      "epoch": 1.1355599214145382,
+      "grad_norm": 0.23055253876770496,
+      "learning_rate": 7.613681091056887e-05,
+      "loss": 0.5027,
+      "step": 578
+    },
+    {
+      "epoch": 1.137524557956778,
+      "grad_norm": 0.21917882492346694,
+      "learning_rate": 7.611324892860355e-05,
+      "loss": 0.5078,
+      "step": 579
+    },
+    {
+      "epoch": 1.1394891944990178,
+      "grad_norm": 0.20294521214747926,
+      "learning_rate": 7.608961898008066e-05,
+      "loss": 0.4943,
+      "step": 580
+    },
+    {
+      "epoch": 1.1414538310412574,
+      "grad_norm": 0.2448022592625579,
+      "learning_rate": 7.606592110947266e-05,
+      "loss": 0.4936,
+      "step": 581
+    },
+    {
+      "epoch": 1.143418467583497,
+      "grad_norm": 0.24233853858478094,
+      "learning_rate": 7.604215536137988e-05,
+      "loss": 0.4868,
+      "step": 582
+    },
+    {
+      "epoch": 1.1453831041257367,
+      "grad_norm": 0.23523141806521639,
+      "learning_rate": 7.601832178053042e-05,
+      "loss": 0.4982,
+      "step": 583
+    },
+    {
+      "epoch": 1.1473477406679764,
+      "grad_norm": 0.30966852074579143,
+      "learning_rate": 7.599442041177998e-05,
+      "loss": 0.5045,
+      "step": 584
+    },
+    {
+      "epoch": 1.149312377210216,
+      "grad_norm": 0.32680714763185564,
+      "learning_rate": 7.597045130011191e-05,
+      "loss": 0.4995,
+      "step": 585
+    },
+    {
+      "epoch": 1.1512770137524557,
+      "grad_norm": 0.3255217067918779,
+      "learning_rate": 7.594641449063702e-05,
+      "loss": 0.5074,
+      "step": 586
+    },
+    {
+      "epoch": 1.1532416502946954,
+      "grad_norm": 0.28643058346324396,
+      "learning_rate": 7.592231002859352e-05,
+      "loss": 0.507,
+      "step": 587
+    },
+    {
+      "epoch": 1.1552062868369353,
+      "grad_norm": 0.2654262297968433,
+      "learning_rate": 7.589813795934697e-05,
+      "loss": 0.4975,
+      "step": 588
+    },
+    {
+      "epoch": 1.157170923379175,
+      "grad_norm": 0.26358347241978675,
+      "learning_rate": 7.587389832839015e-05,
+      "loss": 0.5026,
+      "step": 589
+    },
+    {
+      "epoch": 1.1591355599214146,
+      "grad_norm": 0.2725458056299067,
+      "learning_rate": 7.5849591181343e-05,
+      "loss": 0.5004,
+      "step": 590
+    },
+    {
+      "epoch": 1.1611001964636543,
+      "grad_norm": 0.290925918310358,
+      "learning_rate": 7.582521656395253e-05,
+      "loss": 0.5079,
+      "step": 591
+    },
+    {
+      "epoch": 1.163064833005894,
+      "grad_norm": 0.30124181738568023,
+      "learning_rate": 7.580077452209273e-05,
+      "loss": 0.5058,
+      "step": 592
+    },
+    {
+      "epoch": 1.1650294695481336,
+      "grad_norm": 0.29814387107114926,
+      "learning_rate": 7.577626510176448e-05,
+      "loss": 0.5081,
+      "step": 593
+    },
+    {
+      "epoch": 1.1669941060903732,
+      "grad_norm": 0.309467920078405,
+      "learning_rate": 7.575168834909548e-05,
+      "loss": 0.512,
+      "step": 594
+    },
+    {
+      "epoch": 1.168958742632613,
+      "grad_norm": 0.2777852520646252,
+      "learning_rate": 7.572704431034015e-05,
+      "loss": 0.5058,
+      "step": 595
+    },
+    {
+      "epoch": 1.1709233791748526,
+      "grad_norm": 0.2013465191625076,
+      "learning_rate": 7.570233303187954e-05,
+      "loss": 0.4996,
+      "step": 596
+    },
+    {
+      "epoch": 1.1728880157170924,
+      "grad_norm": 0.18391397324188313,
+      "learning_rate": 7.567755456022124e-05,
+      "loss": 0.5054,
+      "step": 597
+    },
+    {
+      "epoch": 1.174852652259332,
+      "grad_norm": 0.21114464165418464,
+      "learning_rate": 7.565270894199933e-05,
+      "loss": 0.5001,
+      "step": 598
+    },
+    {
+      "epoch": 1.1768172888015718,
+      "grad_norm": 0.22022022882925654,
+      "learning_rate": 7.562779622397424e-05,
+      "loss": 0.4935,
+      "step": 599
+    },
+    {
+      "epoch": 1.1787819253438114,
+      "grad_norm": 0.23185916262601441,
+      "learning_rate": 7.560281645303266e-05,
+      "loss": 0.5082,
+      "step": 600
+    },
+    {
+      "epoch": 1.180746561886051,
+      "grad_norm": 0.2633196299072739,
+      "learning_rate": 7.557776967618754e-05,
+      "loss": 0.5091,
+      "step": 601
+    },
+    {
+      "epoch": 1.1827111984282908,
+      "grad_norm": 0.2344144148341606,
+      "learning_rate": 7.555265594057787e-05,
+      "loss": 0.4949,
+      "step": 602
+    },
+    {
+      "epoch": 1.1846758349705304,
+      "grad_norm": 0.22155093364709255,
+      "learning_rate": 7.552747529346871e-05,
+      "loss": 0.487,
+      "step": 603
+    },
+    {
+      "epoch": 1.18664047151277,
+      "grad_norm": 0.20710023390545712,
+      "learning_rate": 7.550222778225102e-05,
+      "loss": 0.5006,
+      "step": 604
+    },
+    {
+      "epoch": 1.1886051080550097,
+      "grad_norm": 0.19337871234418327,
+      "learning_rate": 7.547691345444162e-05,
+      "loss": 0.4924,
+      "step": 605
+    },
+    {
+      "epoch": 1.1905697445972496,
+      "grad_norm": 0.24107179416452032,
+      "learning_rate": 7.545153235768306e-05,
+      "loss": 0.492,
+      "step": 606
+    },
+    {
+      "epoch": 1.1925343811394893,
+      "grad_norm": 0.3013551971978563,
+      "learning_rate": 7.542608453974359e-05,
+      "loss": 0.5041,
+      "step": 607
+    },
+    {
+      "epoch": 1.194499017681729,
+      "grad_norm": 0.3203763478500961,
+      "learning_rate": 7.540057004851698e-05,
+      "loss": 0.4985,
+      "step": 608
+    },
+    {
+      "epoch": 1.1964636542239686,
+      "grad_norm": 0.2995333655656614,
+      "learning_rate": 7.53749889320225e-05,
+      "loss": 0.5045,
+      "step": 609
+    },
+    {
+      "epoch": 1.1984282907662083,
+      "grad_norm": 0.24642203352534364,
+      "learning_rate": 7.534934123840486e-05,
+      "loss": 0.5017,
+      "step": 610
+    },
+    {
+      "epoch": 1.200392927308448,
+      "grad_norm": 0.21459660829399652,
+      "learning_rate": 7.532362701593398e-05,
+      "loss": 0.4961,
+      "step": 611
+    },
+    {
+      "epoch": 1.2023575638506876,
+      "grad_norm": 0.2797810298424695,
+      "learning_rate": 7.529784631300508e-05,
+      "loss": 0.5034,
+      "step": 612
+    },
+    {
+      "epoch": 1.2043222003929273,
+      "grad_norm": 0.3279773125888662,
+      "learning_rate": 7.527199917813843e-05,
+      "loss": 0.5015,
+      "step": 613
+    },
+    {
+      "epoch": 1.206286836935167,
+      "grad_norm": 0.32340876536841595,
+      "learning_rate": 7.524608565997937e-05,
+      "loss": 0.5048,
+      "step": 614
+    },
+    {
+      "epoch": 1.2082514734774068,
+      "grad_norm": 0.2488199296512009,
+      "learning_rate": 7.522010580729819e-05,
+      "loss": 0.4938,
+      "step": 615
+    },
+    {
+      "epoch": 1.2102161100196462,
+      "grad_norm": 0.18485085040617072,
+      "learning_rate": 7.519405966898995e-05,
+      "loss": 0.5069,
+      "step": 616
+    },
+    {
+      "epoch": 1.2121807465618861,
+      "grad_norm": 0.19057563688809206,
+      "learning_rate": 7.516794729407453e-05,
+      "loss": 0.5063,
+      "step": 617
+    },
+    {
+      "epoch": 1.2141453831041258,
+      "grad_norm": 0.20649099584274924,
+      "learning_rate": 7.514176873169644e-05,
+      "loss": 0.4989,
+      "step": 618
+    },
+    {
+      "epoch": 1.2161100196463654,
+      "grad_norm": 0.21825868309080496,
+      "learning_rate": 7.511552403112477e-05,
+      "loss": 0.501,
+      "step": 619
+    },
+    {
+      "epoch": 1.218074656188605,
+      "grad_norm": 0.23025536289046308,
+      "learning_rate": 7.50892132417531e-05,
+      "loss": 0.5003,
+      "step": 620
+    },
+    {
+      "epoch": 1.2200392927308448,
+      "grad_norm": 0.23355477807391212,
+      "learning_rate": 7.506283641309935e-05,
+      "loss": 0.5005,
+      "step": 621
+    },
+    {
+      "epoch": 1.2220039292730844,
+      "grad_norm": 0.24358558850877407,
+      "learning_rate": 7.503639359480576e-05,
+      "loss": 0.5029,
+      "step": 622
+    },
+    {
+      "epoch": 1.223968565815324,
+      "grad_norm": 0.20488889651553424,
+      "learning_rate": 7.500988483663873e-05,
+      "loss": 0.5025,
+      "step": 623
+    },
+    {
+      "epoch": 1.2259332023575638,
+      "grad_norm": 0.18018080516564614,
+      "learning_rate": 7.498331018848884e-05,
+      "loss": 0.5012,
+      "step": 624
+    },
+    {
+      "epoch": 1.2278978388998034,
+      "grad_norm": 0.2121423568549863,
+      "learning_rate": 7.495666970037058e-05,
+      "loss": 0.4974,
+      "step": 625
+    },
+    {
+      "epoch": 1.2298624754420433,
+      "grad_norm": 0.20566707959779196,
+      "learning_rate": 7.49299634224224e-05,
+      "loss": 0.5003,
+      "step": 626
+    },
+    {
+      "epoch": 1.231827111984283,
+      "grad_norm": 0.18279216109759427,
+      "learning_rate": 7.490319140490661e-05,
+      "loss": 0.5111,
+      "step": 627
+    },
+    {
+      "epoch": 1.2337917485265226,
+      "grad_norm": 0.17319456224626495,
+      "learning_rate": 7.487635369820916e-05,
+      "loss": 0.4906,
+      "step": 628
+    },
+    {
+      "epoch": 1.2357563850687623,
+      "grad_norm": 0.22021759952637948,
+      "learning_rate": 7.484945035283968e-05,
+      "loss": 0.5045,
+      "step": 629
+    },
+    {
+      "epoch": 1.237721021611002,
+      "grad_norm": 0.18639878419676192,
+      "learning_rate": 7.482248141943133e-05,
+      "loss": 0.5048,
+      "step": 630
+    },
+    {
+      "epoch": 1.2396856581532416,
+      "grad_norm": 0.17816070974946244,
+      "learning_rate": 7.479544694874072e-05,
+      "loss": 0.5014,
+      "step": 631
+    },
+    {
+      "epoch": 1.2416502946954813,
+      "grad_norm": 0.20469541783354522,
+      "learning_rate": 7.476834699164776e-05,
+      "loss": 0.5044,
+      "step": 632
+    },
+    {
+      "epoch": 1.243614931237721,
+      "grad_norm": 0.250312145276167,
+      "learning_rate": 7.474118159915566e-05,
+      "loss": 0.5061,
+      "step": 633
+    },
+    {
+      "epoch": 1.2455795677799606,
+      "grad_norm": 0.2627383653568964,
+      "learning_rate": 7.471395082239075e-05,
+      "loss": 0.5018,
+      "step": 634
+    },
+    {
+      "epoch": 1.2475442043222005,
+      "grad_norm": 0.26401376629229883,
+      "learning_rate": 7.468665471260242e-05,
+      "loss": 0.5073,
+      "step": 635
+    },
+    {
+      "epoch": 1.2495088408644401,
+      "grad_norm": 0.2198057279535012,
+      "learning_rate": 7.465929332116304e-05,
+      "loss": 0.4953,
+      "step": 636
+    },
+    {
+      "epoch": 1.2514734774066798,
+      "grad_norm": 0.17649123299673405,
+      "learning_rate": 7.463186669956782e-05,
+      "loss": 0.5017,
+      "step": 637
+    },
+    {
+      "epoch": 1.2534381139489195,
+      "grad_norm": 0.17379627228872085,
+      "learning_rate": 7.460437489943473e-05,
+      "loss": 0.4985,
+      "step": 638
+    },
+    {
+      "epoch": 1.2554027504911591,
+      "grad_norm": 0.22307635758803093,
+      "learning_rate": 7.457681797250445e-05,
+      "loss": 0.4945,
+      "step": 639
+    },
+    {
+      "epoch": 1.2573673870333988,
+      "grad_norm": 0.269152987652741,
+      "learning_rate": 7.454919597064017e-05,
+      "loss": 0.4992,
+      "step": 640
+    },
+    {
+      "epoch": 1.2593320235756384,
+      "grad_norm": 0.2927083244041568,
+      "learning_rate": 7.452150894582761e-05,
+      "loss": 0.504,
+      "step": 641
+    },
+    {
+      "epoch": 1.261296660117878,
+      "grad_norm": 0.271206610817319,
+      "learning_rate": 7.449375695017483e-05,
+      "loss": 0.497,
+      "step": 642
+    },
+    {
+      "epoch": 1.2632612966601178,
+      "grad_norm": 0.26457372593828654,
+      "learning_rate": 7.446594003591216e-05,
+      "loss": 0.5027,
+      "step": 643
+    },
+    {
+      "epoch": 1.2652259332023577,
+      "grad_norm": 0.27502404451434004,
+      "learning_rate": 7.443805825539218e-05,
+      "loss": 0.4996,
+      "step": 644
+    },
+    {
+      "epoch": 1.2671905697445973,
+      "grad_norm": 0.3192521868822564,
+      "learning_rate": 7.441011166108945e-05,
+      "loss": 0.5121,
+      "step": 645
+    },
+    {
+      "epoch": 1.269155206286837,
+      "grad_norm": 0.3354164820375092,
+      "learning_rate": 7.438210030560058e-05,
+      "loss": 0.503,
+      "step": 646
+    },
+    {
+      "epoch": 1.2711198428290766,
+      "grad_norm": 0.3194642221735784,
+      "learning_rate": 7.435402424164405e-05,
+      "loss": 0.4977,
+      "step": 647
+    },
+    {
+      "epoch": 1.2730844793713163,
+      "grad_norm": 0.31872398930183715,
+      "learning_rate": 7.432588352206014e-05,
+      "loss": 0.5033,
+      "step": 648
+    },
+    {
+      "epoch": 1.275049115913556,
+      "grad_norm": 0.2977876536075987,
+      "learning_rate": 7.429767819981073e-05,
+      "loss": 0.5051,
+      "step": 649
+    },
+    {
+      "epoch": 1.2770137524557956,
+      "grad_norm": 0.2350356933003974,
+      "learning_rate": 7.426940832797941e-05,
+      "loss": 0.4959,
+      "step": 650
+    },
+    {
+      "epoch": 1.2789783889980353,
+      "grad_norm": 0.21859195106869267,
+      "learning_rate": 7.424107395977116e-05,
+      "loss": 0.5016,
+      "step": 651
+    },
+    {
+      "epoch": 1.280943025540275,
+      "grad_norm": 0.25123020366634513,
+      "learning_rate": 7.42126751485124e-05,
+      "loss": 0.5042,
+      "step": 652
+    },
+    {
+      "epoch": 1.2829076620825148,
+      "grad_norm": 0.24413947955102935,
+      "learning_rate": 7.418421194765078e-05,
+      "loss": 0.5066,
+      "step": 653
+    },
+    {
+      "epoch": 1.2848722986247545,
+      "grad_norm": 0.1724045259246127,
+      "learning_rate": 7.41556844107552e-05,
+      "loss": 0.4924,
+      "step": 654
+    },
+    {
+      "epoch": 1.2868369351669942,
+      "grad_norm": 0.17004956247604963,
+      "learning_rate": 7.412709259151558e-05,
+      "loss": 0.5057,
+      "step": 655
+    },
+    {
+      "epoch": 1.2888015717092338,
+      "grad_norm": 0.19539382553792004,
+      "learning_rate": 7.409843654374287e-05,
+      "loss": 0.5005,
+      "step": 656
+    },
+    {
+      "epoch": 1.2907662082514735,
+      "grad_norm": 0.1793079377879988,
+      "learning_rate": 7.406971632136888e-05,
+      "loss": 0.5022,
+      "step": 657
+    },
+    {
+      "epoch": 1.2927308447937131,
+      "grad_norm": 0.19643010319486698,
+      "learning_rate": 7.404093197844619e-05,
+      "loss": 0.5054,
+      "step": 658
+    },
+    {
+      "epoch": 1.2946954813359528,
+      "grad_norm": 0.21624651546040458,
+      "learning_rate": 7.401208356914806e-05,
+      "loss": 0.5024,
+      "step": 659
+    },
+    {
+      "epoch": 1.2966601178781925,
+      "grad_norm": 0.20561966864668976,
+      "learning_rate": 7.398317114776836e-05,
+      "loss": 0.496,
+      "step": 660
+    },
+    {
+      "epoch": 1.2986247544204321,
+      "grad_norm": 0.22170028107572878,
+      "learning_rate": 7.395419476872139e-05,
+      "loss": 0.5048,
+      "step": 661
+    },
+    {
+      "epoch": 1.300589390962672,
+      "grad_norm": 0.20181101593721854,
+      "learning_rate": 7.392515448654184e-05,
+      "loss": 0.4964,
+      "step": 662
+    },
+    {
+      "epoch": 1.3025540275049117,
+      "grad_norm": 0.22949767096895235,
+      "learning_rate": 7.389605035588466e-05,
+      "loss": 0.4993,
+      "step": 663
+    },
+    {
+      "epoch": 1.3045186640471513,
+      "grad_norm": 0.2905922119437003,
+      "learning_rate": 7.386688243152499e-05,
+      "loss": 0.5049,
+      "step": 664
+    },
+    {
+      "epoch": 1.306483300589391,
+      "grad_norm": 0.33785870587076716,
+      "learning_rate": 7.383765076835799e-05,
+      "loss": 0.4934,
+      "step": 665
+    },
+    {
+      "epoch": 1.3084479371316307,
+      "grad_norm": 0.35534431557597407,
+      "learning_rate": 7.38083554213988e-05,
+      "loss": 0.4963,
+      "step": 666
+    },
+    {
+      "epoch": 1.3104125736738703,
+      "grad_norm": 0.3982067827547152,
+      "learning_rate": 7.377899644578244e-05,
+      "loss": 0.5082,
+      "step": 667
+    },
+    {
+      "epoch": 1.31237721021611,
+      "grad_norm": 0.4144816794813476,
+      "learning_rate": 7.374957389676364e-05,
+      "loss": 0.5051,
+      "step": 668
+    },
+    {
+      "epoch": 1.3143418467583496,
+      "grad_norm": 0.3984328932237845,
+      "learning_rate": 7.37200878297168e-05,
+      "loss": 0.5051,
+      "step": 669
+    },
+    {
+      "epoch": 1.3163064833005893,
+      "grad_norm": 0.3373945403880092,
+      "learning_rate": 7.369053830013588e-05,
+      "loss": 0.4915,
+      "step": 670
+    },
+    {
+      "epoch": 1.3182711198428292,
+      "grad_norm": 0.29600880741904156,
+      "learning_rate": 7.366092536363423e-05,
+      "loss": 0.4981,
+      "step": 671
+    },
+    {
+      "epoch": 1.3202357563850686,
+      "grad_norm": 0.25946466806597046,
+      "learning_rate": 7.363124907594458e-05,
+      "loss": 0.4964,
+      "step": 672
+    },
+    {
+      "epoch": 1.3222003929273085,
+      "grad_norm": 0.3326278589237823,
+      "learning_rate": 7.360150949291887e-05,
+      "loss": 0.5053,
+      "step": 673
+    },
+    {
+      "epoch": 1.3241650294695482,
+      "grad_norm": 0.3598149234871574,
+      "learning_rate": 7.357170667052816e-05,
+      "loss": 0.5062,
+      "step": 674
+    },
+    {
+      "epoch": 1.3261296660117878,
+      "grad_norm": 0.27179211226460587,
+      "learning_rate": 7.354184066486256e-05,
+      "loss": 0.4969,
+      "step": 675
+    },
+    {
+      "epoch": 1.3280943025540275,
+      "grad_norm": 0.2027764310423967,
+      "learning_rate": 7.351191153213103e-05,
+      "loss": 0.5041,
+      "step": 676
+    },
+    {
+      "epoch": 1.3300589390962672,
+      "grad_norm": 0.23139083120929335,
+      "learning_rate": 7.348191932866141e-05,
+      "loss": 0.4999,
+      "step": 677
+    },
+    {
+      "epoch": 1.3320235756385068,
+      "grad_norm": 0.26075304758398316,
+      "learning_rate": 7.34518641109002e-05,
+      "loss": 0.5027,
+      "step": 678
+    },
+    {
+      "epoch": 1.3339882121807465,
+      "grad_norm": 0.26817433131036333,
+      "learning_rate": 7.34217459354125e-05,
+      "loss": 0.4945,
+      "step": 679
+    },
+    {
+      "epoch": 1.3359528487229864,
+      "grad_norm": 0.2711153442184527,
+      "learning_rate": 7.339156485888191e-05,
+      "loss": 0.4932,
+      "step": 680
+    },
+    {
+      "epoch": 1.3379174852652258,
+      "grad_norm": 0.21504688796360547,
+      "learning_rate": 7.33613209381104e-05,
+      "loss": 0.4993,
+      "step": 681
+    },
+    {
+      "epoch": 1.3398821218074657,
+      "grad_norm": 0.15858984026338746,
+      "learning_rate": 7.333101423001821e-05,
+      "loss": 0.501,
+      "step": 682
+    },
+    {
+      "epoch": 1.3418467583497053,
+      "grad_norm": 0.24593697470695708,
+      "learning_rate": 7.330064479164378e-05,
+      "loss": 0.5043,
+      "step": 683
+    },
+    {
+      "epoch": 1.343811394891945,
+      "grad_norm": 0.290705231807666,
+      "learning_rate": 7.327021268014358e-05,
+      "loss": 0.507,
+      "step": 684
+    },
+    {
+      "epoch": 1.3457760314341847,
+      "grad_norm": 0.26483695569260857,
+      "learning_rate": 7.323971795279205e-05,
+      "loss": 0.5016,
+      "step": 685
+    },
+    {
+      "epoch": 1.3477406679764243,
+      "grad_norm": 0.23361145254995394,
+      "learning_rate": 7.320916066698145e-05,
+      "loss": 0.5037,
+      "step": 686
+    },
+    {
+      "epoch": 1.349705304518664,
+      "grad_norm": 0.2466410521076216,
+      "learning_rate": 7.317854088022182e-05,
+      "loss": 0.4992,
+      "step": 687
+    },
+    {
+      "epoch": 1.3516699410609037,
+      "grad_norm": 0.20082778336072885,
+      "learning_rate": 7.31478586501408e-05,
+      "loss": 0.4996,
+      "step": 688
+    },
+    {
+      "epoch": 1.3536345776031435,
+      "grad_norm": 0.18592807302839148,
+      "learning_rate": 7.311711403448353e-05,
+      "loss": 0.4952,
+      "step": 689
+    },
+    {
+      "epoch": 1.355599214145383,
+      "grad_norm": 0.2003362383779622,
+      "learning_rate": 7.308630709111261e-05,
+      "loss": 0.4993,
+      "step": 690
+    },
+    {
+      "epoch": 1.3575638506876229,
+      "grad_norm": 0.22778064943799148,
+      "learning_rate": 7.305543787800793e-05,
+      "loss": 0.4894,
+      "step": 691
+    },
+    {
+      "epoch": 1.3595284872298625,
+      "grad_norm": 0.23247562712932185,
+      "learning_rate": 7.302450645326654e-05,
+      "loss": 0.502,
+      "step": 692
+    },
+    {
+      "epoch": 1.3614931237721022,
+      "grad_norm": 0.24021811551488106,
+      "learning_rate": 7.299351287510263e-05,
+      "loss": 0.5037,
+      "step": 693
+    },
+    {
+      "epoch": 1.3634577603143418,
+      "grad_norm": 0.30342968391057007,
+      "learning_rate": 7.296245720184731e-05,
+      "loss": 0.5021,
+      "step": 694
+    },
+    {
+      "epoch": 1.3654223968565815,
+      "grad_norm": 0.2570070743281194,
+      "learning_rate": 7.293133949194859e-05,
+      "loss": 0.5013,
+      "step": 695
+    },
+    {
+      "epoch": 1.3673870333988212,
+      "grad_norm": 0.22821345251733705,
+      "learning_rate": 7.290015980397124e-05,
+      "loss": 0.5013,
+      "step": 696
+    },
+    {
+      "epoch": 1.3693516699410608,
+      "grad_norm": 0.2161707796903105,
+      "learning_rate": 7.286891819659664e-05,
+      "loss": 0.4873,
+      "step": 697
+    },
+    {
+      "epoch": 1.3713163064833007,
+      "grad_norm": 0.19550627391599776,
+      "learning_rate": 7.283761472862275e-05,
+      "loss": 0.4978,
+      "step": 698
+    },
+    {
+      "epoch": 1.3732809430255402,
+      "grad_norm": 0.1987231462707916,
+      "learning_rate": 7.28062494589639e-05,
+      "loss": 0.4909,
+      "step": 699
+    },
+    {
+      "epoch": 1.37524557956778,
+      "grad_norm": 0.18737665084896968,
+      "learning_rate": 7.277482244665079e-05,
+      "loss": 0.5008,
+      "step": 700
+    },
+    {
+      "epoch": 1.3772102161100197,
+      "grad_norm": 0.22640728386273765,
+      "learning_rate": 7.274333375083027e-05,
+      "loss": 0.5002,
+      "step": 701
+    },
+    {
+      "epoch": 1.3791748526522594,
+      "grad_norm": 0.23667971415152367,
+      "learning_rate": 7.271178343076532e-05,
+      "loss": 0.4843,
+      "step": 702
+    },
+    {
+      "epoch": 1.381139489194499,
+      "grad_norm": 0.2655555606986464,
+      "learning_rate": 7.26801715458349e-05,
+      "loss": 0.5062,
+      "step": 703
+    },
+    {
+      "epoch": 1.3831041257367387,
+      "grad_norm": 0.26747847407282327,
+      "learning_rate": 7.26484981555338e-05,
+      "loss": 0.5066,
+      "step": 704
+    },
+    {
+      "epoch": 1.3850687622789783,
+      "grad_norm": 0.2352230318642284,
+      "learning_rate": 7.261676331947261e-05,
+      "loss": 0.5016,
+      "step": 705
+    },
+    {
+      "epoch": 1.387033398821218,
+      "grad_norm": 0.20511300315326028,
+      "learning_rate": 7.258496709737753e-05,
+      "loss": 0.5,
+      "step": 706
+    },
+    {
+      "epoch": 1.388998035363458,
+      "grad_norm": 0.18480408233712753,
+      "learning_rate": 7.255310954909028e-05,
+      "loss": 0.4918,
+      "step": 707
+    },
+    {
+      "epoch": 1.3909626719056973,
+      "grad_norm": 0.2016362496553805,
+      "learning_rate": 7.252119073456808e-05,
+      "loss": 0.5019,
+      "step": 708
+    },
+    {
+      "epoch": 1.3929273084479372,
+      "grad_norm": 0.17007773672087503,
+      "learning_rate": 7.248921071388333e-05,
+      "loss": 0.4915,
+      "step": 709
+    },
+    {
+      "epoch": 1.3948919449901769,
+      "grad_norm": 0.19919190697635894,
+      "learning_rate": 7.245716954722374e-05,
+      "loss": 0.4997,
+      "step": 710
+    },
+    {
+      "epoch": 1.3968565815324165,
+      "grad_norm": 0.19958809751728265,
+      "learning_rate": 7.242506729489203e-05,
+      "loss": 0.5006,
+      "step": 711
+    },
+    {
+      "epoch": 1.3988212180746562,
+      "grad_norm": 0.19717737773439412,
+      "learning_rate": 7.23929040173059e-05,
+      "loss": 0.5026,
+      "step": 712
+    },
+    {
+      "epoch": 1.4007858546168959,
+      "grad_norm": 0.2049818636182103,
+      "learning_rate": 7.236067977499791e-05,
+      "loss": 0.5012,
+      "step": 713
+    },
+    {
+      "epoch": 1.4027504911591355,
+      "grad_norm": 0.18600600808249418,
+      "learning_rate": 7.232839462861535e-05,
+      "loss": 0.493,
+      "step": 714
+    },
+    {
+      "epoch": 1.4047151277013752,
+      "grad_norm": 0.2060918433938838,
+      "learning_rate": 7.229604863892014e-05,
+      "loss": 0.5014,
+      "step": 715
+    },
+    {
+      "epoch": 1.4066797642436148,
+      "grad_norm": 0.20603353650440337,
+      "learning_rate": 7.226364186678869e-05,
+      "loss": 0.505,
+      "step": 716
+    },
+    {
+      "epoch": 1.4086444007858545,
+      "grad_norm": 0.1986315228693543,
+      "learning_rate": 7.223117437321187e-05,
+      "loss": 0.497,
+      "step": 717
+    },
+    {
+      "epoch": 1.4106090373280944,
+      "grad_norm": 0.23396493697543497,
+      "learning_rate": 7.219864621929474e-05,
+      "loss": 0.5079,
+      "step": 718
+    },
+    {
+      "epoch": 1.412573673870334,
+      "grad_norm": 0.27276644346923073,
+      "learning_rate": 7.216605746625658e-05,
+      "loss": 0.507,
+      "step": 719
+    },
+    {
+      "epoch": 1.4145383104125737,
+      "grad_norm": 0.2513750311034153,
+      "learning_rate": 7.21334081754307e-05,
+      "loss": 0.5022,
+      "step": 720
+    },
+    {
+      "epoch": 1.4165029469548134,
+      "grad_norm": 0.21387803925191803,
+      "learning_rate": 7.210069840826437e-05,
+      "loss": 0.4998,
+      "step": 721
+    },
+    {
+      "epoch": 1.418467583497053,
+      "grad_norm": 0.21144348501821797,
+      "learning_rate": 7.206792822631864e-05,
+      "loss": 0.4996,
+      "step": 722
+    },
+    {
+      "epoch": 1.4204322200392927,
+      "grad_norm": 0.22881341255330961,
+      "learning_rate": 7.203509769126831e-05,
+      "loss": 0.5011,
+      "step": 723
+    },
+    {
+      "epoch": 1.4223968565815324,
+      "grad_norm": 0.24284789966039744,
+      "learning_rate": 7.20022068649017e-05,
+      "loss": 0.4983,
+      "step": 724
+    },
+    {
+      "epoch": 1.424361493123772,
+      "grad_norm": 0.22082090513787556,
+      "learning_rate": 7.19692558091207e-05,
+      "loss": 0.5042,
+      "step": 725
+    },
+    {
+      "epoch": 1.4263261296660117,
+      "grad_norm": 0.17719310899788887,
+      "learning_rate": 7.193624458594043e-05,
+      "loss": 0.5043,
+      "step": 726
+    },
+    {
+      "epoch": 1.4282907662082516,
+      "grad_norm": 0.2335375959536549,
+      "learning_rate": 7.190317325748936e-05,
+      "loss": 0.4993,
+      "step": 727
+    },
+    {
+      "epoch": 1.4302554027504912,
+      "grad_norm": 0.2962165671209469,
+      "learning_rate": 7.187004188600901e-05,
+      "loss": 0.4973,
+      "step": 728
+    },
+    {
+      "epoch": 1.432220039292731,
+      "grad_norm": 0.2796674669039181,
+      "learning_rate": 7.183685053385393e-05,
+      "loss": 0.5125,
+      "step": 729
+    },
+    {
+      "epoch": 1.4341846758349706,
+      "grad_norm": 0.23121344519218506,
+      "learning_rate": 7.180359926349158e-05,
+      "loss": 0.5009,
+      "step": 730
+    },
+    {
+      "epoch": 1.4361493123772102,
+      "grad_norm": 0.2389600693568357,
+      "learning_rate": 7.177028813750211e-05,
+      "loss": 0.5056,
+      "step": 731
+    },
+    {
+      "epoch": 1.4381139489194499,
+      "grad_norm": 0.2448121185605615,
+      "learning_rate": 7.173691721857841e-05,
+      "loss": 0.4955,
+      "step": 732
+    },
+    {
+      "epoch": 1.4400785854616895,
+      "grad_norm": 0.24418332869514853,
+      "learning_rate": 7.170348656952586e-05,
+      "loss": 0.5001,
+      "step": 733
+    },
+    {
+      "epoch": 1.4420432220039292,
+      "grad_norm": 0.26574129705652666,
+      "learning_rate": 7.166999625326225e-05,
+      "loss": 0.4988,
+      "step": 734
+    },
+    {
+      "epoch": 1.4440078585461689,
+      "grad_norm": 0.2652079528826138,
+      "learning_rate": 7.163644633281769e-05,
+      "loss": 0.4921,
+      "step": 735
+    },
+    {
+      "epoch": 1.4459724950884087,
+      "grad_norm": 0.2444675882922687,
+      "learning_rate": 7.160283687133442e-05,
+      "loss": 0.5044,
+      "step": 736
+    },
+    {
+      "epoch": 1.4479371316306484,
+      "grad_norm": 0.21206492267477348,
+      "learning_rate": 7.15691679320668e-05,
+      "loss": 0.507,
+      "step": 737
+    },
+    {
+      "epoch": 1.449901768172888,
+      "grad_norm": 0.15913104184108803,
+      "learning_rate": 7.15354395783811e-05,
+      "loss": 0.4994,
+      "step": 738
+    },
+    {
+      "epoch": 1.4518664047151277,
+      "grad_norm": 0.14058888555032034,
+      "learning_rate": 7.150165187375539e-05,
+      "loss": 0.4919,
+      "step": 739
+    },
+    {
+      "epoch": 1.4538310412573674,
+      "grad_norm": 0.1648835364448364,
+      "learning_rate": 7.146780488177948e-05,
+      "loss": 0.4977,
+      "step": 740
+    },
+    {
+      "epoch": 1.455795677799607,
+      "grad_norm": 0.1861951835471473,
+      "learning_rate": 7.143389866615475e-05,
+      "loss": 0.5088,
+      "step": 741
+    },
+    {
+      "epoch": 1.4577603143418467,
+      "grad_norm": 0.19586928368135562,
+      "learning_rate": 7.139993329069401e-05,
+      "loss": 0.4942,
+      "step": 742
+    },
+    {
+      "epoch": 1.4597249508840864,
+      "grad_norm": 0.16428366777268139,
+      "learning_rate": 7.136590881932144e-05,
+      "loss": 0.5052,
+      "step": 743
+    },
+    {
+      "epoch": 1.461689587426326,
+      "grad_norm": 0.15169428517978517,
+      "learning_rate": 7.133182531607245e-05,
+      "loss": 0.5009,
+      "step": 744
+    },
+    {
+      "epoch": 1.463654223968566,
+      "grad_norm": 0.1797004324880978,
+      "learning_rate": 7.129768284509352e-05,
+      "loss": 0.495,
+      "step": 745
+    },
+    {
+      "epoch": 1.4656188605108054,
+      "grad_norm": 0.20164430854239104,
+      "learning_rate": 7.126348147064214e-05,
+      "loss": 0.4965,
+      "step": 746
+    },
+    {
+      "epoch": 1.4675834970530452,
+      "grad_norm": 0.2178359593281144,
+      "learning_rate": 7.122922125708665e-05,
+      "loss": 0.5094,
+      "step": 747
+    },
+    {
+      "epoch": 1.469548133595285,
+      "grad_norm": 0.23275433243747815,
+      "learning_rate": 7.119490226890608e-05,
+      "loss": 0.5034,
+      "step": 748
+    },
+    {
+      "epoch": 1.4715127701375246,
+      "grad_norm": 0.22594330783632474,
+      "learning_rate": 7.116052457069018e-05,
+      "loss": 0.5048,
+      "step": 749
+    },
+    {
+      "epoch": 1.4734774066797642,
+      "grad_norm": 0.2196749370366948,
+      "learning_rate": 7.112608822713908e-05,
+      "loss": 0.4955,
+      "step": 750
+    },
+    {
+      "epoch": 1.475442043222004,
+      "grad_norm": 0.18344129836675702,
+      "learning_rate": 7.109159330306338e-05,
+      "loss": 0.4997,
+      "step": 751
+    },
+    {
+      "epoch": 1.4774066797642436,
+      "grad_norm": 0.25549405211767323,
+      "learning_rate": 7.105703986338384e-05,
+      "loss": 0.4882,
+      "step": 752
+    },
+    {
+      "epoch": 1.4793713163064832,
+      "grad_norm": 0.2858279103708264,
+      "learning_rate": 7.102242797313143e-05,
+      "loss": 0.5166,
+      "step": 753
+    },
+    {
+      "epoch": 1.481335952848723,
+      "grad_norm": 0.26047010456018005,
+      "learning_rate": 7.098775769744708e-05,
+      "loss": 0.5051,
+      "step": 754
+    },
+    {
+      "epoch": 1.4833005893909625,
+      "grad_norm": 0.3019055712237542,
+      "learning_rate": 7.095302910158161e-05,
+      "loss": 0.49,
+      "step": 755
+    },
+    {
+      "epoch": 1.4852652259332024,
+      "grad_norm": 0.2700125864849798,
+      "learning_rate": 7.091824225089563e-05,
+      "loss": 0.5044,
+      "step": 756
+    },
+    {
+      "epoch": 1.487229862475442,
+      "grad_norm": 0.23876847145403102,
+      "learning_rate": 7.088339721085933e-05,
+      "loss": 0.5006,
+      "step": 757
+    },
+    {
+      "epoch": 1.4891944990176817,
+      "grad_norm": 0.26176571776077484,
+      "learning_rate": 7.084849404705245e-05,
+      "loss": 0.4988,
+      "step": 758
+    },
+    {
+      "epoch": 1.4911591355599214,
+      "grad_norm": 0.2511971877200919,
+      "learning_rate": 7.081353282516416e-05,
+      "loss": 0.5076,
+      "step": 759
+    },
+    {
+      "epoch": 1.493123772102161,
+      "grad_norm": 0.27817854693799665,
+      "learning_rate": 7.077851361099281e-05,
+      "loss": 0.4921,
+      "step": 760
+    },
+    {
+      "epoch": 1.4950884086444007,
+      "grad_norm": 0.2977415298818136,
+      "learning_rate": 7.074343647044596e-05,
+      "loss": 0.494,
+      "step": 761
+    },
+    {
+      "epoch": 1.4970530451866404,
+      "grad_norm": 0.2537856591516271,
+      "learning_rate": 7.070830146954018e-05,
+      "loss": 0.5044,
+      "step": 762
+    },
+    {
+      "epoch": 1.4990176817288803,
+      "grad_norm": 0.1868485282789405,
+      "learning_rate": 7.067310867440093e-05,
+      "loss": 0.4959,
+      "step": 763
+    },
+    {
+      "epoch": 1.5009823182711197,
+      "grad_norm": 0.16529385274148198,
+      "learning_rate": 7.063785815126241e-05,
+      "loss": 0.4922,
+      "step": 764
+    },
+    {
+      "epoch": 1.5029469548133596,
+      "grad_norm": 0.19831819720216423,
+      "learning_rate": 7.060254996646752e-05,
+      "loss": 0.5032,
+      "step": 765
+    },
+    {
+      "epoch": 1.5049115913555993,
+      "grad_norm": 0.21810588613338394,
+      "learning_rate": 7.056718418646764e-05,
+      "loss": 0.4918,
+      "step": 766
+    },
+    {
+      "epoch": 1.506876227897839,
+      "grad_norm": 0.2511208433104004,
+      "learning_rate": 7.053176087782256e-05,
+      "loss": 0.5039,
+      "step": 767
+    },
+    {
+      "epoch": 1.5088408644400786,
+      "grad_norm": 0.23286048799673723,
+      "learning_rate": 7.049628010720036e-05,
+      "loss": 0.4968,
+      "step": 768
+    },
+    {
+      "epoch": 1.5108055009823183,
+      "grad_norm": 0.1654880132628785,
+      "learning_rate": 7.046074194137723e-05,
+      "loss": 0.5088,
+      "step": 769
+    },
+    {
+      "epoch": 1.512770137524558,
+      "grad_norm": 0.19181457905688512,
+      "learning_rate": 7.042514644723741e-05,
+      "loss": 0.4982,
+      "step": 770
+    },
+    {
+      "epoch": 1.5147347740667976,
+      "grad_norm": 0.21429566492337535,
+      "learning_rate": 7.038949369177303e-05,
+      "loss": 0.4926,
+      "step": 771
+    },
+    {
+      "epoch": 1.5166994106090375,
+      "grad_norm": 0.1841358271398813,
+      "learning_rate": 7.035378374208395e-05,
+      "loss": 0.4909,
+      "step": 772
+    },
+    {
+      "epoch": 1.518664047151277,
+      "grad_norm": 0.194430466688897,
+      "learning_rate": 7.031801666537772e-05,
+      "loss": 0.4931,
+      "step": 773
+    },
+    {
+      "epoch": 1.5206286836935168,
+      "grad_norm": 0.21802683982516954,
+      "learning_rate": 7.02821925289694e-05,
+      "loss": 0.5021,
+      "step": 774
+    },
+    {
+      "epoch": 1.5225933202357562,
+      "grad_norm": 0.26545437930994875,
+      "learning_rate": 7.024631140028142e-05,
+      "loss": 0.4907,
+      "step": 775
+    },
+    {
+      "epoch": 1.524557956777996,
+      "grad_norm": 0.27258155384694754,
+      "learning_rate": 7.021037334684347e-05,
+      "loss": 0.5015,
+      "step": 776
+    },
+    {
+      "epoch": 1.5265225933202358,
+      "grad_norm": 0.2505409411599532,
+      "learning_rate": 7.017437843629239e-05,
+      "loss": 0.4974,
+      "step": 777
+    },
+    {
+      "epoch": 1.5284872298624754,
+      "grad_norm": 0.24361623032603696,
+      "learning_rate": 7.013832673637202e-05,
+      "loss": 0.502,
+      "step": 778
+    },
+    {
+      "epoch": 1.530451866404715,
+      "grad_norm": 0.24289006605631783,
+      "learning_rate": 7.010221831493307e-05,
+      "loss": 0.4985,
+      "step": 779
+    },
+    {
+      "epoch": 1.5324165029469548,
+      "grad_norm": 0.24058127490318496,
+      "learning_rate": 7.006605323993301e-05,
+      "loss": 0.4998,
+      "step": 780
+    },
+    {
+      "epoch": 1.5343811394891946,
+      "grad_norm": 0.2053738189392177,
+      "learning_rate": 7.002983157943596e-05,
+      "loss": 0.4996,
+      "step": 781
+    },
+    {
+      "epoch": 1.536345776031434,
+      "grad_norm": 0.16113661031694754,
+      "learning_rate": 6.99935534016125e-05,
+      "loss": 0.499,
+      "step": 782
+    },
+    {
+      "epoch": 1.538310412573674,
+      "grad_norm": 0.1707962219357705,
+      "learning_rate": 6.995721877473958e-05,
+      "loss": 0.5018,
+      "step": 783
+    },
+    {
+      "epoch": 1.5402750491159134,
+      "grad_norm": 0.2052561331621828,
+      "learning_rate": 6.992082776720041e-05,
+      "loss": 0.5026,
+      "step": 784
+    },
+    {
+      "epoch": 1.5422396856581533,
+      "grad_norm": 0.2124354014126845,
+      "learning_rate": 6.98843804474843e-05,
+      "loss": 0.5011,
+      "step": 785
+    },
+    {
+      "epoch": 1.544204322200393,
+      "grad_norm": 0.23820032198268148,
+      "learning_rate": 6.984787688418654e-05,
+      "loss": 0.4997,
+      "step": 786
+    },
+    {
+      "epoch": 1.5461689587426326,
+      "grad_norm": 0.26323841170139395,
+      "learning_rate": 6.981131714600828e-05,
+      "loss": 0.494,
+      "step": 787
+    },
+    {
+      "epoch": 1.5481335952848723,
+      "grad_norm": 0.23807103202592525,
+      "learning_rate": 6.977470130175637e-05,
+      "loss": 0.4997,
+      "step": 788
+    },
+    {
+      "epoch": 1.550098231827112,
+      "grad_norm": 0.21080965902956658,
+      "learning_rate": 6.973802942034329e-05,
+      "loss": 0.4909,
+      "step": 789
+    },
+    {
+      "epoch": 1.5520628683693518,
+      "grad_norm": 0.23543385161527478,
+      "learning_rate": 6.970130157078697e-05,
+      "loss": 0.5021,
+      "step": 790
+    },
+    {
+      "epoch": 1.5540275049115913,
+      "grad_norm": 0.24198413881065772,
+      "learning_rate": 6.966451782221065e-05,
+      "loss": 0.5086,
+      "step": 791
+    },
+    {
+      "epoch": 1.5559921414538311,
+      "grad_norm": 0.21029351362142404,
+      "learning_rate": 6.96276782438428e-05,
+      "loss": 0.5022,
+      "step": 792
+    },
+    {
+      "epoch": 1.5579567779960706,
+      "grad_norm": 0.19632328680452757,
+      "learning_rate": 6.959078290501698e-05,
+      "loss": 0.5016,
+      "step": 793
+    },
+    {
+      "epoch": 1.5599214145383105,
+      "grad_norm": 0.1767247278370662,
+      "learning_rate": 6.955383187517164e-05,
+      "loss": 0.5046,
+      "step": 794
+    },
+    {
+      "epoch": 1.5618860510805501,
+      "grad_norm": 0.146721035365349,
+      "learning_rate": 6.95168252238501e-05,
+      "loss": 0.499,
+      "step": 795
+    },
+    {
+      "epoch": 1.5638506876227898,
+      "grad_norm": 0.1886395902734468,
+      "learning_rate": 6.947976302070032e-05,
+      "loss": 0.4987,
+      "step": 796
+    },
+    {
+      "epoch": 1.5658153241650294,
+      "grad_norm": 0.2267903243738179,
+      "learning_rate": 6.944264533547484e-05,
+      "loss": 0.4964,
+      "step": 797
+    },
+    {
+      "epoch": 1.567779960707269,
+      "grad_norm": 0.22482671753173566,
+      "learning_rate": 6.940547223803058e-05,
+      "loss": 0.4939,
+      "step": 798
+    },
+    {
+      "epoch": 1.569744597249509,
+      "grad_norm": 0.22196579288570092,
+      "learning_rate": 6.93682437983288e-05,
+      "loss": 0.5012,
+      "step": 799
+    },
+    {
+      "epoch": 1.5717092337917484,
+      "grad_norm": 0.19342442446331487,
+      "learning_rate": 6.933096008643489e-05,
+      "loss": 0.502,
+      "step": 800
+    },
+    {
+      "epoch": 1.5736738703339883,
+      "grad_norm": 0.17134458200760777,
+      "learning_rate": 6.929362117251826e-05,
+      "loss": 0.4954,
+      "step": 801
+    },
+    {
+      "epoch": 1.5756385068762278,
+      "grad_norm": 0.1476458071115812,
+      "learning_rate": 6.925622712685221e-05,
+      "loss": 0.4993,
+      "step": 802
+    },
+    {
+      "epoch": 1.5776031434184676,
+      "grad_norm": 0.172474842560906,
+      "learning_rate": 6.921877801981382e-05,
+      "loss": 0.4947,
+      "step": 803
+    },
+    {
+      "epoch": 1.5795677799607073,
+      "grad_norm": 0.1946751525069121,
+      "learning_rate": 6.918127392188377e-05,
+      "loss": 0.5046,
+      "step": 804
+    },
+    {
+      "epoch": 1.581532416502947,
+      "grad_norm": 0.20512738552116747,
+      "learning_rate": 6.914371490364625e-05,
+      "loss": 0.4914,
+      "step": 805
+    },
+    {
+      "epoch": 1.5834970530451866,
+      "grad_norm": 0.2740885126467242,
+      "learning_rate": 6.910610103578884e-05,
+      "loss": 0.502,
+      "step": 806
+    },
+    {
+      "epoch": 1.5854616895874263,
+      "grad_norm": 0.32297934959405633,
+      "learning_rate": 6.906843238910229e-05,
+      "loss": 0.4989,
+      "step": 807
+    },
+    {
+      "epoch": 1.5874263261296662,
+      "grad_norm": 0.3582435216880345,
+      "learning_rate": 6.90307090344805e-05,
+      "loss": 0.5055,
+      "step": 808
+    },
+    {
+      "epoch": 1.5893909626719056,
+      "grad_norm": 0.39000898188809235,
+      "learning_rate": 6.899293104292028e-05,
+      "loss": 0.5004,
+      "step": 809
+    },
+    {
+      "epoch": 1.5913555992141455,
+      "grad_norm": 0.3681960190079745,
+      "learning_rate": 6.895509848552134e-05,
+      "loss": 0.4897,
+      "step": 810
+    },
+    {
+      "epoch": 1.593320235756385,
+      "grad_norm": 0.3466764390615321,
+      "learning_rate": 6.891721143348601e-05,
+      "loss": 0.4921,
+      "step": 811
+    },
+    {
+      "epoch": 1.5952848722986248,
+      "grad_norm": 0.3197632180761977,
+      "learning_rate": 6.887926995811924e-05,
+      "loss": 0.4957,
+      "step": 812
+    },
+    {
+      "epoch": 1.5972495088408645,
+      "grad_norm": 0.2395005220946794,
+      "learning_rate": 6.884127413082836e-05,
+      "loss": 0.4945,
+      "step": 813
+    },
+    {
+      "epoch": 1.5992141453831041,
+      "grad_norm": 0.20007965979649706,
+      "learning_rate": 6.880322402312302e-05,
+      "loss": 0.5083,
+      "step": 814
+    },
+    {
+      "epoch": 1.6011787819253438,
+      "grad_norm": 0.23659012436301155,
+      "learning_rate": 6.876511970661505e-05,
+      "loss": 0.5053,
+      "step": 815
+    },
+    {
+      "epoch": 1.6031434184675835,
+      "grad_norm": 0.30617017587539447,
+      "learning_rate": 6.872696125301823e-05,
+      "loss": 0.4986,
+      "step": 816
+    },
+    {
+      "epoch": 1.6051080550098233,
+      "grad_norm": 0.2883936898282049,
+      "learning_rate": 6.86887487341483e-05,
+      "loss": 0.5038,
+      "step": 817
+    },
+    {
+      "epoch": 1.6070726915520628,
+      "grad_norm": 0.21674897515034078,
+      "learning_rate": 6.86504822219227e-05,
+      "loss": 0.507,
+      "step": 818
+    },
+    {
+      "epoch": 1.6090373280943027,
+      "grad_norm": 0.20153538060665316,
+      "learning_rate": 6.861216178836055e-05,
+      "loss": 0.5065,
+      "step": 819
+    },
+    {
+      "epoch": 1.611001964636542,
+      "grad_norm": 0.2250399391416745,
+      "learning_rate": 6.85737875055824e-05,
+      "loss": 0.4965,
+      "step": 820
+    },
+    {
+      "epoch": 1.612966601178782,
+      "grad_norm": 0.25041801993208246,
+      "learning_rate": 6.853535944581017e-05,
+      "loss": 0.5036,
+      "step": 821
+    },
+    {
+      "epoch": 1.6149312377210217,
+      "grad_norm": 0.2050118236808713,
+      "learning_rate": 6.849687768136697e-05,
+      "loss": 0.5028,
+      "step": 822
+    },
+    {
+      "epoch": 1.6168958742632613,
+      "grad_norm": 0.19633027646288798,
+      "learning_rate": 6.845834228467702e-05,
+      "loss": 0.5093,
+      "step": 823
+    },
+    {
+      "epoch": 1.618860510805501,
+      "grad_norm": 0.20285092137651603,
+      "learning_rate": 6.841975332826542e-05,
+      "loss": 0.503,
+      "step": 824
+    },
+    {
+      "epoch": 1.6208251473477406,
+      "grad_norm": 0.18327150858981137,
+      "learning_rate": 6.838111088475812e-05,
+      "loss": 0.4999,
+      "step": 825
+    },
+    {
+      "epoch": 1.6227897838899805,
+      "grad_norm": 0.19555081272198188,
+      "learning_rate": 6.834241502688173e-05,
+      "loss": 0.4916,
+      "step": 826
+    },
+    {
+      "epoch": 1.62475442043222,
+      "grad_norm": 0.19095604749502812,
+      "learning_rate": 6.830366582746338e-05,
+      "loss": 0.5042,
+      "step": 827
+    },
+    {
+      "epoch": 1.6267190569744598,
+      "grad_norm": 0.173263919448087,
+      "learning_rate": 6.82648633594306e-05,
+      "loss": 0.5115,
+      "step": 828
+    },
+    {
+      "epoch": 1.6286836935166993,
+      "grad_norm": 0.14205837547020572,
+      "learning_rate": 6.822600769581113e-05,
+      "loss": 0.4947,
+      "step": 829
+    },
+    {
+      "epoch": 1.6306483300589392,
+      "grad_norm": 0.16063390892357368,
+      "learning_rate": 6.818709890973289e-05,
+      "loss": 0.4901,
+      "step": 830
+    },
+    {
+      "epoch": 1.6326129666011788,
+      "grad_norm": 0.18164334214198788,
+      "learning_rate": 6.814813707442372e-05,
+      "loss": 0.501,
+      "step": 831
+    },
+    {
+      "epoch": 1.6345776031434185,
+      "grad_norm": 0.17741413588712257,
+      "learning_rate": 6.810912226321136e-05,
+      "loss": 0.5006,
+      "step": 832
+    },
+    {
+      "epoch": 1.6365422396856582,
+      "grad_norm": 0.18423752528455,
+      "learning_rate": 6.807005454952321e-05,
+      "loss": 0.4963,
+      "step": 833
+    },
+    {
+      "epoch": 1.6385068762278978,
+      "grad_norm": 0.23238103114536207,
+      "learning_rate": 6.803093400688626e-05,
+      "loss": 0.5071,
+      "step": 834
+    },
+    {
+      "epoch": 1.6404715127701375,
+      "grad_norm": 0.2295684530019606,
+      "learning_rate": 6.799176070892689e-05,
+      "loss": 0.5018,
+      "step": 835
+    },
+    {
+      "epoch": 1.6424361493123771,
+      "grad_norm": 0.1956788004609701,
+      "learning_rate": 6.79525347293708e-05,
+      "loss": 0.4975,
+      "step": 836
+    },
+    {
+      "epoch": 1.644400785854617,
+      "grad_norm": 0.15624752744306217,
+      "learning_rate": 6.791325614204284e-05,
+      "loss": 0.4988,
+      "step": 837
+    },
+    {
+      "epoch": 1.6463654223968565,
+      "grad_norm": 0.1763850939757464,
+      "learning_rate": 6.787392502086682e-05,
+      "loss": 0.489,
+      "step": 838
+    },
+    {
+      "epoch": 1.6483300589390963,
+      "grad_norm": 0.24906642043640756,
+      "learning_rate": 6.78345414398655e-05,
+      "loss": 0.5018,
+      "step": 839
+    },
+    {
+      "epoch": 1.650294695481336,
+      "grad_norm": 0.24378414994446534,
+      "learning_rate": 6.779510547316034e-05,
+      "loss": 0.4952,
+      "step": 840
+    },
+    {
+      "epoch": 1.6522593320235757,
+      "grad_norm": 0.19071551341738843,
+      "learning_rate": 6.775561719497135e-05,
+      "loss": 0.5059,
+      "step": 841
+    },
+    {
+      "epoch": 1.6542239685658153,
+      "grad_norm": 0.21836802149451703,
+      "learning_rate": 6.771607667961704e-05,
+      "loss": 0.5004,
+      "step": 842
+    },
+    {
+      "epoch": 1.656188605108055,
+      "grad_norm": 0.2232932731727261,
+      "learning_rate": 6.767648400151421e-05,
+      "loss": 0.4884,
+      "step": 843
+    },
+    {
+      "epoch": 1.6581532416502947,
+      "grad_norm": 0.2187551319348705,
+      "learning_rate": 6.763683923517785e-05,
+      "loss": 0.4968,
+      "step": 844
+    },
+    {
+      "epoch": 1.6601178781925343,
+      "grad_norm": 0.2613763960900517,
+      "learning_rate": 6.759714245522098e-05,
+      "loss": 0.5074,
+      "step": 845
+    },
+    {
+      "epoch": 1.6620825147347742,
+      "grad_norm": 0.2940470465679677,
+      "learning_rate": 6.755739373635449e-05,
+      "loss": 0.4972,
+      "step": 846
+    },
+    {
+      "epoch": 1.6640471512770136,
+      "grad_norm": 0.2703758793038654,
+      "learning_rate": 6.751759315338704e-05,
+      "loss": 0.4934,
+      "step": 847
+    },
+    {
+      "epoch": 1.6660117878192535,
+      "grad_norm": 0.2213105225785526,
+      "learning_rate": 6.747774078122489e-05,
+      "loss": 0.5,
+      "step": 848
+    },
+    {
+      "epoch": 1.667976424361493,
+      "grad_norm": 0.2193403159203196,
+      "learning_rate": 6.743783669487178e-05,
+      "loss": 0.5006,
+      "step": 849
+    },
+    {
+      "epoch": 1.6699410609037328,
+      "grad_norm": 0.25147609638161816,
+      "learning_rate": 6.739788096942879e-05,
+      "loss": 0.5022,
+      "step": 850
+    },
+    {
+      "epoch": 1.6719056974459725,
+      "grad_norm": 0.19736020589663342,
+      "learning_rate": 6.735787368009413e-05,
+      "loss": 0.4925,
+      "step": 851
+    },
+    {
+      "epoch": 1.6738703339882122,
+      "grad_norm": 0.17371790314939126,
+      "learning_rate": 6.731781490216313e-05,
+      "loss": 0.5062,
+      "step": 852
+    },
+    {
+      "epoch": 1.6758349705304518,
+      "grad_norm": 0.18566986006225078,
+      "learning_rate": 6.727770471102796e-05,
+      "loss": 0.4932,
+      "step": 853
+    },
+    {
+      "epoch": 1.6777996070726915,
+      "grad_norm": 0.20724633716574345,
+      "learning_rate": 6.723754318217762e-05,
+      "loss": 0.5013,
+      "step": 854
+    },
+    {
+      "epoch": 1.6797642436149314,
+      "grad_norm": 0.21118434845773004,
+      "learning_rate": 6.719733039119765e-05,
+      "loss": 0.5051,
+      "step": 855
+    },
+    {
+      "epoch": 1.6817288801571708,
+      "grad_norm": 0.2279390628208479,
+      "learning_rate": 6.715706641377015e-05,
+      "loss": 0.4914,
+      "step": 856
+    },
+    {
+      "epoch": 1.6836935166994107,
+      "grad_norm": 0.26548061607811874,
+      "learning_rate": 6.711675132567346e-05,
+      "loss": 0.4971,
+      "step": 857
+    },
+    {
+      "epoch": 1.6856581532416501,
+      "grad_norm": 0.2760960833608819,
+      "learning_rate": 6.70763852027822e-05,
+      "loss": 0.5009,
+      "step": 858
+    },
+    {
+      "epoch": 1.68762278978389,
+      "grad_norm": 0.2794622965370606,
+      "learning_rate": 6.703596812106698e-05,
+      "loss": 0.5047,
+      "step": 859
+    },
+    {
+      "epoch": 1.6895874263261297,
+      "grad_norm": 0.2152336077404678,
+      "learning_rate": 6.699550015659437e-05,
+      "loss": 0.4895,
+      "step": 860
+    },
+    {
+      "epoch": 1.6915520628683693,
+      "grad_norm": 0.17463832848397992,
+      "learning_rate": 6.695498138552666e-05,
+      "loss": 0.5003,
+      "step": 861
+    },
+    {
+      "epoch": 1.693516699410609,
+      "grad_norm": 0.15164661453400033,
+      "learning_rate": 6.691441188412178e-05,
+      "loss": 0.4942,
+      "step": 862
+    },
+    {
+      "epoch": 1.6954813359528487,
+      "grad_norm": 0.1598792875422971,
+      "learning_rate": 6.68737917287331e-05,
+      "loss": 0.4926,
+      "step": 863
+    },
+    {
+      "epoch": 1.6974459724950886,
+      "grad_norm": 0.1742876518372632,
+      "learning_rate": 6.68331209958094e-05,
+      "loss": 0.4968,
+      "step": 864
+    },
+    {
+      "epoch": 1.699410609037328,
+      "grad_norm": 0.20120432665629762,
+      "learning_rate": 6.679239976189457e-05,
+      "loss": 0.5037,
+      "step": 865
+    },
+    {
+      "epoch": 1.7013752455795679,
+      "grad_norm": 0.19254678806986522,
+      "learning_rate": 6.675162810362759e-05,
+      "loss": 0.5006,
+      "step": 866
+    },
+    {
+      "epoch": 1.7033398821218073,
+      "grad_norm": 0.17636170558976336,
+      "learning_rate": 6.671080609774231e-05,
+      "loss": 0.499,
+      "step": 867
+    },
+    {
+      "epoch": 1.7053045186640472,
+      "grad_norm": 0.2803302842794596,
+      "learning_rate": 6.666993382106738e-05,
+      "loss": 0.4968,
+      "step": 868
+    },
+    {
+      "epoch": 1.7072691552062869,
+      "grad_norm": 0.2916359446725541,
+      "learning_rate": 6.662901135052602e-05,
+      "loss": 0.4985,
+      "step": 869
+    },
+    {
+      "epoch": 1.7092337917485265,
+      "grad_norm": 0.19176699493440968,
+      "learning_rate": 6.658803876313592e-05,
+      "loss": 0.502,
+      "step": 870
+    },
+    {
+      "epoch": 1.7111984282907662,
+      "grad_norm": 0.18855058662688448,
+      "learning_rate": 6.654701613600914e-05,
+      "loss": 0.5031,
+      "step": 871
+    },
+    {
+      "epoch": 1.7131630648330058,
+      "grad_norm": 0.1706079959804198,
+      "learning_rate": 6.650594354635185e-05,
+      "loss": 0.4977,
+      "step": 872
+    },
+    {
+      "epoch": 1.7151277013752457,
+      "grad_norm": 0.1670730589141051,
+      "learning_rate": 6.646482107146428e-05,
+      "loss": 0.4984,
+      "step": 873
+    },
+    {
+      "epoch": 1.7170923379174852,
+      "grad_norm": 0.18208579900017552,
+      "learning_rate": 6.642364878874057e-05,
+      "loss": 0.4907,
+      "step": 874
+    },
+    {
+      "epoch": 1.719056974459725,
+      "grad_norm": 0.2052151584348083,
+      "learning_rate": 6.638242677566856e-05,
+      "loss": 0.505,
+      "step": 875
+    },
+    {
+      "epoch": 1.7210216110019645,
+      "grad_norm": 0.22086857790655193,
+      "learning_rate": 6.634115510982973e-05,
+      "loss": 0.4992,
+      "step": 876
+    },
+    {
+      "epoch": 1.7229862475442044,
+      "grad_norm": 0.1852540379797261,
+      "learning_rate": 6.629983386889896e-05,
+      "loss": 0.501,
+      "step": 877
+    },
+    {
+      "epoch": 1.724950884086444,
+      "grad_norm": 0.17493980669399972,
+      "learning_rate": 6.625846313064445e-05,
+      "loss": 0.4908,
+      "step": 878
+    },
+    {
+      "epoch": 1.7269155206286837,
+      "grad_norm": 0.2025932387684324,
+      "learning_rate": 6.621704297292755e-05,
+      "loss": 0.4881,
+      "step": 879
+    },
+    {
+      "epoch": 1.7288801571709234,
+      "grad_norm": 0.19212483758987992,
+      "learning_rate": 6.617557347370266e-05,
+      "loss": 0.493,
+      "step": 880
+    },
+    {
+      "epoch": 1.730844793713163,
+      "grad_norm": 0.211978851409156,
+      "learning_rate": 6.613405471101701e-05,
+      "loss": 0.4985,
+      "step": 881
+    },
+    {
+      "epoch": 1.732809430255403,
+      "grad_norm": 0.21418683490304818,
+      "learning_rate": 6.609248676301052e-05,
+      "loss": 0.4997,
+      "step": 882
+    },
+    {
+      "epoch": 1.7347740667976423,
+      "grad_norm": 0.2028670871278396,
+      "learning_rate": 6.605086970791572e-05,
+      "loss": 0.4966,
+      "step": 883
+    },
+    {
+      "epoch": 1.7367387033398822,
+      "grad_norm": 0.2112859186816638,
+      "learning_rate": 6.600920362405755e-05,
+      "loss": 0.5086,
+      "step": 884
+    },
+    {
+      "epoch": 1.7387033398821217,
+      "grad_norm": 0.22161558923283467,
+      "learning_rate": 6.596748858985323e-05,
+      "loss": 0.495,
+      "step": 885
+    },
+    {
+      "epoch": 1.7406679764243616,
+      "grad_norm": 0.1940400908993572,
+      "learning_rate": 6.592572468381207e-05,
+      "loss": 0.5021,
+      "step": 886
+    },
+    {
+      "epoch": 1.7426326129666012,
+      "grad_norm": 0.18702056500441575,
+      "learning_rate": 6.588391198453542e-05,
+      "loss": 0.5007,
+      "step": 887
+    },
+    {
+      "epoch": 1.7445972495088409,
+      "grad_norm": 0.23755070202432965,
+      "learning_rate": 6.584205057071641e-05,
+      "loss": 0.4964,
+      "step": 888
+    },
+    {
+      "epoch": 1.7465618860510805,
+      "grad_norm": 0.23383633105017934,
+      "learning_rate": 6.580014052113986e-05,
+      "loss": 0.4885,
+      "step": 889
+    },
+    {
+      "epoch": 1.7485265225933202,
+      "grad_norm": 0.19019165339664384,
+      "learning_rate": 6.575818191468217e-05,
+      "loss": 0.4879,
+      "step": 890
+    },
+    {
+      "epoch": 1.75049115913556,
+      "grad_norm": 0.17023725160727154,
+      "learning_rate": 6.571617483031105e-05,
+      "loss": 0.4994,
+      "step": 891
+    },
+    {
+      "epoch": 1.7524557956777995,
+      "grad_norm": 0.17781702551711653,
+      "learning_rate": 6.567411934708553e-05,
+      "loss": 0.5046,
+      "step": 892
+    },
+    {
+      "epoch": 1.7544204322200394,
+      "grad_norm": 0.20008062613492061,
+      "learning_rate": 6.563201554415565e-05,
+      "loss": 0.5022,
+      "step": 893
+    },
+    {
+      "epoch": 1.7563850687622788,
+      "grad_norm": 0.21059268924698948,
+      "learning_rate": 6.558986350076243e-05,
+      "loss": 0.5008,
+      "step": 894
+    },
+    {
+      "epoch": 1.7583497053045187,
+      "grad_norm": 0.22808595220918323,
+      "learning_rate": 6.55476632962377e-05,
+      "loss": 0.4994,
+      "step": 895
+    },
+    {
+      "epoch": 1.7603143418467584,
+      "grad_norm": 0.2390588625568595,
+      "learning_rate": 6.550541501000389e-05,
+      "loss": 0.484,
+      "step": 896
+    },
+    {
+      "epoch": 1.762278978388998,
+      "grad_norm": 0.21397114274430615,
+      "learning_rate": 6.546311872157395e-05,
+      "loss": 0.4942,
+      "step": 897
+    },
+    {
+      "epoch": 1.7642436149312377,
+      "grad_norm": 0.20777807519303235,
+      "learning_rate": 6.542077451055116e-05,
+      "loss": 0.5033,
+      "step": 898
+    },
+    {
+      "epoch": 1.7662082514734774,
+      "grad_norm": 0.20178552927038496,
+      "learning_rate": 6.537838245662897e-05,
+      "loss": 0.4993,
+      "step": 899
+    },
+    {
+      "epoch": 1.768172888015717,
+      "grad_norm": 0.1783938660353064,
+      "learning_rate": 6.533594263959095e-05,
+      "loss": 0.4968,
+      "step": 900
+    },
+    {
+      "epoch": 1.7701375245579567,
+      "grad_norm": 0.1725356543795445,
+      "learning_rate": 6.529345513931044e-05,
+      "loss": 0.4933,
+      "step": 901
+    },
+    {
+      "epoch": 1.7721021611001966,
+      "grad_norm": 0.17352955007702606,
+      "learning_rate": 6.525092003575064e-05,
+      "loss": 0.4997,
+      "step": 902
+    },
+    {
+      "epoch": 1.774066797642436,
+      "grad_norm": 0.16239623795066715,
+      "learning_rate": 6.520833740896427e-05,
+      "loss": 0.5008,
+      "step": 903
+    },
+    {
+      "epoch": 1.776031434184676,
+      "grad_norm": 0.16219867242487346,
+      "learning_rate": 6.516570733909351e-05,
+      "loss": 0.491,
+      "step": 904
+    },
+    {
+      "epoch": 1.7779960707269156,
+      "grad_norm": 0.15782128651237606,
+      "learning_rate": 6.512302990636987e-05,
+      "loss": 0.4977,
+      "step": 905
+    },
+    {
+      "epoch": 1.7799607072691552,
+      "grad_norm": 0.14270734953350722,
+      "learning_rate": 6.50803051911139e-05,
+      "loss": 0.5007,
+      "step": 906
+    },
+    {
+      "epoch": 1.781925343811395,
+      "grad_norm": 0.1642850351494911,
+      "learning_rate": 6.503753327373522e-05,
+      "loss": 0.5016,
+      "step": 907
+    },
+    {
+      "epoch": 1.7838899803536346,
+      "grad_norm": 0.17894335961697497,
+      "learning_rate": 6.499471423473228e-05,
+      "loss": 0.5013,
+      "step": 908
+    },
+    {
+      "epoch": 1.7858546168958742,
+      "grad_norm": 0.1710581206959307,
+      "learning_rate": 6.495184815469219e-05,
+      "loss": 0.4973,
+      "step": 909
+    },
+    {
+      "epoch": 1.7878192534381139,
+      "grad_norm": 0.15840707775473137,
+      "learning_rate": 6.49089351142906e-05,
+      "loss": 0.5055,
+      "step": 910
+    },
+    {
+      "epoch": 1.7897838899803538,
+      "grad_norm": 0.15351898571979672,
+      "learning_rate": 6.486597519429154e-05,
+      "loss": 0.4986,
+      "step": 911
+    },
+    {
+      "epoch": 1.7917485265225932,
+      "grad_norm": 0.1969399579068348,
+      "learning_rate": 6.482296847554729e-05,
+      "loss": 0.4943,
+      "step": 912
+    },
+    {
+      "epoch": 1.793713163064833,
+      "grad_norm": 0.19121110171995043,
+      "learning_rate": 6.477991503899816e-05,
+      "loss": 0.4996,
+      "step": 913
+    },
+    {
+      "epoch": 1.7956777996070727,
+      "grad_norm": 0.20236379749970387,
+      "learning_rate": 6.473681496567243e-05,
+      "loss": 0.5004,
+      "step": 914
+    },
+    {
+      "epoch": 1.7976424361493124,
+      "grad_norm": 0.22865105662914953,
+      "learning_rate": 6.469366833668615e-05,
+      "loss": 0.4937,
+      "step": 915
+    },
+    {
+      "epoch": 1.799607072691552,
+      "grad_norm": 0.2158947905590702,
+      "learning_rate": 6.465047523324295e-05,
+      "loss": 0.5062,
+      "step": 916
+    },
+    {
+      "epoch": 1.8015717092337917,
+      "grad_norm": 0.20124698533758373,
+      "learning_rate": 6.460723573663396e-05,
+      "loss": 0.4932,
+      "step": 917
+    },
+    {
+      "epoch": 1.8035363457760314,
+      "grad_norm": 0.16898222755432327,
+      "learning_rate": 6.456394992823763e-05,
+      "loss": 0.4969,
+      "step": 918
+    },
+    {
+      "epoch": 1.805500982318271,
+      "grad_norm": 0.20279864932380962,
+      "learning_rate": 6.452061788951955e-05,
+      "loss": 0.4932,
+      "step": 919
+    },
+    {
+      "epoch": 1.807465618860511,
+      "grad_norm": 0.205668649298984,
+      "learning_rate": 6.447723970203232e-05,
+      "loss": 0.4944,
+      "step": 920
+    },
+    {
+      "epoch": 1.8094302554027504,
+      "grad_norm": 0.21160168974466426,
+      "learning_rate": 6.44338154474154e-05,
+      "loss": 0.4939,
+      "step": 921
+    },
+    {
+      "epoch": 1.8113948919449903,
+      "grad_norm": 0.2142298036654362,
+      "learning_rate": 6.439034520739495e-05,
+      "loss": 0.5078,
+      "step": 922
+    },
+    {
+      "epoch": 1.8133595284872297,
+      "grad_norm": 0.2074686847577365,
+      "learning_rate": 6.434682906378367e-05,
+      "loss": 0.4987,
+      "step": 923
+    },
+    {
+      "epoch": 1.8153241650294696,
+      "grad_norm": 0.19698160548087817,
+      "learning_rate": 6.430326709848067e-05,
+      "loss": 0.4917,
+      "step": 924
+    },
+    {
+      "epoch": 1.8172888015717092,
+      "grad_norm": 0.18034226089855415,
+      "learning_rate": 6.425965939347127e-05,
+      "loss": 0.496,
+      "step": 925
+    },
+    {
+      "epoch": 1.819253438113949,
+      "grad_norm": 0.16532494709169768,
+      "learning_rate": 6.42160060308269e-05,
+      "loss": 0.4971,
+      "step": 926
+    },
+    {
+      "epoch": 1.8212180746561886,
+      "grad_norm": 0.17204295841785322,
+      "learning_rate": 6.417230709270491e-05,
+      "loss": 0.5035,
+      "step": 927
+    },
+    {
+      "epoch": 1.8231827111984282,
+      "grad_norm": 0.16646169297778157,
+      "learning_rate": 6.41285626613484e-05,
+      "loss": 0.5025,
+      "step": 928
+    },
+    {
+      "epoch": 1.8251473477406681,
+      "grad_norm": 0.1807507044007825,
+      "learning_rate": 6.408477281908616e-05,
+      "loss": 0.4951,
+      "step": 929
+    },
+    {
+      "epoch": 1.8271119842829076,
+      "grad_norm": 0.1636565466055324,
+      "learning_rate": 6.404093764833235e-05,
+      "loss": 0.4892,
+      "step": 930
+    },
+    {
+      "epoch": 1.8290766208251474,
+      "grad_norm": 0.1578839405574448,
+      "learning_rate": 6.399705723158651e-05,
+      "loss": 0.4924,
+      "step": 931
+    },
+    {
+      "epoch": 1.8310412573673869,
+      "grad_norm": 0.17970880518174964,
+      "learning_rate": 6.395313165143333e-05,
+      "loss": 0.4993,
+      "step": 932
+    },
+    {
+      "epoch": 1.8330058939096268,
+      "grad_norm": 0.21130020368553237,
+      "learning_rate": 6.390916099054244e-05,
+      "loss": 0.4903,
+      "step": 933
+    },
+    {
+      "epoch": 1.8349705304518664,
+      "grad_norm": 0.21051079066304001,
+      "learning_rate": 6.386514533166841e-05,
+      "loss": 0.4955,
+      "step": 934
+    },
+    {
+      "epoch": 1.836935166994106,
+      "grad_norm": 0.20380388830810234,
+      "learning_rate": 6.382108475765039e-05,
+      "loss": 0.4947,
+      "step": 935
+    },
+    {
+      "epoch": 1.8388998035363457,
+      "grad_norm": 0.1739356370374381,
+      "learning_rate": 6.377697935141216e-05,
+      "loss": 0.4987,
+      "step": 936
+    },
+    {
+      "epoch": 1.8408644400785854,
+      "grad_norm": 0.15002657964535593,
+      "learning_rate": 6.373282919596182e-05,
+      "loss": 0.5051,
+      "step": 937
+    },
+    {
+      "epoch": 1.8428290766208253,
+      "grad_norm": 0.1808463768238514,
+      "learning_rate": 6.368863437439168e-05,
+      "loss": 0.4966,
+      "step": 938
+    },
+    {
+      "epoch": 1.8447937131630647,
+      "grad_norm": 0.19022920550066025,
+      "learning_rate": 6.364439496987818e-05,
+      "loss": 0.4977,
+      "step": 939
+    },
+    {
+      "epoch": 1.8467583497053046,
+      "grad_norm": 0.1851514162542868,
+      "learning_rate": 6.360011106568158e-05,
+      "loss": 0.4957,
+      "step": 940
+    },
+    {
+      "epoch": 1.848722986247544,
+      "grad_norm": 0.1980584366284314,
+      "learning_rate": 6.355578274514595e-05,
+      "loss": 0.5012,
+      "step": 941
+    },
+    {
+      "epoch": 1.850687622789784,
+      "grad_norm": 0.1823050106734708,
+      "learning_rate": 6.351141009169893e-05,
+      "loss": 0.4919,
+      "step": 942
+    },
+    {
+      "epoch": 1.8526522593320236,
+      "grad_norm": 0.13141856369301622,
+      "learning_rate": 6.346699318885161e-05,
+      "loss": 0.4906,
+      "step": 943
+    },
+    {
+      "epoch": 1.8546168958742633,
+      "grad_norm": 0.17640918084975385,
+      "learning_rate": 6.342253212019836e-05,
+      "loss": 0.4951,
+      "step": 944
+    },
+    {
+      "epoch": 1.856581532416503,
+      "grad_norm": 0.19924669762693184,
+      "learning_rate": 6.337802696941665e-05,
+      "loss": 0.4956,
+      "step": 945
+    },
+    {
+      "epoch": 1.8585461689587426,
+      "grad_norm": 0.20181370881693353,
+      "learning_rate": 6.333347782026693e-05,
+      "loss": 0.4943,
+      "step": 946
+    },
+    {
+      "epoch": 1.8605108055009825,
+      "grad_norm": 0.1840950836520622,
+      "learning_rate": 6.328888475659245e-05,
+      "loss": 0.4922,
+      "step": 947
+    },
+    {
+      "epoch": 1.862475442043222,
+      "grad_norm": 0.1681711394721591,
+      "learning_rate": 6.324424786231912e-05,
+      "loss": 0.4849,
+      "step": 948
+    },
+    {
+      "epoch": 1.8644400785854618,
+      "grad_norm": 0.18735252486316487,
+      "learning_rate": 6.319956722145534e-05,
+      "loss": 0.4863,
+      "step": 949
+    },
+    {
+      "epoch": 1.8664047151277012,
+      "grad_norm": 0.17317904380907806,
+      "learning_rate": 6.315484291809181e-05,
+      "loss": 0.5047,
+      "step": 950
+    },
+    {
+      "epoch": 1.8683693516699411,
+      "grad_norm": 0.11579891903876761,
+      "learning_rate": 6.311007503640147e-05,
+      "loss": 0.4963,
+      "step": 951
+    },
+    {
+      "epoch": 1.8703339882121808,
+      "grad_norm": 0.163413055391194,
+      "learning_rate": 6.306526366063921e-05,
+      "loss": 0.4935,
+      "step": 952
+    },
+    {
+      "epoch": 1.8722986247544204,
+      "grad_norm": 0.17234060023578593,
+      "learning_rate": 6.302040887514182e-05,
+      "loss": 0.4953,
+      "step": 953
+    },
+    {
+      "epoch": 1.87426326129666,
+      "grad_norm": 0.15685700782999856,
+      "learning_rate": 6.297551076432775e-05,
+      "loss": 0.5019,
+      "step": 954
+    },
+    {
+      "epoch": 1.8762278978388998,
+      "grad_norm": 0.18093739573484724,
+      "learning_rate": 6.293056941269703e-05,
+      "loss": 0.4989,
+      "step": 955
+    },
+    {
+      "epoch": 1.8781925343811396,
+      "grad_norm": 0.19153457701660087,
+      "learning_rate": 6.288558490483108e-05,
+      "loss": 0.4995,
+      "step": 956
+    },
+    {
+      "epoch": 1.880157170923379,
+      "grad_norm": 0.2091823532615773,
+      "learning_rate": 6.284055732539246e-05,
+      "loss": 0.5057,
+      "step": 957
+    },
+    {
+      "epoch": 1.882121807465619,
+      "grad_norm": 0.23923890138139592,
+      "learning_rate": 6.279548675912491e-05,
+      "loss": 0.496,
+      "step": 958
+    },
+    {
+      "epoch": 1.8840864440078584,
+      "grad_norm": 0.21760579720448137,
+      "learning_rate": 6.275037329085296e-05,
+      "loss": 0.4968,
+      "step": 959
+    },
+    {
+      "epoch": 1.8860510805500983,
+      "grad_norm": 0.16184755756206792,
+      "learning_rate": 6.270521700548196e-05,
+      "loss": 0.4891,
+      "step": 960
+    },
+    {
+      "epoch": 1.888015717092338,
+      "grad_norm": 0.16189748424387887,
+      "learning_rate": 6.266001798799781e-05,
+      "loss": 0.4937,
+      "step": 961
+    },
+    {
+      "epoch": 1.8899803536345776,
+      "grad_norm": 0.20309990264901226,
+      "learning_rate": 6.261477632346686e-05,
+      "loss": 0.4964,
+      "step": 962
+    },
+    {
+      "epoch": 1.8919449901768173,
+      "grad_norm": 0.2237965447521191,
+      "learning_rate": 6.256949209703572e-05,
+      "loss": 0.498,
+      "step": 963
+    },
+    {
+      "epoch": 1.893909626719057,
+      "grad_norm": 0.20574001666472336,
+      "learning_rate": 6.252416539393101e-05,
+      "loss": 0.4926,
+      "step": 964
+    },
+    {
+      "epoch": 1.8958742632612968,
+      "grad_norm": 0.21595553554835373,
+      "learning_rate": 6.247879629945946e-05,
+      "loss": 0.4886,
+      "step": 965
+    },
+    {
+      "epoch": 1.8978388998035363,
+      "grad_norm": 0.24370133627201063,
+      "learning_rate": 6.243338489900745e-05,
+      "loss": 0.4925,
+      "step": 966
+    },
+    {
+      "epoch": 1.8998035363457761,
+      "grad_norm": 0.19304381252040684,
+      "learning_rate": 6.238793127804101e-05,
+      "loss": 0.5029,
+      "step": 967
+    },
+    {
+      "epoch": 1.9017681728880156,
+      "grad_norm": 0.16562969176103354,
+      "learning_rate": 6.234243552210569e-05,
+      "loss": 0.5028,
+      "step": 968
+    },
+    {
+      "epoch": 1.9037328094302555,
+      "grad_norm": 0.20469025861128032,
+      "learning_rate": 6.229689771682626e-05,
+      "loss": 0.5077,
+      "step": 969
+    },
+    {
+      "epoch": 1.9056974459724951,
+      "grad_norm": 0.2646601187359145,
+      "learning_rate": 6.22513179479067e-05,
+      "loss": 0.503,
+      "step": 970
+    },
+    {
+      "epoch": 1.9076620825147348,
+      "grad_norm": 0.22997374622958316,
+      "learning_rate": 6.220569630112989e-05,
+      "loss": 0.4974,
+      "step": 971
+    },
+    {
+      "epoch": 1.9096267190569745,
+      "grad_norm": 0.15043640767289468,
+      "learning_rate": 6.216003286235758e-05,
+      "loss": 0.4909,
+      "step": 972
+    },
+    {
+      "epoch": 1.9115913555992141,
+      "grad_norm": 0.15541848649552323,
+      "learning_rate": 6.211432771753019e-05,
+      "loss": 0.4968,
+      "step": 973
+    },
+    {
+      "epoch": 1.9135559921414538,
+      "grad_norm": 0.19272155453484002,
+      "learning_rate": 6.206858095266656e-05,
+      "loss": 0.5025,
+      "step": 974
+    },
+    {
+      "epoch": 1.9155206286836934,
+      "grad_norm": 0.18836192545040814,
+      "learning_rate": 6.202279265386394e-05,
+      "loss": 0.4933,
+      "step": 975
+    },
+    {
+      "epoch": 1.9174852652259333,
+      "grad_norm": 0.16223702065997556,
+      "learning_rate": 6.19769629072977e-05,
+      "loss": 0.4993,
+      "step": 976
+    },
+    {
+      "epoch": 1.9194499017681728,
+      "grad_norm": 0.15012151669414617,
+      "learning_rate": 6.19310917992212e-05,
+      "loss": 0.4884,
+      "step": 977
+    },
+    {
+      "epoch": 1.9214145383104126,
+      "grad_norm": 0.15227396812953037,
+      "learning_rate": 6.188517941596573e-05,
+      "loss": 0.4962,
+      "step": 978
+    },
+    {
+      "epoch": 1.9233791748526523,
+      "grad_norm": 0.2018075596838725,
+      "learning_rate": 6.183922584394018e-05,
+      "loss": 0.4997,
+      "step": 979
+    },
+    {
+      "epoch": 1.925343811394892,
+      "grad_norm": 0.21745672138413352,
+      "learning_rate": 6.179323116963095e-05,
+      "loss": 0.4974,
+      "step": 980
+    },
+    {
+      "epoch": 1.9273084479371316,
+      "grad_norm": 0.150688670510925,
+      "learning_rate": 6.174719547960186e-05,
+      "loss": 0.4863,
+      "step": 981
+    },
+    {
+      "epoch": 1.9292730844793713,
+      "grad_norm": 0.134777079403887,
+      "learning_rate": 6.17011188604939e-05,
+      "loss": 0.4899,
+      "step": 982
+    },
+    {
+      "epoch": 1.931237721021611,
+      "grad_norm": 0.13788844461373068,
+      "learning_rate": 6.165500139902507e-05,
+      "loss": 0.4929,
+      "step": 983
+    },
+    {
+      "epoch": 1.9332023575638506,
+      "grad_norm": 0.15296687273669426,
+      "learning_rate": 6.160884318199025e-05,
+      "loss": 0.4969,
+      "step": 984
+    },
+    {
+      "epoch": 1.9351669941060905,
+      "grad_norm": 0.17514694801494543,
+      "learning_rate": 6.156264429626103e-05,
+      "loss": 0.4983,
+      "step": 985
+    },
+    {
+      "epoch": 1.93713163064833,
+      "grad_norm": 0.1690998477623786,
+      "learning_rate": 6.151640482878554e-05,
+      "loss": 0.4953,
+      "step": 986
+    },
+    {
+      "epoch": 1.9390962671905698,
+      "grad_norm": 0.1886104026233183,
+      "learning_rate": 6.147012486658825e-05,
+      "loss": 0.4934,
+      "step": 987
+    },
+    {
+      "epoch": 1.9410609037328095,
+      "grad_norm": 0.1952525031489842,
+      "learning_rate": 6.14238044967699e-05,
+      "loss": 0.502,
+      "step": 988
+    },
+    {
+      "epoch": 1.9430255402750491,
+      "grad_norm": 0.16585381649399858,
+      "learning_rate": 6.137744380650726e-05,
+      "loss": 0.4862,
+      "step": 989
+    },
+    {
+      "epoch": 1.9449901768172888,
+      "grad_norm": 0.13460972718974348,
+      "learning_rate": 6.133104288305293e-05,
+      "loss": 0.4901,
+      "step": 990
+    },
+    {
+      "epoch": 1.9469548133595285,
+      "grad_norm": 0.1644880673539835,
+      "learning_rate": 6.128460181373533e-05,
+      "loss": 0.4937,
+      "step": 991
+    },
+    {
+      "epoch": 1.9489194499017681,
+      "grad_norm": 0.18385580243445426,
+      "learning_rate": 6.123812068595835e-05,
+      "loss": 0.5036,
+      "step": 992
+    },
+    {
+      "epoch": 1.9508840864440078,
+      "grad_norm": 0.14864865300626162,
+      "learning_rate": 6.119159958720132e-05,
+      "loss": 0.4972,
+      "step": 993
+    },
+    {
+      "epoch": 1.9528487229862477,
+      "grad_norm": 0.17408827006905755,
+      "learning_rate": 6.114503860501876e-05,
+      "loss": 0.5028,
+      "step": 994
+    },
+    {
+      "epoch": 1.9548133595284871,
+      "grad_norm": 0.15200870927336663,
+      "learning_rate": 6.109843782704029e-05,
+      "loss": 0.4846,
+      "step": 995
+    },
+    {
+      "epoch": 1.956777996070727,
+      "grad_norm": 0.13577888504246258,
+      "learning_rate": 6.105179734097041e-05,
+      "loss": 0.4969,
+      "step": 996
+    },
+    {
+      "epoch": 1.9587426326129664,
+      "grad_norm": 0.17285281683791986,
+      "learning_rate": 6.100511723458835e-05,
+      "loss": 0.4973,
+      "step": 997
+    },
+    {
+      "epoch": 1.9607072691552063,
+      "grad_norm": 0.19912859126980414,
+      "learning_rate": 6.095839759574791e-05,
+      "loss": 0.5,
+      "step": 998
+    },
+    {
+      "epoch": 1.962671905697446,
+      "grad_norm": 0.19487027371902196,
+      "learning_rate": 6.0911638512377296e-05,
+      "loss": 0.5003,
+      "step": 999
+    },
+    {
+      "epoch": 1.9646365422396856,
+      "grad_norm": 0.18690902909019727,
+      "learning_rate": 6.086484007247894e-05,
+      "loss": 0.5075,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9666011787819253,
+      "grad_norm": 0.20838276435028344,
+      "learning_rate": 6.081800236412935e-05,
+      "loss": 0.4939,
+      "step": 1001
+    },
+    {
+      "epoch": 1.968565815324165,
+      "grad_norm": 0.19092676843251227,
+      "learning_rate": 6.077112547547894e-05,
+      "loss": 0.4932,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9705304518664049,
+      "grad_norm": 0.18535635939755488,
+      "learning_rate": 6.0724209494751865e-05,
+      "loss": 0.4879,
+      "step": 1003
+    },
+    {
+      "epoch": 1.9724950884086443,
+      "grad_norm": 0.1727716545086572,
+      "learning_rate": 6.067725451024583e-05,
+      "loss": 0.4913,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9744597249508842,
+      "grad_norm": 0.1741055629380095,
+      "learning_rate": 6.063026061033201e-05,
+      "loss": 0.5091,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9764243614931236,
+      "grad_norm": 0.17559794880904558,
+      "learning_rate": 6.058322788345473e-05,
+      "loss": 0.4959,
+      "step": 1006
+    },
+    {
+      "epoch": 1.9783889980353635,
+      "grad_norm": 0.17155444709238657,
+      "learning_rate": 6.053615641813146e-05,
+      "loss": 0.501,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9803536345776032,
+      "grad_norm": 0.17697560258127942,
+      "learning_rate": 6.0489046302952544e-05,
+      "loss": 0.4907,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9823182711198428,
+      "grad_norm": 0.16808249865065525,
+      "learning_rate": 6.044189762658107e-05,
+      "loss": 0.4993,
+      "step": 1009
+    },
+    {
+      "epoch": 1.9842829076620825,
+      "grad_norm": 0.15436979494670997,
+      "learning_rate": 6.039471047775273e-05,
+      "loss": 0.4978,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9862475442043221,
+      "grad_norm": 0.15132438996208575,
+      "learning_rate": 6.034748494527557e-05,
+      "loss": 0.5005,
+      "step": 1011
+    },
+    {
+      "epoch": 1.988212180746562,
+      "grad_norm": 0.13379856399531997,
+      "learning_rate": 6.030022111802992e-05,
+      "loss": 0.4966,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9901768172888015,
+      "grad_norm": 0.1442739428935243,
+      "learning_rate": 6.025291908496815e-05,
+      "loss": 0.4949,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9921414538310414,
+      "grad_norm": 0.16156673553743822,
+      "learning_rate": 6.020557893511456e-05,
+      "loss": 0.4952,
+      "step": 1014
+    },
+    {
+      "epoch": 1.9941060903732808,
+      "grad_norm": 0.17357638532186948,
+      "learning_rate": 6.01582007575652e-05,
+      "loss": 0.493,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9960707269155207,
+      "grad_norm": 0.18067623625964732,
+      "learning_rate": 6.011078464148762e-05,
+      "loss": 0.5053,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9980353634577603,
+      "grad_norm": 0.17685360348576856,
+      "learning_rate": 6.006333067612085e-05,
+      "loss": 0.4932,
+      "step": 1017
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.14236230913705675,
+      "learning_rate": 6.0015838950775116e-05,
+      "loss": 0.4954,
+      "step": 1018
+    },
+    {
+      "epoch": 2.00196463654224,
+      "grad_norm": 0.22688241033683693,
+      "learning_rate": 5.9968309554831705e-05,
+      "loss": 0.4466,
+      "step": 1019
+    },
+    {
+      "epoch": 2.0039292730844793,
+      "grad_norm": 0.4117317902489836,
+      "learning_rate": 5.992074257774285e-05,
+      "loss": 0.4471,
+      "step": 1020
+    },
+    {
+      "epoch": 2.005893909626719,
+      "grad_norm": 0.5083450343920801,
+      "learning_rate": 5.9873138109031435e-05,
+      "loss": 0.4564,
+      "step": 1021
+    },
+    {
+      "epoch": 2.0078585461689586,
+      "grad_norm": 0.6281903099965783,
+      "learning_rate": 5.982549623829096e-05,
+      "loss": 0.45,
+      "step": 1022
+    },
+    {
+      "epoch": 2.0098231827111985,
+      "grad_norm": 0.847199856575939,
+      "learning_rate": 5.97778170551853e-05,
+      "loss": 0.464,
+      "step": 1023
+    },
+    {
+      "epoch": 2.011787819253438,
+      "grad_norm": 0.8213778543807939,
+      "learning_rate": 5.9730100649448586e-05,
+      "loss": 0.4566,
+      "step": 1024
+    },
+    {
+      "epoch": 2.013752455795678,
+      "grad_norm": 0.5894869912716305,
+      "learning_rate": 5.9682347110884925e-05,
+      "loss": 0.4578,
+      "step": 1025
+    },
+    {
+      "epoch": 2.0157170923379173,
+      "grad_norm": 0.5943974405041138,
+      "learning_rate": 5.96345565293684e-05,
+      "loss": 0.459,
+      "step": 1026
+    },
+    {
+      "epoch": 2.017681728880157,
+      "grad_norm": 0.46823552571126564,
+      "learning_rate": 5.9586728994842744e-05,
+      "loss": 0.4593,
+      "step": 1027
+    },
+    {
+      "epoch": 2.019646365422397,
+      "grad_norm": 0.4588094678256471,
+      "learning_rate": 5.9538864597321266e-05,
+      "loss": 0.4614,
+      "step": 1028
+    },
+    {
+      "epoch": 2.0216110019646365,
+      "grad_norm": 0.5700432586208983,
+      "learning_rate": 5.9490963426886664e-05,
+      "loss": 0.4589,
+      "step": 1029
+    },
+    {
+      "epoch": 2.0235756385068764,
+      "grad_norm": 0.4609196248144152,
+      "learning_rate": 5.9443025573690795e-05,
+      "loss": 0.4556,
+      "step": 1030
+    },
+    {
+      "epoch": 2.025540275049116,
+      "grad_norm": 0.3639349170652564,
+      "learning_rate": 5.939505112795463e-05,
+      "loss": 0.4487,
+      "step": 1031
+    },
+    {
+      "epoch": 2.0275049115913557,
+      "grad_norm": 0.39946753094341647,
+      "learning_rate": 5.934704017996794e-05,
+      "loss": 0.4464,
+      "step": 1032
+    },
+    {
+      "epoch": 2.029469548133595,
+      "grad_norm": 0.36144765062656514,
+      "learning_rate": 5.929899282008922e-05,
+      "loss": 0.4454,
+      "step": 1033
+    },
+    {
+      "epoch": 2.031434184675835,
+      "grad_norm": 0.2889110767637736,
+      "learning_rate": 5.9250909138745505e-05,
+      "loss": 0.4496,
+      "step": 1034
+    },
+    {
+      "epoch": 2.0333988212180745,
+      "grad_norm": 0.364482930447891,
+      "learning_rate": 5.920278922643217e-05,
+      "loss": 0.4582,
+      "step": 1035
+    },
+    {
+      "epoch": 2.0353634577603144,
+      "grad_norm": 0.2682736676328122,
+      "learning_rate": 5.9154633173712805e-05,
+      "loss": 0.4477,
+      "step": 1036
+    },
+    {
+      "epoch": 2.0373280943025542,
+      "grad_norm": 0.26964745806282964,
+      "learning_rate": 5.910644107121897e-05,
+      "loss": 0.4498,
+      "step": 1037
+    },
+    {
+      "epoch": 2.0392927308447937,
+      "grad_norm": 0.2895270434934501,
+      "learning_rate": 5.905821300965012e-05,
+      "loss": 0.4503,
+      "step": 1038
+    },
+    {
+      "epoch": 2.0412573673870336,
+      "grad_norm": 0.2513388870974083,
+      "learning_rate": 5.9009949079773366e-05,
+      "loss": 0.4531,
+      "step": 1039
+    },
+    {
+      "epoch": 2.043222003929273,
+      "grad_norm": 0.2602622248611014,
+      "learning_rate": 5.896164937242331e-05,
+      "loss": 0.4582,
+      "step": 1040
+    },
+    {
+      "epoch": 2.045186640471513,
+      "grad_norm": 0.21112136660796071,
+      "learning_rate": 5.891331397850193e-05,
+      "loss": 0.447,
+      "step": 1041
+    },
+    {
+      "epoch": 2.0471512770137523,
+      "grad_norm": 0.24695043406728412,
+      "learning_rate": 5.886494298897833e-05,
+      "loss": 0.4526,
+      "step": 1042
+    },
+    {
+      "epoch": 2.049115913555992,
+      "grad_norm": 0.2438405260370881,
+      "learning_rate": 5.881653649488861e-05,
+      "loss": 0.4482,
+      "step": 1043
+    },
+    {
+      "epoch": 2.0510805500982316,
+      "grad_norm": 0.23822813821167246,
+      "learning_rate": 5.8768094587335716e-05,
+      "loss": 0.4522,
+      "step": 1044
+    },
+    {
+      "epoch": 2.0530451866404715,
+      "grad_norm": 0.1694203533969747,
+      "learning_rate": 5.8719617357489204e-05,
+      "loss": 0.4445,
+      "step": 1045
+    },
+    {
+      "epoch": 2.0550098231827114,
+      "grad_norm": 0.24117447660432179,
+      "learning_rate": 5.867110489658516e-05,
+      "loss": 0.447,
+      "step": 1046
+    },
+    {
+      "epoch": 2.056974459724951,
+      "grad_norm": 0.2526453617482501,
+      "learning_rate": 5.862255729592592e-05,
+      "loss": 0.4518,
+      "step": 1047
+    },
+    {
+      "epoch": 2.0589390962671907,
+      "grad_norm": 0.21960832754414483,
+      "learning_rate": 5.8573974646880005e-05,
+      "loss": 0.4428,
+      "step": 1048
+    },
+    {
+      "epoch": 2.06090373280943,
+      "grad_norm": 0.21224444190851757,
+      "learning_rate": 5.852535704088186e-05,
+      "loss": 0.451,
+      "step": 1049
+    },
+    {
+      "epoch": 2.06286836935167,
+      "grad_norm": 0.2146075263359662,
+      "learning_rate": 5.847670456943174e-05,
+      "loss": 0.4541,
+      "step": 1050
+    },
+    {
+      "epoch": 2.0648330058939095,
+      "grad_norm": 0.21063335337411787,
+      "learning_rate": 5.842801732409551e-05,
+      "loss": 0.4447,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0667976424361494,
+      "grad_norm": 0.20028566950373694,
+      "learning_rate": 5.837929539650448e-05,
+      "loss": 0.446,
+      "step": 1052
+    },
+    {
+      "epoch": 2.068762278978389,
+      "grad_norm": 0.20702075264847208,
+      "learning_rate": 5.8330538878355254e-05,
+      "loss": 0.4454,
+      "step": 1053
+    },
+    {
+      "epoch": 2.0707269155206287,
+      "grad_norm": 0.14591753845014643,
+      "learning_rate": 5.828174786140952e-05,
+      "loss": 0.4512,
+      "step": 1054
+    },
+    {
+      "epoch": 2.072691552062868,
+      "grad_norm": 0.16551734048537933,
+      "learning_rate": 5.8232922437493875e-05,
+      "loss": 0.4506,
+      "step": 1055
+    },
+    {
+      "epoch": 2.074656188605108,
+      "grad_norm": 0.1868969988061947,
+      "learning_rate": 5.8184062698499696e-05,
+      "loss": 0.4426,
+      "step": 1056
+    },
+    {
+      "epoch": 2.076620825147348,
+      "grad_norm": 0.17216356140283914,
+      "learning_rate": 5.8135168736382944e-05,
+      "loss": 0.4511,
+      "step": 1057
+    },
+    {
+      "epoch": 2.0785854616895874,
+      "grad_norm": 0.16740802094664745,
+      "learning_rate": 5.808624064316399e-05,
+      "loss": 0.4495,
+      "step": 1058
+    },
+    {
+      "epoch": 2.0805500982318272,
+      "grad_norm": 0.15136261059680803,
+      "learning_rate": 5.803727851092742e-05,
+      "loss": 0.4509,
+      "step": 1059
+    },
+    {
+      "epoch": 2.0825147347740667,
+      "grad_norm": 0.13905257883301586,
+      "learning_rate": 5.798828243182191e-05,
+      "loss": 0.4457,
+      "step": 1060
+    },
+    {
+      "epoch": 2.0844793713163066,
+      "grad_norm": 0.1657361680020093,
+      "learning_rate": 5.7939252498059986e-05,
+      "loss": 0.4519,
+      "step": 1061
+    },
+    {
+      "epoch": 2.086444007858546,
+      "grad_norm": 0.1573397204884581,
+      "learning_rate": 5.7890188801917944e-05,
+      "loss": 0.4472,
+      "step": 1062
+    },
+    {
+      "epoch": 2.088408644400786,
+      "grad_norm": 0.13963488986206557,
+      "learning_rate": 5.784109143573558e-05,
+      "loss": 0.4521,
+      "step": 1063
+    },
+    {
+      "epoch": 2.0903732809430258,
+      "grad_norm": 0.14493373037665766,
+      "learning_rate": 5.779196049191607e-05,
+      "loss": 0.4487,
+      "step": 1064
+    },
+    {
+      "epoch": 2.092337917485265,
+      "grad_norm": 0.14895801076380386,
+      "learning_rate": 5.7742796062925794e-05,
+      "loss": 0.4505,
+      "step": 1065
+    },
+    {
+      "epoch": 2.094302554027505,
+      "grad_norm": 0.14463049086035823,
+      "learning_rate": 5.769359824129414e-05,
+      "loss": 0.4497,
+      "step": 1066
+    },
+    {
+      "epoch": 2.0962671905697445,
+      "grad_norm": 0.1554599172296196,
+      "learning_rate": 5.764436711961334e-05,
+      "loss": 0.4433,
+      "step": 1067
+    },
+    {
+      "epoch": 2.0982318271119844,
+      "grad_norm": 0.1590633495859465,
+      "learning_rate": 5.7595102790538315e-05,
+      "loss": 0.4491,
+      "step": 1068
+    },
+    {
+      "epoch": 2.100196463654224,
+      "grad_norm": 0.12286982036966913,
+      "learning_rate": 5.7545805346786484e-05,
+      "loss": 0.4499,
+      "step": 1069
+    },
+    {
+      "epoch": 2.1021611001964637,
+      "grad_norm": 0.13250480121571445,
+      "learning_rate": 5.749647488113755e-05,
+      "loss": 0.4575,
+      "step": 1070
+    },
+    {
+      "epoch": 2.104125736738703,
+      "grad_norm": 0.1599340263470461,
+      "learning_rate": 5.744711148643342e-05,
+      "loss": 0.4519,
+      "step": 1071
+    },
+    {
+      "epoch": 2.106090373280943,
+      "grad_norm": 0.165680376680774,
+      "learning_rate": 5.739771525557794e-05,
+      "loss": 0.4537,
+      "step": 1072
+    },
+    {
+      "epoch": 2.1080550098231825,
+      "grad_norm": 0.16131646834685986,
+      "learning_rate": 5.734828628153675e-05,
+      "loss": 0.4478,
+      "step": 1073
+    },
+    {
+      "epoch": 2.1100196463654224,
+      "grad_norm": 0.16720277557027755,
+      "learning_rate": 5.7298824657337154e-05,
+      "loss": 0.4471,
+      "step": 1074
+    },
+    {
+      "epoch": 2.1119842829076623,
+      "grad_norm": 0.155070022023178,
+      "learning_rate": 5.724933047606786e-05,
+      "loss": 0.4451,
+      "step": 1075
+    },
+    {
+      "epoch": 2.1139489194499017,
+      "grad_norm": 0.13862138680468228,
+      "learning_rate": 5.719980383087885e-05,
+      "loss": 0.4457,
+      "step": 1076
+    },
+    {
+      "epoch": 2.1159135559921416,
+      "grad_norm": 0.19635275845791597,
+      "learning_rate": 5.715024481498125e-05,
+      "loss": 0.4624,
+      "step": 1077
+    },
+    {
+      "epoch": 2.117878192534381,
+      "grad_norm": 0.17400266994524752,
+      "learning_rate": 5.710065352164709e-05,
+      "loss": 0.4372,
+      "step": 1078
+    },
+    {
+      "epoch": 2.119842829076621,
+      "grad_norm": 0.14712859455106914,
+      "learning_rate": 5.70510300442091e-05,
+      "loss": 0.4404,
+      "step": 1079
+    },
+    {
+      "epoch": 2.1218074656188604,
+      "grad_norm": 0.12243170169977255,
+      "learning_rate": 5.7001374476060646e-05,
+      "loss": 0.4536,
+      "step": 1080
+    },
+    {
+      "epoch": 2.1237721021611002,
+      "grad_norm": 0.1317370473168329,
+      "learning_rate": 5.6951686910655456e-05,
+      "loss": 0.4467,
+      "step": 1081
+    },
+    {
+      "epoch": 2.1257367387033397,
+      "grad_norm": 0.14195173913053702,
+      "learning_rate": 5.690196744150749e-05,
+      "loss": 0.4508,
+      "step": 1082
+    },
+    {
+      "epoch": 2.1277013752455796,
+      "grad_norm": 0.16663601324741345,
+      "learning_rate": 5.6852216162190744e-05,
+      "loss": 0.4462,
+      "step": 1083
+    },
+    {
+      "epoch": 2.1296660117878194,
+      "grad_norm": 0.17293747459939132,
+      "learning_rate": 5.68024331663391e-05,
+      "loss": 0.4521,
+      "step": 1084
+    },
+    {
+      "epoch": 2.131630648330059,
+      "grad_norm": 0.15198003157605436,
+      "learning_rate": 5.6752618547646095e-05,
+      "loss": 0.4516,
+      "step": 1085
+    },
+    {
+      "epoch": 2.1335952848722988,
+      "grad_norm": 0.14766474882804956,
+      "learning_rate": 5.670277239986481e-05,
+      "loss": 0.4504,
+      "step": 1086
+    },
+    {
+      "epoch": 2.135559921414538,
+      "grad_norm": 0.1540990531331929,
+      "learning_rate": 5.665289481680768e-05,
+      "loss": 0.4565,
+      "step": 1087
+    },
+    {
+      "epoch": 2.137524557956778,
+      "grad_norm": 0.1633234390548173,
+      "learning_rate": 5.660298589234627e-05,
+      "loss": 0.4569,
+      "step": 1088
+    },
+    {
+      "epoch": 2.1394891944990175,
+      "grad_norm": 0.1818884382202553,
+      "learning_rate": 5.655304572041114e-05,
+      "loss": 0.4491,
+      "step": 1089
+    },
+    {
+      "epoch": 2.1414538310412574,
+      "grad_norm": 0.17488050589654966,
+      "learning_rate": 5.650307439499164e-05,
+      "loss": 0.4456,
+      "step": 1090
+    },
+    {
+      "epoch": 2.143418467583497,
+      "grad_norm": 0.15800516504664378,
+      "learning_rate": 5.645307201013581e-05,
+      "loss": 0.4518,
+      "step": 1091
+    },
+    {
+      "epoch": 2.1453831041257367,
+      "grad_norm": 0.14565671471382952,
+      "learning_rate": 5.640303865995011e-05,
+      "loss": 0.452,
+      "step": 1092
+    },
+    {
+      "epoch": 2.1473477406679766,
+      "grad_norm": 0.1845095390799896,
+      "learning_rate": 5.635297443859924e-05,
+      "loss": 0.4414,
+      "step": 1093
+    },
+    {
+      "epoch": 2.149312377210216,
+      "grad_norm": 0.20074659506324052,
+      "learning_rate": 5.6302879440306064e-05,
+      "loss": 0.4465,
+      "step": 1094
+    },
+    {
+      "epoch": 2.151277013752456,
+      "grad_norm": 0.1471292208054478,
+      "learning_rate": 5.6252753759351325e-05,
+      "loss": 0.4572,
+      "step": 1095
+    },
+    {
+      "epoch": 2.1532416502946954,
+      "grad_norm": 0.14911819764539958,
+      "learning_rate": 5.620259749007356e-05,
+      "loss": 0.4476,
+      "step": 1096
+    },
+    {
+      "epoch": 2.1552062868369353,
+      "grad_norm": 0.15337016226930938,
+      "learning_rate": 5.615241072686883e-05,
+      "loss": 0.4543,
+      "step": 1097
+    },
+    {
+      "epoch": 2.1571709233791747,
+      "grad_norm": 0.15391138991109343,
+      "learning_rate": 5.610219356419058e-05,
+      "loss": 0.45,
+      "step": 1098
+    },
+    {
+      "epoch": 2.1591355599214146,
+      "grad_norm": 0.15820810306866206,
+      "learning_rate": 5.605194609654952e-05,
+      "loss": 0.4588,
+      "step": 1099
+    },
+    {
+      "epoch": 2.161100196463654,
+      "grad_norm": 0.14083614085962995,
+      "learning_rate": 5.600166841851333e-05,
+      "loss": 0.4516,
+      "step": 1100
+    },
+    {
+      "epoch": 2.163064833005894,
+      "grad_norm": 0.15399967538464324,
+      "learning_rate": 5.5951360624706624e-05,
+      "loss": 0.4452,
+      "step": 1101
+    },
+    {
+      "epoch": 2.165029469548134,
+      "grad_norm": 0.15095294997123276,
+      "learning_rate": 5.5901022809810626e-05,
+      "loss": 0.4505,
+      "step": 1102
+    },
+    {
+      "epoch": 2.1669941060903732,
+      "grad_norm": 0.11876888105808363,
+      "learning_rate": 5.585065506856309e-05,
+      "loss": 0.4412,
+      "step": 1103
+    },
+    {
+      "epoch": 2.168958742632613,
+      "grad_norm": 0.1456824059185586,
+      "learning_rate": 5.580025749575809e-05,
+      "loss": 0.4459,
+      "step": 1104
+    },
+    {
+      "epoch": 2.1709233791748526,
+      "grad_norm": 0.1731878432605739,
+      "learning_rate": 5.574983018624584e-05,
+      "loss": 0.455,
+      "step": 1105
+    },
+    {
+      "epoch": 2.1728880157170924,
+      "grad_norm": 0.15874463610541778,
+      "learning_rate": 5.569937323493253e-05,
+      "loss": 0.4516,
+      "step": 1106
+    },
+    {
+      "epoch": 2.174852652259332,
+      "grad_norm": 0.15401106674236428,
+      "learning_rate": 5.564888673678014e-05,
+      "loss": 0.4458,
+      "step": 1107
+    },
+    {
+      "epoch": 2.1768172888015718,
+      "grad_norm": 0.15960292771562237,
+      "learning_rate": 5.559837078680623e-05,
+      "loss": 0.4511,
+      "step": 1108
+    },
+    {
+      "epoch": 2.178781925343811,
+      "grad_norm": 0.14662111156323565,
+      "learning_rate": 5.5547825480083826e-05,
+      "loss": 0.4509,
+      "step": 1109
+    },
+    {
+      "epoch": 2.180746561886051,
+      "grad_norm": 0.16188814686958933,
+      "learning_rate": 5.549725091174117e-05,
+      "loss": 0.4417,
+      "step": 1110
+    },
+    {
+      "epoch": 2.182711198428291,
+      "grad_norm": 0.1785353851533852,
+      "learning_rate": 5.54466471769616e-05,
+      "loss": 0.4485,
+      "step": 1111
+    },
+    {
+      "epoch": 2.1846758349705304,
+      "grad_norm": 0.17179021375551004,
+      "learning_rate": 5.539601437098335e-05,
+      "loss": 0.4584,
+      "step": 1112
+    },
+    {
+      "epoch": 2.1866404715127703,
+      "grad_norm": 0.1411556390827246,
+      "learning_rate": 5.534535258909934e-05,
+      "loss": 0.4405,
+      "step": 1113
+    },
+    {
+      "epoch": 2.1886051080550097,
+      "grad_norm": 0.13828335829533353,
+      "learning_rate": 5.529466192665704e-05,
+      "loss": 0.4542,
+      "step": 1114
+    },
+    {
+      "epoch": 2.1905697445972496,
+      "grad_norm": 0.16473607873107374,
+      "learning_rate": 5.524394247905829e-05,
+      "loss": 0.452,
+      "step": 1115
+    },
+    {
+      "epoch": 2.192534381139489,
+      "grad_norm": 0.18643979793255938,
+      "learning_rate": 5.5193194341759074e-05,
+      "loss": 0.4512,
+      "step": 1116
+    },
+    {
+      "epoch": 2.194499017681729,
+      "grad_norm": 0.1442363126737495,
+      "learning_rate": 5.51424176102694e-05,
+      "loss": 0.4502,
+      "step": 1117
+    },
+    {
+      "epoch": 2.1964636542239684,
+      "grad_norm": 0.10675501952917513,
+      "learning_rate": 5.509161238015306e-05,
+      "loss": 0.4547,
+      "step": 1118
+    },
+    {
+      "epoch": 2.1984282907662083,
+      "grad_norm": 0.13688612657410515,
+      "learning_rate": 5.5040778747027504e-05,
+      "loss": 0.4444,
+      "step": 1119
+    },
+    {
+      "epoch": 2.2003929273084477,
+      "grad_norm": 0.14118128191931817,
+      "learning_rate": 5.4989916806563655e-05,
+      "loss": 0.4508,
+      "step": 1120
+    },
+    {
+      "epoch": 2.2023575638506876,
+      "grad_norm": 0.13204751266825115,
+      "learning_rate": 5.493902665448566e-05,
+      "loss": 0.449,
+      "step": 1121
+    },
+    {
+      "epoch": 2.2043222003929275,
+      "grad_norm": 0.12245523285978761,
+      "learning_rate": 5.488810838657081e-05,
+      "loss": 0.4492,
+      "step": 1122
+    },
+    {
+      "epoch": 2.206286836935167,
+      "grad_norm": 0.11296406685735665,
+      "learning_rate": 5.483716209864928e-05,
+      "loss": 0.4477,
+      "step": 1123
+    },
+    {
+      "epoch": 2.208251473477407,
+      "grad_norm": 0.11979666546172515,
+      "learning_rate": 5.478618788660398e-05,
+      "loss": 0.4459,
+      "step": 1124
+    },
+    {
+      "epoch": 2.2102161100196462,
+      "grad_norm": 0.13610825584244496,
+      "learning_rate": 5.4735185846370415e-05,
+      "loss": 0.4474,
+      "step": 1125
+    },
+    {
+      "epoch": 2.212180746561886,
+      "grad_norm": 0.15866595313533385,
+      "learning_rate": 5.468415607393643e-05,
+      "loss": 0.4499,
+      "step": 1126
+    },
+    {
+      "epoch": 2.2141453831041256,
+      "grad_norm": 0.137505567424083,
+      "learning_rate": 5.463309866534205e-05,
+      "loss": 0.4433,
+      "step": 1127
+    },
+    {
+      "epoch": 2.2161100196463654,
+      "grad_norm": 0.12358582792626695,
+      "learning_rate": 5.458201371667933e-05,
+      "loss": 0.448,
+      "step": 1128
+    },
+    {
+      "epoch": 2.2180746561886053,
+      "grad_norm": 0.12127152287436908,
+      "learning_rate": 5.453090132409216e-05,
+      "loss": 0.4481,
+      "step": 1129
+    },
+    {
+      "epoch": 2.2200392927308448,
+      "grad_norm": 0.13335351588984473,
+      "learning_rate": 5.447976158377611e-05,
+      "loss": 0.45,
+      "step": 1130
+    },
+    {
+      "epoch": 2.2220039292730847,
+      "grad_norm": 0.11153453648456209,
+      "learning_rate": 5.442859459197814e-05,
+      "loss": 0.448,
+      "step": 1131
+    },
+    {
+      "epoch": 2.223968565815324,
+      "grad_norm": 0.11686217202204696,
+      "learning_rate": 5.437740044499656e-05,
+      "loss": 0.4449,
+      "step": 1132
+    },
+    {
+      "epoch": 2.225933202357564,
+      "grad_norm": 0.12028823976506275,
+      "learning_rate": 5.432617923918077e-05,
+      "loss": 0.4465,
+      "step": 1133
+    },
+    {
+      "epoch": 2.2278978388998034,
+      "grad_norm": 0.12170654744728136,
+      "learning_rate": 5.427493107093112e-05,
+      "loss": 0.441,
+      "step": 1134
+    },
+    {
+      "epoch": 2.2298624754420433,
+      "grad_norm": 0.12865380003005095,
+      "learning_rate": 5.422365603669866e-05,
+      "loss": 0.4523,
+      "step": 1135
+    },
+    {
+      "epoch": 2.2318271119842827,
+      "grad_norm": 0.12486565999532875,
+      "learning_rate": 5.417235423298503e-05,
+      "loss": 0.4589,
+      "step": 1136
+    },
+    {
+      "epoch": 2.2337917485265226,
+      "grad_norm": 0.12358263640271015,
+      "learning_rate": 5.412102575634224e-05,
+      "loss": 0.4458,
+      "step": 1137
+    },
+    {
+      "epoch": 2.235756385068762,
+      "grad_norm": 0.11508190776106345,
+      "learning_rate": 5.406967070337252e-05,
+      "loss": 0.4529,
+      "step": 1138
+    },
+    {
+      "epoch": 2.237721021611002,
+      "grad_norm": 0.11488440142621001,
+      "learning_rate": 5.401828917072809e-05,
+      "loss": 0.4451,
+      "step": 1139
+    },
+    {
+      "epoch": 2.239685658153242,
+      "grad_norm": 0.1110112746899281,
+      "learning_rate": 5.3966881255111024e-05,
+      "loss": 0.4497,
+      "step": 1140
+    },
+    {
+      "epoch": 2.2416502946954813,
+      "grad_norm": 0.1347606270572007,
+      "learning_rate": 5.391544705327306e-05,
+      "loss": 0.4446,
+      "step": 1141
+    },
+    {
+      "epoch": 2.243614931237721,
+      "grad_norm": 0.1502120054856367,
+      "learning_rate": 5.386398666201536e-05,
+      "loss": 0.4536,
+      "step": 1142
+    },
+    {
+      "epoch": 2.2455795677799606,
+      "grad_norm": 0.12234619884234972,
+      "learning_rate": 5.381250017818843e-05,
+      "loss": 0.4499,
+      "step": 1143
+    },
+    {
+      "epoch": 2.2475442043222005,
+      "grad_norm": 0.10979684454165071,
+      "learning_rate": 5.376098769869187e-05,
+      "loss": 0.4506,
+      "step": 1144
+    },
+    {
+      "epoch": 2.24950884086444,
+      "grad_norm": 0.14304937767518403,
+      "learning_rate": 5.370944932047417e-05,
+      "loss": 0.4492,
+      "step": 1145
+    },
+    {
+      "epoch": 2.25147347740668,
+      "grad_norm": 0.12820131657977693,
+      "learning_rate": 5.365788514053262e-05,
+      "loss": 0.4508,
+      "step": 1146
+    },
+    {
+      "epoch": 2.2534381139489197,
+      "grad_norm": 0.11931047966319995,
+      "learning_rate": 5.360629525591301e-05,
+      "loss": 0.4485,
+      "step": 1147
+    },
+    {
+      "epoch": 2.255402750491159,
+      "grad_norm": 0.14468675378802548,
+      "learning_rate": 5.355467976370955e-05,
+      "loss": 0.4474,
+      "step": 1148
+    },
+    {
+      "epoch": 2.257367387033399,
+      "grad_norm": 0.128898043087968,
+      "learning_rate": 5.3503038761064636e-05,
+      "loss": 0.4498,
+      "step": 1149
+    },
+    {
+      "epoch": 2.2593320235756384,
+      "grad_norm": 0.13121697715264086,
+      "learning_rate": 5.345137234516864e-05,
+      "loss": 0.4524,
+      "step": 1150
+    },
+    {
+      "epoch": 2.2612966601178783,
+      "grad_norm": 0.13191461703438628,
+      "learning_rate": 5.3399680613259845e-05,
+      "loss": 0.4477,
+      "step": 1151
+    },
+    {
+      "epoch": 2.2632612966601178,
+      "grad_norm": 0.14731429721345973,
+      "learning_rate": 5.334796366262408e-05,
+      "loss": 0.4508,
+      "step": 1152
+    },
+    {
+      "epoch": 2.2652259332023577,
+      "grad_norm": 0.12474678123886214,
+      "learning_rate": 5.3296221590594695e-05,
+      "loss": 0.4448,
+      "step": 1153
+    },
+    {
+      "epoch": 2.267190569744597,
+      "grad_norm": 0.12443248256864324,
+      "learning_rate": 5.324445449455233e-05,
+      "loss": 0.4475,
+      "step": 1154
+    },
+    {
+      "epoch": 2.269155206286837,
+      "grad_norm": 0.12962388105990724,
+      "learning_rate": 5.3192662471924684e-05,
+      "loss": 0.4415,
+      "step": 1155
+    },
+    {
+      "epoch": 2.2711198428290764,
+      "grad_norm": 0.12737116053993489,
+      "learning_rate": 5.314084562018638e-05,
+      "loss": 0.4555,
+      "step": 1156
+    },
+    {
+      "epoch": 2.2730844793713163,
+      "grad_norm": 0.11383209094183275,
+      "learning_rate": 5.308900403685878e-05,
+      "loss": 0.4502,
+      "step": 1157
+    },
+    {
+      "epoch": 2.275049115913556,
+      "grad_norm": 0.14235402602687858,
+      "learning_rate": 5.3037137819509786e-05,
+      "loss": 0.455,
+      "step": 1158
+    },
+    {
+      "epoch": 2.2770137524557956,
+      "grad_norm": 0.14020675990245546,
+      "learning_rate": 5.298524706575366e-05,
+      "loss": 0.4432,
+      "step": 1159
+    },
+    {
+      "epoch": 2.2789783889980355,
+      "grad_norm": 0.11446588813367851,
+      "learning_rate": 5.293333187325086e-05,
+      "loss": 0.4467,
+      "step": 1160
+    },
+    {
+      "epoch": 2.280943025540275,
+      "grad_norm": 0.14498281134418164,
+      "learning_rate": 5.288139233970781e-05,
+      "loss": 0.4597,
+      "step": 1161
+    },
+    {
+      "epoch": 2.282907662082515,
+      "grad_norm": 0.11596578514890574,
+      "learning_rate": 5.2829428562876754e-05,
+      "loss": 0.446,
+      "step": 1162
+    },
+    {
+      "epoch": 2.2848722986247543,
+      "grad_norm": 0.10203018940997839,
+      "learning_rate": 5.277744064055558e-05,
+      "loss": 0.4598,
+      "step": 1163
+    },
+    {
+      "epoch": 2.286836935166994,
+      "grad_norm": 0.10244083534877517,
+      "learning_rate": 5.272542867058761e-05,
+      "loss": 0.4563,
+      "step": 1164
+    },
+    {
+      "epoch": 2.288801571709234,
+      "grad_norm": 0.10327142855596888,
+      "learning_rate": 5.267339275086141e-05,
+      "loss": 0.4461,
+      "step": 1165
+    },
+    {
+      "epoch": 2.2907662082514735,
+      "grad_norm": 0.11166252514921922,
+      "learning_rate": 5.262133297931064e-05,
+      "loss": 0.4473,
+      "step": 1166
+    },
+    {
+      "epoch": 2.292730844793713,
+      "grad_norm": 0.1002402430544765,
+      "learning_rate": 5.2569249453913815e-05,
+      "loss": 0.4456,
+      "step": 1167
+    },
+    {
+      "epoch": 2.294695481335953,
+      "grad_norm": 0.10610056160176841,
+      "learning_rate": 5.251714227269423e-05,
+      "loss": 0.4473,
+      "step": 1168
+    },
+    {
+      "epoch": 2.2966601178781927,
+      "grad_norm": 0.12452414076839978,
+      "learning_rate": 5.246501153371961e-05,
+      "loss": 0.4518,
+      "step": 1169
+    },
+    {
+      "epoch": 2.298624754420432,
+      "grad_norm": 0.10342633081117969,
+      "learning_rate": 5.241285733510209e-05,
+      "loss": 0.4362,
+      "step": 1170
+    },
+    {
+      "epoch": 2.300589390962672,
+      "grad_norm": 0.09979632149183572,
+      "learning_rate": 5.23606797749979e-05,
+      "loss": 0.4507,
+      "step": 1171
+    },
+    {
+      "epoch": 2.3025540275049114,
+      "grad_norm": 0.10705066132460682,
+      "learning_rate": 5.230847895160729e-05,
+      "loss": 0.4506,
+      "step": 1172
+    },
+    {
+      "epoch": 2.3045186640471513,
+      "grad_norm": 0.10566541296090068,
+      "learning_rate": 5.225625496317426e-05,
+      "loss": 0.452,
+      "step": 1173
+    },
+    {
+      "epoch": 2.3064833005893908,
+      "grad_norm": 0.14037484245171922,
+      "learning_rate": 5.220400790798639e-05,
+      "loss": 0.4567,
+      "step": 1174
+    },
+    {
+      "epoch": 2.3084479371316307,
+      "grad_norm": 0.10980677795041702,
+      "learning_rate": 5.215173788437473e-05,
+      "loss": 0.4508,
+      "step": 1175
+    },
+    {
+      "epoch": 2.3104125736738705,
+      "grad_norm": 0.12892206072017925,
+      "learning_rate": 5.209944499071351e-05,
+      "loss": 0.4446,
+      "step": 1176
+    },
+    {
+      "epoch": 2.31237721021611,
+      "grad_norm": 0.1053371452150924,
+      "learning_rate": 5.2047129325420014e-05,
+      "loss": 0.4566,
+      "step": 1177
+    },
+    {
+      "epoch": 2.31434184675835,
+      "grad_norm": 0.11358713025221705,
+      "learning_rate": 5.199479098695439e-05,
+      "loss": 0.4564,
+      "step": 1178
+    },
+    {
+      "epoch": 2.3163064833005893,
+      "grad_norm": 0.11803685274840986,
+      "learning_rate": 5.194243007381945e-05,
+      "loss": 0.4526,
+      "step": 1179
+    },
+    {
+      "epoch": 2.318271119842829,
+      "grad_norm": 0.12994556855150757,
+      "learning_rate": 5.189004668456051e-05,
+      "loss": 0.4532,
+      "step": 1180
+    },
+    {
+      "epoch": 2.3202357563850686,
+      "grad_norm": 0.10308797442003997,
+      "learning_rate": 5.183764091776515e-05,
+      "loss": 0.4447,
+      "step": 1181
+    },
+    {
+      "epoch": 2.3222003929273085,
+      "grad_norm": 0.14119157372359756,
+      "learning_rate": 5.178521287206309e-05,
+      "loss": 0.4378,
+      "step": 1182
+    },
+    {
+      "epoch": 2.324165029469548,
+      "grad_norm": 0.1529714067430109,
+      "learning_rate": 5.1732762646126005e-05,
+      "loss": 0.4532,
+      "step": 1183
+    },
+    {
+      "epoch": 2.326129666011788,
+      "grad_norm": 0.12824078092397198,
+      "learning_rate": 5.1680290338667265e-05,
+      "loss": 0.4472,
+      "step": 1184
+    },
+    {
+      "epoch": 2.3280943025540273,
+      "grad_norm": 0.11999237713130392,
+      "learning_rate": 5.1627796048441824e-05,
+      "loss": 0.45,
+      "step": 1185
+    },
+    {
+      "epoch": 2.330058939096267,
+      "grad_norm": 0.10431221275976534,
+      "learning_rate": 5.1575279874246005e-05,
+      "loss": 0.4536,
+      "step": 1186
+    },
+    {
+      "epoch": 2.332023575638507,
+      "grad_norm": 0.11396858186309326,
+      "learning_rate": 5.152274191491732e-05,
+      "loss": 0.4621,
+      "step": 1187
+    },
+    {
+      "epoch": 2.3339882121807465,
+      "grad_norm": 0.11865533301499405,
+      "learning_rate": 5.147018226933427e-05,
+      "loss": 0.4499,
+      "step": 1188
+    },
+    {
+      "epoch": 2.3359528487229864,
+      "grad_norm": 0.11399626196193263,
+      "learning_rate": 5.14176010364162e-05,
+      "loss": 0.4418,
+      "step": 1189
+    },
+    {
+      "epoch": 2.337917485265226,
+      "grad_norm": 0.1202883263710369,
+      "learning_rate": 5.1364998315123024e-05,
+      "loss": 0.4572,
+      "step": 1190
+    },
+    {
+      "epoch": 2.3398821218074657,
+      "grad_norm": 0.1244013909048219,
+      "learning_rate": 5.1312374204455175e-05,
+      "loss": 0.4489,
+      "step": 1191
+    },
+    {
+      "epoch": 2.341846758349705,
+      "grad_norm": 0.13939035608101988,
+      "learning_rate": 5.125972880345329e-05,
+      "loss": 0.4538,
+      "step": 1192
+    },
+    {
+      "epoch": 2.343811394891945,
+      "grad_norm": 0.14825506122974003,
+      "learning_rate": 5.120706221119809e-05,
+      "loss": 0.4587,
+      "step": 1193
+    },
+    {
+      "epoch": 2.345776031434185,
+      "grad_norm": 0.11336127084200585,
+      "learning_rate": 5.115437452681018e-05,
+      "loss": 0.4595,
+      "step": 1194
+    },
+    {
+      "epoch": 2.3477406679764243,
+      "grad_norm": 0.15241337285624226,
+      "learning_rate": 5.1101665849449834e-05,
+      "loss": 0.4525,
+      "step": 1195
+    },
+    {
+      "epoch": 2.349705304518664,
+      "grad_norm": 0.1315623141232332,
+      "learning_rate": 5.104893627831688e-05,
+      "loss": 0.4508,
+      "step": 1196
+    },
+    {
+      "epoch": 2.3516699410609037,
+      "grad_norm": 0.11952843749449507,
+      "learning_rate": 5.099618591265046e-05,
+      "loss": 0.4465,
+      "step": 1197
+    },
+    {
+      "epoch": 2.3536345776031435,
+      "grad_norm": 0.1255091838323739,
+      "learning_rate": 5.094341485172882e-05,
+      "loss": 0.4554,
+      "step": 1198
+    },
+    {
+      "epoch": 2.355599214145383,
+      "grad_norm": 0.1244353207450483,
+      "learning_rate": 5.089062319486919e-05,
+      "loss": 0.4591,
+      "step": 1199
+    },
+    {
+      "epoch": 2.357563850687623,
+      "grad_norm": 0.12612232029860765,
+      "learning_rate": 5.0837811041427534e-05,
+      "loss": 0.4527,
+      "step": 1200
+    },
+    {
+      "epoch": 2.3595284872298623,
+      "grad_norm": 0.13146142480736892,
+      "learning_rate": 5.078497849079842e-05,
+      "loss": 0.4418,
+      "step": 1201
+    },
+    {
+      "epoch": 2.361493123772102,
+      "grad_norm": 0.1562047578593488,
+      "learning_rate": 5.073212564241479e-05,
+      "loss": 0.4478,
+      "step": 1202
+    },
+    {
+      "epoch": 2.3634577603143416,
+      "grad_norm": 0.1010501203169563,
+      "learning_rate": 5.067925259574777e-05,
+      "loss": 0.4579,
+      "step": 1203
+    },
+    {
+      "epoch": 2.3654223968565815,
+      "grad_norm": 0.13132015444032671,
+      "learning_rate": 5.062635945030654e-05,
+      "loss": 0.451,
+      "step": 1204
+    },
+    {
+      "epoch": 2.3673870333988214,
+      "grad_norm": 0.12912208736396522,
+      "learning_rate": 5.057344630563807e-05,
+      "loss": 0.4479,
+      "step": 1205
+    },
+    {
+      "epoch": 2.369351669941061,
+      "grad_norm": 0.11878314639448177,
+      "learning_rate": 5.0520513261326985e-05,
+      "loss": 0.4543,
+      "step": 1206
+    },
+    {
+      "epoch": 2.3713163064833007,
+      "grad_norm": 0.11510261042219506,
+      "learning_rate": 5.046756041699537e-05,
+      "loss": 0.4533,
+      "step": 1207
+    },
+    {
+      "epoch": 2.37328094302554,
+      "grad_norm": 0.15243673688769746,
+      "learning_rate": 5.041458787230255e-05,
+      "loss": 0.4416,
+      "step": 1208
+    },
+    {
+      "epoch": 2.37524557956778,
+      "grad_norm": 0.12666362433225628,
+      "learning_rate": 5.0361595726944955e-05,
+      "loss": 0.4494,
+      "step": 1209
+    },
+    {
+      "epoch": 2.3772102161100195,
+      "grad_norm": 0.14381367983126414,
+      "learning_rate": 5.030858408065587e-05,
+      "loss": 0.4586,
+      "step": 1210
+    },
+    {
+      "epoch": 2.3791748526522594,
+      "grad_norm": 0.13985053443203174,
+      "learning_rate": 5.025555303320532e-05,
+      "loss": 0.4492,
+      "step": 1211
+    },
+    {
+      "epoch": 2.3811394891944992,
+      "grad_norm": 0.13813035325653547,
+      "learning_rate": 5.020250268439984e-05,
+      "loss": 0.4522,
+      "step": 1212
+    },
+    {
+      "epoch": 2.3831041257367387,
+      "grad_norm": 0.12547799299547033,
+      "learning_rate": 5.014943313408225e-05,
+      "loss": 0.4516,
+      "step": 1213
+    },
+    {
+      "epoch": 2.3850687622789786,
+      "grad_norm": 0.15613755041867694,
+      "learning_rate": 5.009634448213155e-05,
+      "loss": 0.4544,
+      "step": 1214
+    },
+    {
+      "epoch": 2.387033398821218,
+      "grad_norm": 0.1572849047491199,
+      "learning_rate": 5.0043236828462665e-05,
+      "loss": 0.4627,
+      "step": 1215
+    },
+    {
+      "epoch": 2.388998035363458,
+      "grad_norm": 0.1300685599304094,
+      "learning_rate": 4.999011027302629e-05,
+      "loss": 0.4505,
+      "step": 1216
+    },
+    {
+      "epoch": 2.3909626719056973,
+      "grad_norm": 0.11522847175459784,
+      "learning_rate": 4.993696491580872e-05,
+      "loss": 0.4504,
+      "step": 1217
+    },
+    {
+      "epoch": 2.392927308447937,
+      "grad_norm": 0.13226030090615093,
+      "learning_rate": 4.988380085683158e-05,
+      "loss": 0.4544,
+      "step": 1218
+    },
+    {
+      "epoch": 2.3948919449901767,
+      "grad_norm": 0.14434612906398023,
+      "learning_rate": 4.983061819615174e-05,
+      "loss": 0.4633,
+      "step": 1219
+    },
+    {
+      "epoch": 2.3968565815324165,
+      "grad_norm": 0.12555337118579668,
+      "learning_rate": 4.977741703386107e-05,
+      "loss": 0.457,
+      "step": 1220
+    },
+    {
+      "epoch": 2.398821218074656,
+      "grad_norm": 0.14525444638075538,
+      "learning_rate": 4.9724197470086236e-05,
+      "loss": 0.4442,
+      "step": 1221
+    },
+    {
+      "epoch": 2.400785854616896,
+      "grad_norm": 0.14036858669584848,
+      "learning_rate": 4.9670959604988574e-05,
+      "loss": 0.4544,
+      "step": 1222
+    },
+    {
+      "epoch": 2.4027504911591357,
+      "grad_norm": 0.11508091484340266,
+      "learning_rate": 4.961770353876383e-05,
+      "loss": 0.4509,
+      "step": 1223
+    },
+    {
+      "epoch": 2.404715127701375,
+      "grad_norm": 0.1528470489693015,
+      "learning_rate": 4.9564429371642016e-05,
+      "loss": 0.4482,
+      "step": 1224
+    },
+    {
+      "epoch": 2.406679764243615,
+      "grad_norm": 0.15128889575351595,
+      "learning_rate": 4.951113720388723e-05,
+      "loss": 0.4538,
+      "step": 1225
+    },
+    {
+      "epoch": 2.4086444007858545,
+      "grad_norm": 0.14419499472298036,
+      "learning_rate": 4.9457827135797406e-05,
+      "loss": 0.4513,
+      "step": 1226
+    },
+    {
+      "epoch": 2.4106090373280944,
+      "grad_norm": 0.1283369565648405,
+      "learning_rate": 4.9404499267704236e-05,
+      "loss": 0.4513,
+      "step": 1227
+    },
+    {
+      "epoch": 2.412573673870334,
+      "grad_norm": 0.1338904299389974,
+      "learning_rate": 4.9351153699972836e-05,
+      "loss": 0.4566,
+      "step": 1228
+    },
+    {
+      "epoch": 2.4145383104125737,
+      "grad_norm": 0.14146637052911415,
+      "learning_rate": 4.929779053300166e-05,
+      "loss": 0.4453,
+      "step": 1229
+    },
+    {
+      "epoch": 2.4165029469548136,
+      "grad_norm": 0.13720463180703124,
+      "learning_rate": 4.924440986722231e-05,
+      "loss": 0.4553,
+      "step": 1230
+    },
+    {
+      "epoch": 2.418467583497053,
+      "grad_norm": 0.13829259439453287,
+      "learning_rate": 4.919101180309931e-05,
+      "loss": 0.4449,
+      "step": 1231
+    },
+    {
+      "epoch": 2.4204322200392925,
+      "grad_norm": 0.1376885179226031,
+      "learning_rate": 4.91375964411299e-05,
+      "loss": 0.4399,
+      "step": 1232
+    },
+    {
+      "epoch": 2.4223968565815324,
+      "grad_norm": 0.12228278756750609,
+      "learning_rate": 4.90841638818439e-05,
+      "loss": 0.4545,
+      "step": 1233
+    },
+    {
+      "epoch": 2.4243614931237722,
+      "grad_norm": 0.12155669354775858,
+      "learning_rate": 4.9030714225803496e-05,
+      "loss": 0.4498,
+      "step": 1234
+    },
+    {
+      "epoch": 2.4263261296660117,
+      "grad_norm": 0.112654413790653,
+      "learning_rate": 4.8977247573603046e-05,
+      "loss": 0.454,
+      "step": 1235
+    },
+    {
+      "epoch": 2.4282907662082516,
+      "grad_norm": 0.10348189532531228,
+      "learning_rate": 4.892376402586891e-05,
+      "loss": 0.4528,
+      "step": 1236
+    },
+    {
+      "epoch": 2.430255402750491,
+      "grad_norm": 0.11639434712921885,
+      "learning_rate": 4.887026368325921e-05,
+      "loss": 0.4598,
+      "step": 1237
+    },
+    {
+      "epoch": 2.432220039292731,
+      "grad_norm": 0.10658750121608052,
+      "learning_rate": 4.881674664646371e-05,
+      "loss": 0.4474,
+      "step": 1238
+    },
+    {
+      "epoch": 2.4341846758349703,
+      "grad_norm": 0.11370795025666842,
+      "learning_rate": 4.8763213016203594e-05,
+      "loss": 0.445,
+      "step": 1239
+    },
+    {
+      "epoch": 2.43614931237721,
+      "grad_norm": 0.11757030050060588,
+      "learning_rate": 4.870966289323123e-05,
+      "loss": 0.4507,
+      "step": 1240
+    },
+    {
+      "epoch": 2.43811394891945,
+      "grad_norm": 0.12914916571747023,
+      "learning_rate": 4.86560963783301e-05,
+      "loss": 0.4598,
+      "step": 1241
+    },
+    {
+      "epoch": 2.4400785854616895,
+      "grad_norm": 0.11734917047257462,
+      "learning_rate": 4.860251357231446e-05,
+      "loss": 0.452,
+      "step": 1242
+    },
+    {
+      "epoch": 2.4420432220039294,
+      "grad_norm": 0.101012987794651,
+      "learning_rate": 4.8548914576029306e-05,
+      "loss": 0.453,
+      "step": 1243
+    },
+    {
+      "epoch": 2.444007858546169,
+      "grad_norm": 0.12755640966722173,
+      "learning_rate": 4.849529949035001e-05,
+      "loss": 0.4516,
+      "step": 1244
+    },
+    {
+      "epoch": 2.4459724950884087,
+      "grad_norm": 0.13320687235109077,
+      "learning_rate": 4.84416684161823e-05,
+      "loss": 0.4601,
+      "step": 1245
+    },
+    {
+      "epoch": 2.447937131630648,
+      "grad_norm": 0.12194477612944339,
+      "learning_rate": 4.838802145446199e-05,
+      "loss": 0.4517,
+      "step": 1246
+    },
+    {
+      "epoch": 2.449901768172888,
+      "grad_norm": 0.13082678956981325,
+      "learning_rate": 4.833435870615474e-05,
+      "loss": 0.4514,
+      "step": 1247
+    },
+    {
+      "epoch": 2.4518664047151275,
+      "grad_norm": 0.13202376369627725,
+      "learning_rate": 4.828068027225599e-05,
+      "loss": 0.4451,
+      "step": 1248
+    },
+    {
+      "epoch": 2.4538310412573674,
+      "grad_norm": 0.10818538714490016,
+      "learning_rate": 4.822698625379063e-05,
+      "loss": 0.4581,
+      "step": 1249
+    },
+    {
+      "epoch": 2.455795677799607,
+      "grad_norm": 0.1170522170369757,
+      "learning_rate": 4.817327675181297e-05,
+      "loss": 0.4555,
+      "step": 1250
+    },
+    {
+      "epoch": 2.4577603143418467,
+      "grad_norm": 0.12525604884419714,
+      "learning_rate": 4.811955186740637e-05,
+      "loss": 0.4489,
+      "step": 1251
+    },
+    {
+      "epoch": 2.4597249508840866,
+      "grad_norm": 0.1200907231985152,
+      "learning_rate": 4.806581170168318e-05,
+      "loss": 0.4472,
+      "step": 1252
+    },
+    {
+      "epoch": 2.461689587426326,
+      "grad_norm": 0.10966208429877618,
+      "learning_rate": 4.801205635578453e-05,
+      "loss": 0.462,
+      "step": 1253
+    },
+    {
+      "epoch": 2.463654223968566,
+      "grad_norm": 0.14642332782864595,
+      "learning_rate": 4.795828593088008e-05,
+      "loss": 0.4505,
+      "step": 1254
+    },
+    {
+      "epoch": 2.4656188605108054,
+      "grad_norm": 0.12329053579401238,
+      "learning_rate": 4.79045005281679e-05,
+      "loss": 0.4414,
+      "step": 1255
+    },
+    {
+      "epoch": 2.4675834970530452,
+      "grad_norm": 0.11206769245015581,
+      "learning_rate": 4.7850700248874244e-05,
+      "loss": 0.4491,
+      "step": 1256
+    },
+    {
+      "epoch": 2.4695481335952847,
+      "grad_norm": 0.13453367687380288,
+      "learning_rate": 4.779688519425334e-05,
+      "loss": 0.4562,
+      "step": 1257
+    },
+    {
+      "epoch": 2.4715127701375246,
+      "grad_norm": 0.11838442081824356,
+      "learning_rate": 4.7743055465587256e-05,
+      "loss": 0.4558,
+      "step": 1258
+    },
+    {
+      "epoch": 2.4734774066797645,
+      "grad_norm": 0.10620072722859915,
+      "learning_rate": 4.768921116418565e-05,
+      "loss": 0.4538,
+      "step": 1259
+    },
+    {
+      "epoch": 2.475442043222004,
+      "grad_norm": 0.11162184548376057,
+      "learning_rate": 4.763535239138562e-05,
+      "loss": 0.4552,
+      "step": 1260
+    },
+    {
+      "epoch": 2.477406679764244,
+      "grad_norm": 0.11758577448531718,
+      "learning_rate": 4.758147924855152e-05,
+      "loss": 0.4482,
+      "step": 1261
+    },
+    {
+      "epoch": 2.479371316306483,
+      "grad_norm": 0.11755248832376443,
+      "learning_rate": 4.752759183707471e-05,
+      "loss": 0.4499,
+      "step": 1262
+    },
+    {
+      "epoch": 2.481335952848723,
+      "grad_norm": 0.11624549286680513,
+      "learning_rate": 4.747369025837341e-05,
+      "loss": 0.4445,
+      "step": 1263
+    },
+    {
+      "epoch": 2.4833005893909625,
+      "grad_norm": 0.12409320437774236,
+      "learning_rate": 4.741977461389255e-05,
+      "loss": 0.4469,
+      "step": 1264
+    },
+    {
+      "epoch": 2.4852652259332024,
+      "grad_norm": 0.10927674474762347,
+      "learning_rate": 4.736584500510347e-05,
+      "loss": 0.4558,
+      "step": 1265
+    },
+    {
+      "epoch": 2.487229862475442,
+      "grad_norm": 0.1317671408805185,
+      "learning_rate": 4.7311901533503817e-05,
+      "loss": 0.4451,
+      "step": 1266
+    },
+    {
+      "epoch": 2.4891944990176817,
+      "grad_norm": 0.09896574264130666,
+      "learning_rate": 4.725794430061735e-05,
+      "loss": 0.4567,
+      "step": 1267
+    },
+    {
+      "epoch": 2.491159135559921,
+      "grad_norm": 0.1363421178259206,
+      "learning_rate": 4.720397340799369e-05,
+      "loss": 0.4461,
+      "step": 1268
+    },
+    {
+      "epoch": 2.493123772102161,
+      "grad_norm": 0.13782272669656914,
+      "learning_rate": 4.7149988957208184e-05,
+      "loss": 0.4536,
+      "step": 1269
+    },
+    {
+      "epoch": 2.495088408644401,
+      "grad_norm": 0.11360855161900757,
+      "learning_rate": 4.709599104986172e-05,
+      "loss": 0.4464,
+      "step": 1270
+    },
+    {
+      "epoch": 2.4970530451866404,
+      "grad_norm": 0.12448572824571011,
+      "learning_rate": 4.704197978758046e-05,
+      "loss": 0.4544,
+      "step": 1271
+    },
+    {
+      "epoch": 2.4990176817288803,
+      "grad_norm": 0.10287777146014776,
+      "learning_rate": 4.698795527201574e-05,
+      "loss": 0.4506,
+      "step": 1272
+    },
+    {
+      "epoch": 2.5009823182711197,
+      "grad_norm": 0.11920158906751364,
+      "learning_rate": 4.693391760484382e-05,
+      "loss": 0.4472,
+      "step": 1273
+    },
+    {
+      "epoch": 2.5029469548133596,
+      "grad_norm": 0.11250227155528607,
+      "learning_rate": 4.6879866887765723e-05,
+      "loss": 0.4531,
+      "step": 1274
+    },
+    {
+      "epoch": 2.504911591355599,
+      "grad_norm": 0.10780983455489213,
+      "learning_rate": 4.682580322250705e-05,
+      "loss": 0.4462,
+      "step": 1275
+    },
+    {
+      "epoch": 2.506876227897839,
+      "grad_norm": 0.10680317386290084,
+      "learning_rate": 4.6771726710817706e-05,
+      "loss": 0.4423,
+      "step": 1276
+    },
+    {
+      "epoch": 2.508840864440079,
+      "grad_norm": 0.11206383122743135,
+      "learning_rate": 4.671763745447186e-05,
+      "loss": 0.4506,
+      "step": 1277
+    },
+    {
+      "epoch": 2.5108055009823183,
+      "grad_norm": 0.11501023362140704,
+      "learning_rate": 4.6663535555267584e-05,
+      "loss": 0.4548,
+      "step": 1278
+    },
+    {
+      "epoch": 2.5127701375245577,
+      "grad_norm": 0.1035753915428985,
+      "learning_rate": 4.660942111502681e-05,
+      "loss": 0.4488,
+      "step": 1279
+    },
+    {
+      "epoch": 2.5147347740667976,
+      "grad_norm": 0.12601447370264715,
+      "learning_rate": 4.6555294235595047e-05,
+      "loss": 0.4493,
+      "step": 1280
+    },
+    {
+      "epoch": 2.5166994106090375,
+      "grad_norm": 0.1092528194279918,
+      "learning_rate": 4.650115501884121e-05,
+      "loss": 0.4494,
+      "step": 1281
+    },
+    {
+      "epoch": 2.518664047151277,
+      "grad_norm": 0.1321775322528286,
+      "learning_rate": 4.644700356665744e-05,
+      "loss": 0.4571,
+      "step": 1282
+    },
+    {
+      "epoch": 2.520628683693517,
+      "grad_norm": 0.11421482059533995,
+      "learning_rate": 4.6392839980958886e-05,
+      "loss": 0.4585,
+      "step": 1283
+    },
+    {
+      "epoch": 2.522593320235756,
+      "grad_norm": 0.1273808695373411,
+      "learning_rate": 4.633866436368358e-05,
+      "loss": 0.4501,
+      "step": 1284
+    },
+    {
+      "epoch": 2.524557956777996,
+      "grad_norm": 0.11813238302180325,
+      "learning_rate": 4.628447681679215e-05,
+      "loss": 0.4563,
+      "step": 1285
+    },
+    {
+      "epoch": 2.5265225933202355,
+      "grad_norm": 0.12293069774186503,
+      "learning_rate": 4.623027744226771e-05,
+      "loss": 0.4637,
+      "step": 1286
+    },
+    {
+      "epoch": 2.5284872298624754,
+      "grad_norm": 0.09725777834194362,
+      "learning_rate": 4.61760663421156e-05,
+      "loss": 0.4575,
+      "step": 1287
+    },
+    {
+      "epoch": 2.5304518664047153,
+      "grad_norm": 0.10584095101844429,
+      "learning_rate": 4.612184361836325e-05,
+      "loss": 0.449,
+      "step": 1288
+    },
+    {
+      "epoch": 2.5324165029469548,
+      "grad_norm": 0.10168012595197144,
+      "learning_rate": 4.606760937305997e-05,
+      "loss": 0.4521,
+      "step": 1289
+    },
+    {
+      "epoch": 2.5343811394891946,
+      "grad_norm": 0.10971134740773912,
+      "learning_rate": 4.6013363708276754e-05,
+      "loss": 0.4506,
+      "step": 1290
+    },
+    {
+      "epoch": 2.536345776031434,
+      "grad_norm": 0.10374847064760194,
+      "learning_rate": 4.595910672610607e-05,
+      "loss": 0.4489,
+      "step": 1291
+    },
+    {
+      "epoch": 2.538310412573674,
+      "grad_norm": 0.12051543701530733,
+      "learning_rate": 4.590483852866168e-05,
+      "loss": 0.4502,
+      "step": 1292
+    },
+    {
+      "epoch": 2.5402750491159134,
+      "grad_norm": 0.12510820800363648,
+      "learning_rate": 4.585055921807851e-05,
+      "loss": 0.4564,
+      "step": 1293
+    },
+    {
+      "epoch": 2.5422396856581533,
+      "grad_norm": 0.12667088086584108,
+      "learning_rate": 4.579626889651233e-05,
+      "loss": 0.4612,
+      "step": 1294
+    },
+    {
+      "epoch": 2.544204322200393,
+      "grad_norm": 0.12677939565018706,
+      "learning_rate": 4.5741967666139676e-05,
+      "loss": 0.4449,
+      "step": 1295
+    },
+    {
+      "epoch": 2.5461689587426326,
+      "grad_norm": 0.17358287767367722,
+      "learning_rate": 4.568765562915761e-05,
+      "loss": 0.4477,
+      "step": 1296
+    },
+    {
+      "epoch": 2.548133595284872,
+      "grad_norm": 0.10979934086499384,
+      "learning_rate": 4.56333328877835e-05,
+      "loss": 0.4512,
+      "step": 1297
+    },
+    {
+      "epoch": 2.550098231827112,
+      "grad_norm": 0.13234705879987207,
+      "learning_rate": 4.557899954425492e-05,
+      "loss": 0.4437,
+      "step": 1298
+    },
+    {
+      "epoch": 2.552062868369352,
+      "grad_norm": 0.1346799973283053,
+      "learning_rate": 4.5524655700829346e-05,
+      "loss": 0.459,
+      "step": 1299
+    },
+    {
+      "epoch": 2.5540275049115913,
+      "grad_norm": 0.11407257171484887,
+      "learning_rate": 4.5470301459784026e-05,
+      "loss": 0.4633,
+      "step": 1300
+    },
+    {
+      "epoch": 2.555992141453831,
+      "grad_norm": 0.12289565984628645,
+      "learning_rate": 4.5415936923415795e-05,
+      "loss": 0.4541,
+      "step": 1301
+    },
+    {
+      "epoch": 2.5579567779960706,
+      "grad_norm": 0.10870359351561994,
+      "learning_rate": 4.536156219404084e-05,
+      "loss": 0.4448,
+      "step": 1302
+    },
+    {
+      "epoch": 2.5599214145383105,
+      "grad_norm": 0.11872434514042761,
+      "learning_rate": 4.5307177373994566e-05,
+      "loss": 0.4541,
+      "step": 1303
+    },
+    {
+      "epoch": 2.56188605108055,
+      "grad_norm": 0.11826277095929981,
+      "learning_rate": 4.525278256563134e-05,
+      "loss": 0.4504,
+      "step": 1304
+    },
+    {
+      "epoch": 2.56385068762279,
+      "grad_norm": 0.11908195851241922,
+      "learning_rate": 4.5198377871324315e-05,
+      "loss": 0.4581,
+      "step": 1305
+    },
+    {
+      "epoch": 2.5658153241650297,
+      "grad_norm": 0.10613067536050273,
+      "learning_rate": 4.51439633934653e-05,
+      "loss": 0.4518,
+      "step": 1306
+    },
+    {
+      "epoch": 2.567779960707269,
+      "grad_norm": 0.12135857632154885,
+      "learning_rate": 4.508953923446445e-05,
+      "loss": 0.4543,
+      "step": 1307
+    },
+    {
+      "epoch": 2.569744597249509,
+      "grad_norm": 0.12174063503970278,
+      "learning_rate": 4.5035105496750216e-05,
+      "loss": 0.4507,
+      "step": 1308
+    },
+    {
+      "epoch": 2.5717092337917484,
+      "grad_norm": 0.0970595608118759,
+      "learning_rate": 4.4980662282769024e-05,
+      "loss": 0.4447,
+      "step": 1309
+    },
+    {
+      "epoch": 2.5736738703339883,
+      "grad_norm": 0.129894016835485,
+      "learning_rate": 4.4926209694985134e-05,
+      "loss": 0.4643,
+      "step": 1310
+    },
+    {
+      "epoch": 2.5756385068762278,
+      "grad_norm": 0.12290458882836702,
+      "learning_rate": 4.4871747835880466e-05,
+      "loss": 0.4521,
+      "step": 1311
+    },
+    {
+      "epoch": 2.5776031434184676,
+      "grad_norm": 0.11990299782379676,
+      "learning_rate": 4.481727680795441e-05,
+      "loss": 0.4613,
+      "step": 1312
+    },
+    {
+      "epoch": 2.5795677799607075,
+      "grad_norm": 0.1153754236521546,
+      "learning_rate": 4.4762796713723555e-05,
+      "loss": 0.4523,
+      "step": 1313
+    },
+    {
+      "epoch": 2.581532416502947,
+      "grad_norm": 0.1189782054763776,
+      "learning_rate": 4.4708307655721604e-05,
+      "loss": 0.4463,
+      "step": 1314
+    },
+    {
+      "epoch": 2.5834970530451864,
+      "grad_norm": 0.1285914540083342,
+      "learning_rate": 4.465380973649909e-05,
+      "loss": 0.4573,
+      "step": 1315
+    },
+    {
+      "epoch": 2.5854616895874263,
+      "grad_norm": 0.10439894735578753,
+      "learning_rate": 4.459930305862327e-05,
+      "loss": 0.4575,
+      "step": 1316
+    },
+    {
+      "epoch": 2.587426326129666,
+      "grad_norm": 0.1141517453981676,
+      "learning_rate": 4.454478772467786e-05,
+      "loss": 0.4593,
+      "step": 1317
+    },
+    {
+      "epoch": 2.5893909626719056,
+      "grad_norm": 0.11482028078892331,
+      "learning_rate": 4.449026383726284e-05,
+      "loss": 0.457,
+      "step": 1318
+    },
+    {
+      "epoch": 2.5913555992141455,
+      "grad_norm": 0.13037988259965066,
+      "learning_rate": 4.443573149899434e-05,
+      "loss": 0.4527,
+      "step": 1319
+    },
+    {
+      "epoch": 2.593320235756385,
+      "grad_norm": 0.11073948715272945,
+      "learning_rate": 4.4381190812504375e-05,
+      "loss": 0.4524,
+      "step": 1320
+    },
+    {
+      "epoch": 2.595284872298625,
+      "grad_norm": 0.10218588041008418,
+      "learning_rate": 4.432664188044061e-05,
+      "loss": 0.4483,
+      "step": 1321
+    },
+    {
+      "epoch": 2.5972495088408643,
+      "grad_norm": 0.12186130808696936,
+      "learning_rate": 4.4272084805466366e-05,
+      "loss": 0.46,
+      "step": 1322
+    },
+    {
+      "epoch": 2.599214145383104,
+      "grad_norm": 0.10594543870886707,
+      "learning_rate": 4.421751969026016e-05,
+      "loss": 0.453,
+      "step": 1323
+    },
+    {
+      "epoch": 2.601178781925344,
+      "grad_norm": 0.10908309310550862,
+      "learning_rate": 4.416294663751569e-05,
+      "loss": 0.4557,
+      "step": 1324
+    },
+    {
+      "epoch": 2.6031434184675835,
+      "grad_norm": 0.10330986581468594,
+      "learning_rate": 4.4108365749941616e-05,
+      "loss": 0.4517,
+      "step": 1325
+    },
+    {
+      "epoch": 2.6051080550098233,
+      "grad_norm": 0.10397146491179449,
+      "learning_rate": 4.405377713026128e-05,
+      "loss": 0.4455,
+      "step": 1326
+    },
+    {
+      "epoch": 2.607072691552063,
+      "grad_norm": 0.10228199996978345,
+      "learning_rate": 4.399918088121266e-05,
+      "loss": 0.4509,
+      "step": 1327
+    },
+    {
+      "epoch": 2.6090373280943027,
+      "grad_norm": 0.0969282485100846,
+      "learning_rate": 4.394457710554801e-05,
+      "loss": 0.4535,
+      "step": 1328
+    },
+    {
+      "epoch": 2.611001964636542,
+      "grad_norm": 0.10569132998847348,
+      "learning_rate": 4.388996590603381e-05,
+      "loss": 0.4508,
+      "step": 1329
+    },
+    {
+      "epoch": 2.612966601178782,
+      "grad_norm": 0.11480708113895728,
+      "learning_rate": 4.3835347385450487e-05,
+      "loss": 0.4564,
+      "step": 1330
+    },
+    {
+      "epoch": 2.614931237721022,
+      "grad_norm": 0.10902554270941398,
+      "learning_rate": 4.378072164659224e-05,
+      "loss": 0.4385,
+      "step": 1331
+    },
+    {
+      "epoch": 2.6168958742632613,
+      "grad_norm": 0.10224832787481691,
+      "learning_rate": 4.372608879226688e-05,
+      "loss": 0.4498,
+      "step": 1332
+    },
+    {
+      "epoch": 2.6188605108055008,
+      "grad_norm": 0.1063144416160829,
+      "learning_rate": 4.367144892529558e-05,
+      "loss": 0.4463,
+      "step": 1333
+    },
+    {
+      "epoch": 2.6208251473477406,
+      "grad_norm": 0.11245570346920764,
+      "learning_rate": 4.3616802148512703e-05,
+      "loss": 0.448,
+      "step": 1334
+    },
+    {
+      "epoch": 2.6227897838899805,
+      "grad_norm": 0.11262181628618531,
+      "learning_rate": 4.356214856476567e-05,
+      "loss": 0.4466,
+      "step": 1335
+    },
+    {
+      "epoch": 2.62475442043222,
+      "grad_norm": 0.10642202038631403,
+      "learning_rate": 4.350748827691465e-05,
+      "loss": 0.4547,
+      "step": 1336
+    },
+    {
+      "epoch": 2.62671905697446,
+      "grad_norm": 0.11743196019142667,
+      "learning_rate": 4.3452821387832496e-05,
+      "loss": 0.4541,
+      "step": 1337
+    },
+    {
+      "epoch": 2.6286836935166993,
+      "grad_norm": 0.11954175585732266,
+      "learning_rate": 4.339814800040442e-05,
+      "loss": 0.4443,
+      "step": 1338
+    },
+    {
+      "epoch": 2.630648330058939,
+      "grad_norm": 0.10001849011752642,
+      "learning_rate": 4.334346821752789e-05,
+      "loss": 0.4461,
+      "step": 1339
+    },
+    {
+      "epoch": 2.6326129666011786,
+      "grad_norm": 0.10924304921245696,
+      "learning_rate": 4.3288782142112425e-05,
+      "loss": 0.4496,
+      "step": 1340
+    },
+    {
+      "epoch": 2.6345776031434185,
+      "grad_norm": 0.12176622708601406,
+      "learning_rate": 4.323408987707937e-05,
+      "loss": 0.4531,
+      "step": 1341
+    },
+    {
+      "epoch": 2.6365422396856584,
+      "grad_norm": 0.10754298077289823,
+      "learning_rate": 4.317939152536171e-05,
+      "loss": 0.4511,
+      "step": 1342
+    },
+    {
+      "epoch": 2.638506876227898,
+      "grad_norm": 0.10386588274994592,
+      "learning_rate": 4.312468718990392e-05,
+      "loss": 0.4574,
+      "step": 1343
+    },
+    {
+      "epoch": 2.6404715127701373,
+      "grad_norm": 0.11046825752423814,
+      "learning_rate": 4.3069976973661697e-05,
+      "loss": 0.4549,
+      "step": 1344
+    },
+    {
+      "epoch": 2.642436149312377,
+      "grad_norm": 0.09955435793630453,
+      "learning_rate": 4.3015260979601835e-05,
+      "loss": 0.4411,
+      "step": 1345
+    },
+    {
+      "epoch": 2.644400785854617,
+      "grad_norm": 0.0968243242671388,
+      "learning_rate": 4.296053931070199e-05,
+      "loss": 0.445,
+      "step": 1346
+    },
+    {
+      "epoch": 2.6463654223968565,
+      "grad_norm": 0.10330258695749844,
+      "learning_rate": 4.290581206995049e-05,
+      "loss": 0.4501,
+      "step": 1347
+    },
+    {
+      "epoch": 2.6483300589390963,
+      "grad_norm": 0.10568502875579827,
+      "learning_rate": 4.285107936034618e-05,
+      "loss": 0.4491,
+      "step": 1348
+    },
+    {
+      "epoch": 2.6502946954813362,
+      "grad_norm": 0.1068111845565804,
+      "learning_rate": 4.2796341284898144e-05,
+      "loss": 0.4522,
+      "step": 1349
+    },
+    {
+      "epoch": 2.6522593320235757,
+      "grad_norm": 0.10813072501004757,
+      "learning_rate": 4.274159794662563e-05,
+      "loss": 0.4476,
+      "step": 1350
+    },
+    {
+      "epoch": 2.654223968565815,
+      "grad_norm": 0.11011743477253844,
+      "learning_rate": 4.2686849448557736e-05,
+      "loss": 0.4542,
+      "step": 1351
+    },
+    {
+      "epoch": 2.656188605108055,
+      "grad_norm": 0.13530020291385766,
+      "learning_rate": 4.26320958937333e-05,
+      "loss": 0.4496,
+      "step": 1352
+    },
+    {
+      "epoch": 2.658153241650295,
+      "grad_norm": 0.1148079213258007,
+      "learning_rate": 4.2577337385200665e-05,
+      "loss": 0.4571,
+      "step": 1353
+    },
+    {
+      "epoch": 2.6601178781925343,
+      "grad_norm": 0.08931464508601375,
+      "learning_rate": 4.252257402601752e-05,
+      "loss": 0.4546,
+      "step": 1354
+    },
+    {
+      "epoch": 2.662082514734774,
+      "grad_norm": 0.10184627433139114,
+      "learning_rate": 4.246780591925063e-05,
+      "loss": 0.4482,
+      "step": 1355
+    },
+    {
+      "epoch": 2.6640471512770136,
+      "grad_norm": 0.12115401811525842,
+      "learning_rate": 4.2413033167975774e-05,
+      "loss": 0.447,
+      "step": 1356
+    },
+    {
+      "epoch": 2.6660117878192535,
+      "grad_norm": 0.12519391970924196,
+      "learning_rate": 4.235825587527739e-05,
+      "loss": 0.4517,
+      "step": 1357
+    },
+    {
+      "epoch": 2.667976424361493,
+      "grad_norm": 0.13849231139273213,
+      "learning_rate": 4.230347414424853e-05,
+      "loss": 0.4528,
+      "step": 1358
+    },
+    {
+      "epoch": 2.669941060903733,
+      "grad_norm": 0.11268031288116975,
+      "learning_rate": 4.2248688077990546e-05,
+      "loss": 0.4489,
+      "step": 1359
+    },
+    {
+      "epoch": 2.6719056974459727,
+      "grad_norm": 0.13940926709406085,
+      "learning_rate": 4.219389777961299e-05,
+      "loss": 0.4488,
+      "step": 1360
+    },
+    {
+      "epoch": 2.673870333988212,
+      "grad_norm": 0.11377240863784405,
+      "learning_rate": 4.2139103352233365e-05,
+      "loss": 0.4519,
+      "step": 1361
+    },
+    {
+      "epoch": 2.6758349705304516,
+      "grad_norm": 0.14512477335934879,
+      "learning_rate": 4.2084304898976935e-05,
+      "loss": 0.454,
+      "step": 1362
+    },
+    {
+      "epoch": 2.6777996070726915,
+      "grad_norm": 0.13571908939978114,
+      "learning_rate": 4.202950252297654e-05,
+      "loss": 0.4485,
+      "step": 1363
+    },
+    {
+      "epoch": 2.6797642436149314,
+      "grad_norm": 0.12673712904405332,
+      "learning_rate": 4.1974696327372426e-05,
+      "loss": 0.4479,
+      "step": 1364
+    },
+    {
+      "epoch": 2.681728880157171,
+      "grad_norm": 0.10897119916958024,
+      "learning_rate": 4.1919886415312004e-05,
+      "loss": 0.4479,
+      "step": 1365
+    },
+    {
+      "epoch": 2.6836935166994107,
+      "grad_norm": 0.1352703393109706,
+      "learning_rate": 4.186507288994969e-05,
+      "loss": 0.4661,
+      "step": 1366
+    },
+    {
+      "epoch": 2.68565815324165,
+      "grad_norm": 0.11795776255005916,
+      "learning_rate": 4.181025585444672e-05,
+      "loss": 0.4557,
+      "step": 1367
+    },
+    {
+      "epoch": 2.68762278978389,
+      "grad_norm": 0.13029081392676764,
+      "learning_rate": 4.175543541197088e-05,
+      "loss": 0.4519,
+      "step": 1368
+    },
+    {
+      "epoch": 2.6895874263261295,
+      "grad_norm": 0.12200539806055082,
+      "learning_rate": 4.170061166569641e-05,
+      "loss": 0.4506,
+      "step": 1369
+    },
+    {
+      "epoch": 2.6915520628683693,
+      "grad_norm": 0.11727806404902322,
+      "learning_rate": 4.164578471880378e-05,
+      "loss": 0.4532,
+      "step": 1370
+    },
+    {
+      "epoch": 2.6935166994106092,
+      "grad_norm": 0.11149117772395238,
+      "learning_rate": 4.159095467447945e-05,
+      "loss": 0.4438,
+      "step": 1371
+    },
+    {
+      "epoch": 2.6954813359528487,
+      "grad_norm": 0.09549885021314664,
+      "learning_rate": 4.153612163591574e-05,
+      "loss": 0.4453,
+      "step": 1372
+    },
+    {
+      "epoch": 2.6974459724950886,
+      "grad_norm": 0.11234054421279305,
+      "learning_rate": 4.1481285706310564e-05,
+      "loss": 0.452,
+      "step": 1373
+    },
+    {
+      "epoch": 2.699410609037328,
+      "grad_norm": 0.11007088557880927,
+      "learning_rate": 4.142644698886732e-05,
+      "loss": 0.4545,
+      "step": 1374
+    },
+    {
+      "epoch": 2.701375245579568,
+      "grad_norm": 0.11255601207348559,
+      "learning_rate": 4.137160558679463e-05,
+      "loss": 0.4483,
+      "step": 1375
+    },
+    {
+      "epoch": 2.7033398821218073,
+      "grad_norm": 0.11220930378359184,
+      "learning_rate": 4.131676160330618e-05,
+      "loss": 0.4516,
+      "step": 1376
+    },
+    {
+      "epoch": 2.705304518664047,
+      "grad_norm": 0.1015377817951622,
+      "learning_rate": 4.126191514162051e-05,
+      "loss": 0.452,
+      "step": 1377
+    },
+    {
+      "epoch": 2.707269155206287,
+      "grad_norm": 0.10692611265574811,
+      "learning_rate": 4.1207066304960797e-05,
+      "loss": 0.4509,
+      "step": 1378
+    },
+    {
+      "epoch": 2.7092337917485265,
+      "grad_norm": 0.09369608614282308,
+      "learning_rate": 4.115221519655474e-05,
+      "loss": 0.4544,
+      "step": 1379
+    },
+    {
+      "epoch": 2.711198428290766,
+      "grad_norm": 0.11603169131937906,
+      "learning_rate": 4.109736191963427e-05,
+      "loss": 0.456,
+      "step": 1380
+    },
+    {
+      "epoch": 2.713163064833006,
+      "grad_norm": 0.11033352350284896,
+      "learning_rate": 4.104250657743544e-05,
+      "loss": 0.4449,
+      "step": 1381
+    },
+    {
+      "epoch": 2.7151277013752457,
+      "grad_norm": 0.10172302108929263,
+      "learning_rate": 4.0987649273198145e-05,
+      "loss": 0.4493,
+      "step": 1382
+    },
+    {
+      "epoch": 2.717092337917485,
+      "grad_norm": 0.1355770120308079,
+      "learning_rate": 4.093279011016599e-05,
+      "loss": 0.4616,
+      "step": 1383
+    },
+    {
+      "epoch": 2.719056974459725,
+      "grad_norm": 0.12595643230430992,
+      "learning_rate": 4.087792919158608e-05,
+      "loss": 0.4489,
+      "step": 1384
+    },
+    {
+      "epoch": 2.7210216110019645,
+      "grad_norm": 0.13062717110995545,
+      "learning_rate": 4.082306662070887e-05,
+      "loss": 0.4467,
+      "step": 1385
+    },
+    {
+      "epoch": 2.7229862475442044,
+      "grad_norm": 0.13156374823442019,
+      "learning_rate": 4.0768202500787806e-05,
+      "loss": 0.4552,
+      "step": 1386
+    },
+    {
+      "epoch": 2.724950884086444,
+      "grad_norm": 0.14202562918295158,
+      "learning_rate": 4.0713336935079393e-05,
+      "loss": 0.4575,
+      "step": 1387
+    },
+    {
+      "epoch": 2.7269155206286837,
+      "grad_norm": 0.13102261933888473,
+      "learning_rate": 4.065847002684274e-05,
+      "loss": 0.4591,
+      "step": 1388
+    },
+    {
+      "epoch": 2.7288801571709236,
+      "grad_norm": 0.14284096785429903,
+      "learning_rate": 4.060360187933955e-05,
+      "loss": 0.4518,
+      "step": 1389
+    },
+    {
+      "epoch": 2.730844793713163,
+      "grad_norm": 0.146447478385908,
+      "learning_rate": 4.054873259583385e-05,
+      "loss": 0.4456,
+      "step": 1390
+    },
+    {
+      "epoch": 2.732809430255403,
+      "grad_norm": 0.10955186932104283,
+      "learning_rate": 4.0493862279591764e-05,
+      "loss": 0.441,
+      "step": 1391
+    },
+    {
+      "epoch": 2.7347740667976423,
+      "grad_norm": 0.11557584333462476,
+      "learning_rate": 4.0438991033881425e-05,
+      "loss": 0.4405,
+      "step": 1392
+    },
+    {
+      "epoch": 2.7367387033398822,
+      "grad_norm": 0.13047104967618472,
+      "learning_rate": 4.038411896197264e-05,
+      "loss": 0.4505,
+      "step": 1393
+    },
+    {
+      "epoch": 2.7387033398821217,
+      "grad_norm": 0.14557980111848967,
+      "learning_rate": 4.0329246167136845e-05,
+      "loss": 0.4508,
+      "step": 1394
+    },
+    {
+      "epoch": 2.7406679764243616,
+      "grad_norm": 0.11573602902526445,
+      "learning_rate": 4.027437275264679e-05,
+      "loss": 0.4581,
+      "step": 1395
+    },
+    {
+      "epoch": 2.7426326129666014,
+      "grad_norm": 0.13966097953941176,
+      "learning_rate": 4.02194988217764e-05,
+      "loss": 0.4508,
+      "step": 1396
+    },
+    {
+      "epoch": 2.744597249508841,
+      "grad_norm": 0.13309599576931744,
+      "learning_rate": 4.0164624477800576e-05,
+      "loss": 0.4434,
+      "step": 1397
+    },
+    {
+      "epoch": 2.7465618860510803,
+      "grad_norm": 0.12625827770174874,
+      "learning_rate": 4.010974982399499e-05,
+      "loss": 0.4586,
+      "step": 1398
+    },
+    {
+      "epoch": 2.74852652259332,
+      "grad_norm": 0.13317693448784454,
+      "learning_rate": 4.005487496363592e-05,
+      "loss": 0.4571,
+      "step": 1399
+    },
+    {
+      "epoch": 2.75049115913556,
+      "grad_norm": 0.10038735303436275,
+      "learning_rate": 4e-05,
+      "loss": 0.4502,
+      "step": 1400
+    },
+    {
+      "epoch": 2.7524557956777995,
+      "grad_norm": 0.12898368854865636,
+      "learning_rate": 3.9945125036364086e-05,
+      "loss": 0.4452,
+      "step": 1401
+    },
+    {
+      "epoch": 2.7544204322200394,
+      "grad_norm": 0.10410302837101625,
+      "learning_rate": 3.989025017600502e-05,
+      "loss": 0.4413,
+      "step": 1402
+    },
+    {
+      "epoch": 2.756385068762279,
+      "grad_norm": 0.11503844738898163,
+      "learning_rate": 3.983537552219944e-05,
+      "loss": 0.4545,
+      "step": 1403
+    },
+    {
+      "epoch": 2.7583497053045187,
+      "grad_norm": 0.1189473096357218,
+      "learning_rate": 3.9780501178223605e-05,
+      "loss": 0.4547,
+      "step": 1404
+    },
+    {
+      "epoch": 2.760314341846758,
+      "grad_norm": 0.10734786835827485,
+      "learning_rate": 3.972562724735323e-05,
+      "loss": 0.4512,
+      "step": 1405
+    },
+    {
+      "epoch": 2.762278978388998,
+      "grad_norm": 0.12524428225254686,
+      "learning_rate": 3.967075383286317e-05,
+      "loss": 0.4469,
+      "step": 1406
+    },
+    {
+      "epoch": 2.764243614931238,
+      "grad_norm": 0.11241276952926854,
+      "learning_rate": 3.961588103802737e-05,
+      "loss": 0.4443,
+      "step": 1407
+    },
+    {
+      "epoch": 2.7662082514734774,
+      "grad_norm": 0.11246906611876296,
+      "learning_rate": 3.956100896611859e-05,
+      "loss": 0.4474,
+      "step": 1408
+    },
+    {
+      "epoch": 2.768172888015717,
+      "grad_norm": 0.13672617109155705,
+      "learning_rate": 3.950613772040825e-05,
+      "loss": 0.4504,
+      "step": 1409
+    },
+    {
+      "epoch": 2.7701375245579567,
+      "grad_norm": 0.11695065789558451,
+      "learning_rate": 3.9451267404166164e-05,
+      "loss": 0.4443,
+      "step": 1410
+    },
+    {
+      "epoch": 2.7721021611001966,
+      "grad_norm": 0.1257321014995227,
+      "learning_rate": 3.939639812066045e-05,
+      "loss": 0.448,
+      "step": 1411
+    },
+    {
+      "epoch": 2.774066797642436,
+      "grad_norm": 0.11592192590605466,
+      "learning_rate": 3.9341529973157266e-05,
+      "loss": 0.4467,
+      "step": 1412
+    },
+    {
+      "epoch": 2.776031434184676,
+      "grad_norm": 0.11509834221449342,
+      "learning_rate": 3.928666306492062e-05,
+      "loss": 0.4539,
+      "step": 1413
+    },
+    {
+      "epoch": 2.777996070726916,
+      "grad_norm": 0.09776360619609353,
+      "learning_rate": 3.923179749921219e-05,
+      "loss": 0.4426,
+      "step": 1414
+    },
+    {
+      "epoch": 2.7799607072691552,
+      "grad_norm": 0.11629927597625905,
+      "learning_rate": 3.917693337929116e-05,
+      "loss": 0.456,
+      "step": 1415
+    },
+    {
+      "epoch": 2.7819253438113947,
+      "grad_norm": 0.10789102122908144,
+      "learning_rate": 3.9122070808413924e-05,
+      "loss": 0.4458,
+      "step": 1416
+    },
+    {
+      "epoch": 2.7838899803536346,
+      "grad_norm": 0.11919726268799906,
+      "learning_rate": 3.906720988983402e-05,
+      "loss": 0.4577,
+      "step": 1417
+    },
+    {
+      "epoch": 2.7858546168958744,
+      "grad_norm": 0.10691019544171991,
+      "learning_rate": 3.9012350726801875e-05,
+      "loss": 0.4489,
+      "step": 1418
+    },
+    {
+      "epoch": 2.787819253438114,
+      "grad_norm": 0.13783082143303516,
+      "learning_rate": 3.895749342256457e-05,
+      "loss": 0.4506,
+      "step": 1419
+    },
+    {
+      "epoch": 2.7897838899803538,
+      "grad_norm": 0.1229835115765077,
+      "learning_rate": 3.8902638080365734e-05,
+      "loss": 0.4432,
+      "step": 1420
+    },
+    {
+      "epoch": 2.791748526522593,
+      "grad_norm": 0.1230682965227979,
+      "learning_rate": 3.884778480344527e-05,
+      "loss": 0.4554,
+      "step": 1421
+    },
+    {
+      "epoch": 2.793713163064833,
+      "grad_norm": 0.11036657831505155,
+      "learning_rate": 3.879293369503922e-05,
+      "loss": 0.4496,
+      "step": 1422
+    },
+    {
+      "epoch": 2.7956777996070725,
+      "grad_norm": 0.14020452220530366,
+      "learning_rate": 3.873808485837951e-05,
+      "loss": 0.4533,
+      "step": 1423
+    },
+    {
+      "epoch": 2.7976424361493124,
+      "grad_norm": 0.11407113419221702,
+      "learning_rate": 3.868323839669383e-05,
+      "loss": 0.4552,
+      "step": 1424
+    },
+    {
+      "epoch": 2.7996070726915523,
+      "grad_norm": 0.11549964512165785,
+      "learning_rate": 3.862839441320539e-05,
+      "loss": 0.459,
+      "step": 1425
+    },
+    {
+      "epoch": 2.8015717092337917,
+      "grad_norm": 0.14557353448658092,
+      "learning_rate": 3.8573553011132694e-05,
+      "loss": 0.4584,
+      "step": 1426
+    },
+    {
+      "epoch": 2.803536345776031,
+      "grad_norm": 0.12701388916425213,
+      "learning_rate": 3.851871429368945e-05,
+      "loss": 0.4579,
+      "step": 1427
+    },
+    {
+      "epoch": 2.805500982318271,
+      "grad_norm": 0.11286886617369254,
+      "learning_rate": 3.846387836408428e-05,
+      "loss": 0.4588,
+      "step": 1428
+    },
+    {
+      "epoch": 2.807465618860511,
+      "grad_norm": 0.11088949711659635,
+      "learning_rate": 3.8409045325520556e-05,
+      "loss": 0.4598,
+      "step": 1429
+    },
+    {
+      "epoch": 2.8094302554027504,
+      "grad_norm": 0.13523141804542396,
+      "learning_rate": 3.835421528119623e-05,
+      "loss": 0.4497,
+      "step": 1430
+    },
+    {
+      "epoch": 2.8113948919449903,
+      "grad_norm": 0.10573352500820175,
+      "learning_rate": 3.8299388334303604e-05,
+      "loss": 0.4489,
+      "step": 1431
+    },
+    {
+      "epoch": 2.8133595284872297,
+      "grad_norm": 0.12547341128911474,
+      "learning_rate": 3.8244564588029135e-05,
+      "loss": 0.4471,
+      "step": 1432
+    },
+    {
+      "epoch": 2.8153241650294696,
+      "grad_norm": 0.12428591073022045,
+      "learning_rate": 3.818974414555329e-05,
+      "loss": 0.4524,
+      "step": 1433
+    },
+    {
+      "epoch": 2.817288801571709,
+      "grad_norm": 0.11101658190876541,
+      "learning_rate": 3.813492711005033e-05,
+      "loss": 0.448,
+      "step": 1434
+    },
+    {
+      "epoch": 2.819253438113949,
+      "grad_norm": 0.10971083845653325,
+      "learning_rate": 3.808011358468801e-05,
+      "loss": 0.4527,
+      "step": 1435
+    },
+    {
+      "epoch": 2.821218074656189,
+      "grad_norm": 0.10211830655625893,
+      "learning_rate": 3.802530367262759e-05,
+      "loss": 0.4517,
+      "step": 1436
+    },
+    {
+      "epoch": 2.8231827111984282,
+      "grad_norm": 0.11089279306205296,
+      "learning_rate": 3.797049747702346e-05,
+      "loss": 0.4474,
+      "step": 1437
+    },
+    {
+      "epoch": 2.825147347740668,
+      "grad_norm": 0.10870688608439892,
+      "learning_rate": 3.7915695101023085e-05,
+      "loss": 0.4434,
+      "step": 1438
+    },
+    {
+      "epoch": 2.8271119842829076,
+      "grad_norm": 0.10574224766306652,
+      "learning_rate": 3.786089664776665e-05,
+      "loss": 0.4477,
+      "step": 1439
+    },
+    {
+      "epoch": 2.8290766208251474,
+      "grad_norm": 0.12234046180185697,
+      "learning_rate": 3.780610222038701e-05,
+      "loss": 0.4473,
+      "step": 1440
+    },
+    {
+      "epoch": 2.831041257367387,
+      "grad_norm": 0.11078634402155785,
+      "learning_rate": 3.775131192200947e-05,
+      "loss": 0.4525,
+      "step": 1441
+    },
+    {
+      "epoch": 2.8330058939096268,
+      "grad_norm": 0.12999435554710348,
+      "learning_rate": 3.769652585575148e-05,
+      "loss": 0.4505,
+      "step": 1442
+    },
+    {
+      "epoch": 2.8349705304518666,
+      "grad_norm": 0.12159932258377415,
+      "learning_rate": 3.764174412472261e-05,
+      "loss": 0.4548,
+      "step": 1443
+    },
+    {
+      "epoch": 2.836935166994106,
+      "grad_norm": 0.10851283411236887,
+      "learning_rate": 3.7586966832024247e-05,
+      "loss": 0.4456,
+      "step": 1444
+    },
+    {
+      "epoch": 2.8388998035363455,
+      "grad_norm": 0.10777954722879636,
+      "learning_rate": 3.7532194080749375e-05,
+      "loss": 0.4523,
+      "step": 1445
+    },
+    {
+      "epoch": 2.8408644400785854,
+      "grad_norm": 0.11405662965483065,
+      "learning_rate": 3.74774259739825e-05,
+      "loss": 0.4476,
+      "step": 1446
+    },
+    {
+      "epoch": 2.8428290766208253,
+      "grad_norm": 0.10115990027651815,
+      "learning_rate": 3.742266261479935e-05,
+      "loss": 0.4478,
+      "step": 1447
+    },
+    {
+      "epoch": 2.8447937131630647,
+      "grad_norm": 0.09974627139176301,
+      "learning_rate": 3.736790410626672e-05,
+      "loss": 0.4436,
+      "step": 1448
+    },
+    {
+      "epoch": 2.8467583497053046,
+      "grad_norm": 0.12488631428567294,
+      "learning_rate": 3.731315055144228e-05,
+      "loss": 0.445,
+      "step": 1449
+    },
+    {
+      "epoch": 2.848722986247544,
+      "grad_norm": 0.09225599734118611,
+      "learning_rate": 3.725840205337438e-05,
+      "loss": 0.4466,
+      "step": 1450
+    },
+    {
+      "epoch": 2.850687622789784,
+      "grad_norm": 0.10569321204138098,
+      "learning_rate": 3.720365871510186e-05,
+      "loss": 0.4571,
+      "step": 1451
+    },
+    {
+      "epoch": 2.8526522593320234,
+      "grad_norm": 0.10622050793892152,
+      "learning_rate": 3.714892063965383e-05,
+      "loss": 0.4472,
+      "step": 1452
+    },
+    {
+      "epoch": 2.8546168958742633,
+      "grad_norm": 0.09646053638538776,
+      "learning_rate": 3.709418793004951e-05,
+      "loss": 0.4564,
+      "step": 1453
+    },
+    {
+      "epoch": 2.856581532416503,
+      "grad_norm": 0.11543256352994033,
+      "learning_rate": 3.7039460689298025e-05,
+      "loss": 0.4534,
+      "step": 1454
+    },
+    {
+      "epoch": 2.8585461689587426,
+      "grad_norm": 0.10616162353041407,
+      "learning_rate": 3.698473902039818e-05,
+      "loss": 0.4481,
+      "step": 1455
+    },
+    {
+      "epoch": 2.8605108055009825,
+      "grad_norm": 0.09321367102826979,
+      "learning_rate": 3.69300230263383e-05,
+      "loss": 0.4524,
+      "step": 1456
+    },
+    {
+      "epoch": 2.862475442043222,
+      "grad_norm": 0.098935202598614,
+      "learning_rate": 3.68753128100961e-05,
+      "loss": 0.4551,
+      "step": 1457
+    },
+    {
+      "epoch": 2.864440078585462,
+      "grad_norm": 0.08546926203571314,
+      "learning_rate": 3.6820608474638307e-05,
+      "loss": 0.4521,
+      "step": 1458
+    },
+    {
+      "epoch": 2.8664047151277012,
+      "grad_norm": 0.11025984360012024,
+      "learning_rate": 3.676591012292065e-05,
+      "loss": 0.4515,
+      "step": 1459
+    },
+    {
+      "epoch": 2.868369351669941,
+      "grad_norm": 0.1005201506505379,
+      "learning_rate": 3.671121785788759e-05,
+      "loss": 0.4605,
+      "step": 1460
+    },
+    {
+      "epoch": 2.870333988212181,
+      "grad_norm": 0.11293908761265267,
+      "learning_rate": 3.665653178247212e-05,
+      "loss": 0.4644,
+      "step": 1461
+    },
+    {
+      "epoch": 2.8722986247544204,
+      "grad_norm": 0.11424041602091137,
+      "learning_rate": 3.660185199959559e-05,
+      "loss": 0.4511,
+      "step": 1462
+    },
+    {
+      "epoch": 2.87426326129666,
+      "grad_norm": 0.10364212060761462,
+      "learning_rate": 3.6547178612167504e-05,
+      "loss": 0.4526,
+      "step": 1463
+    },
+    {
+      "epoch": 2.8762278978388998,
+      "grad_norm": 0.11378999442738272,
+      "learning_rate": 3.6492511723085356e-05,
+      "loss": 0.4523,
+      "step": 1464
+    },
+    {
+      "epoch": 2.8781925343811396,
+      "grad_norm": 0.102361736307065,
+      "learning_rate": 3.6437851435234346e-05,
+      "loss": 0.4532,
+      "step": 1465
+    },
+    {
+      "epoch": 2.880157170923379,
+      "grad_norm": 0.10146698354738314,
+      "learning_rate": 3.6383197851487296e-05,
+      "loss": 0.4518,
+      "step": 1466
+    },
+    {
+      "epoch": 2.882121807465619,
+      "grad_norm": 0.10213440004769589,
+      "learning_rate": 3.6328551074704446e-05,
+      "loss": 0.4579,
+      "step": 1467
+    },
+    {
+      "epoch": 2.8840864440078584,
+      "grad_norm": 0.12011035689618098,
+      "learning_rate": 3.627391120773313e-05,
+      "loss": 0.4443,
+      "step": 1468
+    },
+    {
+      "epoch": 2.8860510805500983,
+      "grad_norm": 0.10369716127729628,
+      "learning_rate": 3.621927835340776e-05,
+      "loss": 0.4491,
+      "step": 1469
+    },
+    {
+      "epoch": 2.8880157170923377,
+      "grad_norm": 0.10162162761948966,
+      "learning_rate": 3.616465261454953e-05,
+      "loss": 0.4576,
+      "step": 1470
+    },
+    {
+      "epoch": 2.8899803536345776,
+      "grad_norm": 0.09963247858111783,
+      "learning_rate": 3.6110034093966194e-05,
+      "loss": 0.4515,
+      "step": 1471
+    },
+    {
+      "epoch": 2.8919449901768175,
+      "grad_norm": 0.10258441869370113,
+      "learning_rate": 3.6055422894451994e-05,
+      "loss": 0.4475,
+      "step": 1472
+    },
+    {
+      "epoch": 2.893909626719057,
+      "grad_norm": 0.1120442814331307,
+      "learning_rate": 3.600081911878736e-05,
+      "loss": 0.4523,
+      "step": 1473
+    },
+    {
+      "epoch": 2.895874263261297,
+      "grad_norm": 0.10217897621911382,
+      "learning_rate": 3.5946222869738734e-05,
+      "loss": 0.4529,
+      "step": 1474
+    },
+    {
+      "epoch": 2.8978388998035363,
+      "grad_norm": 0.1000686110327807,
+      "learning_rate": 3.58916342500584e-05,
+      "loss": 0.4485,
+      "step": 1475
+    },
+    {
+      "epoch": 2.899803536345776,
+      "grad_norm": 0.09898035632081392,
+      "learning_rate": 3.58370533624843e-05,
+      "loss": 0.4455,
+      "step": 1476
+    },
+    {
+      "epoch": 2.9017681728880156,
+      "grad_norm": 0.09577935150761908,
+      "learning_rate": 3.5782480309739854e-05,
+      "loss": 0.447,
+      "step": 1477
+    },
+    {
+      "epoch": 2.9037328094302555,
+      "grad_norm": 0.09474679814951903,
+      "learning_rate": 3.572791519453364e-05,
+      "loss": 0.4562,
+      "step": 1478
+    },
+    {
+      "epoch": 2.9056974459724954,
+      "grad_norm": 0.102740589196727,
+      "learning_rate": 3.567335811955938e-05,
+      "loss": 0.4491,
+      "step": 1479
+    },
+    {
+      "epoch": 2.907662082514735,
+      "grad_norm": 0.105011059349859,
+      "learning_rate": 3.561880918749565e-05,
+      "loss": 0.4507,
+      "step": 1480
+    },
+    {
+      "epoch": 2.9096267190569742,
+      "grad_norm": 0.10001945461106931,
+      "learning_rate": 3.556426850100567e-05,
+      "loss": 0.4478,
+      "step": 1481
+    },
+    {
+      "epoch": 2.911591355599214,
+      "grad_norm": 0.09804028208337448,
+      "learning_rate": 3.550973616273716e-05,
+      "loss": 0.4445,
+      "step": 1482
+    },
+    {
+      "epoch": 2.913555992141454,
+      "grad_norm": 0.10304002323419238,
+      "learning_rate": 3.545521227532216e-05,
+      "loss": 0.4524,
+      "step": 1483
+    },
+    {
+      "epoch": 2.9155206286836934,
+      "grad_norm": 0.10932594404005899,
+      "learning_rate": 3.540069694137674e-05,
+      "loss": 0.463,
+      "step": 1484
+    },
+    {
+      "epoch": 2.9174852652259333,
+      "grad_norm": 0.11199398078223533,
+      "learning_rate": 3.534619026350091e-05,
+      "loss": 0.4487,
+      "step": 1485
+    },
+    {
+      "epoch": 2.9194499017681728,
+      "grad_norm": 0.12179148890578316,
+      "learning_rate": 3.529169234427842e-05,
+      "loss": 0.4562,
+      "step": 1486
+    },
+    {
+      "epoch": 2.9214145383104126,
+      "grad_norm": 0.10304186053956522,
+      "learning_rate": 3.5237203286276465e-05,
+      "loss": 0.4494,
+      "step": 1487
+    },
+    {
+      "epoch": 2.923379174852652,
+      "grad_norm": 0.10373721276738954,
+      "learning_rate": 3.5182723192045606e-05,
+      "loss": 0.4551,
+      "step": 1488
+    },
+    {
+      "epoch": 2.925343811394892,
+      "grad_norm": 0.10149983033167854,
+      "learning_rate": 3.512825216411954e-05,
+      "loss": 0.4636,
+      "step": 1489
+    },
+    {
+      "epoch": 2.927308447937132,
+      "grad_norm": 0.10059074524200104,
+      "learning_rate": 3.507379030501488e-05,
+      "loss": 0.4514,
+      "step": 1490
+    },
+    {
+      "epoch": 2.9292730844793713,
+      "grad_norm": 0.104233092847209,
+      "learning_rate": 3.501933771723099e-05,
+      "loss": 0.4503,
+      "step": 1491
+    },
+    {
+      "epoch": 2.9312377210216107,
+      "grad_norm": 0.10020870681151846,
+      "learning_rate": 3.496489450324978e-05,
+      "loss": 0.4495,
+      "step": 1492
+    },
+    {
+      "epoch": 2.9332023575638506,
+      "grad_norm": 0.09524949058999518,
+      "learning_rate": 3.491046076553556e-05,
+      "loss": 0.4533,
+      "step": 1493
+    },
+    {
+      "epoch": 2.9351669941060905,
+      "grad_norm": 0.11580742311574914,
+      "learning_rate": 3.485603660653471e-05,
+      "loss": 0.4596,
+      "step": 1494
+    },
+    {
+      "epoch": 2.93713163064833,
+      "grad_norm": 0.10905483187801966,
+      "learning_rate": 3.4801622128675685e-05,
+      "loss": 0.4508,
+      "step": 1495
+    },
+    {
+      "epoch": 2.93909626719057,
+      "grad_norm": 0.09856809425038444,
+      "learning_rate": 3.474721743436868e-05,
+      "loss": 0.4377,
+      "step": 1496
+    },
+    {
+      "epoch": 2.9410609037328097,
+      "grad_norm": 0.11116221374793439,
+      "learning_rate": 3.469282262600544e-05,
+      "loss": 0.4504,
+      "step": 1497
+    },
+    {
+      "epoch": 2.943025540275049,
+      "grad_norm": 0.09336544037447073,
+      "learning_rate": 3.463843780595916e-05,
+      "loss": 0.4484,
+      "step": 1498
+    },
+    {
+      "epoch": 2.9449901768172886,
+      "grad_norm": 0.11286842859250225,
+      "learning_rate": 3.458406307658422e-05,
+      "loss": 0.4559,
+      "step": 1499
+    },
+    {
+      "epoch": 2.9469548133595285,
+      "grad_norm": 0.09658225428605906,
+      "learning_rate": 3.452969854021599e-05,
+      "loss": 0.4592,
+      "step": 1500
+    },
+    {
+      "epoch": 2.9489194499017684,
+      "grad_norm": 0.1076098827495761,
+      "learning_rate": 3.447534429917067e-05,
+      "loss": 0.4488,
+      "step": 1501
+    },
+    {
+      "epoch": 2.950884086444008,
+      "grad_norm": 0.11144694914758617,
+      "learning_rate": 3.4421000455745096e-05,
+      "loss": 0.4503,
+      "step": 1502
+    },
+    {
+      "epoch": 2.9528487229862477,
+      "grad_norm": 0.08928338969634193,
+      "learning_rate": 3.436666711221651e-05,
+      "loss": 0.4466,
+      "step": 1503
+    },
+    {
+      "epoch": 2.954813359528487,
+      "grad_norm": 0.12886156731260268,
+      "learning_rate": 3.4312344370842395e-05,
+      "loss": 0.4516,
+      "step": 1504
+    },
+    {
+      "epoch": 2.956777996070727,
+      "grad_norm": 0.09783915724142281,
+      "learning_rate": 3.425803233386032e-05,
+      "loss": 0.4498,
+      "step": 1505
+    },
+    {
+      "epoch": 2.9587426326129664,
+      "grad_norm": 0.09891870222563559,
+      "learning_rate": 3.4203731103487684e-05,
+      "loss": 0.442,
+      "step": 1506
+    },
+    {
+      "epoch": 2.9607072691552063,
+      "grad_norm": 0.10571356062912722,
+      "learning_rate": 3.41494407819215e-05,
+      "loss": 0.4416,
+      "step": 1507
+    },
+    {
+      "epoch": 2.962671905697446,
+      "grad_norm": 0.09535660636696403,
+      "learning_rate": 3.409516147133832e-05,
+      "loss": 0.4539,
+      "step": 1508
+    },
+    {
+      "epoch": 2.9646365422396856,
+      "grad_norm": 0.10688096923220688,
+      "learning_rate": 3.404089327389395e-05,
+      "loss": 0.4422,
+      "step": 1509
+    },
+    {
+      "epoch": 2.966601178781925,
+      "grad_norm": 0.13119779958985628,
+      "learning_rate": 3.398663629172326e-05,
+      "loss": 0.4496,
+      "step": 1510
+    },
+    {
+      "epoch": 2.968565815324165,
+      "grad_norm": 0.10830213378125064,
+      "learning_rate": 3.393239062694003e-05,
+      "loss": 0.4507,
+      "step": 1511
+    },
+    {
+      "epoch": 2.970530451866405,
+      "grad_norm": 0.08891303884822001,
+      "learning_rate": 3.387815638163677e-05,
+      "loss": 0.4493,
+      "step": 1512
+    },
+    {
+      "epoch": 2.9724950884086443,
+      "grad_norm": 0.12531669152302594,
+      "learning_rate": 3.382393365788441e-05,
+      "loss": 0.4552,
+      "step": 1513
+    },
+    {
+      "epoch": 2.974459724950884,
+      "grad_norm": 0.09085972614627834,
+      "learning_rate": 3.3769722557732296e-05,
+      "loss": 0.454,
+      "step": 1514
+    },
+    {
+      "epoch": 2.9764243614931236,
+      "grad_norm": 0.10447742764637687,
+      "learning_rate": 3.371552318320786e-05,
+      "loss": 0.4534,
+      "step": 1515
+    },
+    {
+      "epoch": 2.9783889980353635,
+      "grad_norm": 0.1111949923133436,
+      "learning_rate": 3.366133563631644e-05,
+      "loss": 0.4531,
+      "step": 1516
+    },
+    {
+      "epoch": 2.980353634577603,
+      "grad_norm": 0.10277676185072455,
+      "learning_rate": 3.360716001904112e-05,
+      "loss": 0.4486,
+      "step": 1517
+    },
+    {
+      "epoch": 2.982318271119843,
+      "grad_norm": 0.11163067361918634,
+      "learning_rate": 3.3552996433342575e-05,
+      "loss": 0.4521,
+      "step": 1518
+    },
+    {
+      "epoch": 2.9842829076620827,
+      "grad_norm": 0.0967411251878006,
+      "learning_rate": 3.3498844981158804e-05,
+      "loss": 0.4502,
+      "step": 1519
+    },
+    {
+      "epoch": 2.986247544204322,
+      "grad_norm": 0.09834236910096412,
+      "learning_rate": 3.344470576440496e-05,
+      "loss": 0.4508,
+      "step": 1520
+    },
+    {
+      "epoch": 2.988212180746562,
+      "grad_norm": 0.09627590390967528,
+      "learning_rate": 3.339057888497319e-05,
+      "loss": 0.4475,
+      "step": 1521
+    },
+    {
+      "epoch": 2.9901768172888015,
+      "grad_norm": 0.11175528706846488,
+      "learning_rate": 3.3336464444732436e-05,
+      "loss": 0.4437,
+      "step": 1522
+    },
+    {
+      "epoch": 2.9921414538310414,
+      "grad_norm": 0.08892370456641928,
+      "learning_rate": 3.3282362545528156e-05,
+      "loss": 0.448,
+      "step": 1523
+    },
+    {
+      "epoch": 2.994106090373281,
+      "grad_norm": 0.12223964247417532,
+      "learning_rate": 3.3228273289182294e-05,
+      "loss": 0.4518,
+      "step": 1524
+    },
+    {
+      "epoch": 2.9960707269155207,
+      "grad_norm": 0.09529877701120519,
+      "learning_rate": 3.317419677749297e-05,
+      "loss": 0.4484,
+      "step": 1525
+    },
+    {
+      "epoch": 2.9980353634577606,
+      "grad_norm": 0.09610264535621751,
+      "learning_rate": 3.312013311223428e-05,
+      "loss": 0.4498,
+      "step": 1526
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.10637265147263308,
+      "learning_rate": 3.306608239515619e-05,
+      "loss": 0.4588,
+      "step": 1527
+    },
+    {
+      "epoch": 3.00196463654224,
+      "grad_norm": 0.21682582266591255,
+      "learning_rate": 3.301204472798428e-05,
+      "loss": 0.3969,
+      "step": 1528
+    },
+    {
+      "epoch": 3.0039292730844793,
+      "grad_norm": 0.19290040385771245,
+      "learning_rate": 3.295802021241955e-05,
+      "loss": 0.4036,
+      "step": 1529
+    },
+    {
+      "epoch": 3.005893909626719,
+      "grad_norm": 0.21956269765384548,
+      "learning_rate": 3.2904008950138294e-05,
+      "loss": 0.4064,
+      "step": 1530
+    },
+    {
+      "epoch": 3.0078585461689586,
+      "grad_norm": 0.22007744569772608,
+      "learning_rate": 3.285001104279183e-05,
+      "loss": 0.3957,
+      "step": 1531
+    },
+    {
+      "epoch": 3.0098231827111985,
+      "grad_norm": 0.2012556287045933,
+      "learning_rate": 3.279602659200632e-05,
+      "loss": 0.4008,
+      "step": 1532
+    },
+    {
+      "epoch": 3.011787819253438,
+      "grad_norm": 0.1981964928858252,
+      "learning_rate": 3.274205569938266e-05,
+      "loss": 0.4016,
+      "step": 1533
+    },
+    {
+      "epoch": 3.013752455795678,
+      "grad_norm": 0.18581910241626012,
+      "learning_rate": 3.268809846649619e-05,
+      "loss": 0.4011,
+      "step": 1534
+    },
+    {
+      "epoch": 3.0157170923379173,
+      "grad_norm": 0.18494879662579314,
+      "learning_rate": 3.2634154994896556e-05,
+      "loss": 0.4084,
+      "step": 1535
+    },
+    {
+      "epoch": 3.017681728880157,
+      "grad_norm": 0.16983166742570616,
+      "learning_rate": 3.258022538610747e-05,
+      "loss": 0.4035,
+      "step": 1536
+    },
+    {
+      "epoch": 3.019646365422397,
+      "grad_norm": 0.16030553491519503,
+      "learning_rate": 3.2526309741626594e-05,
+      "loss": 0.3982,
+      "step": 1537
+    },
+    {
+      "epoch": 3.0216110019646365,
+      "grad_norm": 0.17980951763634828,
+      "learning_rate": 3.247240816292531e-05,
+      "loss": 0.4,
+      "step": 1538
+    },
+    {
+      "epoch": 3.0235756385068764,
+      "grad_norm": 0.12536797851387355,
+      "learning_rate": 3.241852075144849e-05,
+      "loss": 0.3978,
+      "step": 1539
+    },
+    {
+      "epoch": 3.025540275049116,
+      "grad_norm": 0.17293951136829216,
+      "learning_rate": 3.2364647608614376e-05,
+      "loss": 0.4022,
+      "step": 1540
+    },
+    {
+      "epoch": 3.0275049115913557,
+      "grad_norm": 0.13598070341688462,
+      "learning_rate": 3.231078883581437e-05,
+      "loss": 0.3943,
+      "step": 1541
+    },
+    {
+      "epoch": 3.029469548133595,
+      "grad_norm": 0.14714756905499207,
+      "learning_rate": 3.225694453441276e-05,
+      "loss": 0.3999,
+      "step": 1542
+    },
+    {
+      "epoch": 3.031434184675835,
+      "grad_norm": 0.1293595447388648,
+      "learning_rate": 3.2203114805746666e-05,
+      "loss": 0.4016,
+      "step": 1543
+    },
+    {
+      "epoch": 3.0333988212180745,
+      "grad_norm": 0.1447451707562068,
+      "learning_rate": 3.2149299751125776e-05,
+      "loss": 0.4046,
+      "step": 1544
+    },
+    {
+      "epoch": 3.0353634577603144,
+      "grad_norm": 0.12652086238095528,
+      "learning_rate": 3.209549947183211e-05,
+      "loss": 0.4006,
+      "step": 1545
+    },
+    {
+      "epoch": 3.0373280943025542,
+      "grad_norm": 0.1395062375669137,
+      "learning_rate": 3.204171406911993e-05,
+      "loss": 0.4069,
+      "step": 1546
+    },
+    {
+      "epoch": 3.0392927308447937,
+      "grad_norm": 0.11413237473520936,
+      "learning_rate": 3.1987943644215486e-05,
+      "loss": 0.404,
+      "step": 1547
+    },
+    {
+      "epoch": 3.0412573673870336,
+      "grad_norm": 0.13225564215391405,
+      "learning_rate": 3.1934188298316836e-05,
+      "loss": 0.4062,
+      "step": 1548
+    },
+    {
+      "epoch": 3.043222003929273,
+      "grad_norm": 0.12628869150016486,
+      "learning_rate": 3.188044813259365e-05,
+      "loss": 0.3942,
+      "step": 1549
+    },
+    {
+      "epoch": 3.045186640471513,
+      "grad_norm": 0.1504843029318896,
+      "learning_rate": 3.182672324818704e-05,
+      "loss": 0.3985,
+      "step": 1550
+    },
+    {
+      "epoch": 3.0471512770137523,
+      "grad_norm": 0.13294788877267877,
+      "learning_rate": 3.177301374620937e-05,
+      "loss": 0.3954,
+      "step": 1551
+    },
+    {
+      "epoch": 3.049115913555992,
+      "grad_norm": 0.11621841080604199,
+      "learning_rate": 3.171931972774402e-05,
+      "loss": 0.3895,
+      "step": 1552
+    },
+    {
+      "epoch": 3.0510805500982316,
+      "grad_norm": 0.11728181663725619,
+      "learning_rate": 3.166564129384526e-05,
+      "loss": 0.3992,
+      "step": 1553
+    },
+    {
+      "epoch": 3.0530451866404715,
+      "grad_norm": 0.11485680910605385,
+      "learning_rate": 3.161197854553803e-05,
+      "loss": 0.4013,
+      "step": 1554
+    },
+    {
+      "epoch": 3.0550098231827114,
+      "grad_norm": 0.1287326618971316,
+      "learning_rate": 3.155833158381771e-05,
+      "loss": 0.4013,
+      "step": 1555
+    },
+    {
+      "epoch": 3.056974459724951,
+      "grad_norm": 0.13170956005088366,
+      "learning_rate": 3.1504700509650004e-05,
+      "loss": 0.4014,
+      "step": 1556
+    },
+    {
+      "epoch": 3.0589390962671907,
+      "grad_norm": 0.1303364350141836,
+      "learning_rate": 3.145108542397072e-05,
+      "loss": 0.4003,
+      "step": 1557
+    },
+    {
+      "epoch": 3.06090373280943,
+      "grad_norm": 0.10973053988206542,
+      "learning_rate": 3.139748642768554e-05,
+      "loss": 0.3999,
+      "step": 1558
+    },
+    {
+      "epoch": 3.06286836935167,
+      "grad_norm": 0.14981006948333944,
+      "learning_rate": 3.134390362166991e-05,
+      "loss": 0.401,
+      "step": 1559
+    },
+    {
+      "epoch": 3.0648330058939095,
+      "grad_norm": 0.10850304233906277,
+      "learning_rate": 3.129033710676877e-05,
+      "loss": 0.3981,
+      "step": 1560
+    },
+    {
+      "epoch": 3.0667976424361494,
+      "grad_norm": 0.11576228660211359,
+      "learning_rate": 3.1236786983796426e-05,
+      "loss": 0.3957,
+      "step": 1561
+    },
+    {
+      "epoch": 3.068762278978389,
+      "grad_norm": 0.1249709059384307,
+      "learning_rate": 3.1183253353536294e-05,
+      "loss": 0.402,
+      "step": 1562
+    },
+    {
+      "epoch": 3.0707269155206287,
+      "grad_norm": 0.11683113187958709,
+      "learning_rate": 3.1129736316740794e-05,
+      "loss": 0.3955,
+      "step": 1563
+    },
+    {
+      "epoch": 3.072691552062868,
+      "grad_norm": 0.12034258518599446,
+      "learning_rate": 3.107623597413111e-05,
+      "loss": 0.399,
+      "step": 1564
+    },
+    {
+      "epoch": 3.074656188605108,
+      "grad_norm": 0.10750925653327374,
+      "learning_rate": 3.102275242639696e-05,
+      "loss": 0.3959,
+      "step": 1565
+    },
+    {
+      "epoch": 3.076620825147348,
+      "grad_norm": 0.14095708301728171,
+      "learning_rate": 3.096928577419651e-05,
+      "loss": 0.3978,
+      "step": 1566
+    },
+    {
+      "epoch": 3.0785854616895874,
+      "grad_norm": 0.10379507016298843,
+      "learning_rate": 3.091583611815612e-05,
+      "loss": 0.3986,
+      "step": 1567
+    },
+    {
+      "epoch": 3.0805500982318272,
+      "grad_norm": 0.12898608963087518,
+      "learning_rate": 3.0862403558870115e-05,
+      "loss": 0.4106,
+      "step": 1568
+    },
+    {
+      "epoch": 3.0825147347740667,
+      "grad_norm": 0.10290007736065089,
+      "learning_rate": 3.08089881969007e-05,
+      "loss": 0.3978,
+      "step": 1569
+    },
+    {
+      "epoch": 3.0844793713163066,
+      "grad_norm": 0.10036717627139395,
+      "learning_rate": 3.075559013277769e-05,
+      "loss": 0.3947,
+      "step": 1570
+    },
+    {
+      "epoch": 3.086444007858546,
+      "grad_norm": 0.1031712179178971,
+      "learning_rate": 3.0702209466998344e-05,
+      "loss": 0.4099,
+      "step": 1571
+    },
+    {
+      "epoch": 3.088408644400786,
+      "grad_norm": 0.10485438931525144,
+      "learning_rate": 3.064884630002718e-05,
+      "loss": 0.4,
+      "step": 1572
+    },
+    {
+      "epoch": 3.0903732809430258,
+      "grad_norm": 0.09826424335254015,
+      "learning_rate": 3.0595500732295764e-05,
+      "loss": 0.3979,
+      "step": 1573
+    },
+    {
+      "epoch": 3.092337917485265,
+      "grad_norm": 0.09559412668984901,
+      "learning_rate": 3.05421728642026e-05,
+      "loss": 0.4085,
+      "step": 1574
+    },
+    {
+      "epoch": 3.094302554027505,
+      "grad_norm": 0.1076043828509243,
+      "learning_rate": 3.0488862796112784e-05,
+      "loss": 0.3964,
+      "step": 1575
+    },
+    {
+      "epoch": 3.0962671905697445,
+      "grad_norm": 0.08971497049371964,
+      "learning_rate": 3.043557062835799e-05,
+      "loss": 0.4082,
+      "step": 1576
+    },
+    {
+      "epoch": 3.0982318271119844,
+      "grad_norm": 0.09788151148294727,
+      "learning_rate": 3.0382296461236187e-05,
+      "loss": 0.3995,
+      "step": 1577
+    },
+    {
+      "epoch": 3.100196463654224,
+      "grad_norm": 0.08769428277027409,
+      "learning_rate": 3.0329040395011436e-05,
+      "loss": 0.3889,
+      "step": 1578
+    },
+    {
+      "epoch": 3.1021611001964637,
+      "grad_norm": 0.09569226931351063,
+      "learning_rate": 3.027580252991377e-05,
+      "loss": 0.3953,
+      "step": 1579
+    },
+    {
+      "epoch": 3.104125736738703,
+      "grad_norm": 0.10198657753226509,
+      "learning_rate": 3.0222582966138944e-05,
+      "loss": 0.3979,
+      "step": 1580
+    },
+    {
+      "epoch": 3.106090373280943,
+      "grad_norm": 0.0898857548308694,
+      "learning_rate": 3.0169381803848264e-05,
+      "loss": 0.412,
+      "step": 1581
+    },
+    {
+      "epoch": 3.1080550098231825,
+      "grad_norm": 0.09522072057241268,
+      "learning_rate": 3.0116199143168425e-05,
+      "loss": 0.4072,
+      "step": 1582
+    },
+    {
+      "epoch": 3.1100196463654224,
+      "grad_norm": 0.10542659651790569,
+      "learning_rate": 3.00630350841913e-05,
+      "loss": 0.398,
+      "step": 1583
+    },
+    {
+      "epoch": 3.1119842829076623,
+      "grad_norm": 0.08888265539928412,
+      "learning_rate": 3.0009889726973723e-05,
+      "loss": 0.4096,
+      "step": 1584
+    },
+    {
+      "epoch": 3.1139489194499017,
+      "grad_norm": 0.09205868500428324,
+      "learning_rate": 2.995676317153735e-05,
+      "loss": 0.3993,
+      "step": 1585
+    },
+    {
+      "epoch": 3.1159135559921416,
+      "grad_norm": 0.09013584331067682,
+      "learning_rate": 2.990365551786847e-05,
+      "loss": 0.3937,
+      "step": 1586
+    },
+    {
+      "epoch": 3.117878192534381,
+      "grad_norm": 0.0880255536016394,
+      "learning_rate": 2.985056686591776e-05,
+      "loss": 0.409,
+      "step": 1587
+    },
+    {
+      "epoch": 3.119842829076621,
+      "grad_norm": 0.0886598128428135,
+      "learning_rate": 2.979749731560017e-05,
+      "loss": 0.4093,
+      "step": 1588
+    },
+    {
+      "epoch": 3.1218074656188604,
+      "grad_norm": 0.08617111148003839,
+      "learning_rate": 2.9744446966794673e-05,
+      "loss": 0.3963,
+      "step": 1589
+    },
+    {
+      "epoch": 3.1237721021611002,
+      "grad_norm": 0.08795611416608787,
+      "learning_rate": 2.969141591934414e-05,
+      "loss": 0.4107,
+      "step": 1590
+    },
+    {
+      "epoch": 3.1257367387033397,
+      "grad_norm": 0.0973074082186057,
+      "learning_rate": 2.9638404273055062e-05,
+      "loss": 0.4042,
+      "step": 1591
+    },
+    {
+      "epoch": 3.1277013752455796,
+      "grad_norm": 0.09804497891520347,
+      "learning_rate": 2.9585412127697456e-05,
+      "loss": 0.3973,
+      "step": 1592
+    },
+    {
+      "epoch": 3.1296660117878194,
+      "grad_norm": 0.09354960922528358,
+      "learning_rate": 2.953243958300465e-05,
+      "loss": 0.397,
+      "step": 1593
+    },
+    {
+      "epoch": 3.131630648330059,
+      "grad_norm": 0.09089408275917497,
+      "learning_rate": 2.9479486738673028e-05,
+      "loss": 0.3946,
+      "step": 1594
+    },
+    {
+      "epoch": 3.1335952848722988,
+      "grad_norm": 0.09132878261728027,
+      "learning_rate": 2.9426553694361942e-05,
+      "loss": 0.4037,
+      "step": 1595
+    },
+    {
+      "epoch": 3.135559921414538,
+      "grad_norm": 0.10030807947559753,
+      "learning_rate": 2.9373640549693475e-05,
+      "loss": 0.4118,
+      "step": 1596
+    },
+    {
+      "epoch": 3.137524557956778,
+      "grad_norm": 0.09463563369319991,
+      "learning_rate": 2.9320747404252238e-05,
+      "loss": 0.3994,
+      "step": 1597
+    },
+    {
+      "epoch": 3.1394891944990175,
+      "grad_norm": 0.10323380137495797,
+      "learning_rate": 2.9267874357585218e-05,
+      "loss": 0.4116,
+      "step": 1598
+    },
+    {
+      "epoch": 3.1414538310412574,
+      "grad_norm": 0.08526695942134707,
+      "learning_rate": 2.921502150920159e-05,
+      "loss": 0.391,
+      "step": 1599
+    },
+    {
+      "epoch": 3.143418467583497,
+      "grad_norm": 0.09417343419443681,
+      "learning_rate": 2.916218895857247e-05,
+      "loss": 0.4042,
+      "step": 1600
+    },
+    {
+      "epoch": 3.1453831041257367,
+      "grad_norm": 0.09836772342110066,
+      "learning_rate": 2.9109376805130817e-05,
+      "loss": 0.4046,
+      "step": 1601
+    },
+    {
+      "epoch": 3.1473477406679766,
+      "grad_norm": 0.08636447384240543,
+      "learning_rate": 2.905658514827118e-05,
+      "loss": 0.4035,
+      "step": 1602
+    },
+    {
+      "epoch": 3.149312377210216,
+      "grad_norm": 0.11170696288455495,
+      "learning_rate": 2.9003814087349558e-05,
+      "loss": 0.4083,
+      "step": 1603
+    },
+    {
+      "epoch": 3.151277013752456,
+      "grad_norm": 0.08435668609577023,
+      "learning_rate": 2.8951063721683127e-05,
+      "loss": 0.4007,
+      "step": 1604
+    },
+    {
+      "epoch": 3.1532416502946954,
+      "grad_norm": 0.10222185434335862,
+      "learning_rate": 2.8898334150550176e-05,
+      "loss": 0.4094,
+      "step": 1605
+    },
+    {
+      "epoch": 3.1552062868369353,
+      "grad_norm": 0.09071158105332494,
+      "learning_rate": 2.8845625473189845e-05,
+      "loss": 0.4069,
+      "step": 1606
+    },
+    {
+      "epoch": 3.1571709233791747,
+      "grad_norm": 0.10953933017718308,
+      "learning_rate": 2.8792937788801924e-05,
+      "loss": 0.3962,
+      "step": 1607
+    },
+    {
+      "epoch": 3.1591355599214146,
+      "grad_norm": 0.09825240866193617,
+      "learning_rate": 2.874027119654671e-05,
+      "loss": 0.3994,
+      "step": 1608
+    },
+    {
+      "epoch": 3.161100196463654,
+      "grad_norm": 0.09268644138010858,
+      "learning_rate": 2.868762579554483e-05,
+      "loss": 0.4062,
+      "step": 1609
+    },
+    {
+      "epoch": 3.163064833005894,
+      "grad_norm": 0.09014040570561463,
+      "learning_rate": 2.863500168487698e-05,
+      "loss": 0.3957,
+      "step": 1610
+    },
+    {
+      "epoch": 3.165029469548134,
+      "grad_norm": 0.0848810926452704,
+      "learning_rate": 2.8582398963583817e-05,
+      "loss": 0.3936,
+      "step": 1611
+    },
+    {
+      "epoch": 3.1669941060903732,
+      "grad_norm": 0.08895712761831184,
+      "learning_rate": 2.852981773066575e-05,
+      "loss": 0.3949,
+      "step": 1612
+    },
+    {
+      "epoch": 3.168958742632613,
+      "grad_norm": 0.08307537209938264,
+      "learning_rate": 2.84772580850827e-05,
+      "loss": 0.4025,
+      "step": 1613
+    },
+    {
+      "epoch": 3.1709233791748526,
+      "grad_norm": 0.09008406251783706,
+      "learning_rate": 2.8424720125754012e-05,
+      "loss": 0.3977,
+      "step": 1614
+    },
+    {
+      "epoch": 3.1728880157170924,
+      "grad_norm": 0.0931940381429735,
+      "learning_rate": 2.837220395155819e-05,
+      "loss": 0.3989,
+      "step": 1615
+    },
+    {
+      "epoch": 3.174852652259332,
+      "grad_norm": 0.08703368900469856,
+      "learning_rate": 2.831970966133275e-05,
+      "loss": 0.4067,
+      "step": 1616
+    },
+    {
+      "epoch": 3.1768172888015718,
+      "grad_norm": 0.09281183611029649,
+      "learning_rate": 2.8267237353874005e-05,
+      "loss": 0.4043,
+      "step": 1617
+    },
+    {
+      "epoch": 3.178781925343811,
+      "grad_norm": 0.0947906418593917,
+      "learning_rate": 2.8214787127936907e-05,
+      "loss": 0.4019,
+      "step": 1618
+    },
+    {
+      "epoch": 3.180746561886051,
+      "grad_norm": 0.09304387839085337,
+      "learning_rate": 2.8162359082234866e-05,
+      "loss": 0.4039,
+      "step": 1619
+    },
+    {
+      "epoch": 3.182711198428291,
+      "grad_norm": 0.09720232832471747,
+      "learning_rate": 2.8109953315439503e-05,
+      "loss": 0.4016,
+      "step": 1620
+    },
+    {
+      "epoch": 3.1846758349705304,
+      "grad_norm": 0.09389640576946691,
+      "learning_rate": 2.8057569926180552e-05,
+      "loss": 0.3937,
+      "step": 1621
+    },
+    {
+      "epoch": 3.1866404715127703,
+      "grad_norm": 0.0953622072962431,
+      "learning_rate": 2.8005209013045624e-05,
+      "loss": 0.4,
+      "step": 1622
+    },
+    {
+      "epoch": 3.1886051080550097,
+      "grad_norm": 0.09178947578923707,
+      "learning_rate": 2.795287067458e-05,
+      "loss": 0.3999,
+      "step": 1623
+    },
+    {
+      "epoch": 3.1905697445972496,
+      "grad_norm": 0.10170741685029298,
+      "learning_rate": 2.7900555009286505e-05,
+      "loss": 0.4025,
+      "step": 1624
+    },
+    {
+      "epoch": 3.192534381139489,
+      "grad_norm": 0.0897256301601962,
+      "learning_rate": 2.7848262115625285e-05,
+      "loss": 0.3946,
+      "step": 1625
+    },
+    {
+      "epoch": 3.194499017681729,
+      "grad_norm": 0.10468274258342364,
+      "learning_rate": 2.779599209201362e-05,
+      "loss": 0.3992,
+      "step": 1626
+    },
+    {
+      "epoch": 3.1964636542239684,
+      "grad_norm": 0.09776170937152256,
+      "learning_rate": 2.7743745036825757e-05,
+      "loss": 0.3929,
+      "step": 1627
+    },
+    {
+      "epoch": 3.1984282907662083,
+      "grad_norm": 0.10298926679363389,
+      "learning_rate": 2.769152104839272e-05,
+      "loss": 0.4023,
+      "step": 1628
+    },
+    {
+      "epoch": 3.2003929273084477,
+      "grad_norm": 0.09320360993086167,
+      "learning_rate": 2.7639320225002108e-05,
+      "loss": 0.4116,
+      "step": 1629
+    },
+    {
+      "epoch": 3.2023575638506876,
+      "grad_norm": 0.0862874223987401,
+      "learning_rate": 2.7587142664897926e-05,
+      "loss": 0.4021,
+      "step": 1630
+    },
+    {
+      "epoch": 3.2043222003929275,
+      "grad_norm": 0.0891541213441066,
+      "learning_rate": 2.7534988466280396e-05,
+      "loss": 0.4006,
+      "step": 1631
+    },
+    {
+      "epoch": 3.206286836935167,
+      "grad_norm": 0.09835856885801847,
+      "learning_rate": 2.7482857727305795e-05,
+      "loss": 0.4039,
+      "step": 1632
+    },
+    {
+      "epoch": 3.208251473477407,
+      "grad_norm": 0.08059251040377874,
+      "learning_rate": 2.7430750546086195e-05,
+      "loss": 0.4031,
+      "step": 1633
+    },
+    {
+      "epoch": 3.2102161100196462,
+      "grad_norm": 0.09396157835423938,
+      "learning_rate": 2.737866702068938e-05,
+      "loss": 0.3984,
+      "step": 1634
+    },
+    {
+      "epoch": 3.212180746561886,
+      "grad_norm": 0.09125202947596524,
+      "learning_rate": 2.7326607249138607e-05,
+      "loss": 0.4133,
+      "step": 1635
+    },
+    {
+      "epoch": 3.2141453831041256,
+      "grad_norm": 0.09064972166245613,
+      "learning_rate": 2.7274571329412403e-05,
+      "loss": 0.4017,
+      "step": 1636
+    },
+    {
+      "epoch": 3.2161100196463654,
+      "grad_norm": 0.08171263240687743,
+      "learning_rate": 2.7222559359444422e-05,
+      "loss": 0.4004,
+      "step": 1637
+    },
+    {
+      "epoch": 3.2180746561886053,
+      "grad_norm": 0.08386498138655814,
+      "learning_rate": 2.7170571437123253e-05,
+      "loss": 0.3992,
+      "step": 1638
+    },
+    {
+      "epoch": 3.2200392927308448,
+      "grad_norm": 0.08763481779127934,
+      "learning_rate": 2.7118607660292205e-05,
+      "loss": 0.3951,
+      "step": 1639
+    },
+    {
+      "epoch": 3.2220039292730847,
+      "grad_norm": 0.0821965655848571,
+      "learning_rate": 2.7066668126749145e-05,
+      "loss": 0.4035,
+      "step": 1640
+    },
+    {
+      "epoch": 3.223968565815324,
+      "grad_norm": 0.09059551504200736,
+      "learning_rate": 2.7014752934246353e-05,
+      "loss": 0.4027,
+      "step": 1641
+    },
+    {
+      "epoch": 3.225933202357564,
+      "grad_norm": 0.10079958787248712,
+      "learning_rate": 2.696286218049023e-05,
+      "loss": 0.4072,
+      "step": 1642
+    },
+    {
+      "epoch": 3.2278978388998034,
+      "grad_norm": 0.08110163615208314,
+      "learning_rate": 2.6910995963141235e-05,
+      "loss": 0.3983,
+      "step": 1643
+    },
+    {
+      "epoch": 3.2298624754420433,
+      "grad_norm": 0.09811274113859714,
+      "learning_rate": 2.6859154379813622e-05,
+      "loss": 0.4095,
+      "step": 1644
+    },
+    {
+      "epoch": 3.2318271119842827,
+      "grad_norm": 0.08613889375204886,
+      "learning_rate": 2.680733752807533e-05,
+      "loss": 0.3975,
+      "step": 1645
+    },
+    {
+      "epoch": 3.2337917485265226,
+      "grad_norm": 0.08711918610889645,
+      "learning_rate": 2.6755545505447678e-05,
+      "loss": 0.4025,
+      "step": 1646
+    },
+    {
+      "epoch": 3.235756385068762,
+      "grad_norm": 0.10435577692771354,
+      "learning_rate": 2.67037784094053e-05,
+      "loss": 0.3943,
+      "step": 1647
+    },
+    {
+      "epoch": 3.237721021611002,
+      "grad_norm": 0.09398880222368369,
+      "learning_rate": 2.6652036337375933e-05,
+      "loss": 0.4085,
+      "step": 1648
+    },
+    {
+      "epoch": 3.239685658153242,
+      "grad_norm": 0.09810345010662071,
+      "learning_rate": 2.660031938674017e-05,
+      "loss": 0.4083,
+      "step": 1649
+    },
+    {
+      "epoch": 3.2416502946954813,
+      "grad_norm": 0.08438204289478642,
+      "learning_rate": 2.6548627654831355e-05,
+      "loss": 0.399,
+      "step": 1650
+    },
+    {
+      "epoch": 3.243614931237721,
+      "grad_norm": 0.1027581838086227,
+      "learning_rate": 2.6496961238935388e-05,
+      "loss": 0.4009,
+      "step": 1651
+    },
+    {
+      "epoch": 3.2455795677799606,
+      "grad_norm": 0.08827079553937367,
+      "learning_rate": 2.6445320236290463e-05,
+      "loss": 0.3968,
+      "step": 1652
+    },
+    {
+      "epoch": 3.2475442043222005,
+      "grad_norm": 0.09949789575968311,
+      "learning_rate": 2.6393704744086998e-05,
+      "loss": 0.399,
+      "step": 1653
+    },
+    {
+      "epoch": 3.24950884086444,
+      "grad_norm": 0.08567189522659219,
+      "learning_rate": 2.6342114859467396e-05,
+      "loss": 0.3978,
+      "step": 1654
+    },
+    {
+      "epoch": 3.25147347740668,
+      "grad_norm": 0.10862137427834641,
+      "learning_rate": 2.6290550679525834e-05,
+      "loss": 0.4049,
+      "step": 1655
+    },
+    {
+      "epoch": 3.2534381139489197,
+      "grad_norm": 0.08408914928379652,
+      "learning_rate": 2.6239012301308142e-05,
+      "loss": 0.3929,
+      "step": 1656
+    },
+    {
+      "epoch": 3.255402750491159,
+      "grad_norm": 0.0953219986113377,
+      "learning_rate": 2.618749982181157e-05,
+      "loss": 0.4018,
+      "step": 1657
+    },
+    {
+      "epoch": 3.257367387033399,
+      "grad_norm": 0.07948665257008214,
+      "learning_rate": 2.6136013337984646e-05,
+      "loss": 0.3983,
+      "step": 1658
+    },
+    {
+      "epoch": 3.2593320235756384,
+      "grad_norm": 0.09486283941830691,
+      "learning_rate": 2.6084552946726956e-05,
+      "loss": 0.4031,
+      "step": 1659
+    },
+    {
+      "epoch": 3.2612966601178783,
+      "grad_norm": 0.08777663875919817,
+      "learning_rate": 2.6033118744888972e-05,
+      "loss": 0.4007,
+      "step": 1660
+    },
+    {
+      "epoch": 3.2632612966601178,
+      "grad_norm": 0.09625875202261766,
+      "learning_rate": 2.5981710829271923e-05,
+      "loss": 0.4029,
+      "step": 1661
+    },
+    {
+      "epoch": 3.2652259332023577,
+      "grad_norm": 0.08992868771689806,
+      "learning_rate": 2.5930329296627497e-05,
+      "loss": 0.4068,
+      "step": 1662
+    },
+    {
+      "epoch": 3.267190569744597,
+      "grad_norm": 0.10148805629330133,
+      "learning_rate": 2.5878974243657758e-05,
+      "loss": 0.3969,
+      "step": 1663
+    },
+    {
+      "epoch": 3.269155206286837,
+      "grad_norm": 0.08213883129537239,
+      "learning_rate": 2.5827645767014988e-05,
+      "loss": 0.3984,
+      "step": 1664
+    },
+    {
+      "epoch": 3.2711198428290764,
+      "grad_norm": 0.08715199441272636,
+      "learning_rate": 2.5776343963301352e-05,
+      "loss": 0.3931,
+      "step": 1665
+    },
+    {
+      "epoch": 3.2730844793713163,
+      "grad_norm": 0.09130424013377329,
+      "learning_rate": 2.572506892906889e-05,
+      "loss": 0.395,
+      "step": 1666
+    },
+    {
+      "epoch": 3.275049115913556,
+      "grad_norm": 0.08087017552410891,
+      "learning_rate": 2.5673820760819236e-05,
+      "loss": 0.4063,
+      "step": 1667
+    },
+    {
+      "epoch": 3.2770137524557956,
+      "grad_norm": 0.08744055043566604,
+      "learning_rate": 2.562259955500345e-05,
+      "loss": 0.4089,
+      "step": 1668
+    },
+    {
+      "epoch": 3.2789783889980355,
+      "grad_norm": 0.08032910373283983,
+      "learning_rate": 2.5571405408021874e-05,
+      "loss": 0.3994,
+      "step": 1669
+    },
+    {
+      "epoch": 3.280943025540275,
+      "grad_norm": 0.08838040537392357,
+      "learning_rate": 2.55202384162239e-05,
+      "loss": 0.3957,
+      "step": 1670
+    },
+    {
+      "epoch": 3.282907662082515,
+      "grad_norm": 0.08412603979538674,
+      "learning_rate": 2.5469098675907848e-05,
+      "loss": 0.4112,
+      "step": 1671
+    },
+    {
+      "epoch": 3.2848722986247543,
+      "grad_norm": 0.08644421599046753,
+      "learning_rate": 2.5417986283320684e-05,
+      "loss": 0.4026,
+      "step": 1672
+    },
+    {
+      "epoch": 3.286836935166994,
+      "grad_norm": 0.0817308321411826,
+      "learning_rate": 2.536690133465796e-05,
+      "loss": 0.4033,
+      "step": 1673
+    },
+    {
+      "epoch": 3.288801571709234,
+      "grad_norm": 0.09959881058718023,
+      "learning_rate": 2.5315843926063588e-05,
+      "loss": 0.3982,
+      "step": 1674
+    },
+    {
+      "epoch": 3.2907662082514735,
+      "grad_norm": 0.07983455182637308,
+      "learning_rate": 2.526481415362959e-05,
+      "loss": 0.39,
+      "step": 1675
+    },
+    {
+      "epoch": 3.292730844793713,
+      "grad_norm": 0.09845399817125618,
+      "learning_rate": 2.5213812113396023e-05,
+      "loss": 0.4147,
+      "step": 1676
+    },
+    {
+      "epoch": 3.294695481335953,
+      "grad_norm": 0.08569516454529881,
+      "learning_rate": 2.516283790135074e-05,
+      "loss": 0.4022,
+      "step": 1677
+    },
+    {
+      "epoch": 3.2966601178781927,
+      "grad_norm": 0.09753477077046269,
+      "learning_rate": 2.5111891613429203e-05,
+      "loss": 0.3999,
+      "step": 1678
+    },
+    {
+      "epoch": 3.298624754420432,
+      "grad_norm": 0.09027531028209472,
+      "learning_rate": 2.5060973345514343e-05,
+      "loss": 0.4022,
+      "step": 1679
+    },
+    {
+      "epoch": 3.300589390962672,
+      "grad_norm": 0.08565972933554944,
+      "learning_rate": 2.501008319343637e-05,
+      "loss": 0.4071,
+      "step": 1680
+    },
+    {
+      "epoch": 3.3025540275049114,
+      "grad_norm": 0.09396254065856466,
+      "learning_rate": 2.4959221252972503e-05,
+      "loss": 0.4,
+      "step": 1681
+    },
+    {
+      "epoch": 3.3045186640471513,
+      "grad_norm": 0.08863592047963394,
+      "learning_rate": 2.4908387619846942e-05,
+      "loss": 0.3942,
+      "step": 1682
+    },
+    {
+      "epoch": 3.3064833005893908,
+      "grad_norm": 0.09161884058370111,
+      "learning_rate": 2.4857582389730604e-05,
+      "loss": 0.3972,
+      "step": 1683
+    },
+    {
+      "epoch": 3.3084479371316307,
+      "grad_norm": 0.08930857811942466,
+      "learning_rate": 2.4806805658240936e-05,
+      "loss": 0.396,
+      "step": 1684
+    },
+    {
+      "epoch": 3.3104125736738705,
+      "grad_norm": 0.08853365940620588,
+      "learning_rate": 2.475605752094172e-05,
+      "loss": 0.4063,
+      "step": 1685
+    },
+    {
+      "epoch": 3.31237721021611,
+      "grad_norm": 0.0884168077258335,
+      "learning_rate": 2.4705338073342963e-05,
+      "loss": 0.4056,
+      "step": 1686
+    },
+    {
+      "epoch": 3.31434184675835,
+      "grad_norm": 0.09967929202787662,
+      "learning_rate": 2.4654647410900675e-05,
+      "loss": 0.3957,
+      "step": 1687
+    },
+    {
+      "epoch": 3.3163064833005893,
+      "grad_norm": 0.07385813136180545,
+      "learning_rate": 2.4603985629016664e-05,
+      "loss": 0.41,
+      "step": 1688
+    },
+    {
+      "epoch": 3.318271119842829,
+      "grad_norm": 0.0964231780384055,
+      "learning_rate": 2.4553352823038404e-05,
+      "loss": 0.3957,
+      "step": 1689
+    },
+    {
+      "epoch": 3.3202357563850686,
+      "grad_norm": 0.08582024851626455,
+      "learning_rate": 2.4502749088258854e-05,
+      "loss": 0.395,
+      "step": 1690
+    },
+    {
+      "epoch": 3.3222003929273085,
+      "grad_norm": 0.10239622042042244,
+      "learning_rate": 2.4452174519916184e-05,
+      "loss": 0.3965,
+      "step": 1691
+    },
+    {
+      "epoch": 3.324165029469548,
+      "grad_norm": 0.08185376552628378,
+      "learning_rate": 2.4401629213193774e-05,
+      "loss": 0.3987,
+      "step": 1692
+    },
+    {
+      "epoch": 3.326129666011788,
+      "grad_norm": 0.09015713919035646,
+      "learning_rate": 2.435111326321988e-05,
+      "loss": 0.4013,
+      "step": 1693
+    },
+    {
+      "epoch": 3.3280943025540273,
+      "grad_norm": 0.08503209296110423,
+      "learning_rate": 2.430062676506748e-05,
+      "loss": 0.4016,
+      "step": 1694
+    },
+    {
+      "epoch": 3.330058939096267,
+      "grad_norm": 0.08301104043942985,
+      "learning_rate": 2.425016981375417e-05,
+      "loss": 0.3979,
+      "step": 1695
+    },
+    {
+      "epoch": 3.332023575638507,
+      "grad_norm": 0.0904384283482352,
+      "learning_rate": 2.419974250424193e-05,
+      "loss": 0.407,
+      "step": 1696
+    },
+    {
+      "epoch": 3.3339882121807465,
+      "grad_norm": 0.09617426848022617,
+      "learning_rate": 2.4149344931436927e-05,
+      "loss": 0.4098,
+      "step": 1697
+    },
+    {
+      "epoch": 3.3359528487229864,
+      "grad_norm": 0.08235054135906761,
+      "learning_rate": 2.4098977190189388e-05,
+      "loss": 0.3992,
+      "step": 1698
+    },
+    {
+      "epoch": 3.337917485265226,
+      "grad_norm": 0.10004764018270244,
+      "learning_rate": 2.4048639375293382e-05,
+      "loss": 0.3988,
+      "step": 1699
+    },
+    {
+      "epoch": 3.3398821218074657,
+      "grad_norm": 0.11004939107277846,
+      "learning_rate": 2.3998331581486677e-05,
+      "loss": 0.4,
+      "step": 1700
+    },
+    {
+      "epoch": 3.341846758349705,
+      "grad_norm": 0.08906954063685758,
+      "learning_rate": 2.394805390345049e-05,
+      "loss": 0.396,
+      "step": 1701
+    },
+    {
+      "epoch": 3.343811394891945,
+      "grad_norm": 0.11228820746285419,
+      "learning_rate": 2.3897806435809425e-05,
+      "loss": 0.4047,
+      "step": 1702
+    },
+    {
+      "epoch": 3.345776031434185,
+      "grad_norm": 0.09158424398671879,
+      "learning_rate": 2.3847589273131197e-05,
+      "loss": 0.4062,
+      "step": 1703
+    },
+    {
+      "epoch": 3.3477406679764243,
+      "grad_norm": 0.07960452727134876,
+      "learning_rate": 2.379740250992645e-05,
+      "loss": 0.4119,
+      "step": 1704
+    },
+    {
+      "epoch": 3.349705304518664,
+      "grad_norm": 0.10139531641960543,
+      "learning_rate": 2.3747246240648675e-05,
+      "loss": 0.3909,
+      "step": 1705
+    },
+    {
+      "epoch": 3.3516699410609037,
+      "grad_norm": 0.07534926984141815,
+      "learning_rate": 2.3697120559693952e-05,
+      "loss": 0.4069,
+      "step": 1706
+    },
+    {
+      "epoch": 3.3536345776031435,
+      "grad_norm": 0.08236141437957568,
+      "learning_rate": 2.3647025561400777e-05,
+      "loss": 0.4018,
+      "step": 1707
+    },
+    {
+      "epoch": 3.355599214145383,
+      "grad_norm": 0.08565252891381465,
+      "learning_rate": 2.359696134004991e-05,
+      "loss": 0.4002,
+      "step": 1708
+    },
+    {
+      "epoch": 3.357563850687623,
+      "grad_norm": 0.08091219083054894,
+      "learning_rate": 2.3546927989864204e-05,
+      "loss": 0.3947,
+      "step": 1709
+    },
+    {
+      "epoch": 3.3595284872298623,
+      "grad_norm": 0.08261784739276652,
+      "learning_rate": 2.3496925605008363e-05,
+      "loss": 0.4045,
+      "step": 1710
+    },
+    {
+      "epoch": 3.361493123772102,
+      "grad_norm": 0.08027351285625156,
+      "learning_rate": 2.3446954279588878e-05,
+      "loss": 0.4073,
+      "step": 1711
+    },
+    {
+      "epoch": 3.3634577603143416,
+      "grad_norm": 0.08081890199548152,
+      "learning_rate": 2.339701410765374e-05,
+      "loss": 0.3973,
+      "step": 1712
+    },
+    {
+      "epoch": 3.3654223968565815,
+      "grad_norm": 0.08201592610043835,
+      "learning_rate": 2.3347105183192324e-05,
+      "loss": 0.4005,
+      "step": 1713
+    },
+    {
+      "epoch": 3.3673870333988214,
+      "grad_norm": 0.08905723453465593,
+      "learning_rate": 2.3297227600135184e-05,
+      "loss": 0.4003,
+      "step": 1714
+    },
+    {
+      "epoch": 3.369351669941061,
+      "grad_norm": 0.07989937708298293,
+      "learning_rate": 2.324738145235391e-05,
+      "loss": 0.3969,
+      "step": 1715
+    },
+    {
+      "epoch": 3.3713163064833007,
+      "grad_norm": 0.08897038724221723,
+      "learning_rate": 2.319756683366092e-05,
+      "loss": 0.3932,
+      "step": 1716
+    },
+    {
+      "epoch": 3.37328094302554,
+      "grad_norm": 0.08236653644250487,
+      "learning_rate": 2.314778383780927e-05,
+      "loss": 0.4105,
+      "step": 1717
+    },
+    {
+      "epoch": 3.37524557956778,
+      "grad_norm": 0.09575876154410462,
+      "learning_rate": 2.3098032558492518e-05,
+      "loss": 0.391,
+      "step": 1718
+    },
+    {
+      "epoch": 3.3772102161100195,
+      "grad_norm": 0.08288245871484613,
+      "learning_rate": 2.304831308934457e-05,
+      "loss": 0.3988,
+      "step": 1719
+    },
+    {
+      "epoch": 3.3791748526522594,
+      "grad_norm": 0.08661494110510691,
+      "learning_rate": 2.2998625523939374e-05,
+      "loss": 0.3919,
+      "step": 1720
+    },
+    {
+      "epoch": 3.3811394891944992,
+      "grad_norm": 0.08351221501778944,
+      "learning_rate": 2.294896995579092e-05,
+      "loss": 0.3982,
+      "step": 1721
+    },
+    {
+      "epoch": 3.3831041257367387,
+      "grad_norm": 0.0778690212313438,
+      "learning_rate": 2.2899346478352928e-05,
+      "loss": 0.4114,
+      "step": 1722
+    },
+    {
+      "epoch": 3.3850687622789786,
+      "grad_norm": 0.08798561575341488,
+      "learning_rate": 2.2849755185018748e-05,
+      "loss": 0.4049,
+      "step": 1723
+    },
+    {
+      "epoch": 3.387033398821218,
+      "grad_norm": 0.09207162603141585,
+      "learning_rate": 2.2800196169121148e-05,
+      "loss": 0.3876,
+      "step": 1724
+    },
+    {
+      "epoch": 3.388998035363458,
+      "grad_norm": 0.08757959867507971,
+      "learning_rate": 2.275066952393215e-05,
+      "loss": 0.4012,
+      "step": 1725
+    },
+    {
+      "epoch": 3.3909626719056973,
+      "grad_norm": 0.09386077427980156,
+      "learning_rate": 2.270117534266286e-05,
+      "loss": 0.4087,
+      "step": 1726
+    },
+    {
+      "epoch": 3.392927308447937,
+      "grad_norm": 0.081105119272709,
+      "learning_rate": 2.265171371846325e-05,
+      "loss": 0.4057,
+      "step": 1727
+    },
+    {
+      "epoch": 3.3948919449901767,
+      "grad_norm": 0.08476866366755689,
+      "learning_rate": 2.2602284744422063e-05,
+      "loss": 0.4079,
+      "step": 1728
+    },
+    {
+      "epoch": 3.3968565815324165,
+      "grad_norm": 0.08317409185573939,
+      "learning_rate": 2.2552888513566592e-05,
+      "loss": 0.4025,
+      "step": 1729
+    },
+    {
+      "epoch": 3.398821218074656,
+      "grad_norm": 0.07988712864131135,
+      "learning_rate": 2.250352511886246e-05,
+      "loss": 0.4053,
+      "step": 1730
+    },
+    {
+      "epoch": 3.400785854616896,
+      "grad_norm": 0.08839840130075274,
+      "learning_rate": 2.2454194653213533e-05,
+      "loss": 0.4035,
+      "step": 1731
+    },
+    {
+      "epoch": 3.4027504911591357,
+      "grad_norm": 0.08142362986658283,
+      "learning_rate": 2.2404897209461688e-05,
+      "loss": 0.3949,
+      "step": 1732
+    },
+    {
+      "epoch": 3.404715127701375,
+      "grad_norm": 0.07268806368333983,
+      "learning_rate": 2.2355632880386663e-05,
+      "loss": 0.4085,
+      "step": 1733
+    },
+    {
+      "epoch": 3.406679764243615,
+      "grad_norm": 0.08150576469469907,
+      "learning_rate": 2.230640175870587e-05,
+      "loss": 0.4023,
+      "step": 1734
+    },
+    {
+      "epoch": 3.4086444007858545,
+      "grad_norm": 0.07454696850457224,
+      "learning_rate": 2.2257203937074222e-05,
+      "loss": 0.4007,
+      "step": 1735
+    },
+    {
+      "epoch": 3.4106090373280944,
+      "grad_norm": 0.08124841979526694,
+      "learning_rate": 2.2208039508083946e-05,
+      "loss": 0.4026,
+      "step": 1736
+    },
+    {
+      "epoch": 3.412573673870334,
+      "grad_norm": 0.07874657642271078,
+      "learning_rate": 2.2158908564264436e-05,
+      "loss": 0.3976,
+      "step": 1737
+    },
+    {
+      "epoch": 3.4145383104125737,
+      "grad_norm": 0.09491012144170076,
+      "learning_rate": 2.2109811198082066e-05,
+      "loss": 0.4037,
+      "step": 1738
+    },
+    {
+      "epoch": 3.4165029469548136,
+      "grad_norm": 0.08146865730726072,
+      "learning_rate": 2.206074750194003e-05,
+      "loss": 0.4054,
+      "step": 1739
+    },
+    {
+      "epoch": 3.418467583497053,
+      "grad_norm": 0.08423863554010623,
+      "learning_rate": 2.201171756817811e-05,
+      "loss": 0.4133,
+      "step": 1740
+    },
+    {
+      "epoch": 3.4204322200392925,
+      "grad_norm": 0.07941667629255293,
+      "learning_rate": 2.1962721489072592e-05,
+      "loss": 0.404,
+      "step": 1741
+    },
+    {
+      "epoch": 3.4223968565815324,
+      "grad_norm": 0.08389545768794017,
+      "learning_rate": 2.191375935683602e-05,
+      "loss": 0.3975,
+      "step": 1742
+    },
+    {
+      "epoch": 3.4243614931237722,
+      "grad_norm": 0.07936127828882825,
+      "learning_rate": 2.1864831263617056e-05,
+      "loss": 0.4097,
+      "step": 1743
+    },
+    {
+      "epoch": 3.4263261296660117,
+      "grad_norm": 0.08231983544723506,
+      "learning_rate": 2.1815937301500307e-05,
+      "loss": 0.4025,
+      "step": 1744
+    },
+    {
+      "epoch": 3.4282907662082516,
+      "grad_norm": 0.07822134392538403,
+      "learning_rate": 2.1767077562506148e-05,
+      "loss": 0.3998,
+      "step": 1745
+    },
+    {
+      "epoch": 3.430255402750491,
+      "grad_norm": 0.0845078123113617,
+      "learning_rate": 2.1718252138590498e-05,
+      "loss": 0.3973,
+      "step": 1746
+    },
+    {
+      "epoch": 3.432220039292731,
+      "grad_norm": 0.08635978889781967,
+      "learning_rate": 2.166946112164475e-05,
+      "loss": 0.3997,
+      "step": 1747
+    },
+    {
+      "epoch": 3.4341846758349703,
+      "grad_norm": 0.08210457312206339,
+      "learning_rate": 2.1620704603495534e-05,
+      "loss": 0.4026,
+      "step": 1748
+    },
+    {
+      "epoch": 3.43614931237721,
+      "grad_norm": 0.08095818755335647,
+      "learning_rate": 2.1571982675904506e-05,
+      "loss": 0.4022,
+      "step": 1749
+    },
+    {
+      "epoch": 3.43811394891945,
+      "grad_norm": 0.07425506512527098,
+      "learning_rate": 2.1523295430568278e-05,
+      "loss": 0.4043,
+      "step": 1750
+    },
+    {
+      "epoch": 3.4400785854616895,
+      "grad_norm": 0.08103788235430726,
+      "learning_rate": 2.147464295911815e-05,
+      "loss": 0.4077,
+      "step": 1751
+    },
+    {
+      "epoch": 3.4420432220039294,
+      "grad_norm": 0.08121853369657873,
+      "learning_rate": 2.1426025353120002e-05,
+      "loss": 0.4075,
+      "step": 1752
+    },
+    {
+      "epoch": 3.444007858546169,
+      "grad_norm": 0.07804370045819682,
+      "learning_rate": 2.137744270407408e-05,
+      "loss": 0.3993,
+      "step": 1753
+    },
+    {
+      "epoch": 3.4459724950884087,
+      "grad_norm": 0.08101341384817454,
+      "learning_rate": 2.1328895103414843e-05,
+      "loss": 0.4031,
+      "step": 1754
+    },
+    {
+      "epoch": 3.447937131630648,
+      "grad_norm": 0.07615975123816222,
+      "learning_rate": 2.1280382642510802e-05,
+      "loss": 0.3979,
+      "step": 1755
+    },
+    {
+      "epoch": 3.449901768172888,
+      "grad_norm": 0.08190158750128435,
+      "learning_rate": 2.1231905412664297e-05,
+      "loss": 0.3962,
+      "step": 1756
+    },
+    {
+      "epoch": 3.4518664047151275,
+      "grad_norm": 0.07665213183343984,
+      "learning_rate": 2.118346350511139e-05,
+      "loss": 0.4036,
+      "step": 1757
+    },
+    {
+      "epoch": 3.4538310412573674,
+      "grad_norm": 0.07469989838375338,
+      "learning_rate": 2.1135057011021694e-05,
+      "loss": 0.3984,
+      "step": 1758
+    },
+    {
+      "epoch": 3.455795677799607,
+      "grad_norm": 0.07810935208254827,
+      "learning_rate": 2.1086686021498083e-05,
+      "loss": 0.4024,
+      "step": 1759
+    },
+    {
+      "epoch": 3.4577603143418467,
+      "grad_norm": 0.08165756409066073,
+      "learning_rate": 2.1038350627576697e-05,
+      "loss": 0.4091,
+      "step": 1760
+    },
+    {
+      "epoch": 3.4597249508840866,
+      "grad_norm": 0.07930430261834914,
+      "learning_rate": 2.0990050920226644e-05,
+      "loss": 0.4019,
+      "step": 1761
+    },
+    {
+      "epoch": 3.461689587426326,
+      "grad_norm": 0.07463246784003785,
+      "learning_rate": 2.0941786990349887e-05,
+      "loss": 0.4102,
+      "step": 1762
+    },
+    {
+      "epoch": 3.463654223968566,
+      "grad_norm": 0.07985564087395768,
+      "learning_rate": 2.0893558928781033e-05,
+      "loss": 0.408,
+      "step": 1763
+    },
+    {
+      "epoch": 3.4656188605108054,
+      "grad_norm": 0.08660853204595166,
+      "learning_rate": 2.0845366826287215e-05,
+      "loss": 0.3949,
+      "step": 1764
+    },
+    {
+      "epoch": 3.4675834970530452,
+      "grad_norm": 0.07228424068830637,
+      "learning_rate": 2.0797210773567834e-05,
+      "loss": 0.3942,
+      "step": 1765
+    },
+    {
+      "epoch": 3.4695481335952847,
+      "grad_norm": 0.08231120794154967,
+      "learning_rate": 2.07490908612545e-05,
+      "loss": 0.3932,
+      "step": 1766
+    },
+    {
+      "epoch": 3.4715127701375246,
+      "grad_norm": 0.07951598178894063,
+      "learning_rate": 2.0701007179910784e-05,
+      "loss": 0.3984,
+      "step": 1767
+    },
+    {
+      "epoch": 3.4734774066797645,
+      "grad_norm": 0.07551154087385774,
+      "learning_rate": 2.0652959820032083e-05,
+      "loss": 0.4085,
+      "step": 1768
+    },
+    {
+      "epoch": 3.475442043222004,
+      "grad_norm": 0.08347212367743614,
+      "learning_rate": 2.0604948872045388e-05,
+      "loss": 0.4085,
+      "step": 1769
+    },
+    {
+      "epoch": 3.477406679764244,
+      "grad_norm": 0.07535397017354029,
+      "learning_rate": 2.0556974426309208e-05,
+      "loss": 0.3991,
+      "step": 1770
+    },
+    {
+      "epoch": 3.479371316306483,
+      "grad_norm": 0.07833133290198357,
+      "learning_rate": 2.0509036573113353e-05,
+      "loss": 0.4016,
+      "step": 1771
+    },
+    {
+      "epoch": 3.481335952848723,
+      "grad_norm": 0.08906188161604638,
+      "learning_rate": 2.0461135402678737e-05,
+      "loss": 0.3985,
+      "step": 1772
+    },
+    {
+      "epoch": 3.4833005893909625,
+      "grad_norm": 0.09045872829251493,
+      "learning_rate": 2.041327100515726e-05,
+      "loss": 0.409,
+      "step": 1773
+    },
+    {
+      "epoch": 3.4852652259332024,
+      "grad_norm": 0.08909575492320418,
+      "learning_rate": 2.0365443470631615e-05,
+      "loss": 0.4058,
+      "step": 1774
+    },
+    {
+      "epoch": 3.487229862475442,
+      "grad_norm": 0.09129255219813269,
+      "learning_rate": 2.031765288911508e-05,
+      "loss": 0.408,
+      "step": 1775
+    },
+    {
+      "epoch": 3.4891944990176817,
+      "grad_norm": 0.08203201307745563,
+      "learning_rate": 2.0269899350551427e-05,
+      "loss": 0.3961,
+      "step": 1776
+    },
+    {
+      "epoch": 3.491159135559921,
+      "grad_norm": 0.08503946982808311,
+      "learning_rate": 2.022218294481471e-05,
+      "loss": 0.4045,
+      "step": 1777
+    },
+    {
+      "epoch": 3.493123772102161,
+      "grad_norm": 0.07595218037842626,
+      "learning_rate": 2.0174503761709058e-05,
+      "loss": 0.3935,
+      "step": 1778
+    },
+    {
+      "epoch": 3.495088408644401,
+      "grad_norm": 0.07833015390450979,
+      "learning_rate": 2.0126861890968582e-05,
+      "loss": 0.4024,
+      "step": 1779
+    },
+    {
+      "epoch": 3.4970530451866404,
+      "grad_norm": 0.0788499495039049,
+      "learning_rate": 2.0079257422257168e-05,
+      "loss": 0.3972,
+      "step": 1780
+    },
+    {
+      "epoch": 3.4990176817288803,
+      "grad_norm": 0.07487841060761408,
+      "learning_rate": 2.0031690445168292e-05,
+      "loss": 0.4037,
+      "step": 1781
+    },
+    {
+      "epoch": 3.5009823182711197,
+      "grad_norm": 0.07952094049876654,
+      "learning_rate": 1.9984161049224894e-05,
+      "loss": 0.3963,
+      "step": 1782
+    },
+    {
+      "epoch": 3.5029469548133596,
+      "grad_norm": 0.07867730312121303,
+      "learning_rate": 1.9936669323879155e-05,
+      "loss": 0.4096,
+      "step": 1783
+    },
+    {
+      "epoch": 3.504911591355599,
+      "grad_norm": 0.07641144602067207,
+      "learning_rate": 1.98892153585124e-05,
+      "loss": 0.4016,
+      "step": 1784
+    },
+    {
+      "epoch": 3.506876227897839,
+      "grad_norm": 0.08637477235801638,
+      "learning_rate": 1.984179924243482e-05,
+      "loss": 0.399,
+      "step": 1785
+    },
+    {
+      "epoch": 3.508840864440079,
+      "grad_norm": 0.07488655938903717,
+      "learning_rate": 1.979442106488544e-05,
+      "loss": 0.4034,
+      "step": 1786
+    },
+    {
+      "epoch": 3.5108055009823183,
+      "grad_norm": 0.07901984906038066,
+      "learning_rate": 1.9747080915031866e-05,
+      "loss": 0.4023,
+      "step": 1787
+    },
+    {
+      "epoch": 3.5127701375245577,
+      "grad_norm": 0.07732090650274363,
+      "learning_rate": 1.96997788819701e-05,
+      "loss": 0.4018,
+      "step": 1788
+    },
+    {
+      "epoch": 3.5147347740667976,
+      "grad_norm": 0.07796957496020122,
+      "learning_rate": 1.9652515054724444e-05,
+      "loss": 0.4055,
+      "step": 1789
+    },
+    {
+      "epoch": 3.5166994106090375,
+      "grad_norm": 0.09124246698232427,
+      "learning_rate": 1.960528952224728e-05,
+      "loss": 0.4,
+      "step": 1790
+    },
+    {
+      "epoch": 3.518664047151277,
+      "grad_norm": 0.07546581600532903,
+      "learning_rate": 1.955810237341893e-05,
+      "loss": 0.4017,
+      "step": 1791
+    },
+    {
+      "epoch": 3.520628683693517,
+      "grad_norm": 0.07435226373441858,
+      "learning_rate": 1.9510953697047466e-05,
+      "loss": 0.4027,
+      "step": 1792
+    },
+    {
+      "epoch": 3.522593320235756,
+      "grad_norm": 0.09382924104867246,
+      "learning_rate": 1.946384358186856e-05,
+      "loss": 0.4133,
+      "step": 1793
+    },
+    {
+      "epoch": 3.524557956777996,
+      "grad_norm": 0.07679696900807588,
+      "learning_rate": 1.9416772116545286e-05,
+      "loss": 0.4079,
+      "step": 1794
+    },
+    {
+      "epoch": 3.5265225933202355,
+      "grad_norm": 0.08368383659262482,
+      "learning_rate": 1.9369739389668006e-05,
+      "loss": 0.4012,
+      "step": 1795
+    },
+    {
+      "epoch": 3.5284872298624754,
+      "grad_norm": 0.07616734611672425,
+      "learning_rate": 1.932274548975417e-05,
+      "loss": 0.4005,
+      "step": 1796
+    },
+    {
+      "epoch": 3.5304518664047153,
+      "grad_norm": 0.07585562247209893,
+      "learning_rate": 1.927579050524816e-05,
+      "loss": 0.3987,
+      "step": 1797
+    },
+    {
+      "epoch": 3.5324165029469548,
+      "grad_norm": 0.08634091854572687,
+      "learning_rate": 1.9228874524521077e-05,
+      "loss": 0.4008,
+      "step": 1798
+    },
+    {
+      "epoch": 3.5343811394891946,
+      "grad_norm": 0.07676322491777511,
+      "learning_rate": 1.918199763587066e-05,
+      "loss": 0.4035,
+      "step": 1799
+    },
+    {
+      "epoch": 3.536345776031434,
+      "grad_norm": 0.0839884527863537,
+      "learning_rate": 1.9135159927521068e-05,
+      "loss": 0.4016,
+      "step": 1800
+    },
+    {
+      "epoch": 3.538310412573674,
+      "grad_norm": 0.07839742835085586,
+      "learning_rate": 1.9088361487622707e-05,
+      "loss": 0.4022,
+      "step": 1801
+    },
+    {
+      "epoch": 3.5402750491159134,
+      "grad_norm": 0.08162726972089206,
+      "learning_rate": 1.9041602404252088e-05,
+      "loss": 0.3978,
+      "step": 1802
+    },
+    {
+      "epoch": 3.5422396856581533,
+      "grad_norm": 0.08516541473429448,
+      "learning_rate": 1.899488276541166e-05,
+      "loss": 0.4009,
+      "step": 1803
+    },
+    {
+      "epoch": 3.544204322200393,
+      "grad_norm": 0.07810589253201565,
+      "learning_rate": 1.89482026590296e-05,
+      "loss": 0.3945,
+      "step": 1804
+    },
+    {
+      "epoch": 3.5461689587426326,
+      "grad_norm": 0.08328953775191177,
+      "learning_rate": 1.8901562172959718e-05,
+      "loss": 0.3972,
+      "step": 1805
+    },
+    {
+      "epoch": 3.548133595284872,
+      "grad_norm": 0.07536326372857173,
+      "learning_rate": 1.8854961394981263e-05,
+      "loss": 0.4064,
+      "step": 1806
+    },
+    {
+      "epoch": 3.550098231827112,
+      "grad_norm": 0.0802025674237676,
+      "learning_rate": 1.880840041279871e-05,
+      "loss": 0.3982,
+      "step": 1807
+    },
+    {
+      "epoch": 3.552062868369352,
+      "grad_norm": 0.0791669009113214,
+      "learning_rate": 1.8761879314041666e-05,
+      "loss": 0.408,
+      "step": 1808
+    },
+    {
+      "epoch": 3.5540275049115913,
+      "grad_norm": 0.08209723000219013,
+      "learning_rate": 1.8715398186264685e-05,
+      "loss": 0.408,
+      "step": 1809
+    },
+    {
+      "epoch": 3.555992141453831,
+      "grad_norm": 0.0777747709453421,
+      "learning_rate": 1.8668957116947074e-05,
+      "loss": 0.4014,
+      "step": 1810
+    },
+    {
+      "epoch": 3.5579567779960706,
+      "grad_norm": 0.08491613006225522,
+      "learning_rate": 1.8622556193492754e-05,
+      "loss": 0.3967,
+      "step": 1811
+    },
+    {
+      "epoch": 3.5599214145383105,
+      "grad_norm": 0.07547844441466134,
+      "learning_rate": 1.8576195503230097e-05,
+      "loss": 0.4053,
+      "step": 1812
+    },
+    {
+      "epoch": 3.56188605108055,
+      "grad_norm": 0.08500321202051749,
+      "learning_rate": 1.852987513341176e-05,
+      "loss": 0.4039,
+      "step": 1813
+    },
+    {
+      "epoch": 3.56385068762279,
+      "grad_norm": 0.07628785438360357,
+      "learning_rate": 1.8483595171214473e-05,
+      "loss": 0.4051,
+      "step": 1814
+    },
+    {
+      "epoch": 3.5658153241650297,
+      "grad_norm": 0.07468767215240342,
+      "learning_rate": 1.843735570373897e-05,
+      "loss": 0.4057,
+      "step": 1815
+    },
+    {
+      "epoch": 3.567779960707269,
+      "grad_norm": 0.07258861169063384,
+      "learning_rate": 1.8391156818009764e-05,
+      "loss": 0.3952,
+      "step": 1816
+    },
+    {
+      "epoch": 3.569744597249509,
+      "grad_norm": 0.07600727676002517,
+      "learning_rate": 1.8344998600974944e-05,
+      "loss": 0.4005,
+      "step": 1817
+    },
+    {
+      "epoch": 3.5717092337917484,
+      "grad_norm": 0.07430819222350618,
+      "learning_rate": 1.8298881139506113e-05,
+      "loss": 0.3968,
+      "step": 1818
+    },
+    {
+      "epoch": 3.5736738703339883,
+      "grad_norm": 0.07948635325851114,
+      "learning_rate": 1.8252804520398147e-05,
+      "loss": 0.4028,
+      "step": 1819
+    },
+    {
+      "epoch": 3.5756385068762278,
+      "grad_norm": 0.07237988422730401,
+      "learning_rate": 1.820676883036906e-05,
+      "loss": 0.4027,
+      "step": 1820
+    },
+    {
+      "epoch": 3.5776031434184676,
+      "grad_norm": 0.07944109921477316,
+      "learning_rate": 1.816077415605984e-05,
+      "loss": 0.4034,
+      "step": 1821
+    },
+    {
+      "epoch": 3.5795677799607075,
+      "grad_norm": 0.07437704468644644,
+      "learning_rate": 1.8114820584034268e-05,
+      "loss": 0.4026,
+      "step": 1822
+    },
+    {
+      "epoch": 3.581532416502947,
+      "grad_norm": 0.08060910401757315,
+      "learning_rate": 1.80689082007788e-05,
+      "loss": 0.4021,
+      "step": 1823
+    },
+    {
+      "epoch": 3.5834970530451864,
+      "grad_norm": 0.07929738827961239,
+      "learning_rate": 1.802303709270232e-05,
+      "loss": 0.4036,
+      "step": 1824
+    },
+    {
+      "epoch": 3.5854616895874263,
+      "grad_norm": 0.07610858400517262,
+      "learning_rate": 1.7977207346136064e-05,
+      "loss": 0.3975,
+      "step": 1825
+    },
+    {
+      "epoch": 3.587426326129666,
+      "grad_norm": 0.0758434175359333,
+      "learning_rate": 1.7931419047333452e-05,
+      "loss": 0.391,
+      "step": 1826
+    },
+    {
+      "epoch": 3.5893909626719056,
+      "grad_norm": 0.08092356632931848,
+      "learning_rate": 1.788567228246983e-05,
+      "loss": 0.4028,
+      "step": 1827
+    },
+    {
+      "epoch": 3.5913555992141455,
+      "grad_norm": 0.07577874351708982,
+      "learning_rate": 1.7839967137642422e-05,
+      "loss": 0.4121,
+      "step": 1828
+    },
+    {
+      "epoch": 3.593320235756385,
+      "grad_norm": 0.07776247926346612,
+      "learning_rate": 1.779430369887012e-05,
+      "loss": 0.4048,
+      "step": 1829
+    },
+    {
+      "epoch": 3.595284872298625,
+      "grad_norm": 0.08997808511291669,
+      "learning_rate": 1.774868205209331e-05,
+      "loss": 0.4092,
+      "step": 1830
+    },
+    {
+      "epoch": 3.5972495088408643,
+      "grad_norm": 0.07776605405085447,
+      "learning_rate": 1.7703102283173735e-05,
+      "loss": 0.4068,
+      "step": 1831
+    },
+    {
+      "epoch": 3.599214145383104,
+      "grad_norm": 0.08592683578885112,
+      "learning_rate": 1.7657564477894325e-05,
+      "loss": 0.399,
+      "step": 1832
+    },
+    {
+      "epoch": 3.601178781925344,
+      "grad_norm": 0.07303687429946173,
+      "learning_rate": 1.7612068721958998e-05,
+      "loss": 0.4014,
+      "step": 1833
+    },
+    {
+      "epoch": 3.6031434184675835,
+      "grad_norm": 0.08790429648763147,
+      "learning_rate": 1.756661510099257e-05,
+      "loss": 0.404,
+      "step": 1834
+    },
+    {
+      "epoch": 3.6051080550098233,
+      "grad_norm": 0.06963916065864258,
+      "learning_rate": 1.7521203700540553e-05,
+      "loss": 0.4113,
+      "step": 1835
+    },
+    {
+      "epoch": 3.607072691552063,
+      "grad_norm": 0.08440378914792816,
+      "learning_rate": 1.7475834606069004e-05,
+      "loss": 0.3976,
+      "step": 1836
+    },
+    {
+      "epoch": 3.6090373280943027,
+      "grad_norm": 0.07603665597199238,
+      "learning_rate": 1.7430507902964313e-05,
+      "loss": 0.3995,
+      "step": 1837
+    },
+    {
+      "epoch": 3.611001964636542,
+      "grad_norm": 0.07615640713431732,
+      "learning_rate": 1.7385223676533144e-05,
+      "loss": 0.403,
+      "step": 1838
+    },
+    {
+      "epoch": 3.612966601178782,
+      "grad_norm": 0.07631853865213062,
+      "learning_rate": 1.733998201200219e-05,
+      "loss": 0.4001,
+      "step": 1839
+    },
+    {
+      "epoch": 3.614931237721022,
+      "grad_norm": 0.08149585537234863,
+      "learning_rate": 1.7294782994518046e-05,
+      "loss": 0.3966,
+      "step": 1840
+    },
+    {
+      "epoch": 3.6168958742632613,
+      "grad_norm": 0.07962632736068805,
+      "learning_rate": 1.724962670914705e-05,
+      "loss": 0.4088,
+      "step": 1841
+    },
+    {
+      "epoch": 3.6188605108055008,
+      "grad_norm": 0.07525548128719053,
+      "learning_rate": 1.720451324087512e-05,
+      "loss": 0.4016,
+      "step": 1842
+    },
+    {
+      "epoch": 3.6208251473477406,
+      "grad_norm": 0.07802552991834835,
+      "learning_rate": 1.7159442674607548e-05,
+      "loss": 0.408,
+      "step": 1843
+    },
+    {
+      "epoch": 3.6227897838899805,
+      "grad_norm": 0.08131274522134428,
+      "learning_rate": 1.711441509516894e-05,
+      "loss": 0.398,
+      "step": 1844
+    },
+    {
+      "epoch": 3.62475442043222,
+      "grad_norm": 0.0742134848121323,
+      "learning_rate": 1.7069430587302984e-05,
+      "loss": 0.3986,
+      "step": 1845
+    },
+    {
+      "epoch": 3.62671905697446,
+      "grad_norm": 0.08068562347937643,
+      "learning_rate": 1.7024489235672273e-05,
+      "loss": 0.3976,
+      "step": 1846
+    },
+    {
+      "epoch": 3.6286836935166993,
+      "grad_norm": 0.07421612510479146,
+      "learning_rate": 1.6979591124858204e-05,
+      "loss": 0.4015,
+      "step": 1847
+    },
+    {
+      "epoch": 3.630648330058939,
+      "grad_norm": 0.07373094256039563,
+      "learning_rate": 1.6934736339360803e-05,
+      "loss": 0.4062,
+      "step": 1848
+    },
+    {
+      "epoch": 3.6326129666011786,
+      "grad_norm": 0.07493153582681775,
+      "learning_rate": 1.688992496359854e-05,
+      "loss": 0.3996,
+      "step": 1849
+    },
+    {
+      "epoch": 3.6345776031434185,
+      "grad_norm": 0.07321487739367057,
+      "learning_rate": 1.684515708190819e-05,
+      "loss": 0.4011,
+      "step": 1850
+    },
+    {
+      "epoch": 3.6365422396856584,
+      "grad_norm": 0.0747292607538307,
+      "learning_rate": 1.6800432778544665e-05,
+      "loss": 0.4022,
+      "step": 1851
+    },
+    {
+      "epoch": 3.638506876227898,
+      "grad_norm": 0.07189234233842018,
+      "learning_rate": 1.675575213768089e-05,
+      "loss": 0.4094,
+      "step": 1852
+    },
+    {
+      "epoch": 3.6404715127701373,
+      "grad_norm": 0.07537489719753057,
+      "learning_rate": 1.671111524340756e-05,
+      "loss": 0.4129,
+      "step": 1853
+    },
+    {
+      "epoch": 3.642436149312377,
+      "grad_norm": 0.07295296834263842,
+      "learning_rate": 1.666652217973308e-05,
+      "loss": 0.3989,
+      "step": 1854
+    },
+    {
+      "epoch": 3.644400785854617,
+      "grad_norm": 0.07259431368635465,
+      "learning_rate": 1.662197303058337e-05,
+      "loss": 0.4027,
+      "step": 1855
+    },
+    {
+      "epoch": 3.6463654223968565,
+      "grad_norm": 0.07517849352303735,
+      "learning_rate": 1.657746787980165e-05,
+      "loss": 0.4094,
+      "step": 1856
+    },
+    {
+      "epoch": 3.6483300589390963,
+      "grad_norm": 0.07634111863571573,
+      "learning_rate": 1.6533006811148395e-05,
+      "loss": 0.4042,
+      "step": 1857
+    },
+    {
+      "epoch": 3.6502946954813362,
+      "grad_norm": 0.07197304546742307,
+      "learning_rate": 1.648858990830108e-05,
+      "loss": 0.4068,
+      "step": 1858
+    },
+    {
+      "epoch": 3.6522593320235757,
+      "grad_norm": 0.07372133332197302,
+      "learning_rate": 1.6444217254854064e-05,
+      "loss": 0.3986,
+      "step": 1859
+    },
+    {
+      "epoch": 3.654223968565815,
+      "grad_norm": 0.0723666156412117,
+      "learning_rate": 1.639988893431843e-05,
+      "loss": 0.402,
+      "step": 1860
+    },
+    {
+      "epoch": 3.656188605108055,
+      "grad_norm": 0.07788631810261307,
+      "learning_rate": 1.6355605030121844e-05,
+      "loss": 0.3995,
+      "step": 1861
+    },
+    {
+      "epoch": 3.658153241650295,
+      "grad_norm": 0.06978806015355328,
+      "learning_rate": 1.631136562560833e-05,
+      "loss": 0.3954,
+      "step": 1862
+    },
+    {
+      "epoch": 3.6601178781925343,
+      "grad_norm": 0.07223055700646738,
+      "learning_rate": 1.6267170804038196e-05,
+      "loss": 0.3978,
+      "step": 1863
+    },
+    {
+      "epoch": 3.662082514734774,
+      "grad_norm": 0.07679533668314709,
+      "learning_rate": 1.6223020648587848e-05,
+      "loss": 0.3985,
+      "step": 1864
+    },
+    {
+      "epoch": 3.6640471512770136,
+      "grad_norm": 0.07349859511955957,
+      "learning_rate": 1.6178915242349624e-05,
+      "loss": 0.4026,
+      "step": 1865
+    },
+    {
+      "epoch": 3.6660117878192535,
+      "grad_norm": 0.07831080888684497,
+      "learning_rate": 1.6134854668331616e-05,
+      "loss": 0.4028,
+      "step": 1866
+    },
+    {
+      "epoch": 3.667976424361493,
+      "grad_norm": 0.07365375797305107,
+      "learning_rate": 1.609083900945757e-05,
+      "loss": 0.4056,
+      "step": 1867
+    },
+    {
+      "epoch": 3.669941060903733,
+      "grad_norm": 0.07298425518463603,
+      "learning_rate": 1.604686834856669e-05,
+      "loss": 0.4096,
+      "step": 1868
+    },
+    {
+      "epoch": 3.6719056974459727,
+      "grad_norm": 0.07350415445198857,
+      "learning_rate": 1.6002942768413494e-05,
+      "loss": 0.4056,
+      "step": 1869
+    },
+    {
+      "epoch": 3.673870333988212,
+      "grad_norm": 0.07363117550122496,
+      "learning_rate": 1.5959062351667656e-05,
+      "loss": 0.3996,
+      "step": 1870
+    },
+    {
+      "epoch": 3.6758349705304516,
+      "grad_norm": 0.07203582232285262,
+      "learning_rate": 1.591522718091386e-05,
+      "loss": 0.406,
+      "step": 1871
+    },
+    {
+      "epoch": 3.6777996070726915,
+      "grad_norm": 0.0789435965583153,
+      "learning_rate": 1.5871437338651602e-05,
+      "loss": 0.4005,
+      "step": 1872
+    },
+    {
+      "epoch": 3.6797642436149314,
+      "grad_norm": 0.07047014143311248,
+      "learning_rate": 1.5827692907295102e-05,
+      "loss": 0.4005,
+      "step": 1873
+    },
+    {
+      "epoch": 3.681728880157171,
+      "grad_norm": 0.07572636652543344,
+      "learning_rate": 1.5783993969173117e-05,
+      "loss": 0.3991,
+      "step": 1874
+    },
+    {
+      "epoch": 3.6836935166994107,
+      "grad_norm": 0.07643629199596653,
+      "learning_rate": 1.574034060652874e-05,
+      "loss": 0.3983,
+      "step": 1875
+    },
+    {
+      "epoch": 3.68565815324165,
+      "grad_norm": 0.07114084469262785,
+      "learning_rate": 1.5696732901519344e-05,
+      "loss": 0.3991,
+      "step": 1876
+    },
+    {
+      "epoch": 3.68762278978389,
+      "grad_norm": 0.0773394933851934,
+      "learning_rate": 1.5653170936216334e-05,
+      "loss": 0.3884,
+      "step": 1877
+    },
+    {
+      "epoch": 3.6895874263261295,
+      "grad_norm": 0.08150509446443001,
+      "learning_rate": 1.560965479260506e-05,
+      "loss": 0.3987,
+      "step": 1878
+    },
+    {
+      "epoch": 3.6915520628683693,
+      "grad_norm": 0.07659942671411148,
+      "learning_rate": 1.5566184552584606e-05,
+      "loss": 0.404,
+      "step": 1879
+    },
+    {
+      "epoch": 3.6935166994106092,
+      "grad_norm": 0.08975929184648901,
+      "learning_rate": 1.5522760297967682e-05,
+      "loss": 0.4056,
+      "step": 1880
+    },
+    {
+      "epoch": 3.6954813359528487,
+      "grad_norm": 0.07486799407443681,
+      "learning_rate": 1.5479382110480466e-05,
+      "loss": 0.3974,
+      "step": 1881
+    },
+    {
+      "epoch": 3.6974459724950886,
+      "grad_norm": 0.08741225509527935,
+      "learning_rate": 1.543605007176238e-05,
+      "loss": 0.3953,
+      "step": 1882
+    },
+    {
+      "epoch": 3.699410609037328,
+      "grad_norm": 0.07899677487315653,
+      "learning_rate": 1.5392764263366044e-05,
+      "loss": 0.3999,
+      "step": 1883
+    },
+    {
+      "epoch": 3.701375245579568,
+      "grad_norm": 0.07830691754622726,
+      "learning_rate": 1.5349524766757072e-05,
+      "loss": 0.4029,
+      "step": 1884
+    },
+    {
+      "epoch": 3.7033398821218073,
+      "grad_norm": 0.07746438888043361,
+      "learning_rate": 1.5306331663313874e-05,
+      "loss": 0.4041,
+      "step": 1885
+    },
+    {
+      "epoch": 3.705304518664047,
+      "grad_norm": 0.0801803920154972,
+      "learning_rate": 1.526318503432758e-05,
+      "loss": 0.4002,
+      "step": 1886
+    },
+    {
+      "epoch": 3.707269155206287,
+      "grad_norm": 0.06907639254363208,
+      "learning_rate": 1.522008496100185e-05,
+      "loss": 0.4015,
+      "step": 1887
+    },
+    {
+      "epoch": 3.7092337917485265,
+      "grad_norm": 0.08081277168928173,
+      "learning_rate": 1.5177031524452721e-05,
+      "loss": 0.4067,
+      "step": 1888
+    },
+    {
+      "epoch": 3.711198428290766,
+      "grad_norm": 0.07441765834022883,
+      "learning_rate": 1.5134024805708461e-05,
+      "loss": 0.4027,
+      "step": 1889
+    },
+    {
+      "epoch": 3.713163064833006,
+      "grad_norm": 0.07569699695791177,
+      "learning_rate": 1.5091064885709398e-05,
+      "loss": 0.4038,
+      "step": 1890
+    },
+    {
+      "epoch": 3.7151277013752457,
+      "grad_norm": 0.07515516064730056,
+      "learning_rate": 1.5048151845307821e-05,
+      "loss": 0.4057,
+      "step": 1891
+    },
+    {
+      "epoch": 3.717092337917485,
+      "grad_norm": 0.0713391946267384,
+      "learning_rate": 1.5005285765267728e-05,
+      "loss": 0.3997,
+      "step": 1892
+    },
+    {
+      "epoch": 3.719056974459725,
+      "grad_norm": 0.07617524243061924,
+      "learning_rate": 1.4962466726264783e-05,
+      "loss": 0.4048,
+      "step": 1893
+    },
+    {
+      "epoch": 3.7210216110019645,
+      "grad_norm": 0.07579753484728168,
+      "learning_rate": 1.4919694808886126e-05,
+      "loss": 0.3924,
+      "step": 1894
+    },
+    {
+      "epoch": 3.7229862475442044,
+      "grad_norm": 0.07149550711459805,
+      "learning_rate": 1.4876970093630156e-05,
+      "loss": 0.4041,
+      "step": 1895
+    },
+    {
+      "epoch": 3.724950884086444,
+      "grad_norm": 0.07854913344159423,
+      "learning_rate": 1.483429266090649e-05,
+      "loss": 0.4188,
+      "step": 1896
+    },
+    {
+      "epoch": 3.7269155206286837,
+      "grad_norm": 0.07351678898491422,
+      "learning_rate": 1.4791662591035735e-05,
+      "loss": 0.4001,
+      "step": 1897
+    },
+    {
+      "epoch": 3.7288801571709236,
+      "grad_norm": 0.06724862807878827,
+      "learning_rate": 1.4749079964249364e-05,
+      "loss": 0.3966,
+      "step": 1898
+    },
+    {
+      "epoch": 3.730844793713163,
+      "grad_norm": 0.07524902156562395,
+      "learning_rate": 1.470654486068956e-05,
+      "loss": 0.4022,
+      "step": 1899
+    },
+    {
+      "epoch": 3.732809430255403,
+      "grad_norm": 0.07240726578503226,
+      "learning_rate": 1.4664057360409075e-05,
+      "loss": 0.3989,
+      "step": 1900
+    },
+    {
+      "epoch": 3.7347740667976423,
+      "grad_norm": 0.06882883292881418,
+      "learning_rate": 1.4621617543371031e-05,
+      "loss": 0.4002,
+      "step": 1901
+    },
+    {
+      "epoch": 3.7367387033398822,
+      "grad_norm": 0.07123693381854591,
+      "learning_rate": 1.4579225489448852e-05,
+      "loss": 0.4055,
+      "step": 1902
+    },
+    {
+      "epoch": 3.7387033398821217,
+      "grad_norm": 0.06923852160818676,
+      "learning_rate": 1.4536881278426065e-05,
+      "loss": 0.3984,
+      "step": 1903
+    },
+    {
+      "epoch": 3.7406679764243616,
+      "grad_norm": 0.07780674498604481,
+      "learning_rate": 1.4494584989996118e-05,
+      "loss": 0.4018,
+      "step": 1904
+    },
+    {
+      "epoch": 3.7426326129666014,
+      "grad_norm": 0.06819928031240537,
+      "learning_rate": 1.445233670376231e-05,
+      "loss": 0.4097,
+      "step": 1905
+    },
+    {
+      "epoch": 3.744597249508841,
+      "grad_norm": 0.06972984196472067,
+      "learning_rate": 1.4410136499237575e-05,
+      "loss": 0.4041,
+      "step": 1906
+    },
+    {
+      "epoch": 3.7465618860510803,
+      "grad_norm": 0.06579356610874487,
+      "learning_rate": 1.4367984455844365e-05,
+      "loss": 0.4054,
+      "step": 1907
+    },
+    {
+      "epoch": 3.74852652259332,
+      "grad_norm": 0.07032413850810303,
+      "learning_rate": 1.4325880652914483e-05,
+      "loss": 0.4002,
+      "step": 1908
+    },
+    {
+      "epoch": 3.75049115913556,
+      "grad_norm": 0.06573358295103715,
+      "learning_rate": 1.4283825169688945e-05,
+      "loss": 0.4045,
+      "step": 1909
+    },
+    {
+      "epoch": 3.7524557956777995,
+      "grad_norm": 0.07147748828586879,
+      "learning_rate": 1.4241818085317847e-05,
+      "loss": 0.413,
+      "step": 1910
+    },
+    {
+      "epoch": 3.7544204322200394,
+      "grad_norm": 0.07663875825767029,
+      "learning_rate": 1.4199859478860143e-05,
+      "loss": 0.4008,
+      "step": 1911
+    },
+    {
+      "epoch": 3.756385068762279,
+      "grad_norm": 0.0682233934507516,
+      "learning_rate": 1.41579494292836e-05,
+      "loss": 0.406,
+      "step": 1912
+    },
+    {
+      "epoch": 3.7583497053045187,
+      "grad_norm": 0.07537582730253191,
+      "learning_rate": 1.4116088015464601e-05,
+      "loss": 0.3977,
+      "step": 1913
+    },
+    {
+      "epoch": 3.760314341846758,
+      "grad_norm": 0.07823829620902766,
+      "learning_rate": 1.4074275316187941e-05,
+      "loss": 0.3957,
+      "step": 1914
+    },
+    {
+      "epoch": 3.762278978388998,
+      "grad_norm": 0.07482676776720608,
+      "learning_rate": 1.4032511410146792e-05,
+      "loss": 0.405,
+      "step": 1915
+    },
+    {
+      "epoch": 3.764243614931238,
+      "grad_norm": 0.08069350942084583,
+      "learning_rate": 1.3990796375942459e-05,
+      "loss": 0.4065,
+      "step": 1916
+    },
+    {
+      "epoch": 3.7662082514734774,
+      "grad_norm": 0.07366999306886934,
+      "learning_rate": 1.3949130292084286e-05,
+      "loss": 0.4054,
+      "step": 1917
+    },
+    {
+      "epoch": 3.768172888015717,
+      "grad_norm": 0.0746183550671485,
+      "learning_rate": 1.3907513236989486e-05,
+      "loss": 0.3967,
+      "step": 1918
+    },
+    {
+      "epoch": 3.7701375245579567,
+      "grad_norm": 0.06910021146414977,
+      "learning_rate": 1.3865945288982997e-05,
+      "loss": 0.3999,
+      "step": 1919
+    },
+    {
+      "epoch": 3.7721021611001966,
+      "grad_norm": 0.06637558300253697,
+      "learning_rate": 1.3824426526297346e-05,
+      "loss": 0.3972,
+      "step": 1920
+    },
+    {
+      "epoch": 3.774066797642436,
+      "grad_norm": 0.07655544763898874,
+      "learning_rate": 1.378295702707245e-05,
+      "loss": 0.4017,
+      "step": 1921
+    },
+    {
+      "epoch": 3.776031434184676,
+      "grad_norm": 0.0703210191201227,
+      "learning_rate": 1.3741536869355563e-05,
+      "loss": 0.4025,
+      "step": 1922
+    },
+    {
+      "epoch": 3.777996070726916,
+      "grad_norm": 0.07012933026806083,
+      "learning_rate": 1.3700166131101064e-05,
+      "loss": 0.3956,
+      "step": 1923
+    },
+    {
+      "epoch": 3.7799607072691552,
+      "grad_norm": 0.06796945969826204,
+      "learning_rate": 1.3658844890170281e-05,
+      "loss": 0.4126,
+      "step": 1924
+    },
+    {
+      "epoch": 3.7819253438113947,
+      "grad_norm": 0.0755489677265984,
+      "learning_rate": 1.3617573224331442e-05,
+      "loss": 0.4049,
+      "step": 1925
+    },
+    {
+      "epoch": 3.7838899803536346,
+      "grad_norm": 0.06955970350811949,
+      "learning_rate": 1.3576351211259438e-05,
+      "loss": 0.4032,
+      "step": 1926
+    },
+    {
+      "epoch": 3.7858546168958744,
+      "grad_norm": 0.07248888162530481,
+      "learning_rate": 1.3535178928535721e-05,
+      "loss": 0.3933,
+      "step": 1927
+    },
+    {
+      "epoch": 3.787819253438114,
+      "grad_norm": 0.0723222956654959,
+      "learning_rate": 1.349405645364816e-05,
+      "loss": 0.4024,
+      "step": 1928
+    },
+    {
+      "epoch": 3.7897838899803538,
+      "grad_norm": 0.06662237267842523,
+      "learning_rate": 1.345298386399088e-05,
+      "loss": 0.405,
+      "step": 1929
+    },
+    {
+      "epoch": 3.791748526522593,
+      "grad_norm": 0.07063824166057403,
+      "learning_rate": 1.3411961236864088e-05,
+      "loss": 0.4042,
+      "step": 1930
+    },
+    {
+      "epoch": 3.793713163064833,
+      "grad_norm": 0.07162335305533662,
+      "learning_rate": 1.3370988649473996e-05,
+      "loss": 0.3956,
+      "step": 1931
+    },
+    {
+      "epoch": 3.7956777996070725,
+      "grad_norm": 0.07059687568136831,
+      "learning_rate": 1.333006617893263e-05,
+      "loss": 0.4041,
+      "step": 1932
+    },
+    {
+      "epoch": 3.7976424361493124,
+      "grad_norm": 0.07238601225764041,
+      "learning_rate": 1.3289193902257704e-05,
+      "loss": 0.3943,
+      "step": 1933
+    },
+    {
+      "epoch": 3.7996070726915523,
+      "grad_norm": 0.06822601466367134,
+      "learning_rate": 1.3248371896372429e-05,
+      "loss": 0.4035,
+      "step": 1934
+    },
+    {
+      "epoch": 3.8015717092337917,
+      "grad_norm": 0.06742055680403077,
+      "learning_rate": 1.3207600238105429e-05,
+      "loss": 0.4027,
+      "step": 1935
+    },
+    {
+      "epoch": 3.803536345776031,
+      "grad_norm": 0.07112890468488242,
+      "learning_rate": 1.3166879004190608e-05,
+      "loss": 0.4035,
+      "step": 1936
+    },
+    {
+      "epoch": 3.805500982318271,
+      "grad_norm": 0.06862371359144873,
+      "learning_rate": 1.3126208271266897e-05,
+      "loss": 0.407,
+      "step": 1937
+    },
+    {
+      "epoch": 3.807465618860511,
+      "grad_norm": 0.07052422292327054,
+      "learning_rate": 1.308558811587823e-05,
+      "loss": 0.4049,
+      "step": 1938
+    },
+    {
+      "epoch": 3.8094302554027504,
+      "grad_norm": 0.07172398018225133,
+      "learning_rate": 1.304501861447335e-05,
+      "loss": 0.4061,
+      "step": 1939
+    },
+    {
+      "epoch": 3.8113948919449903,
+      "grad_norm": 0.06730334750073662,
+      "learning_rate": 1.3004499843405633e-05,
+      "loss": 0.402,
+      "step": 1940
+    },
+    {
+      "epoch": 3.8133595284872297,
+      "grad_norm": 0.0689306560971264,
+      "learning_rate": 1.2964031878933021e-05,
+      "loss": 0.4013,
+      "step": 1941
+    },
+    {
+      "epoch": 3.8153241650294696,
+      "grad_norm": 0.06798083772111306,
+      "learning_rate": 1.2923614797217824e-05,
+      "loss": 0.4057,
+      "step": 1942
+    },
+    {
+      "epoch": 3.817288801571709,
+      "grad_norm": 0.07038527218644648,
+      "learning_rate": 1.2883248674326558e-05,
+      "loss": 0.3994,
+      "step": 1943
+    },
+    {
+      "epoch": 3.819253438113949,
+      "grad_norm": 0.0708407293877681,
+      "learning_rate": 1.2842933586229877e-05,
+      "loss": 0.4007,
+      "step": 1944
+    },
+    {
+      "epoch": 3.821218074656189,
+      "grad_norm": 0.06960854396631978,
+      "learning_rate": 1.2802669608802343e-05,
+      "loss": 0.396,
+      "step": 1945
+    },
+    {
+      "epoch": 3.8231827111984282,
+      "grad_norm": 0.06675043609072423,
+      "learning_rate": 1.276245681782239e-05,
+      "loss": 0.4023,
+      "step": 1946
+    },
+    {
+      "epoch": 3.825147347740668,
+      "grad_norm": 0.0799635656865461,
+      "learning_rate": 1.2722295288972037e-05,
+      "loss": 0.402,
+      "step": 1947
+    },
+    {
+      "epoch": 3.8271119842829076,
+      "grad_norm": 0.0705906609537971,
+      "learning_rate": 1.2682185097836879e-05,
+      "loss": 0.4012,
+      "step": 1948
+    },
+    {
+      "epoch": 3.8290766208251474,
+      "grad_norm": 0.06718020415103504,
+      "learning_rate": 1.2642126319905886e-05,
+      "loss": 0.4037,
+      "step": 1949
+    },
+    {
+      "epoch": 3.831041257367387,
+      "grad_norm": 0.08308777113750684,
+      "learning_rate": 1.260211903057123e-05,
+      "loss": 0.3972,
+      "step": 1950
+    },
+    {
+      "epoch": 3.8330058939096268,
+      "grad_norm": 0.0684548861843251,
+      "learning_rate": 1.2562163305128224e-05,
+      "loss": 0.403,
+      "step": 1951
+    },
+    {
+      "epoch": 3.8349705304518666,
+      "grad_norm": 0.06686341472879632,
+      "learning_rate": 1.2522259218775129e-05,
+      "loss": 0.4,
+      "step": 1952
+    },
+    {
+      "epoch": 3.836935166994106,
+      "grad_norm": 0.07615091092430859,
+      "learning_rate": 1.248240684661298e-05,
+      "loss": 0.3978,
+      "step": 1953
+    },
+    {
+      "epoch": 3.8388998035363455,
+      "grad_norm": 0.07272178013391459,
+      "learning_rate": 1.2442606263645512e-05,
+      "loss": 0.4023,
+      "step": 1954
+    },
+    {
+      "epoch": 3.8408644400785854,
+      "grad_norm": 0.07024401727956676,
+      "learning_rate": 1.240285754477903e-05,
+      "loss": 0.4063,
+      "step": 1955
+    },
+    {
+      "epoch": 3.8428290766208253,
+      "grad_norm": 0.07201946508681054,
+      "learning_rate": 1.2363160764822148e-05,
+      "loss": 0.4002,
+      "step": 1956
+    },
+    {
+      "epoch": 3.8447937131630647,
+      "grad_norm": 0.06251473145460106,
+      "learning_rate": 1.2323515998485793e-05,
+      "loss": 0.3921,
+      "step": 1957
+    },
+    {
+      "epoch": 3.8467583497053046,
+      "grad_norm": 0.06902111555371625,
+      "learning_rate": 1.2283923320382978e-05,
+      "loss": 0.3971,
+      "step": 1958
+    },
+    {
+      "epoch": 3.848722986247544,
+      "grad_norm": 0.07009005813125245,
+      "learning_rate": 1.2244382805028665e-05,
+      "loss": 0.4022,
+      "step": 1959
+    },
+    {
+      "epoch": 3.850687622789784,
+      "grad_norm": 0.06725281577965755,
+      "learning_rate": 1.220489452683967e-05,
+      "loss": 0.4017,
+      "step": 1960
+    },
+    {
+      "epoch": 3.8526522593320234,
+      "grad_norm": 0.0694954536823525,
+      "learning_rate": 1.2165458560134495e-05,
+      "loss": 0.3956,
+      "step": 1961
+    },
+    {
+      "epoch": 3.8546168958742633,
+      "grad_norm": 0.06748361079631171,
+      "learning_rate": 1.2126074979133193e-05,
+      "loss": 0.3968,
+      "step": 1962
+    },
+    {
+      "epoch": 3.856581532416503,
+      "grad_norm": 0.06969698539458563,
+      "learning_rate": 1.2086743857957189e-05,
+      "loss": 0.4074,
+      "step": 1963
+    },
+    {
+      "epoch": 3.8585461689587426,
+      "grad_norm": 0.07396453732520927,
+      "learning_rate": 1.2047465270629202e-05,
+      "loss": 0.4029,
+      "step": 1964
+    },
+    {
+      "epoch": 3.8605108055009825,
+      "grad_norm": 0.07057855767714646,
+      "learning_rate": 1.2008239291073118e-05,
+      "loss": 0.4105,
+      "step": 1965
+    },
+    {
+      "epoch": 3.862475442043222,
+      "grad_norm": 0.06784958345703423,
+      "learning_rate": 1.1969065993113746e-05,
+      "loss": 0.4019,
+      "step": 1966
+    },
+    {
+      "epoch": 3.864440078585462,
+      "grad_norm": 0.07136856841831889,
+      "learning_rate": 1.1929945450476782e-05,
+      "loss": 0.3986,
+      "step": 1967
+    },
+    {
+      "epoch": 3.8664047151277012,
+      "grad_norm": 0.06953375099742057,
+      "learning_rate": 1.1890877736788645e-05,
+      "loss": 0.3952,
+      "step": 1968
+    },
+    {
+      "epoch": 3.868369351669941,
+      "grad_norm": 0.07032984921426026,
+      "learning_rate": 1.1851862925576283e-05,
+      "loss": 0.4058,
+      "step": 1969
+    },
+    {
+      "epoch": 3.870333988212181,
+      "grad_norm": 0.07629261407566977,
+      "learning_rate": 1.1812901090267128e-05,
+      "loss": 0.4044,
+      "step": 1970
+    },
+    {
+      "epoch": 3.8722986247544204,
+      "grad_norm": 0.06558931908208238,
+      "learning_rate": 1.1773992304188892e-05,
+      "loss": 0.408,
+      "step": 1971
+    },
+    {
+      "epoch": 3.87426326129666,
+      "grad_norm": 0.06856933741381056,
+      "learning_rate": 1.1735136640569426e-05,
+      "loss": 0.3989,
+      "step": 1972
+    },
+    {
+      "epoch": 3.8762278978388998,
+      "grad_norm": 0.0757827898125962,
+      "learning_rate": 1.1696334172536617e-05,
+      "loss": 0.4079,
+      "step": 1973
+    },
+    {
+      "epoch": 3.8781925343811396,
+      "grad_norm": 0.06601386210752483,
+      "learning_rate": 1.1657584973118263e-05,
+      "loss": 0.405,
+      "step": 1974
+    },
+    {
+      "epoch": 3.880157170923379,
+      "grad_norm": 0.07245952896872787,
+      "learning_rate": 1.1618889115241885e-05,
+      "loss": 0.4044,
+      "step": 1975
+    },
+    {
+      "epoch": 3.882121807465619,
+      "grad_norm": 0.06652621443910917,
+      "learning_rate": 1.158024667173459e-05,
+      "loss": 0.3993,
+      "step": 1976
+    },
+    {
+      "epoch": 3.8840864440078584,
+      "grad_norm": 0.07316845320922626,
+      "learning_rate": 1.1541657715322994e-05,
+      "loss": 0.4028,
+      "step": 1977
+    },
+    {
+      "epoch": 3.8860510805500983,
+      "grad_norm": 0.0614488538689494,
+      "learning_rate": 1.1503122318633038e-05,
+      "loss": 0.403,
+      "step": 1978
+    },
+    {
+      "epoch": 3.8880157170923377,
+      "grad_norm": 0.06994278963060006,
+      "learning_rate": 1.1464640554189837e-05,
+      "loss": 0.3991,
+      "step": 1979
+    },
+    {
+      "epoch": 3.8899803536345776,
+      "grad_norm": 0.0682877010183069,
+      "learning_rate": 1.1426212494417599e-05,
+      "loss": 0.3982,
+      "step": 1980
+    },
+    {
+      "epoch": 3.8919449901768175,
+      "grad_norm": 0.06330673283997267,
+      "learning_rate": 1.1387838211639464e-05,
+      "loss": 0.399,
+      "step": 1981
+    },
+    {
+      "epoch": 3.893909626719057,
+      "grad_norm": 0.06761874734281212,
+      "learning_rate": 1.13495177780773e-05,
+      "loss": 0.4072,
+      "step": 1982
+    },
+    {
+      "epoch": 3.895874263261297,
+      "grad_norm": 0.06751855925975371,
+      "learning_rate": 1.1311251265851713e-05,
+      "loss": 0.3973,
+      "step": 1983
+    },
+    {
+      "epoch": 3.8978388998035363,
+      "grad_norm": 0.07369534096848362,
+      "learning_rate": 1.1273038746981787e-05,
+      "loss": 0.4109,
+      "step": 1984
+    },
+    {
+      "epoch": 3.899803536345776,
+      "grad_norm": 0.06621918861864144,
+      "learning_rate": 1.123488029338497e-05,
+      "loss": 0.3974,
+      "step": 1985
+    },
+    {
+      "epoch": 3.9017681728880156,
+      "grad_norm": 0.06349792209974058,
+      "learning_rate": 1.1196775976876978e-05,
+      "loss": 0.4065,
+      "step": 1986
+    },
+    {
+      "epoch": 3.9037328094302555,
+      "grad_norm": 0.0684839152463357,
+      "learning_rate": 1.1158725869171643e-05,
+      "loss": 0.4067,
+      "step": 1987
+    },
+    {
+      "epoch": 3.9056974459724954,
+      "grad_norm": 0.0707538033054012,
+      "learning_rate": 1.1120730041880772e-05,
+      "loss": 0.4023,
+      "step": 1988
+    },
+    {
+      "epoch": 3.907662082514735,
+      "grad_norm": 0.06703270998255086,
+      "learning_rate": 1.1082788566513991e-05,
+      "loss": 0.4032,
+      "step": 1989
+    },
+    {
+      "epoch": 3.9096267190569742,
+      "grad_norm": 0.06716712736843278,
+      "learning_rate": 1.1044901514478665e-05,
+      "loss": 0.4003,
+      "step": 1990
+    },
+    {
+      "epoch": 3.911591355599214,
+      "grad_norm": 0.06500499117280914,
+      "learning_rate": 1.100706895707973e-05,
+      "loss": 0.3973,
+      "step": 1991
+    },
+    {
+      "epoch": 3.913555992141454,
+      "grad_norm": 0.06068052826101439,
+      "learning_rate": 1.0969290965519508e-05,
+      "loss": 0.397,
+      "step": 1992
+    },
+    {
+      "epoch": 3.9155206286836934,
+      "grad_norm": 0.06532992212558597,
+      "learning_rate": 1.0931567610897709e-05,
+      "loss": 0.391,
+      "step": 1993
+    },
+    {
+      "epoch": 3.9174852652259333,
+      "grad_norm": 0.06400419521418571,
+      "learning_rate": 1.0893898964211172e-05,
+      "loss": 0.3997,
+      "step": 1994
+    },
+    {
+      "epoch": 3.9194499017681728,
+      "grad_norm": 0.07073122449821673,
+      "learning_rate": 1.0856285096353752e-05,
+      "loss": 0.3945,
+      "step": 1995
+    },
+    {
+      "epoch": 3.9214145383104126,
+      "grad_norm": 0.3101590914120924,
+      "learning_rate": 1.081872607811624e-05,
+      "loss": 0.4059,
+      "step": 1996
+    },
+    {
+      "epoch": 3.923379174852652,
+      "grad_norm": 0.06137134325408761,
+      "learning_rate": 1.0781221980186198e-05,
+      "loss": 0.4016,
+      "step": 1997
+    },
+    {
+      "epoch": 3.925343811394892,
+      "grad_norm": 0.06779246171333435,
+      "learning_rate": 1.0743772873147802e-05,
+      "loss": 0.3961,
+      "step": 1998
+    },
+    {
+      "epoch": 3.927308447937132,
+      "grad_norm": 0.06851790195642792,
+      "learning_rate": 1.0706378827481748e-05,
+      "loss": 0.4,
+      "step": 1999
+    },
+    {
+      "epoch": 3.9292730844793713,
+      "grad_norm": 0.06640107939089517,
+      "learning_rate": 1.0669039913565124e-05,
+      "loss": 0.3998,
+      "step": 2000
+    },
+    {
+      "epoch": 3.9312377210216107,
+      "grad_norm": 0.07600172337274451,
+      "learning_rate": 1.0631756201671198e-05,
+      "loss": 0.3983,
+      "step": 2001
+    },
+    {
+      "epoch": 3.9332023575638506,
+      "grad_norm": 0.06678932975525444,
+      "learning_rate": 1.0594527761969418e-05,
+      "loss": 0.4029,
+      "step": 2002
+    },
+    {
+      "epoch": 3.9351669941060905,
+      "grad_norm": 0.06746933208115946,
+      "learning_rate": 1.0557354664525166e-05,
+      "loss": 0.4121,
+      "step": 2003
+    },
+    {
+      "epoch": 3.93713163064833,
+      "grad_norm": 0.07196069556421492,
+      "learning_rate": 1.0520236979299688e-05,
+      "loss": 0.4075,
+      "step": 2004
+    },
+    {
+      "epoch": 3.93909626719057,
+      "grad_norm": 0.06689599508700109,
+      "learning_rate": 1.0483174776149907e-05,
+      "loss": 0.4018,
+      "step": 2005
+    },
+    {
+      "epoch": 3.9410609037328097,
+      "grad_norm": 0.06746466576300976,
+      "learning_rate": 1.0446168124828358e-05,
+      "loss": 0.4031,
+      "step": 2006
+    },
+    {
+      "epoch": 3.943025540275049,
+      "grad_norm": 0.06615249822965788,
+      "learning_rate": 1.0409217094983032e-05,
+      "loss": 0.4084,
+      "step": 2007
+    },
+    {
+      "epoch": 3.9449901768172886,
+      "grad_norm": 0.06860209789349127,
+      "learning_rate": 1.0372321756157203e-05,
+      "loss": 0.4058,
+      "step": 2008
+    },
+    {
+      "epoch": 3.9469548133595285,
+      "grad_norm": 0.07523869931111538,
+      "learning_rate": 1.033548217778936e-05,
+      "loss": 0.4053,
+      "step": 2009
+    },
+    {
+      "epoch": 3.9489194499017684,
+      "grad_norm": 0.06762314248889623,
+      "learning_rate": 1.029869842921305e-05,
+      "loss": 0.4034,
+      "step": 2010
+    },
+    {
+      "epoch": 3.950884086444008,
+      "grad_norm": 0.06632922920171817,
+      "learning_rate": 1.0261970579656708e-05,
+      "loss": 0.3975,
+      "step": 2011
+    },
+    {
+      "epoch": 3.9528487229862477,
+      "grad_norm": 0.07492194548960043,
+      "learning_rate": 1.0225298698243629e-05,
+      "loss": 0.4067,
+      "step": 2012
+    },
+    {
+      "epoch": 3.954813359528487,
+      "grad_norm": 0.07373912409884645,
+      "learning_rate": 1.0188682853991737e-05,
+      "loss": 0.4104,
+      "step": 2013
+    },
+    {
+      "epoch": 3.956777996070727,
+      "grad_norm": 0.06386513521407051,
+      "learning_rate": 1.0152123115813467e-05,
+      "loss": 0.4015,
+      "step": 2014
+    },
+    {
+      "epoch": 3.9587426326129664,
+      "grad_norm": 0.06624996271003052,
+      "learning_rate": 1.0115619552515702e-05,
+      "loss": 0.3995,
+      "step": 2015
+    },
+    {
+      "epoch": 3.9607072691552063,
+      "grad_norm": 0.07409773341359095,
+      "learning_rate": 1.007917223279959e-05,
+      "loss": 0.398,
+      "step": 2016
+    },
+    {
+      "epoch": 3.962671905697446,
+      "grad_norm": 0.06352908190972788,
+      "learning_rate": 1.004278122526043e-05,
+      "loss": 0.4046,
+      "step": 2017
+    },
+    {
+      "epoch": 3.9646365422396856,
+      "grad_norm": 0.06698929433250807,
+      "learning_rate": 1.0006446598387512e-05,
+      "loss": 0.4019,
+      "step": 2018
+    },
+    {
+      "epoch": 3.966601178781925,
+      "grad_norm": 0.07335686904033532,
+      "learning_rate": 9.970168420564046e-06,
+      "loss": 0.4027,
+      "step": 2019
+    },
+    {
+      "epoch": 3.968565815324165,
+      "grad_norm": 0.06962923600723228,
+      "learning_rate": 9.933946760066991e-06,
+      "loss": 0.4022,
+      "step": 2020
+    },
+    {
+      "epoch": 3.970530451866405,
+      "grad_norm": 0.06556660148743444,
+      "learning_rate": 9.89778168506694e-06,
+      "loss": 0.3979,
+      "step": 2021
+    },
+    {
+      "epoch": 3.9724950884086443,
+      "grad_norm": 0.06585221761694499,
+      "learning_rate": 9.861673263627987e-06,
+      "loss": 0.4006,
+      "step": 2022
+    },
+    {
+      "epoch": 3.974459724950884,
+      "grad_norm": 0.06834583082237543,
+      "learning_rate": 9.825621563707623e-06,
+      "loss": 0.4016,
+      "step": 2023
+    },
+    {
+      "epoch": 3.9764243614931236,
+      "grad_norm": 0.06555645553189127,
+      "learning_rate": 9.789626653156538e-06,
+      "loss": 0.3921,
+      "step": 2024
+    },
+    {
+      "epoch": 3.9783889980353635,
+      "grad_norm": 0.06695662286281458,
+      "learning_rate": 9.753688599718586e-06,
+      "loss": 0.4001,
+      "step": 2025
+    },
+    {
+      "epoch": 3.980353634577603,
+      "grad_norm": 0.06997598172557186,
+      "learning_rate": 9.717807471030612e-06,
+      "loss": 0.3944,
+      "step": 2026
+    },
+    {
+      "epoch": 3.982318271119843,
+      "grad_norm": 0.06713552825951437,
+      "learning_rate": 9.681983334622287e-06,
+      "loss": 0.3978,
+      "step": 2027
+    },
+    {
+      "epoch": 3.9842829076620827,
+      "grad_norm": 0.06975238024026192,
+      "learning_rate": 9.646216257916064e-06,
+      "loss": 0.3982,
+      "step": 2028
+    },
+    {
+      "epoch": 3.986247544204322,
+      "grad_norm": 0.06348556239332637,
+      "learning_rate": 9.610506308226987e-06,
+      "loss": 0.4102,
+      "step": 2029
+    },
+    {
+      "epoch": 3.988212180746562,
+      "grad_norm": 0.06101886805928113,
+      "learning_rate": 9.574853552762594e-06,
+      "loss": 0.3975,
+      "step": 2030
+    },
+    {
+      "epoch": 3.9901768172888015,
+      "grad_norm": 0.06581732806887909,
+      "learning_rate": 9.53925805862277e-06,
+      "loss": 0.3998,
+      "step": 2031
+    },
+    {
+      "epoch": 3.9921414538310414,
+      "grad_norm": 0.06600197026967601,
+      "learning_rate": 9.503719892799644e-06,
+      "loss": 0.4084,
+      "step": 2032
+    },
+    {
+      "epoch": 3.994106090373281,
+      "grad_norm": 0.05808415294823111,
+      "learning_rate": 9.468239122177447e-06,
+      "loss": 0.3934,
+      "step": 2033
+    },
+    {
+      "epoch": 3.9960707269155207,
+      "grad_norm": 0.06947338492270719,
+      "learning_rate": 9.43281581353237e-06,
+      "loss": 0.4007,
+      "step": 2034
+    },
+    {
+      "epoch": 3.9980353634577606,
+      "grad_norm": 0.0656035126435929,
+      "learning_rate": 9.397450033532487e-06,
+      "loss": 0.3991,
+      "step": 2035
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.06280738802178301,
+      "learning_rate": 9.3621418487376e-06,
+      "loss": 0.3984,
+      "step": 2036
+    },
+    {
+      "epoch": 4.001964636542239,
+      "grad_norm": 0.18746317941877436,
+      "learning_rate": 9.326891325599083e-06,
+      "loss": 0.3784,
+      "step": 2037
+    },
+    {
+      "epoch": 4.00392927308448,
+      "grad_norm": 0.1300593890529019,
+      "learning_rate": 9.291698530459824e-06,
+      "loss": 0.368,
+      "step": 2038
+    },
+    {
+      "epoch": 4.005893909626719,
+      "grad_norm": 0.106458166517229,
+      "learning_rate": 9.25656352955404e-06,
+      "loss": 0.3702,
+      "step": 2039
+    },
+    {
+      "epoch": 4.007858546168959,
+      "grad_norm": 0.16475760076414164,
+      "learning_rate": 9.221486389007199e-06,
+      "loss": 0.3722,
+      "step": 2040
+    },
+    {
+      "epoch": 4.009823182711198,
+      "grad_norm": 0.15003540544978353,
+      "learning_rate": 9.186467174835854e-06,
+      "loss": 0.3715,
+      "step": 2041
+    },
+    {
+      "epoch": 4.011787819253438,
+      "grad_norm": 0.12821221942185956,
+      "learning_rate": 9.151505952947545e-06,
+      "loss": 0.3722,
+      "step": 2042
+    },
+    {
+      "epoch": 4.013752455795678,
+      "grad_norm": 0.14097065431309846,
+      "learning_rate": 9.116602789140692e-06,
+      "loss": 0.369,
+      "step": 2043
+    },
+    {
+      "epoch": 4.015717092337917,
+      "grad_norm": 0.12694691770151958,
+      "learning_rate": 9.081757749104389e-06,
+      "loss": 0.3677,
+      "step": 2044
+    },
+    {
+      "epoch": 4.017681728880157,
+      "grad_norm": 0.11738601018629359,
+      "learning_rate": 9.04697089841839e-06,
+      "loss": 0.3625,
+      "step": 2045
+    },
+    {
+      "epoch": 4.019646365422397,
+      "grad_norm": 0.12384496457960344,
+      "learning_rate": 9.012242302552937e-06,
+      "loss": 0.3669,
+      "step": 2046
+    },
+    {
+      "epoch": 4.0216110019646365,
+      "grad_norm": 0.11824303225794958,
+      "learning_rate": 8.97757202686858e-06,
+      "loss": 0.3682,
+      "step": 2047
+    },
+    {
+      "epoch": 4.023575638506876,
+      "grad_norm": 0.10913334248853608,
+      "learning_rate": 8.942960136616165e-06,
+      "loss": 0.3701,
+      "step": 2048
+    },
+    {
+      "epoch": 4.025540275049116,
+      "grad_norm": 0.11318396757035935,
+      "learning_rate": 8.908406696936636e-06,
+      "loss": 0.3627,
+      "step": 2049
+    },
+    {
+      "epoch": 4.027504911591356,
+      "grad_norm": 0.10156363278068824,
+      "learning_rate": 8.873911772860916e-06,
+      "loss": 0.3806,
+      "step": 2050
+    },
+    {
+      "epoch": 4.029469548133595,
+      "grad_norm": 0.100777910897909,
+      "learning_rate": 8.839475429309826e-06,
+      "loss": 0.3613,
+      "step": 2051
+    },
+    {
+      "epoch": 4.031434184675835,
+      "grad_norm": 0.10232356715353746,
+      "learning_rate": 8.805097731093922e-06,
+      "loss": 0.3659,
+      "step": 2052
+    },
+    {
+      "epoch": 4.033398821218075,
+      "grad_norm": 0.08949314928592908,
+      "learning_rate": 8.770778742913371e-06,
+      "loss": 0.3582,
+      "step": 2053
+    },
+    {
+      "epoch": 4.035363457760314,
+      "grad_norm": 0.08517244355994466,
+      "learning_rate": 8.736518529357866e-06,
+      "loss": 0.3643,
+      "step": 2054
+    },
+    {
+      "epoch": 4.037328094302554,
+      "grad_norm": 0.09009106105624268,
+      "learning_rate": 8.702317154906495e-06,
+      "loss": 0.3677,
+      "step": 2055
+    },
+    {
+      "epoch": 4.039292730844794,
+      "grad_norm": 0.08709632361882444,
+      "learning_rate": 8.668174683927568e-06,
+      "loss": 0.3595,
+      "step": 2056
+    },
+    {
+      "epoch": 4.041257367387034,
+      "grad_norm": 0.0814517612630799,
+      "learning_rate": 8.634091180678571e-06,
+      "loss": 0.369,
+      "step": 2057
+    },
+    {
+      "epoch": 4.043222003929273,
+      "grad_norm": 0.08491003830043559,
+      "learning_rate": 8.600066709306008e-06,
+      "loss": 0.3796,
+      "step": 2058
+    },
+    {
+      "epoch": 4.045186640471512,
+      "grad_norm": 0.08041381466329592,
+      "learning_rate": 8.56610133384526e-06,
+      "loss": 0.3568,
+      "step": 2059
+    },
+    {
+      "epoch": 4.047151277013753,
+      "grad_norm": 0.0819015635820511,
+      "learning_rate": 8.532195118220516e-06,
+      "loss": 0.3647,
+      "step": 2060
+    },
+    {
+      "epoch": 4.049115913555992,
+      "grad_norm": 0.0747416485445965,
+      "learning_rate": 8.498348126244607e-06,
+      "loss": 0.3608,
+      "step": 2061
+    },
+    {
+      "epoch": 4.051080550098232,
+      "grad_norm": 0.07396571202166831,
+      "learning_rate": 8.464560421618912e-06,
+      "loss": 0.3604,
+      "step": 2062
+    },
+    {
+      "epoch": 4.053045186640471,
+      "grad_norm": 0.08192165870436205,
+      "learning_rate": 8.430832067933203e-06,
+      "loss": 0.3731,
+      "step": 2063
+    },
+    {
+      "epoch": 4.055009823182711,
+      "grad_norm": 0.07634660315504181,
+      "learning_rate": 8.397163128665582e-06,
+      "loss": 0.3637,
+      "step": 2064
+    },
+    {
+      "epoch": 4.056974459724951,
+      "grad_norm": 0.07015029980705169,
+      "learning_rate": 8.363553667182329e-06,
+      "loss": 0.3756,
+      "step": 2065
+    },
+    {
+      "epoch": 4.05893909626719,
+      "grad_norm": 0.06944684441916105,
+      "learning_rate": 8.330003746737758e-06,
+      "loss": 0.3668,
+      "step": 2066
+    },
+    {
+      "epoch": 4.060903732809431,
+      "grad_norm": 0.0754986273640299,
+      "learning_rate": 8.296513430474146e-06,
+      "loss": 0.3686,
+      "step": 2067
+    },
+    {
+      "epoch": 4.06286836935167,
+      "grad_norm": 0.07138665177977778,
+      "learning_rate": 8.263082781421592e-06,
+      "loss": 0.3644,
+      "step": 2068
+    },
+    {
+      "epoch": 4.0648330058939095,
+      "grad_norm": 0.07447607560836032,
+      "learning_rate": 8.229711862497889e-06,
+      "loss": 0.361,
+      "step": 2069
+    },
+    {
+      "epoch": 4.066797642436149,
+      "grad_norm": 0.07039449797144805,
+      "learning_rate": 8.196400736508429e-06,
+      "loss": 0.361,
+      "step": 2070
+    },
+    {
+      "epoch": 4.068762278978389,
+      "grad_norm": 0.07325399917329171,
+      "learning_rate": 8.163149466146065e-06,
+      "loss": 0.365,
+      "step": 2071
+    },
+    {
+      "epoch": 4.070726915520629,
+      "grad_norm": 0.06600862788859567,
+      "learning_rate": 8.129958113990999e-06,
+      "loss": 0.3704,
+      "step": 2072
+    },
+    {
+      "epoch": 4.072691552062868,
+      "grad_norm": 0.06625432726151605,
+      "learning_rate": 8.096826742510653e-06,
+      "loss": 0.3526,
+      "step": 2073
+    },
+    {
+      "epoch": 4.0746561886051085,
+      "grad_norm": 0.06914221441229995,
+      "learning_rate": 8.063755414059575e-06,
+      "loss": 0.362,
+      "step": 2074
+    },
+    {
+      "epoch": 4.076620825147348,
+      "grad_norm": 0.06699143028205767,
+      "learning_rate": 8.030744190879325e-06,
+      "loss": 0.3666,
+      "step": 2075
+    },
+    {
+      "epoch": 4.078585461689587,
+      "grad_norm": 0.0626202109534912,
+      "learning_rate": 7.997793135098302e-06,
+      "loss": 0.3628,
+      "step": 2076
+    },
+    {
+      "epoch": 4.080550098231827,
+      "grad_norm": 0.06520523668873579,
+      "learning_rate": 7.964902308731703e-06,
+      "loss": 0.361,
+      "step": 2077
+    },
+    {
+      "epoch": 4.082514734774067,
+      "grad_norm": 0.06669396837913025,
+      "learning_rate": 7.932071773681365e-06,
+      "loss": 0.361,
+      "step": 2078
+    },
+    {
+      "epoch": 4.084479371316307,
+      "grad_norm": 0.06831749596995798,
+      "learning_rate": 7.899301591735638e-06,
+      "loss": 0.3788,
+      "step": 2079
+    },
+    {
+      "epoch": 4.086444007858546,
+      "grad_norm": 0.0634021157756623,
+      "learning_rate": 7.866591824569299e-06,
+      "loss": 0.3577,
+      "step": 2080
+    },
+    {
+      "epoch": 4.088408644400785,
+      "grad_norm": 0.06446322489425511,
+      "learning_rate": 7.833942533743437e-06,
+      "loss": 0.3608,
+      "step": 2081
+    },
+    {
+      "epoch": 4.090373280943026,
+      "grad_norm": 0.06233449804631122,
+      "learning_rate": 7.801353780705274e-06,
+      "loss": 0.3706,
+      "step": 2082
+    },
+    {
+      "epoch": 4.092337917485265,
+      "grad_norm": 0.0673675311175845,
+      "learning_rate": 7.76882562678814e-06,
+      "loss": 0.3601,
+      "step": 2083
+    },
+    {
+      "epoch": 4.094302554027505,
+      "grad_norm": 0.06520778285482241,
+      "learning_rate": 7.736358133211306e-06,
+      "loss": 0.3623,
+      "step": 2084
+    },
+    {
+      "epoch": 4.096267190569745,
+      "grad_norm": 0.06271883973761705,
+      "learning_rate": 7.703951361079881e-06,
+      "loss": 0.3606,
+      "step": 2085
+    },
+    {
+      "epoch": 4.098231827111984,
+      "grad_norm": 0.0647871266179983,
+      "learning_rate": 7.671605371384667e-06,
+      "loss": 0.3703,
+      "step": 2086
+    },
+    {
+      "epoch": 4.100196463654224,
+      "grad_norm": 0.0642579229491418,
+      "learning_rate": 7.639320225002106e-06,
+      "loss": 0.364,
+      "step": 2087
+    },
+    {
+      "epoch": 4.102161100196463,
+      "grad_norm": 0.0638185242587459,
+      "learning_rate": 7.607095982694108e-06,
+      "loss": 0.3582,
+      "step": 2088
+    },
+    {
+      "epoch": 4.104125736738704,
+      "grad_norm": 0.06367686858728912,
+      "learning_rate": 7.574932705107976e-06,
+      "loss": 0.356,
+      "step": 2089
+    },
+    {
+      "epoch": 4.106090373280943,
+      "grad_norm": 0.059697451641839115,
+      "learning_rate": 7.542830452776257e-06,
+      "loss": 0.3681,
+      "step": 2090
+    },
+    {
+      "epoch": 4.1080550098231825,
+      "grad_norm": 0.06454526790473719,
+      "learning_rate": 7.510789286116673e-06,
+      "loss": 0.3556,
+      "step": 2091
+    },
+    {
+      "epoch": 4.110019646365423,
+      "grad_norm": 0.0650580354943275,
+      "learning_rate": 7.478809265431932e-06,
+      "loss": 0.3688,
+      "step": 2092
+    },
+    {
+      "epoch": 4.111984282907662,
+      "grad_norm": 0.06469191276737123,
+      "learning_rate": 7.4468904509097164e-06,
+      "loss": 0.3653,
+      "step": 2093
+    },
+    {
+      "epoch": 4.113948919449902,
+      "grad_norm": 0.06405960783722763,
+      "learning_rate": 7.4150329026224924e-06,
+      "loss": 0.3662,
+      "step": 2094
+    },
+    {
+      "epoch": 4.115913555992141,
+      "grad_norm": 0.05901958565166774,
+      "learning_rate": 7.383236680527402e-06,
+      "loss": 0.3607,
+      "step": 2095
+    },
+    {
+      "epoch": 4.1178781925343815,
+      "grad_norm": 0.061734337658915554,
+      "learning_rate": 7.351501844466202e-06,
+      "loss": 0.3725,
+      "step": 2096
+    },
+    {
+      "epoch": 4.119842829076621,
+      "grad_norm": 0.05951994937027903,
+      "learning_rate": 7.319828454165105e-06,
+      "loss": 0.358,
+      "step": 2097
+    },
+    {
+      "epoch": 4.12180746561886,
+      "grad_norm": 0.06275784072884763,
+      "learning_rate": 7.288216569234681e-06,
+      "loss": 0.362,
+      "step": 2098
+    },
+    {
+      "epoch": 4.1237721021611,
+      "grad_norm": 0.06642883313505578,
+      "learning_rate": 7.256666249169737e-06,
+      "loss": 0.3634,
+      "step": 2099
+    },
+    {
+      "epoch": 4.12573673870334,
+      "grad_norm": 0.06531018185997445,
+      "learning_rate": 7.225177553349221e-06,
+      "loss": 0.3681,
+      "step": 2100
+    },
+    {
+      "epoch": 4.12770137524558,
+      "grad_norm": 0.06027320727812834,
+      "learning_rate": 7.193750541036113e-06,
+      "loss": 0.3669,
+      "step": 2101
+    },
+    {
+      "epoch": 4.129666011787819,
+      "grad_norm": 0.05727514553121101,
+      "learning_rate": 7.162385271377266e-06,
+      "loss": 0.3559,
+      "step": 2102
+    },
+    {
+      "epoch": 4.131630648330059,
+      "grad_norm": 0.05753940248840264,
+      "learning_rate": 7.131081803403362e-06,
+      "loss": 0.3553,
+      "step": 2103
+    },
+    {
+      "epoch": 4.133595284872299,
+      "grad_norm": 0.06171066495301514,
+      "learning_rate": 7.099840196028771e-06,
+      "loss": 0.3576,
+      "step": 2104
+    },
+    {
+      "epoch": 4.135559921414538,
+      "grad_norm": 0.06096894822262158,
+      "learning_rate": 7.0686605080514124e-06,
+      "loss": 0.3717,
+      "step": 2105
+    },
+    {
+      "epoch": 4.137524557956778,
+      "grad_norm": 0.06263141047063535,
+      "learning_rate": 7.037542798152697e-06,
+      "loss": 0.3592,
+      "step": 2106
+    },
+    {
+      "epoch": 4.139489194499018,
+      "grad_norm": 0.06108261725226068,
+      "learning_rate": 7.006487124897376e-06,
+      "loss": 0.36,
+      "step": 2107
+    },
+    {
+      "epoch": 4.141453831041257,
+      "grad_norm": 0.06228074077161129,
+      "learning_rate": 6.975493546733458e-06,
+      "loss": 0.3533,
+      "step": 2108
+    },
+    {
+      "epoch": 4.143418467583497,
+      "grad_norm": 0.06458766627190418,
+      "learning_rate": 6.944562121992078e-06,
+      "loss": 0.3856,
+      "step": 2109
+    },
+    {
+      "epoch": 4.145383104125736,
+      "grad_norm": 0.061000135797814055,
+      "learning_rate": 6.913692908887401e-06,
+      "loss": 0.3642,
+      "step": 2110
+    },
+    {
+      "epoch": 4.147347740667977,
+      "grad_norm": 0.06131949213933078,
+      "learning_rate": 6.882885965516481e-06,
+      "loss": 0.3714,
+      "step": 2111
+    },
+    {
+      "epoch": 4.149312377210216,
+      "grad_norm": 0.058209858911145884,
+      "learning_rate": 6.852141349859218e-06,
+      "loss": 0.3621,
+      "step": 2112
+    },
+    {
+      "epoch": 4.1512770137524555,
+      "grad_norm": 0.05913262692026409,
+      "learning_rate": 6.821459119778185e-06,
+      "loss": 0.3618,
+      "step": 2113
+    },
+    {
+      "epoch": 4.153241650294696,
+      "grad_norm": 0.06091378277333175,
+      "learning_rate": 6.790839333018557e-06,
+      "loss": 0.3634,
+      "step": 2114
+    },
+    {
+      "epoch": 4.155206286836935,
+      "grad_norm": 0.05961983811619346,
+      "learning_rate": 6.76028204720796e-06,
+      "loss": 0.3604,
+      "step": 2115
+    },
+    {
+      "epoch": 4.157170923379175,
+      "grad_norm": 0.06085988183277799,
+      "learning_rate": 6.729787319856425e-06,
+      "loss": 0.3652,
+      "step": 2116
+    },
+    {
+      "epoch": 4.159135559921414,
+      "grad_norm": 0.05908274398140844,
+      "learning_rate": 6.699355208356224e-06,
+      "loss": 0.3565,
+      "step": 2117
+    },
+    {
+      "epoch": 4.1611001964636545,
+      "grad_norm": 0.06271280485170534,
+      "learning_rate": 6.668985769981789e-06,
+      "loss": 0.3657,
+      "step": 2118
+    },
+    {
+      "epoch": 4.163064833005894,
+      "grad_norm": 0.05719882847807065,
+      "learning_rate": 6.6386790618896106e-06,
+      "loss": 0.3601,
+      "step": 2119
+    },
+    {
+      "epoch": 4.165029469548133,
+      "grad_norm": 0.060057002859305264,
+      "learning_rate": 6.608435141118104e-06,
+      "loss": 0.3711,
+      "step": 2120
+    },
+    {
+      "epoch": 4.166994106090374,
+      "grad_norm": 0.06603410138900485,
+      "learning_rate": 6.57825406458751e-06,
+      "loss": 0.3695,
+      "step": 2121
+    },
+    {
+      "epoch": 4.168958742632613,
+      "grad_norm": 0.06382802147436592,
+      "learning_rate": 6.548135889099807e-06,
+      "loss": 0.3524,
+      "step": 2122
+    },
+    {
+      "epoch": 4.170923379174853,
+      "grad_norm": 0.06275261090949763,
+      "learning_rate": 6.5180806713386025e-06,
+      "loss": 0.3641,
+      "step": 2123
+    },
+    {
+      "epoch": 4.172888015717092,
+      "grad_norm": 0.06733131778464349,
+      "learning_rate": 6.4880884678689825e-06,
+      "loss": 0.372,
+      "step": 2124
+    },
+    {
+      "epoch": 4.174852652259332,
+      "grad_norm": 0.06475786454993836,
+      "learning_rate": 6.458159335137461e-06,
+      "loss": 0.3665,
+      "step": 2125
+    },
+    {
+      "epoch": 4.176817288801572,
+      "grad_norm": 0.06425967096127662,
+      "learning_rate": 6.428293329471849e-06,
+      "loss": 0.3653,
+      "step": 2126
+    },
+    {
+      "epoch": 4.178781925343811,
+      "grad_norm": 0.06098687271214388,
+      "learning_rate": 6.398490507081141e-06,
+      "loss": 0.365,
+      "step": 2127
+    },
+    {
+      "epoch": 4.1807465618860515,
+      "grad_norm": 0.06281720124894108,
+      "learning_rate": 6.368750924055427e-06,
+      "loss": 0.3679,
+      "step": 2128
+    },
+    {
+      "epoch": 4.182711198428291,
+      "grad_norm": 0.0653241867218168,
+      "learning_rate": 6.339074636365778e-06,
+      "loss": 0.3733,
+      "step": 2129
+    },
+    {
+      "epoch": 4.18467583497053,
+      "grad_norm": 0.06436926846487889,
+      "learning_rate": 6.309461699864137e-06,
+      "loss": 0.3587,
+      "step": 2130
+    },
+    {
+      "epoch": 4.18664047151277,
+      "grad_norm": 0.06031089498227457,
+      "learning_rate": 6.279912170283204e-06,
+      "loss": 0.3638,
+      "step": 2131
+    },
+    {
+      "epoch": 4.18860510805501,
+      "grad_norm": 0.06678973427486706,
+      "learning_rate": 6.250426103236371e-06,
+      "loss": 0.3631,
+      "step": 2132
+    },
+    {
+      "epoch": 4.19056974459725,
+      "grad_norm": 0.06258669934823113,
+      "learning_rate": 6.221003554217579e-06,
+      "loss": 0.3595,
+      "step": 2133
+    },
+    {
+      "epoch": 4.192534381139489,
+      "grad_norm": 0.06048408843112766,
+      "learning_rate": 6.1916445786012145e-06,
+      "loss": 0.3644,
+      "step": 2134
+    },
+    {
+      "epoch": 4.1944990176817285,
+      "grad_norm": 0.06060406116328243,
+      "learning_rate": 6.162349231642033e-06,
+      "loss": 0.3659,
+      "step": 2135
+    },
+    {
+      "epoch": 4.196463654223969,
+      "grad_norm": 0.06572786186146773,
+      "learning_rate": 6.133117568475029e-06,
+      "loss": 0.3646,
+      "step": 2136
+    },
+    {
+      "epoch": 4.198428290766208,
+      "grad_norm": 0.0635731988750601,
+      "learning_rate": 6.1039496441153415e-06,
+      "loss": 0.3664,
+      "step": 2137
+    },
+    {
+      "epoch": 4.200392927308448,
+      "grad_norm": 0.0612122645290111,
+      "learning_rate": 6.074845513458165e-06,
+      "loss": 0.357,
+      "step": 2138
+    },
+    {
+      "epoch": 4.202357563850688,
+      "grad_norm": 0.0585950847930313,
+      "learning_rate": 6.04580523127861e-06,
+      "loss": 0.3612,
+      "step": 2139
+    },
+    {
+      "epoch": 4.2043222003929275,
+      "grad_norm": 0.06310811246342059,
+      "learning_rate": 6.016828852231649e-06,
+      "loss": 0.3685,
+      "step": 2140
+    },
+    {
+      "epoch": 4.206286836935167,
+      "grad_norm": 0.061522519959194416,
+      "learning_rate": 5.987916430851944e-06,
+      "loss": 0.3677,
+      "step": 2141
+    },
+    {
+      "epoch": 4.208251473477406,
+      "grad_norm": 0.05909968787646219,
+      "learning_rate": 5.959068021553824e-06,
+      "loss": 0.3609,
+      "step": 2142
+    },
+    {
+      "epoch": 4.210216110019647,
+      "grad_norm": 0.0606788567453581,
+      "learning_rate": 5.930283678631137e-06,
+      "loss": 0.3596,
+      "step": 2143
+    },
+    {
+      "epoch": 4.212180746561886,
+      "grad_norm": 0.05948395290203327,
+      "learning_rate": 5.901563456257142e-06,
+      "loss": 0.36,
+      "step": 2144
+    },
+    {
+      "epoch": 4.214145383104126,
+      "grad_norm": 0.06185687502259067,
+      "learning_rate": 5.87290740848443e-06,
+      "loss": 0.3633,
+      "step": 2145
+    },
+    {
+      "epoch": 4.216110019646365,
+      "grad_norm": 0.062387295474085745,
+      "learning_rate": 5.8443155892448135e-06,
+      "loss": 0.3569,
+      "step": 2146
+    },
+    {
+      "epoch": 4.218074656188605,
+      "grad_norm": 0.06013815712228769,
+      "learning_rate": 5.815788052349223e-06,
+      "loss": 0.3681,
+      "step": 2147
+    },
+    {
+      "epoch": 4.220039292730845,
+      "grad_norm": 0.059596696515686894,
+      "learning_rate": 5.787324851487612e-06,
+      "loss": 0.3676,
+      "step": 2148
+    },
+    {
+      "epoch": 4.222003929273084,
+      "grad_norm": 0.0635597892797539,
+      "learning_rate": 5.758926040228847e-06,
+      "loss": 0.3658,
+      "step": 2149
+    },
+    {
+      "epoch": 4.2239685658153245,
+      "grad_norm": 0.06042976482012168,
+      "learning_rate": 5.7305916720206e-06,
+      "loss": 0.3565,
+      "step": 2150
+    },
+    {
+      "epoch": 4.225933202357564,
+      "grad_norm": 0.05683821437946257,
+      "learning_rate": 5.702321800189272e-06,
+      "loss": 0.3715,
+      "step": 2151
+    },
+    {
+      "epoch": 4.227897838899803,
+      "grad_norm": 0.061887954596087306,
+      "learning_rate": 5.674116477939877e-06,
+      "loss": 0.3587,
+      "step": 2152
+    },
+    {
+      "epoch": 4.229862475442043,
+      "grad_norm": 0.05926144742349446,
+      "learning_rate": 5.645975758355957e-06,
+      "loss": 0.3539,
+      "step": 2153
+    },
+    {
+      "epoch": 4.231827111984283,
+      "grad_norm": 0.0570806244440597,
+      "learning_rate": 5.617899694399427e-06,
+      "loss": 0.3702,
+      "step": 2154
+    },
+    {
+      "epoch": 4.233791748526523,
+      "grad_norm": 0.06082171458702773,
+      "learning_rate": 5.589888338910565e-06,
+      "loss": 0.366,
+      "step": 2155
+    },
+    {
+      "epoch": 4.235756385068762,
+      "grad_norm": 0.061030459684122275,
+      "learning_rate": 5.5619417446078375e-06,
+      "loss": 0.3569,
+      "step": 2156
+    },
+    {
+      "epoch": 4.237721021611002,
+      "grad_norm": 0.06030161173759712,
+      "learning_rate": 5.534059964087841e-06,
+      "loss": 0.3699,
+      "step": 2157
+    },
+    {
+      "epoch": 4.239685658153242,
+      "grad_norm": 0.05665764963200982,
+      "learning_rate": 5.5062430498251885e-06,
+      "loss": 0.3677,
+      "step": 2158
+    },
+    {
+      "epoch": 4.241650294695481,
+      "grad_norm": 0.06308025811948946,
+      "learning_rate": 5.47849105417241e-06,
+      "loss": 0.3609,
+      "step": 2159
+    },
+    {
+      "epoch": 4.243614931237721,
+      "grad_norm": 0.05821993198399712,
+      "learning_rate": 5.450804029359842e-06,
+      "loss": 0.3592,
+      "step": 2160
+    },
+    {
+      "epoch": 4.245579567779961,
+      "grad_norm": 0.05730662077546451,
+      "learning_rate": 5.423182027495562e-06,
+      "loss": 0.3582,
+      "step": 2161
+    },
+    {
+      "epoch": 4.2475442043222005,
+      "grad_norm": 0.0583192783640768,
+      "learning_rate": 5.395625100565278e-06,
+      "loss": 0.3666,
+      "step": 2162
+    },
+    {
+      "epoch": 4.24950884086444,
+      "grad_norm": 0.06177504012422726,
+      "learning_rate": 5.36813330043219e-06,
+      "loss": 0.3621,
+      "step": 2163
+    },
+    {
+      "epoch": 4.251473477406679,
+      "grad_norm": 0.06188066605998403,
+      "learning_rate": 5.340706678836966e-06,
+      "loss": 0.3638,
+      "step": 2164
+    },
+    {
+      "epoch": 4.25343811394892,
+      "grad_norm": 0.05967555946147971,
+      "learning_rate": 5.313345287397585e-06,
+      "loss": 0.3696,
+      "step": 2165
+    },
+    {
+      "epoch": 4.255402750491159,
+      "grad_norm": 0.05761023336220951,
+      "learning_rate": 5.286049177609256e-06,
+      "loss": 0.3593,
+      "step": 2166
+    },
+    {
+      "epoch": 4.257367387033399,
+      "grad_norm": 0.06489660580610669,
+      "learning_rate": 5.2588184008443455e-06,
+      "loss": 0.3641,
+      "step": 2167
+    },
+    {
+      "epoch": 4.259332023575639,
+      "grad_norm": 0.06343210539131612,
+      "learning_rate": 5.231653008352245e-06,
+      "loss": 0.368,
+      "step": 2168
+    },
+    {
+      "epoch": 4.261296660117878,
+      "grad_norm": 0.056319332857859346,
+      "learning_rate": 5.204553051259296e-06,
+      "loss": 0.3677,
+      "step": 2169
+    },
+    {
+      "epoch": 4.263261296660118,
+      "grad_norm": 0.06127796958082013,
+      "learning_rate": 5.177518580568674e-06,
+      "loss": 0.366,
+      "step": 2170
+    },
+    {
+      "epoch": 4.265225933202357,
+      "grad_norm": 0.05959516072840952,
+      "learning_rate": 5.15054964716033e-06,
+      "loss": 0.3627,
+      "step": 2171
+    },
+    {
+      "epoch": 4.2671905697445975,
+      "grad_norm": 0.061566191025567624,
+      "learning_rate": 5.123646301790861e-06,
+      "loss": 0.3695,
+      "step": 2172
+    },
+    {
+      "epoch": 4.269155206286837,
+      "grad_norm": 0.06287956122004656,
+      "learning_rate": 5.096808595093406e-06,
+      "loss": 0.3596,
+      "step": 2173
+    },
+    {
+      "epoch": 4.271119842829076,
+      "grad_norm": 0.06076831337842104,
+      "learning_rate": 5.070036577577599e-06,
+      "loss": 0.357,
+      "step": 2174
+    },
+    {
+      "epoch": 4.273084479371317,
+      "grad_norm": 0.06550334859024433,
+      "learning_rate": 5.0433302996294365e-06,
+      "loss": 0.3744,
+      "step": 2175
+    },
+    {
+      "epoch": 4.275049115913556,
+      "grad_norm": 0.0627992278102481,
+      "learning_rate": 5.016689811511173e-06,
+      "loss": 0.3624,
+      "step": 2176
+    },
+    {
+      "epoch": 4.277013752455796,
+      "grad_norm": 0.06181548142811297,
+      "learning_rate": 4.990115163361271e-06,
+      "loss": 0.3629,
+      "step": 2177
+    },
+    {
+      "epoch": 4.278978388998035,
+      "grad_norm": 0.060159791257008016,
+      "learning_rate": 4.9636064051942616e-06,
+      "loss": 0.364,
+      "step": 2178
+    },
+    {
+      "epoch": 4.280943025540275,
+      "grad_norm": 0.05946362220308484,
+      "learning_rate": 4.937163586900662e-06,
+      "loss": 0.3594,
+      "step": 2179
+    },
+    {
+      "epoch": 4.282907662082515,
+      "grad_norm": 0.06330356435562126,
+      "learning_rate": 4.9107867582469085e-06,
+      "loss": 0.355,
+      "step": 2180
+    },
+    {
+      "epoch": 4.284872298624754,
+      "grad_norm": 0.0612199807022724,
+      "learning_rate": 4.884475968875229e-06,
+      "loss": 0.3745,
+      "step": 2181
+    },
+    {
+      "epoch": 4.286836935166994,
+      "grad_norm": 0.05717538104096797,
+      "learning_rate": 4.858231268303572e-06,
+      "loss": 0.3531,
+      "step": 2182
+    },
+    {
+      "epoch": 4.288801571709234,
+      "grad_norm": 0.05982898592429428,
+      "learning_rate": 4.832052705925487e-06,
+      "loss": 0.3726,
+      "step": 2183
+    },
+    {
+      "epoch": 4.2907662082514735,
+      "grad_norm": 0.06087372941112121,
+      "learning_rate": 4.8059403310100635e-06,
+      "loss": 0.3685,
+      "step": 2184
+    },
+    {
+      "epoch": 4.292730844793713,
+      "grad_norm": 0.05705204764926255,
+      "learning_rate": 4.779894192701826e-06,
+      "loss": 0.3705,
+      "step": 2185
+    },
+    {
+      "epoch": 4.294695481335953,
+      "grad_norm": 0.061413851206836415,
+      "learning_rate": 4.753914340020625e-06,
+      "loss": 0.3649,
+      "step": 2186
+    },
+    {
+      "epoch": 4.296660117878193,
+      "grad_norm": 0.0611475442241838,
+      "learning_rate": 4.728000821861574e-06,
+      "loss": 0.3588,
+      "step": 2187
+    },
+    {
+      "epoch": 4.298624754420432,
+      "grad_norm": 0.059993847786793515,
+      "learning_rate": 4.702153686994941e-06,
+      "loss": 0.3562,
+      "step": 2188
+    },
+    {
+      "epoch": 4.300589390962672,
+      "grad_norm": 0.05830421153652512,
+      "learning_rate": 4.676372984066033e-06,
+      "loss": 0.3604,
+      "step": 2189
+    },
+    {
+      "epoch": 4.302554027504912,
+      "grad_norm": 0.06355563554404006,
+      "learning_rate": 4.650658761595162e-06,
+      "loss": 0.3617,
+      "step": 2190
+    },
+    {
+      "epoch": 4.304518664047151,
+      "grad_norm": 0.058125379440603205,
+      "learning_rate": 4.625011067977516e-06,
+      "loss": 0.3719,
+      "step": 2191
+    },
+    {
+      "epoch": 4.306483300589391,
+      "grad_norm": 0.05938390485178116,
+      "learning_rate": 4.599429951483041e-06,
+      "loss": 0.3585,
+      "step": 2192
+    },
+    {
+      "epoch": 4.30844793713163,
+      "grad_norm": 0.06004732410736976,
+      "learning_rate": 4.573915460256424e-06,
+      "loss": 0.3694,
+      "step": 2193
+    },
+    {
+      "epoch": 4.3104125736738705,
+      "grad_norm": 0.06343235126535983,
+      "learning_rate": 4.548467642316943e-06,
+      "loss": 0.3702,
+      "step": 2194
+    },
+    {
+      "epoch": 4.31237721021611,
+      "grad_norm": 0.057639568886207686,
+      "learning_rate": 4.523086545558384e-06,
+      "loss": 0.3702,
+      "step": 2195
+    },
+    {
+      "epoch": 4.314341846758349,
+      "grad_norm": 0.058794421843137004,
+      "learning_rate": 4.497772217748981e-06,
+      "loss": 0.3621,
+      "step": 2196
+    },
+    {
+      "epoch": 4.31630648330059,
+      "grad_norm": 0.05816901678030251,
+      "learning_rate": 4.472524706531292e-06,
+      "loss": 0.3672,
+      "step": 2197
+    },
+    {
+      "epoch": 4.318271119842829,
+      "grad_norm": 0.058671873462326926,
+      "learning_rate": 4.4473440594221405e-06,
+      "loss": 0.3665,
+      "step": 2198
+    },
+    {
+      "epoch": 4.320235756385069,
+      "grad_norm": 0.06506802651491145,
+      "learning_rate": 4.422230323812473e-06,
+      "loss": 0.3633,
+      "step": 2199
+    },
+    {
+      "epoch": 4.322200392927308,
+      "grad_norm": 0.058778678430237825,
+      "learning_rate": 4.397183546967343e-06,
+      "loss": 0.3619,
+      "step": 2200
+    },
+    {
+      "epoch": 4.324165029469548,
+      "grad_norm": 0.058224887093293,
+      "learning_rate": 4.372203776025781e-06,
+      "loss": 0.3635,
+      "step": 2201
+    },
+    {
+      "epoch": 4.326129666011788,
+      "grad_norm": 0.059655379825925295,
+      "learning_rate": 4.347291058000678e-06,
+      "loss": 0.3629,
+      "step": 2202
+    },
+    {
+      "epoch": 4.328094302554027,
+      "grad_norm": 0.05939327876286086,
+      "learning_rate": 4.322445439778764e-06,
+      "loss": 0.3614,
+      "step": 2203
+    },
+    {
+      "epoch": 4.330058939096268,
+      "grad_norm": 0.05856112533494615,
+      "learning_rate": 4.297666968120471e-06,
+      "loss": 0.365,
+      "step": 2204
+    },
+    {
+      "epoch": 4.332023575638507,
+      "grad_norm": 0.05923117426695095,
+      "learning_rate": 4.272955689659859e-06,
+      "loss": 0.3676,
+      "step": 2205
+    },
+    {
+      "epoch": 4.3339882121807465,
+      "grad_norm": 0.059755474002157985,
+      "learning_rate": 4.2483116509045306e-06,
+      "loss": 0.3599,
+      "step": 2206
+    },
+    {
+      "epoch": 4.335952848722986,
+      "grad_norm": 0.059671302736430995,
+      "learning_rate": 4.223734898235536e-06,
+      "loss": 0.3599,
+      "step": 2207
+    },
+    {
+      "epoch": 4.337917485265226,
+      "grad_norm": 0.06405421597175118,
+      "learning_rate": 4.199225477907285e-06,
+      "loss": 0.3613,
+      "step": 2208
+    },
+    {
+      "epoch": 4.339882121807466,
+      "grad_norm": 0.05585963649037604,
+      "learning_rate": 4.174783436047483e-06,
+      "loss": 0.3692,
+      "step": 2209
+    },
+    {
+      "epoch": 4.341846758349705,
+      "grad_norm": 0.05932390703029273,
+      "learning_rate": 4.150408818657008e-06,
+      "loss": 0.3632,
+      "step": 2210
+    },
+    {
+      "epoch": 4.3438113948919455,
+      "grad_norm": 0.06136525867463179,
+      "learning_rate": 4.126101671609859e-06,
+      "loss": 0.3597,
+      "step": 2211
+    },
+    {
+      "epoch": 4.345776031434185,
+      "grad_norm": 0.06043407534671945,
+      "learning_rate": 4.101862040653033e-06,
+      "loss": 0.3619,
+      "step": 2212
+    },
+    {
+      "epoch": 4.347740667976424,
+      "grad_norm": 0.05781677569298015,
+      "learning_rate": 4.077689971406478e-06,
+      "loss": 0.3662,
+      "step": 2213
+    },
+    {
+      "epoch": 4.349705304518664,
+      "grad_norm": 0.05523730673126981,
+      "learning_rate": 4.05358550936298e-06,
+      "loss": 0.358,
+      "step": 2214
+    },
+    {
+      "epoch": 4.351669941060904,
+      "grad_norm": 0.055028149587792036,
+      "learning_rate": 4.0295486998880885e-06,
+      "loss": 0.372,
+      "step": 2215
+    },
+    {
+      "epoch": 4.3536345776031435,
+      "grad_norm": 0.0583467154824755,
+      "learning_rate": 4.005579588220023e-06,
+      "loss": 0.3663,
+      "step": 2216
+    },
+    {
+      "epoch": 4.355599214145383,
+      "grad_norm": 0.05604871825331791,
+      "learning_rate": 3.981678219469602e-06,
+      "loss": 0.3683,
+      "step": 2217
+    },
+    {
+      "epoch": 4.357563850687622,
+      "grad_norm": 0.056519431963238845,
+      "learning_rate": 3.957844638620132e-06,
+      "loss": 0.3595,
+      "step": 2218
+    },
+    {
+      "epoch": 4.359528487229863,
+      "grad_norm": 0.057112468942990144,
+      "learning_rate": 3.934078890527357e-06,
+      "loss": 0.3673,
+      "step": 2219
+    },
+    {
+      "epoch": 4.361493123772102,
+      "grad_norm": 0.06159105734526072,
+      "learning_rate": 3.9103810199193625e-06,
+      "loss": 0.3609,
+      "step": 2220
+    },
+    {
+      "epoch": 4.363457760314342,
+      "grad_norm": 0.059549363153168854,
+      "learning_rate": 3.886751071396457e-06,
+      "loss": 0.3637,
+      "step": 2221
+    },
+    {
+      "epoch": 4.365422396856582,
+      "grad_norm": 0.058449024914002064,
+      "learning_rate": 3.863189089431143e-06,
+      "loss": 0.3642,
+      "step": 2222
+    },
+    {
+      "epoch": 4.367387033398821,
+      "grad_norm": 0.05558480264663322,
+      "learning_rate": 3.839695118367996e-06,
+      "loss": 0.3635,
+      "step": 2223
+    },
+    {
+      "epoch": 4.369351669941061,
+      "grad_norm": 0.05579236309327851,
+      "learning_rate": 3.8162692024235996e-06,
+      "loss": 0.3628,
+      "step": 2224
+    },
+    {
+      "epoch": 4.3713163064833,
+      "grad_norm": 0.05572008694792393,
+      "learning_rate": 3.7929113856864485e-06,
+      "loss": 0.357,
+      "step": 2225
+    },
+    {
+      "epoch": 4.373280943025541,
+      "grad_norm": 0.05591081483940887,
+      "learning_rate": 3.7696217121168777e-06,
+      "loss": 0.3558,
+      "step": 2226
+    },
+    {
+      "epoch": 4.37524557956778,
+      "grad_norm": 0.054679742980199904,
+      "learning_rate": 3.7464002255469712e-06,
+      "loss": 0.3689,
+      "step": 2227
+    },
+    {
+      "epoch": 4.3772102161100195,
+      "grad_norm": 0.05776966084315899,
+      "learning_rate": 3.7232469696804718e-06,
+      "loss": 0.369,
+      "step": 2228
+    },
+    {
+      "epoch": 4.379174852652259,
+      "grad_norm": 0.055505473538715044,
+      "learning_rate": 3.7001619880927276e-06,
+      "loss": 0.3644,
+      "step": 2229
+    },
+    {
+      "epoch": 4.381139489194499,
+      "grad_norm": 0.0560157489930515,
+      "learning_rate": 3.6771453242305886e-06,
+      "loss": 0.3617,
+      "step": 2230
+    },
+    {
+      "epoch": 4.383104125736739,
+      "grad_norm": 0.060670206476946195,
+      "learning_rate": 3.6541970214123066e-06,
+      "loss": 0.3707,
+      "step": 2231
+    },
+    {
+      "epoch": 4.385068762278978,
+      "grad_norm": 0.05679554270683078,
+      "learning_rate": 3.6313171228275066e-06,
+      "loss": 0.3654,
+      "step": 2232
+    },
+    {
+      "epoch": 4.3870333988212185,
+      "grad_norm": 0.056555725246185275,
+      "learning_rate": 3.6085056715370504e-06,
+      "loss": 0.3624,
+      "step": 2233
+    },
+    {
+      "epoch": 4.388998035363458,
+      "grad_norm": 0.0564208142270732,
+      "learning_rate": 3.585762710472995e-06,
+      "loss": 0.3632,
+      "step": 2234
+    },
+    {
+      "epoch": 4.390962671905697,
+      "grad_norm": 0.056124051588307675,
+      "learning_rate": 3.5630882824384893e-06,
+      "loss": 0.3757,
+      "step": 2235
+    },
+    {
+      "epoch": 4.392927308447937,
+      "grad_norm": 0.0545617604912797,
+      "learning_rate": 3.5404824301076946e-06,
+      "loss": 0.3616,
+      "step": 2236
+    },
+    {
+      "epoch": 4.394891944990177,
+      "grad_norm": 0.05873017174862043,
+      "learning_rate": 3.517945196025725e-06,
+      "loss": 0.3678,
+      "step": 2237
+    },
+    {
+      "epoch": 4.3968565815324165,
+      "grad_norm": 0.05948383615418931,
+      "learning_rate": 3.4954766226085355e-06,
+      "loss": 0.3634,
+      "step": 2238
+    },
+    {
+      "epoch": 4.398821218074656,
+      "grad_norm": 0.05741810642800084,
+      "learning_rate": 3.4730767521428654e-06,
+      "loss": 0.3735,
+      "step": 2239
+    },
+    {
+      "epoch": 4.400785854616895,
+      "grad_norm": 0.05658668215549469,
+      "learning_rate": 3.4507456267861605e-06,
+      "loss": 0.3637,
+      "step": 2240
+    },
+    {
+      "epoch": 4.402750491159136,
+      "grad_norm": 0.05560807037470537,
+      "learning_rate": 3.428483288566473e-06,
+      "loss": 0.3642,
+      "step": 2241
+    },
+    {
+      "epoch": 4.404715127701375,
+      "grad_norm": 0.056313575562318474,
+      "learning_rate": 3.406289779382399e-06,
+      "loss": 0.3558,
+      "step": 2242
+    },
+    {
+      "epoch": 4.406679764243615,
+      "grad_norm": 0.05538141830418359,
+      "learning_rate": 3.3841651410030063e-06,
+      "loss": 0.3566,
+      "step": 2243
+    },
+    {
+      "epoch": 4.408644400785855,
+      "grad_norm": 0.056200157553777254,
+      "learning_rate": 3.362109415067725e-06,
+      "loss": 0.363,
+      "step": 2244
+    },
+    {
+      "epoch": 4.410609037328094,
+      "grad_norm": 0.055492750459060054,
+      "learning_rate": 3.3401226430863098e-06,
+      "loss": 0.3638,
+      "step": 2245
+    },
+    {
+      "epoch": 4.412573673870334,
+      "grad_norm": 0.055510914627766805,
+      "learning_rate": 3.3182048664387367e-06,
+      "loss": 0.3643,
+      "step": 2246
+    },
+    {
+      "epoch": 4.414538310412573,
+      "grad_norm": 0.05836767784205766,
+      "learning_rate": 3.2963561263751066e-06,
+      "loss": 0.3685,
+      "step": 2247
+    },
+    {
+      "epoch": 4.416502946954814,
+      "grad_norm": 0.055303043295745175,
+      "learning_rate": 3.27457646401562e-06,
+      "loss": 0.374,
+      "step": 2248
+    },
+    {
+      "epoch": 4.418467583497053,
+      "grad_norm": 0.056517634388274604,
+      "learning_rate": 3.252865920350461e-06,
+      "loss": 0.3673,
+      "step": 2249
+    },
+    {
+      "epoch": 4.4204322200392925,
+      "grad_norm": 0.05452838261228001,
+      "learning_rate": 3.231224536239732e-06,
+      "loss": 0.3693,
+      "step": 2250
+    },
+    {
+      "epoch": 4.422396856581533,
+      "grad_norm": 0.05424048918202045,
+      "learning_rate": 3.209652352413359e-06,
+      "loss": 0.3745,
+      "step": 2251
+    },
+    {
+      "epoch": 4.424361493123772,
+      "grad_norm": 0.056010840847504746,
+      "learning_rate": 3.1881494094710486e-06,
+      "loss": 0.3553,
+      "step": 2252
+    },
+    {
+      "epoch": 4.426326129666012,
+      "grad_norm": 0.05523781162950163,
+      "learning_rate": 3.1667157478821875e-06,
+      "loss": 0.3612,
+      "step": 2253
+    },
+    {
+      "epoch": 4.428290766208251,
+      "grad_norm": 0.05455705391143089,
+      "learning_rate": 3.1453514079857706e-06,
+      "loss": 0.3557,
+      "step": 2254
+    },
+    {
+      "epoch": 4.4302554027504915,
+      "grad_norm": 0.05537964477621436,
+      "learning_rate": 3.1240564299903277e-06,
+      "loss": 0.368,
+      "step": 2255
+    },
+    {
+      "epoch": 4.432220039292731,
+      "grad_norm": 0.05561453992032201,
+      "learning_rate": 3.1028308539738526e-06,
+      "loss": 0.3623,
+      "step": 2256
+    },
+    {
+      "epoch": 4.43418467583497,
+      "grad_norm": 0.05524580767439124,
+      "learning_rate": 3.0816747198837026e-06,
+      "loss": 0.3706,
+      "step": 2257
+    },
+    {
+      "epoch": 4.436149312377211,
+      "grad_norm": 0.053899928693339184,
+      "learning_rate": 3.0605880675365606e-06,
+      "loss": 0.3681,
+      "step": 2258
+    },
+    {
+      "epoch": 4.43811394891945,
+      "grad_norm": 0.05510295998248114,
+      "learning_rate": 3.0395709366183477e-06,
+      "loss": 0.3604,
+      "step": 2259
+    },
+    {
+      "epoch": 4.4400785854616895,
+      "grad_norm": 0.05543280679514751,
+      "learning_rate": 3.018623366684117e-06,
+      "loss": 0.3664,
+      "step": 2260
+    },
+    {
+      "epoch": 4.442043222003929,
+      "grad_norm": 0.05668019919381218,
+      "learning_rate": 2.9977453971580296e-06,
+      "loss": 0.3633,
+      "step": 2261
+    },
+    {
+      "epoch": 4.444007858546169,
+      "grad_norm": 0.05595788600712484,
+      "learning_rate": 2.976937067333241e-06,
+      "loss": 0.3553,
+      "step": 2262
+    },
+    {
+      "epoch": 4.445972495088409,
+      "grad_norm": 0.057018459862077306,
+      "learning_rate": 2.956198416371856e-06,
+      "loss": 0.3682,
+      "step": 2263
+    },
+    {
+      "epoch": 4.447937131630648,
+      "grad_norm": 0.057131173819310155,
+      "learning_rate": 2.9355294833048264e-06,
+      "loss": 0.3661,
+      "step": 2264
+    },
+    {
+      "epoch": 4.449901768172888,
+      "grad_norm": 0.0546536943266307,
+      "learning_rate": 2.9149303070319024e-06,
+      "loss": 0.3609,
+      "step": 2265
+    },
+    {
+      "epoch": 4.451866404715128,
+      "grad_norm": 0.05256252157070986,
+      "learning_rate": 2.8944009263215523e-06,
+      "loss": 0.3641,
+      "step": 2266
+    },
+    {
+      "epoch": 4.453831041257367,
+      "grad_norm": 0.05398610955269697,
+      "learning_rate": 2.873941379810865e-06,
+      "loss": 0.3651,
+      "step": 2267
+    },
+    {
+      "epoch": 4.455795677799607,
+      "grad_norm": 0.055923063870381866,
+      "learning_rate": 2.853551706005515e-06,
+      "loss": 0.368,
+      "step": 2268
+    },
+    {
+      "epoch": 4.457760314341847,
+      "grad_norm": 0.05410913885907504,
+      "learning_rate": 2.8332319432796905e-06,
+      "loss": 0.3547,
+      "step": 2269
+    },
+    {
+      "epoch": 4.459724950884087,
+      "grad_norm": 0.05528464642144032,
+      "learning_rate": 2.8129821298759607e-06,
+      "loss": 0.363,
+      "step": 2270
+    },
+    {
+      "epoch": 4.461689587426326,
+      "grad_norm": 0.05465099743694136,
+      "learning_rate": 2.7928023039052844e-06,
+      "loss": 0.3598,
+      "step": 2271
+    },
+    {
+      "epoch": 4.4636542239685655,
+      "grad_norm": 0.05462126575898459,
+      "learning_rate": 2.77269250334689e-06,
+      "loss": 0.3635,
+      "step": 2272
+    },
+    {
+      "epoch": 4.465618860510806,
+      "grad_norm": 0.055466835417171526,
+      "learning_rate": 2.7526527660482096e-06,
+      "loss": 0.3657,
+      "step": 2273
+    },
+    {
+      "epoch": 4.467583497053045,
+      "grad_norm": 0.054198999703560465,
+      "learning_rate": 2.7326831297248245e-06,
+      "loss": 0.371,
+      "step": 2274
+    },
+    {
+      "epoch": 4.469548133595285,
+      "grad_norm": 0.05474822757187929,
+      "learning_rate": 2.712783631960383e-06,
+      "loss": 0.3697,
+      "step": 2275
+    },
+    {
+      "epoch": 4.471512770137524,
+      "grad_norm": 0.05531454997492565,
+      "learning_rate": 2.6929543102065083e-06,
+      "loss": 0.3741,
+      "step": 2276
+    },
+    {
+      "epoch": 4.4734774066797645,
+      "grad_norm": 0.0536743927317602,
+      "learning_rate": 2.67319520178277e-06,
+      "loss": 0.366,
+      "step": 2277
+    },
+    {
+      "epoch": 4.475442043222004,
+      "grad_norm": 0.055845166757260874,
+      "learning_rate": 2.653506343876595e-06,
+      "loss": 0.3638,
+      "step": 2278
+    },
+    {
+      "epoch": 4.477406679764243,
+      "grad_norm": 0.05507417292869993,
+      "learning_rate": 2.633887773543191e-06,
+      "loss": 0.3729,
+      "step": 2279
+    },
+    {
+      "epoch": 4.479371316306484,
+      "grad_norm": 0.05372157448027472,
+      "learning_rate": 2.6143395277054717e-06,
+      "loss": 0.3617,
+      "step": 2280
+    },
+    {
+      "epoch": 4.481335952848723,
+      "grad_norm": 0.05298674703462983,
+      "learning_rate": 2.5948616431540073e-06,
+      "loss": 0.362,
+      "step": 2281
+    },
+    {
+      "epoch": 4.4833005893909625,
+      "grad_norm": 0.05679617275496257,
+      "learning_rate": 2.5754541565469502e-06,
+      "loss": 0.3707,
+      "step": 2282
+    },
+    {
+      "epoch": 4.485265225933202,
+      "grad_norm": 0.05545867244945176,
+      "learning_rate": 2.5561171044099497e-06,
+      "loss": 0.3687,
+      "step": 2283
+    },
+    {
+      "epoch": 4.487229862475442,
+      "grad_norm": 0.05492449144789259,
+      "learning_rate": 2.536850523136103e-06,
+      "loss": 0.3622,
+      "step": 2284
+    },
+    {
+      "epoch": 4.489194499017682,
+      "grad_norm": 0.05501296731672487,
+      "learning_rate": 2.5176544489858803e-06,
+      "loss": 0.3789,
+      "step": 2285
+    },
+    {
+      "epoch": 4.491159135559921,
+      "grad_norm": 0.05644655258057678,
+      "learning_rate": 2.4985289180870356e-06,
+      "loss": 0.3638,
+      "step": 2286
+    },
+    {
+      "epoch": 4.493123772102161,
+      "grad_norm": 0.05410199585327173,
+      "learning_rate": 2.4794739664345846e-06,
+      "loss": 0.363,
+      "step": 2287
+    },
+    {
+      "epoch": 4.495088408644401,
+      "grad_norm": 0.0561059714727828,
+      "learning_rate": 2.4604896298906945e-06,
+      "loss": 0.3622,
+      "step": 2288
+    },
+    {
+      "epoch": 4.49705304518664,
+      "grad_norm": 0.05367646178964447,
+      "learning_rate": 2.4415759441846287e-06,
+      "loss": 0.3643,
+      "step": 2289
+    },
+    {
+      "epoch": 4.49901768172888,
+      "grad_norm": 0.05532847423975451,
+      "learning_rate": 2.4227329449126913e-06,
+      "loss": 0.3654,
+      "step": 2290
+    },
+    {
+      "epoch": 4.50098231827112,
+      "grad_norm": 0.053972869159675425,
+      "learning_rate": 2.4039606675381454e-06,
+      "loss": 0.3648,
+      "step": 2291
+    },
+    {
+      "epoch": 4.50294695481336,
+      "grad_norm": 0.05453920740558763,
+      "learning_rate": 2.385259147391179e-06,
+      "loss": 0.3751,
+      "step": 2292
+    },
+    {
+      "epoch": 4.504911591355599,
+      "grad_norm": 0.05487903306949807,
+      "learning_rate": 2.3666284196687615e-06,
+      "loss": 0.3603,
+      "step": 2293
+    },
+    {
+      "epoch": 4.506876227897839,
+      "grad_norm": 0.05411220452513569,
+      "learning_rate": 2.3480685194346717e-06,
+      "loss": 0.3564,
+      "step": 2294
+    },
+    {
+      "epoch": 4.508840864440079,
+      "grad_norm": 0.05952001105911974,
+      "learning_rate": 2.3295794816193773e-06,
+      "loss": 0.3714,
+      "step": 2295
+    },
+    {
+      "epoch": 4.510805500982318,
+      "grad_norm": 0.05753736206449744,
+      "learning_rate": 2.3111613410199585e-06,
+      "loss": 0.3599,
+      "step": 2296
+    },
+    {
+      "epoch": 4.512770137524558,
+      "grad_norm": 0.053600396460302686,
+      "learning_rate": 2.2928141323000965e-06,
+      "loss": 0.3668,
+      "step": 2297
+    },
+    {
+      "epoch": 4.514734774066798,
+      "grad_norm": 0.057849420945566596,
+      "learning_rate": 2.2745378899899605e-06,
+      "loss": 0.3611,
+      "step": 2298
+    },
+    {
+      "epoch": 4.5166994106090375,
+      "grad_norm": 0.056068716053103054,
+      "learning_rate": 2.2563326484861526e-06,
+      "loss": 0.3523,
+      "step": 2299
+    },
+    {
+      "epoch": 4.518664047151277,
+      "grad_norm": 0.054588355278735366,
+      "learning_rate": 2.2381984420516555e-06,
+      "loss": 0.3718,
+      "step": 2300
+    },
+    {
+      "epoch": 4.520628683693516,
+      "grad_norm": 0.0545374747024262,
+      "learning_rate": 2.2201353048157604e-06,
+      "loss": 0.3594,
+      "step": 2301
+    },
+    {
+      "epoch": 4.522593320235757,
+      "grad_norm": 0.05654837544130785,
+      "learning_rate": 2.2021432707740065e-06,
+      "loss": 0.3603,
+      "step": 2302
+    },
+    {
+      "epoch": 4.524557956777996,
+      "grad_norm": 0.06009989750264815,
+      "learning_rate": 2.1842223737881073e-06,
+      "loss": 0.3645,
+      "step": 2303
+    },
+    {
+      "epoch": 4.5265225933202355,
+      "grad_norm": 0.05472026153429305,
+      "learning_rate": 2.1663726475859017e-06,
+      "loss": 0.3629,
+      "step": 2304
+    },
+    {
+      "epoch": 4.528487229862476,
+      "grad_norm": 0.054717055418390295,
+      "learning_rate": 2.148594125761281e-06,
+      "loss": 0.3604,
+      "step": 2305
+    },
+    {
+      "epoch": 4.530451866404715,
+      "grad_norm": 0.053561518711871856,
+      "learning_rate": 2.1308868417741113e-06,
+      "loss": 0.3694,
+      "step": 2306
+    },
+    {
+      "epoch": 4.532416502946955,
+      "grad_norm": 0.055810031598279436,
+      "learning_rate": 2.1132508289502063e-06,
+      "loss": 0.3688,
+      "step": 2307
+    },
+    {
+      "epoch": 4.534381139489194,
+      "grad_norm": 0.053797279276052386,
+      "learning_rate": 2.095686120481242e-06,
+      "loss": 0.3594,
+      "step": 2308
+    },
+    {
+      "epoch": 4.5363457760314345,
+      "grad_norm": 0.053740080546123566,
+      "learning_rate": 2.0781927494246813e-06,
+      "loss": 0.3676,
+      "step": 2309
+    },
+    {
+      "epoch": 4.538310412573674,
+      "grad_norm": 0.05868095974762451,
+      "learning_rate": 2.06077074870374e-06,
+      "loss": 0.3611,
+      "step": 2310
+    },
+    {
+      "epoch": 4.540275049115913,
+      "grad_norm": 0.05388263428076354,
+      "learning_rate": 2.0434201511073184e-06,
+      "loss": 0.3693,
+      "step": 2311
+    },
+    {
+      "epoch": 4.542239685658153,
+      "grad_norm": 0.05477790706729364,
+      "learning_rate": 2.0261409892899177e-06,
+      "loss": 0.3587,
+      "step": 2312
+    },
+    {
+      "epoch": 4.544204322200393,
+      "grad_norm": 0.05625354505592831,
+      "learning_rate": 2.0089332957716044e-06,
+      "loss": 0.3647,
+      "step": 2313
+    },
+    {
+      "epoch": 4.546168958742633,
+      "grad_norm": 0.05515962686840251,
+      "learning_rate": 1.9917971029379403e-06,
+      "loss": 0.3699,
+      "step": 2314
+    },
+    {
+      "epoch": 4.548133595284872,
+      "grad_norm": 0.0546458990184582,
+      "learning_rate": 1.9747324430398995e-06,
+      "loss": 0.367,
+      "step": 2315
+    },
+    {
+      "epoch": 4.550098231827112,
+      "grad_norm": 0.05780968156701987,
+      "learning_rate": 1.957739348193859e-06,
+      "loss": 0.3645,
+      "step": 2316
+    },
+    {
+      "epoch": 4.552062868369352,
+      "grad_norm": 0.05704050549019201,
+      "learning_rate": 1.9408178503814933e-06,
+      "loss": 0.365,
+      "step": 2317
+    },
+    {
+      "epoch": 4.554027504911591,
+      "grad_norm": 0.0519628258156792,
+      "learning_rate": 1.9239679814497194e-06,
+      "loss": 0.3579,
+      "step": 2318
+    },
+    {
+      "epoch": 4.555992141453831,
+      "grad_norm": 0.05425716894411006,
+      "learning_rate": 1.9071897731106536e-06,
+      "loss": 0.3651,
+      "step": 2319
+    },
+    {
+      "epoch": 4.557956777996071,
+      "grad_norm": 0.054085374069409144,
+      "learning_rate": 1.8904832569415487e-06,
+      "loss": 0.3727,
+      "step": 2320
+    },
+    {
+      "epoch": 4.5599214145383105,
+      "grad_norm": 0.052762689920875155,
+      "learning_rate": 1.873848464384742e-06,
+      "loss": 0.3575,
+      "step": 2321
+    },
+    {
+      "epoch": 4.56188605108055,
+      "grad_norm": 0.05452627061668984,
+      "learning_rate": 1.857285426747546e-06,
+      "loss": 0.3717,
+      "step": 2322
+    },
+    {
+      "epoch": 4.563850687622789,
+      "grad_norm": 0.05454102218317235,
+      "learning_rate": 1.8407941752022562e-06,
+      "loss": 0.3699,
+      "step": 2323
+    },
+    {
+      "epoch": 4.56581532416503,
+      "grad_norm": 0.053867546215188605,
+      "learning_rate": 1.8243747407860723e-06,
+      "loss": 0.3617,
+      "step": 2324
+    },
+    {
+      "epoch": 4.567779960707269,
+      "grad_norm": 0.05736246075322174,
+      "learning_rate": 1.8080271544009953e-06,
+      "loss": 0.3618,
+      "step": 2325
+    },
+    {
+      "epoch": 4.5697445972495085,
+      "grad_norm": 0.05393907363819538,
+      "learning_rate": 1.7917514468138409e-06,
+      "loss": 0.3714,
+      "step": 2326
+    },
+    {
+      "epoch": 4.571709233791749,
+      "grad_norm": 0.056644212769744705,
+      "learning_rate": 1.7755476486561285e-06,
+      "loss": 0.3621,
+      "step": 2327
+    },
+    {
+      "epoch": 4.573673870333988,
+      "grad_norm": 0.05595661055552878,
+      "learning_rate": 1.7594157904240416e-06,
+      "loss": 0.3671,
+      "step": 2328
+    },
+    {
+      "epoch": 4.575638506876228,
+      "grad_norm": 0.05391126496900479,
+      "learning_rate": 1.7433559024783697e-06,
+      "loss": 0.3633,
+      "step": 2329
+    },
+    {
+      "epoch": 4.577603143418468,
+      "grad_norm": 0.05574977431023451,
+      "learning_rate": 1.7273680150444727e-06,
+      "loss": 0.3613,
+      "step": 2330
+    },
+    {
+      "epoch": 4.5795677799607075,
+      "grad_norm": 0.05469928667547112,
+      "learning_rate": 1.7114521582121702e-06,
+      "loss": 0.3547,
+      "step": 2331
+    },
+    {
+      "epoch": 4.581532416502947,
+      "grad_norm": 0.0541937914869702,
+      "learning_rate": 1.6956083619357322e-06,
+      "loss": 0.3633,
+      "step": 2332
+    },
+    {
+      "epoch": 4.583497053045186,
+      "grad_norm": 0.05359306919628198,
+      "learning_rate": 1.6798366560338131e-06,
+      "loss": 0.3672,
+      "step": 2333
+    },
+    {
+      "epoch": 4.585461689587426,
+      "grad_norm": 0.05576173769614119,
+      "learning_rate": 1.6641370701893844e-06,
+      "loss": 0.3597,
+      "step": 2334
+    },
+    {
+      "epoch": 4.587426326129666,
+      "grad_norm": 0.055726323878637546,
+      "learning_rate": 1.6485096339496775e-06,
+      "loss": 0.3589,
+      "step": 2335
+    },
+    {
+      "epoch": 4.589390962671906,
+      "grad_norm": 0.05334453197879455,
+      "learning_rate": 1.632954376726148e-06,
+      "loss": 0.3598,
+      "step": 2336
+    },
+    {
+      "epoch": 4.591355599214145,
+      "grad_norm": 0.05192136623813475,
+      "learning_rate": 1.6174713277944132e-06,
+      "loss": 0.3608,
+      "step": 2337
+    },
+    {
+      "epoch": 4.593320235756385,
+      "grad_norm": 0.05172974577170931,
+      "learning_rate": 1.6020605162941637e-06,
+      "loss": 0.3684,
+      "step": 2338
+    },
+    {
+      "epoch": 4.595284872298625,
+      "grad_norm": 0.05411999593909359,
+      "learning_rate": 1.5867219712291592e-06,
+      "loss": 0.3649,
+      "step": 2339
+    },
+    {
+      "epoch": 4.597249508840864,
+      "grad_norm": 0.051996972276567764,
+      "learning_rate": 1.5714557214671566e-06,
+      "loss": 0.3714,
+      "step": 2340
+    },
+    {
+      "epoch": 4.599214145383105,
+      "grad_norm": 0.05365243443168699,
+      "learning_rate": 1.5562617957398264e-06,
+      "loss": 0.3714,
+      "step": 2341
+    },
+    {
+      "epoch": 4.601178781925344,
+      "grad_norm": 0.05103466540681832,
+      "learning_rate": 1.541140222642743e-06,
+      "loss": 0.3643,
+      "step": 2342
+    },
+    {
+      "epoch": 4.6031434184675835,
+      "grad_norm": 0.05393747825074902,
+      "learning_rate": 1.526091030635306e-06,
+      "loss": 0.3625,
+      "step": 2343
+    },
+    {
+      "epoch": 4.605108055009823,
+      "grad_norm": 0.055383041451770296,
+      "learning_rate": 1.5111142480406816e-06,
+      "loss": 0.3603,
+      "step": 2344
+    },
+    {
+      "epoch": 4.607072691552063,
+      "grad_norm": 0.05315657314997448,
+      "learning_rate": 1.4962099030457622e-06,
+      "loss": 0.3599,
+      "step": 2345
+    },
+    {
+      "epoch": 4.609037328094303,
+      "grad_norm": 0.05204775429247545,
+      "learning_rate": 1.4813780237011143e-06,
+      "loss": 0.3624,
+      "step": 2346
+    },
+    {
+      "epoch": 4.611001964636542,
+      "grad_norm": 0.05486674656303605,
+      "learning_rate": 1.466618637920929e-06,
+      "loss": 0.3677,
+      "step": 2347
+    },
+    {
+      "epoch": 4.6129666011787815,
+      "grad_norm": 0.05288659223477048,
+      "learning_rate": 1.451931773482933e-06,
+      "loss": 0.3638,
+      "step": 2348
+    },
+    {
+      "epoch": 4.614931237721022,
+      "grad_norm": 0.05567408404372174,
+      "learning_rate": 1.437317458028402e-06,
+      "loss": 0.3683,
+      "step": 2349
+    },
+    {
+      "epoch": 4.616895874263261,
+      "grad_norm": 0.052565889859067925,
+      "learning_rate": 1.422775719062046e-06,
+      "loss": 0.3663,
+      "step": 2350
+    },
+    {
+      "epoch": 4.618860510805501,
+      "grad_norm": 0.05364140334206432,
+      "learning_rate": 1.408306583951995e-06,
+      "loss": 0.3587,
+      "step": 2351
+    },
+    {
+      "epoch": 4.620825147347741,
+      "grad_norm": 0.05318153363198574,
+      "learning_rate": 1.3939100799297234e-06,
+      "loss": 0.3635,
+      "step": 2352
+    },
+    {
+      "epoch": 4.6227897838899805,
+      "grad_norm": 0.05139596166231168,
+      "learning_rate": 1.379586234090029e-06,
+      "loss": 0.353,
+      "step": 2353
+    },
+    {
+      "epoch": 4.62475442043222,
+      "grad_norm": 0.053998816803498086,
+      "learning_rate": 1.3653350733909431e-06,
+      "loss": 0.37,
+      "step": 2354
+    },
+    {
+      "epoch": 4.626719056974459,
+      "grad_norm": 0.05569363978331786,
+      "learning_rate": 1.3511566246537179e-06,
+      "loss": 0.3659,
+      "step": 2355
+    },
+    {
+      "epoch": 4.6286836935167,
+      "grad_norm": 0.052176016472608375,
+      "learning_rate": 1.3370509145627542e-06,
+      "loss": 0.3585,
+      "step": 2356
+    },
+    {
+      "epoch": 4.630648330058939,
+      "grad_norm": 0.05362155821094613,
+      "learning_rate": 1.32301796966555e-06,
+      "loss": 0.3678,
+      "step": 2357
+    },
+    {
+      "epoch": 4.632612966601179,
+      "grad_norm": 0.05322449903324194,
+      "learning_rate": 1.3090578163726587e-06,
+      "loss": 0.3595,
+      "step": 2358
+    },
+    {
+      "epoch": 4.634577603143418,
+      "grad_norm": 0.052077301098551,
+      "learning_rate": 1.2951704809576503e-06,
+      "loss": 0.3679,
+      "step": 2359
+    },
+    {
+      "epoch": 4.636542239685658,
+      "grad_norm": 0.05235466991569094,
+      "learning_rate": 1.2813559895570316e-06,
+      "loss": 0.3592,
+      "step": 2360
+    },
+    {
+      "epoch": 4.638506876227898,
+      "grad_norm": 0.05215022560482393,
+      "learning_rate": 1.2676143681702136e-06,
+      "loss": 0.3732,
+      "step": 2361
+    },
+    {
+      "epoch": 4.640471512770137,
+      "grad_norm": 0.0659993971937782,
+      "learning_rate": 1.2539456426594731e-06,
+      "loss": 0.3688,
+      "step": 2362
+    },
+    {
+      "epoch": 4.642436149312378,
+      "grad_norm": 0.05390603871057192,
+      "learning_rate": 1.2403498387498947e-06,
+      "loss": 0.3708,
+      "step": 2363
+    },
+    {
+      "epoch": 4.644400785854617,
+      "grad_norm": 0.053215188475558,
+      "learning_rate": 1.226826982029312e-06,
+      "loss": 0.3699,
+      "step": 2364
+    },
+    {
+      "epoch": 4.6463654223968565,
+      "grad_norm": 0.054783576049190826,
+      "learning_rate": 1.213377097948274e-06,
+      "loss": 0.3584,
+      "step": 2365
+    },
+    {
+      "epoch": 4.648330058939096,
+      "grad_norm": 0.05441125145509875,
+      "learning_rate": 1.2000002118199982e-06,
+      "loss": 0.361,
+      "step": 2366
+    },
+    {
+      "epoch": 4.650294695481336,
+      "grad_norm": 0.054112736982522516,
+      "learning_rate": 1.1866963488202975e-06,
+      "loss": 0.363,
+      "step": 2367
+    },
+    {
+      "epoch": 4.652259332023576,
+      "grad_norm": 0.05417344872346497,
+      "learning_rate": 1.1734655339875833e-06,
+      "loss": 0.3569,
+      "step": 2368
+    },
+    {
+      "epoch": 4.654223968565815,
+      "grad_norm": 0.05254973713538347,
+      "learning_rate": 1.160307792222768e-06,
+      "loss": 0.3622,
+      "step": 2369
+    },
+    {
+      "epoch": 4.6561886051080545,
+      "grad_norm": 0.05140252558219777,
+      "learning_rate": 1.1472231482892293e-06,
+      "loss": 0.3676,
+      "step": 2370
+    },
+    {
+      "epoch": 4.658153241650295,
+      "grad_norm": 0.05358827257666933,
+      "learning_rate": 1.134211626812789e-06,
+      "loss": 0.3711,
+      "step": 2371
+    },
+    {
+      "epoch": 4.660117878192534,
+      "grad_norm": 0.05828954600157734,
+      "learning_rate": 1.1212732522816494e-06,
+      "loss": 0.3735,
+      "step": 2372
+    },
+    {
+      "epoch": 4.662082514734774,
+      "grad_norm": 0.0539539656492362,
+      "learning_rate": 1.1084080490463278e-06,
+      "loss": 0.3658,
+      "step": 2373
+    },
+    {
+      "epoch": 4.664047151277014,
+      "grad_norm": 0.05481065582853787,
+      "learning_rate": 1.0956160413196558e-06,
+      "loss": 0.3591,
+      "step": 2374
+    },
+    {
+      "epoch": 4.6660117878192535,
+      "grad_norm": 0.051076930930903375,
+      "learning_rate": 1.0828972531766868e-06,
+      "loss": 0.3573,
+      "step": 2375
+    },
+    {
+      "epoch": 4.667976424361493,
+      "grad_norm": 0.053837030703844206,
+      "learning_rate": 1.0702517085546904e-06,
+      "loss": 0.3593,
+      "step": 2376
+    },
+    {
+      "epoch": 4.669941060903733,
+      "grad_norm": 0.05185204382960603,
+      "learning_rate": 1.0576794312530647e-06,
+      "loss": 0.3719,
+      "step": 2377
+    },
+    {
+      "epoch": 4.671905697445973,
+      "grad_norm": 0.0537508578211132,
+      "learning_rate": 1.0451804449333492e-06,
+      "loss": 0.369,
+      "step": 2378
+    },
+    {
+      "epoch": 4.673870333988212,
+      "grad_norm": 0.05592908794042757,
+      "learning_rate": 1.0327547731191267e-06,
+      "loss": 0.3642,
+      "step": 2379
+    },
+    {
+      "epoch": 4.675834970530452,
+      "grad_norm": 0.050903185018322826,
+      "learning_rate": 1.0204024391959844e-06,
+      "loss": 0.3608,
+      "step": 2380
+    },
+    {
+      "epoch": 4.677799607072692,
+      "grad_norm": 0.05236396503979154,
+      "learning_rate": 1.008123466411517e-06,
+      "loss": 0.3677,
+      "step": 2381
+    },
+    {
+      "epoch": 4.679764243614931,
+      "grad_norm": 0.054003315274791026,
+      "learning_rate": 9.959178778752343e-07,
+      "loss": 0.3624,
+      "step": 2382
+    },
+    {
+      "epoch": 4.681728880157171,
+      "grad_norm": 0.05533769782171305,
+      "learning_rate": 9.837856965585301e-07,
+      "loss": 0.3699,
+      "step": 2383
+    },
+    {
+      "epoch": 4.68369351669941,
+      "grad_norm": 0.05144539853872817,
+      "learning_rate": 9.717269452946465e-07,
+      "loss": 0.363,
+      "step": 2384
+    },
+    {
+      "epoch": 4.685658153241651,
+      "grad_norm": 0.052497715421969224,
+      "learning_rate": 9.597416467786335e-07,
+      "loss": 0.3666,
+      "step": 2385
+    },
+    {
+      "epoch": 4.68762278978389,
+      "grad_norm": 0.0516537183333692,
+      "learning_rate": 9.478298235672879e-07,
+      "loss": 0.3619,
+      "step": 2386
+    },
+    {
+      "epoch": 4.6895874263261295,
+      "grad_norm": 0.05220649811563567,
+      "learning_rate": 9.35991498079134e-07,
+      "loss": 0.3674,
+      "step": 2387
+    },
+    {
+      "epoch": 4.69155206286837,
+      "grad_norm": 0.05284614857188527,
+      "learning_rate": 9.242266925943722e-07,
+      "loss": 0.3554,
+      "step": 2388
+    },
+    {
+      "epoch": 4.693516699410609,
+      "grad_norm": 0.050491610385062696,
+      "learning_rate": 9.125354292548283e-07,
+      "loss": 0.3674,
+      "step": 2389
+    },
+    {
+      "epoch": 4.695481335952849,
+      "grad_norm": 0.05072290521041454,
+      "learning_rate": 9.00917730063915e-07,
+      "loss": 0.3694,
+      "step": 2390
+    },
+    {
+      "epoch": 4.697445972495088,
+      "grad_norm": 0.05353590906438111,
+      "learning_rate": 8.89373616886604e-07,
+      "loss": 0.3729,
+      "step": 2391
+    },
+    {
+      "epoch": 4.699410609037328,
+      "grad_norm": 0.05131804204029378,
+      "learning_rate": 8.779031114493741e-07,
+      "loss": 0.3669,
+      "step": 2392
+    },
+    {
+      "epoch": 4.701375245579568,
+      "grad_norm": 0.056536470834000756,
+      "learning_rate": 8.66506235340161e-07,
+      "loss": 0.3807,
+      "step": 2393
+    },
+    {
+      "epoch": 4.703339882121807,
+      "grad_norm": 0.05346506513443442,
+      "learning_rate": 8.551830100083403e-07,
+      "loss": 0.3664,
+      "step": 2394
+    },
+    {
+      "epoch": 4.705304518664047,
+      "grad_norm": 0.052240826280668326,
+      "learning_rate": 8.439334567646651e-07,
+      "loss": 0.3614,
+      "step": 2395
+    },
+    {
+      "epoch": 4.707269155206287,
+      "grad_norm": 0.05235972163870441,
+      "learning_rate": 8.327575967812351e-07,
+      "loss": 0.362,
+      "step": 2396
+    },
+    {
+      "epoch": 4.7092337917485265,
+      "grad_norm": 0.05123642702444419,
+      "learning_rate": 8.216554510914698e-07,
+      "loss": 0.3597,
+      "step": 2397
+    },
+    {
+      "epoch": 4.711198428290766,
+      "grad_norm": 0.05219624527592893,
+      "learning_rate": 8.106270405900374e-07,
+      "loss": 0.3608,
+      "step": 2398
+    },
+    {
+      "epoch": 4.713163064833006,
+      "grad_norm": 0.0522575075100819,
+      "learning_rate": 7.996723860328415e-07,
+      "loss": 0.3655,
+      "step": 2399
+    },
+    {
+      "epoch": 4.715127701375246,
+      "grad_norm": 0.05334832845841247,
+      "learning_rate": 7.887915080369723e-07,
+      "loss": 0.3696,
+      "step": 2400
+    },
+    {
+      "epoch": 4.717092337917485,
+      "grad_norm": 0.052298828372464384,
+      "learning_rate": 7.779844270806758e-07,
+      "loss": 0.3651,
+      "step": 2401
+    },
+    {
+      "epoch": 4.719056974459725,
+      "grad_norm": 0.053562845875749554,
+      "learning_rate": 7.672511635033041e-07,
+      "loss": 0.3692,
+      "step": 2402
+    },
+    {
+      "epoch": 4.721021611001965,
+      "grad_norm": 0.05164939508539694,
+      "learning_rate": 7.565917375052723e-07,
+      "loss": 0.3631,
+      "step": 2403
+    },
+    {
+      "epoch": 4.722986247544204,
+      "grad_norm": 0.06594701419981276,
+      "learning_rate": 7.460061691480524e-07,
+      "loss": 0.3646,
+      "step": 2404
+    },
+    {
+      "epoch": 4.724950884086444,
+      "grad_norm": 0.05159977891638924,
+      "learning_rate": 7.354944783540951e-07,
+      "loss": 0.3618,
+      "step": 2405
+    },
+    {
+      "epoch": 4.726915520628683,
+      "grad_norm": 0.051061433605478906,
+      "learning_rate": 7.250566849068152e-07,
+      "loss": 0.3507,
+      "step": 2406
+    },
+    {
+      "epoch": 4.728880157170924,
+      "grad_norm": 0.05188499556391478,
+      "learning_rate": 7.146928084505567e-07,
+      "loss": 0.3578,
+      "step": 2407
+    },
+    {
+      "epoch": 4.730844793713163,
+      "grad_norm": 0.05006083027148109,
+      "learning_rate": 7.044028684905435e-07,
+      "loss": 0.3721,
+      "step": 2408
+    },
+    {
+      "epoch": 4.7328094302554025,
+      "grad_norm": 0.05447550846168743,
+      "learning_rate": 6.941868843928401e-07,
+      "loss": 0.3728,
+      "step": 2409
+    },
+    {
+      "epoch": 4.734774066797643,
+      "grad_norm": 0.051103237127904086,
+      "learning_rate": 6.840448753843376e-07,
+      "loss": 0.3735,
+      "step": 2410
+    },
+    {
+      "epoch": 4.736738703339882,
+      "grad_norm": 0.053411480802578154,
+      "learning_rate": 6.739768605527008e-07,
+      "loss": 0.3652,
+      "step": 2411
+    },
+    {
+      "epoch": 4.738703339882122,
+      "grad_norm": 0.052727777493117405,
+      "learning_rate": 6.639828588463193e-07,
+      "loss": 0.3588,
+      "step": 2412
+    },
+    {
+      "epoch": 4.740667976424361,
+      "grad_norm": 0.05307151629383048,
+      "learning_rate": 6.540628890743073e-07,
+      "loss": 0.3618,
+      "step": 2413
+    },
+    {
+      "epoch": 4.742632612966601,
+      "grad_norm": 0.050755865456910965,
+      "learning_rate": 6.442169699064327e-07,
+      "loss": 0.3637,
+      "step": 2414
+    },
+    {
+      "epoch": 4.744597249508841,
+      "grad_norm": 0.05151708231005432,
+      "learning_rate": 6.344451198731039e-07,
+      "loss": 0.3752,
+      "step": 2415
+    },
+    {
+      "epoch": 4.74656188605108,
+      "grad_norm": 0.051675704964272436,
+      "learning_rate": 6.24747357365334e-07,
+      "loss": 0.3651,
+      "step": 2416
+    },
+    {
+      "epoch": 4.74852652259332,
+      "grad_norm": 0.05092821515860216,
+      "learning_rate": 6.151237006346878e-07,
+      "loss": 0.3614,
+      "step": 2417
+    },
+    {
+      "epoch": 4.75049115913556,
+      "grad_norm": 0.05350251527964596,
+      "learning_rate": 6.05574167793268e-07,
+      "loss": 0.3664,
+      "step": 2418
+    },
+    {
+      "epoch": 4.7524557956777995,
+      "grad_norm": 0.050879126350618094,
+      "learning_rate": 5.960987768136762e-07,
+      "loss": 0.3698,
+      "step": 2419
+    },
+    {
+      "epoch": 4.754420432220039,
+      "grad_norm": 0.05567969277720775,
+      "learning_rate": 5.866975455289626e-07,
+      "loss": 0.3696,
+      "step": 2420
+    },
+    {
+      "epoch": 4.756385068762279,
+      "grad_norm": 0.05224975414370104,
+      "learning_rate": 5.773704916326273e-07,
+      "loss": 0.3531,
+      "step": 2421
+    },
+    {
+      "epoch": 4.758349705304519,
+      "grad_norm": 0.05031990301800233,
+      "learning_rate": 5.681176326785398e-07,
+      "loss": 0.3559,
+      "step": 2422
+    },
+    {
+      "epoch": 4.760314341846758,
+      "grad_norm": 0.05246399347790176,
+      "learning_rate": 5.589389860809569e-07,
+      "loss": 0.3612,
+      "step": 2423
+    },
+    {
+      "epoch": 4.7622789783889985,
+      "grad_norm": 0.05240860705916298,
+      "learning_rate": 5.498345691144425e-07,
+      "loss": 0.3601,
+      "step": 2424
+    },
+    {
+      "epoch": 4.764243614931238,
+      "grad_norm": 0.0509545582719829,
+      "learning_rate": 5.408043989138767e-07,
+      "loss": 0.3608,
+      "step": 2425
+    },
+    {
+      "epoch": 4.766208251473477,
+      "grad_norm": 0.05320369886952839,
+      "learning_rate": 5.318484924743939e-07,
+      "loss": 0.3651,
+      "step": 2426
+    },
+    {
+      "epoch": 4.768172888015717,
+      "grad_norm": 0.051464141009342,
+      "learning_rate": 5.229668666513599e-07,
+      "loss": 0.366,
+      "step": 2427
+    },
+    {
+      "epoch": 4.770137524557957,
+      "grad_norm": 0.05256070466619038,
+      "learning_rate": 5.141595381603459e-07,
+      "loss": 0.3619,
+      "step": 2428
+    },
+    {
+      "epoch": 4.772102161100197,
+      "grad_norm": 0.17645566900238785,
+      "learning_rate": 5.054265235770928e-07,
+      "loss": 0.3554,
+      "step": 2429
+    },
+    {
+      "epoch": 4.774066797642436,
+      "grad_norm": 0.05238632589061052,
+      "learning_rate": 4.967678393374753e-07,
+      "loss": 0.3605,
+      "step": 2430
+    },
+    {
+      "epoch": 4.7760314341846755,
+      "grad_norm": 0.05209539623474717,
+      "learning_rate": 4.881835017374803e-07,
+      "loss": 0.367,
+      "step": 2431
+    },
+    {
+      "epoch": 4.777996070726916,
+      "grad_norm": 0.05160322781535277,
+      "learning_rate": 4.796735269331709e-07,
+      "loss": 0.3619,
+      "step": 2432
+    },
+    {
+      "epoch": 4.779960707269155,
+      "grad_norm": 0.05070194045212126,
+      "learning_rate": 4.712379309406512e-07,
+      "loss": 0.3556,
+      "step": 2433
+    },
+    {
+      "epoch": 4.781925343811395,
+      "grad_norm": 0.05255552328828293,
+      "learning_rate": 4.628767296360481e-07,
+      "loss": 0.3646,
+      "step": 2434
+    },
+    {
+      "epoch": 4.783889980353635,
+      "grad_norm": 0.052043628638104224,
+      "learning_rate": 4.545899387554675e-07,
+      "loss": 0.3616,
+      "step": 2435
+    },
+    {
+      "epoch": 4.785854616895874,
+      "grad_norm": 0.05108440552221833,
+      "learning_rate": 4.4637757389498495e-07,
+      "loss": 0.3583,
+      "step": 2436
+    },
+    {
+      "epoch": 4.787819253438114,
+      "grad_norm": 0.055842095219215994,
+      "learning_rate": 4.382396505105879e-07,
+      "loss": 0.3678,
+      "step": 2437
+    },
+    {
+      "epoch": 4.789783889980353,
+      "grad_norm": 0.05246495040735323,
+      "learning_rate": 4.301761839181717e-07,
+      "loss": 0.3605,
+      "step": 2438
+    },
+    {
+      "epoch": 4.791748526522594,
+      "grad_norm": 0.05178049703742096,
+      "learning_rate": 4.221871892934992e-07,
+      "loss": 0.3686,
+      "step": 2439
+    },
+    {
+      "epoch": 4.793713163064833,
+      "grad_norm": 0.05175489640577532,
+      "learning_rate": 4.142726816721698e-07,
+      "loss": 0.3612,
+      "step": 2440
+    },
+    {
+      "epoch": 4.7956777996070725,
+      "grad_norm": 0.05111125849226848,
+      "learning_rate": 4.0643267594959734e-07,
+      "loss": 0.3643,
+      "step": 2441
+    },
+    {
+      "epoch": 4.797642436149312,
+      "grad_norm": 0.051598815587862654,
+      "learning_rate": 3.986671868809877e-07,
+      "loss": 0.3696,
+      "step": 2442
+    },
+    {
+      "epoch": 4.799607072691552,
+      "grad_norm": 0.05141947793084467,
+      "learning_rate": 3.909762290812902e-07,
+      "loss": 0.3591,
+      "step": 2443
+    },
+    {
+      "epoch": 4.801571709233792,
+      "grad_norm": 0.052818540942146636,
+      "learning_rate": 3.833598170251973e-07,
+      "loss": 0.3654,
+      "step": 2444
+    },
+    {
+      "epoch": 4.803536345776031,
+      "grad_norm": 0.05318356124276527,
+      "learning_rate": 3.7581796504709166e-07,
+      "loss": 0.3627,
+      "step": 2445
+    },
+    {
+      "epoch": 4.8055009823182715,
+      "grad_norm": 0.061143159379318114,
+      "learning_rate": 3.6835068734103695e-07,
+      "loss": 0.3617,
+      "step": 2446
+    },
+    {
+      "epoch": 4.807465618860511,
+      "grad_norm": 0.05191880100132736,
+      "learning_rate": 3.6095799796075137e-07,
+      "loss": 0.3673,
+      "step": 2447
+    },
+    {
+      "epoch": 4.80943025540275,
+      "grad_norm": 0.051334736602822124,
+      "learning_rate": 3.5363991081955427e-07,
+      "loss": 0.3609,
+      "step": 2448
+    },
+    {
+      "epoch": 4.81139489194499,
+      "grad_norm": 0.051053219797478774,
+      "learning_rate": 3.463964396903841e-07,
+      "loss": 0.3582,
+      "step": 2449
+    },
+    {
+      "epoch": 4.81335952848723,
+      "grad_norm": 0.04956835391826528,
+      "learning_rate": 3.392275982057358e-07,
+      "loss": 0.3662,
+      "step": 2450
+    },
+    {
+      "epoch": 4.81532416502947,
+      "grad_norm": 0.05093843109544226,
+      "learning_rate": 3.3213339985765256e-07,
+      "loss": 0.3709,
+      "step": 2451
+    },
+    {
+      "epoch": 4.817288801571709,
+      "grad_norm": 0.051465742067237474,
+      "learning_rate": 3.251138579976898e-07,
+      "loss": 0.3672,
+      "step": 2452
+    },
+    {
+      "epoch": 4.8192534381139485,
+      "grad_norm": 0.04983993628936402,
+      "learning_rate": 3.181689858369108e-07,
+      "loss": 0.3709,
+      "step": 2453
+    },
+    {
+      "epoch": 4.821218074656189,
+      "grad_norm": 0.0530218433555467,
+      "learning_rate": 3.112987964458292e-07,
+      "loss": 0.3613,
+      "step": 2454
+    },
+    {
+      "epoch": 4.823182711198428,
+      "grad_norm": 0.05312771066584817,
+      "learning_rate": 3.04503302754422e-07,
+      "loss": 0.3644,
+      "step": 2455
+    },
+    {
+      "epoch": 4.825147347740668,
+      "grad_norm": 0.05251789773099509,
+      "learning_rate": 2.977825175520677e-07,
+      "loss": 0.3648,
+      "step": 2456
+    },
+    {
+      "epoch": 4.827111984282908,
+      "grad_norm": 0.05285380249347576,
+      "learning_rate": 2.9113645348755936e-07,
+      "loss": 0.3618,
+      "step": 2457
+    },
+    {
+      "epoch": 4.829076620825147,
+      "grad_norm": 0.05210002528998325,
+      "learning_rate": 2.8456512306903827e-07,
+      "loss": 0.3641,
+      "step": 2458
+    },
+    {
+      "epoch": 4.831041257367387,
+      "grad_norm": 0.05037782817991046,
+      "learning_rate": 2.7806853866401585e-07,
+      "loss": 0.3666,
+      "step": 2459
+    },
+    {
+      "epoch": 4.833005893909627,
+      "grad_norm": 0.05355504983020299,
+      "learning_rate": 2.7164671249932053e-07,
+      "loss": 0.3576,
+      "step": 2460
+    },
+    {
+      "epoch": 4.834970530451867,
+      "grad_norm": 0.052183501286187275,
+      "learning_rate": 2.652996566610755e-07,
+      "loss": 0.3661,
+      "step": 2461
+    },
+    {
+      "epoch": 4.836935166994106,
+      "grad_norm": 0.0518288017542531,
+      "learning_rate": 2.590273830946943e-07,
+      "loss": 0.3702,
+      "step": 2462
+    },
+    {
+      "epoch": 4.8388998035363455,
+      "grad_norm": 0.05201072215035939,
+      "learning_rate": 2.5282990360484096e-07,
+      "loss": 0.3644,
+      "step": 2463
+    },
+    {
+      "epoch": 4.840864440078585,
+      "grad_norm": 0.05411589617031124,
+      "learning_rate": 2.4670722985541627e-07,
+      "loss": 0.3628,
+      "step": 2464
+    },
+    {
+      "epoch": 4.842829076620825,
+      "grad_norm": 0.052086222100697376,
+      "learning_rate": 2.406593733695406e-07,
+      "loss": 0.359,
+      "step": 2465
+    },
+    {
+      "epoch": 4.844793713163065,
+      "grad_norm": 0.05037152594097684,
+      "learning_rate": 2.3468634552950898e-07,
+      "loss": 0.3718,
+      "step": 2466
+    },
+    {
+      "epoch": 4.846758349705304,
+      "grad_norm": 0.052115158733952636,
+      "learning_rate": 2.2878815757680028e-07,
+      "loss": 0.3698,
+      "step": 2467
+    },
+    {
+      "epoch": 4.8487229862475445,
+      "grad_norm": 0.051837189272304046,
+      "learning_rate": 2.2296482061203273e-07,
+      "loss": 0.3713,
+      "step": 2468
+    },
+    {
+      "epoch": 4.850687622789784,
+      "grad_norm": 0.050939422809552534,
+      "learning_rate": 2.1721634559496384e-07,
+      "loss": 0.3646,
+      "step": 2469
+    },
+    {
+      "epoch": 4.852652259332023,
+      "grad_norm": 0.049236543649758255,
+      "learning_rate": 2.1154274334443726e-07,
+      "loss": 0.3614,
+      "step": 2470
+    },
+    {
+      "epoch": 4.854616895874264,
+      "grad_norm": 0.05159031552102578,
+      "learning_rate": 2.0594402453840478e-07,
+      "loss": 0.3666,
+      "step": 2471
+    },
+    {
+      "epoch": 4.856581532416503,
+      "grad_norm": 0.05196916110035245,
+      "learning_rate": 2.0042019971387772e-07,
+      "loss": 0.3585,
+      "step": 2472
+    },
+    {
+      "epoch": 4.858546168958743,
+      "grad_norm": 0.0532627958239236,
+      "learning_rate": 1.9497127926690451e-07,
+      "loss": 0.3561,
+      "step": 2473
+    },
+    {
+      "epoch": 4.860510805500982,
+      "grad_norm": 0.05096100124277097,
+      "learning_rate": 1.8959727345257527e-07,
+      "loss": 0.3715,
+      "step": 2474
+    },
+    {
+      "epoch": 4.862475442043222,
+      "grad_norm": 0.05163904829277873,
+      "learning_rate": 1.8429819238497738e-07,
+      "loss": 0.3586,
+      "step": 2475
+    },
+    {
+      "epoch": 4.864440078585462,
+      "grad_norm": 0.05202674617742681,
+      "learning_rate": 1.7907404603719093e-07,
+      "loss": 0.3636,
+      "step": 2476
+    },
+    {
+      "epoch": 4.866404715127701,
+      "grad_norm": 0.05000120936336375,
+      "learning_rate": 1.7392484424126665e-07,
+      "loss": 0.3633,
+      "step": 2477
+    },
+    {
+      "epoch": 4.868369351669941,
+      "grad_norm": 0.0514746662986566,
+      "learning_rate": 1.6885059668819926e-07,
+      "loss": 0.3602,
+      "step": 2478
+    },
+    {
+      "epoch": 4.870333988212181,
+      "grad_norm": 0.0507079380504819,
+      "learning_rate": 1.6385131292793176e-07,
+      "loss": 0.3588,
+      "step": 2479
+    },
+    {
+      "epoch": 4.87229862475442,
+      "grad_norm": 0.04974239921733917,
+      "learning_rate": 1.5892700236930235e-07,
+      "loss": 0.363,
+      "step": 2480
+    },
+    {
+      "epoch": 4.87426326129666,
+      "grad_norm": 0.05176113456981624,
+      "learning_rate": 1.540776742800665e-07,
+      "loss": 0.3627,
+      "step": 2481
+    },
+    {
+      "epoch": 4.8762278978389,
+      "grad_norm": 0.05023133960329087,
+      "learning_rate": 1.4930333778684803e-07,
+      "loss": 0.3577,
+      "step": 2482
+    },
+    {
+      "epoch": 4.87819253438114,
+      "grad_norm": 0.051828495576896685,
+      "learning_rate": 1.446040018751349e-07,
+      "loss": 0.361,
+      "step": 2483
+    },
+    {
+      "epoch": 4.880157170923379,
+      "grad_norm": 0.05198563064107819,
+      "learning_rate": 1.3997967538926572e-07,
+      "loss": 0.3595,
+      "step": 2484
+    },
+    {
+      "epoch": 4.8821218074656185,
+      "grad_norm": 0.050342233781507956,
+      "learning_rate": 1.3543036703239864e-07,
+      "loss": 0.3672,
+      "step": 2485
+    },
+    {
+      "epoch": 4.884086444007859,
+      "grad_norm": 0.05164721709625265,
+      "learning_rate": 1.3095608536652038e-07,
+      "loss": 0.3627,
+      "step": 2486
+    },
+    {
+      "epoch": 4.886051080550098,
+      "grad_norm": 0.04997933439840741,
+      "learning_rate": 1.2655683881240167e-07,
+      "loss": 0.3556,
+      "step": 2487
+    },
+    {
+      "epoch": 4.888015717092338,
+      "grad_norm": 0.05132929314827621,
+      "learning_rate": 1.2223263564959286e-07,
+      "loss": 0.3718,
+      "step": 2488
+    },
+    {
+      "epoch": 4.889980353634577,
+      "grad_norm": 0.05089777581624961,
+      "learning_rate": 1.1798348401641958e-07,
+      "loss": 0.3665,
+      "step": 2489
+    },
+    {
+      "epoch": 4.8919449901768175,
+      "grad_norm": 0.052078848360301244,
+      "learning_rate": 1.1380939190995587e-07,
+      "loss": 0.3573,
+      "step": 2490
+    },
+    {
+      "epoch": 4.893909626719057,
+      "grad_norm": 0.04987539736894915,
+      "learning_rate": 1.0971036718599781e-07,
+      "loss": 0.3635,
+      "step": 2491
+    },
+    {
+      "epoch": 4.895874263261296,
+      "grad_norm": 0.05010040102968093,
+      "learning_rate": 1.0568641755908104e-07,
+      "loss": 0.3587,
+      "step": 2492
+    },
+    {
+      "epoch": 4.897838899803537,
+      "grad_norm": 0.05142640218465388,
+      "learning_rate": 1.0173755060243207e-07,
+      "loss": 0.3649,
+      "step": 2493
+    },
+    {
+      "epoch": 4.899803536345776,
+      "grad_norm": 0.050547973377159716,
+      "learning_rate": 9.786377374797262e-08,
+      "loss": 0.3697,
+      "step": 2494
+    },
+    {
+      "epoch": 4.901768172888016,
+      "grad_norm": 0.05140641686085488,
+      "learning_rate": 9.406509428631083e-08,
+      "loss": 0.3641,
+      "step": 2495
+    },
+    {
+      "epoch": 4.903732809430255,
+      "grad_norm": 0.05011205633306078,
+      "learning_rate": 9.034151936670566e-08,
+      "loss": 0.3685,
+      "step": 2496
+    },
+    {
+      "epoch": 4.905697445972495,
+      "grad_norm": 0.0502135343046472,
+      "learning_rate": 8.669305599707134e-08,
+      "loss": 0.3758,
+      "step": 2497
+    },
+    {
+      "epoch": 4.907662082514735,
+      "grad_norm": 0.051535380166162326,
+      "learning_rate": 8.31197110439641e-08,
+      "loss": 0.3631,
+      "step": 2498
+    },
+    {
+      "epoch": 4.909626719056974,
+      "grad_norm": 0.05198458906856649,
+      "learning_rate": 7.962149123255547e-08,
+      "loss": 0.3688,
+      "step": 2499
+    },
+    {
+      "epoch": 4.911591355599214,
+      "grad_norm": 0.050913416568651094,
+      "learning_rate": 7.619840314663229e-08,
+      "loss": 0.3748,
+      "step": 2500
+    },
+    {
+      "epoch": 4.913555992141454,
+      "grad_norm": 0.05135186873211408,
+      "learning_rate": 7.285045322858342e-08,
+      "loss": 0.3653,
+      "step": 2501
+    },
+    {
+      "epoch": 4.915520628683693,
+      "grad_norm": 0.05214645440035884,
+      "learning_rate": 6.957764777938636e-08,
+      "loss": 0.3649,
+      "step": 2502
+    },
+    {
+      "epoch": 4.917485265225933,
+      "grad_norm": 0.050521432864935184,
+      "learning_rate": 6.637999295858511e-08,
+      "loss": 0.3601,
+      "step": 2503
+    },
+    {
+      "epoch": 4.919449901768173,
+      "grad_norm": 0.051260588183810976,
+      "learning_rate": 6.325749478429011e-08,
+      "loss": 0.3635,
+      "step": 2504
+    },
+    {
+      "epoch": 4.921414538310413,
+      "grad_norm": 0.05122943528364674,
+      "learning_rate": 6.021015913316941e-08,
+      "loss": 0.3643,
+      "step": 2505
+    },
+    {
+      "epoch": 4.923379174852652,
+      "grad_norm": 0.051575941713372694,
+      "learning_rate": 5.723799174043087e-08,
+      "loss": 0.3627,
+      "step": 2506
+    },
+    {
+      "epoch": 4.925343811394892,
+      "grad_norm": 0.05097823297542795,
+      "learning_rate": 5.434099819980887e-08,
+      "loss": 0.3658,
+      "step": 2507
+    },
+    {
+      "epoch": 4.927308447937132,
+      "grad_norm": 0.0499977001809955,
+      "learning_rate": 5.1519183963555374e-08,
+      "loss": 0.3608,
+      "step": 2508
+    },
+    {
+      "epoch": 4.929273084479371,
+      "grad_norm": 0.050201212042074596,
+      "learning_rate": 4.8772554342439995e-08,
+      "loss": 0.367,
+      "step": 2509
+    },
+    {
+      "epoch": 4.931237721021611,
+      "grad_norm": 0.050673023829474695,
+      "learning_rate": 4.610111450572774e-08,
+      "loss": 0.3656,
+      "step": 2510
+    },
+    {
+      "epoch": 4.933202357563851,
+      "grad_norm": 0.050250466817106064,
+      "learning_rate": 4.350486948117461e-08,
+      "loss": 0.3683,
+      "step": 2511
+    },
+    {
+      "epoch": 4.9351669941060905,
+      "grad_norm": 0.04902248519453472,
+      "learning_rate": 4.098382415500979e-08,
+      "loss": 0.363,
+      "step": 2512
+    },
+    {
+      "epoch": 4.93713163064833,
+      "grad_norm": 0.05148383991540384,
+      "learning_rate": 3.853798327194902e-08,
+      "loss": 0.3599,
+      "step": 2513
+    },
+    {
+      "epoch": 4.939096267190569,
+      "grad_norm": 0.052276384645874696,
+      "learning_rate": 3.61673514351546e-08,
+      "loss": 0.3688,
+      "step": 2514
+    },
+    {
+      "epoch": 4.94106090373281,
+      "grad_norm": 0.0505421110968789,
+      "learning_rate": 3.387193310625314e-08,
+      "loss": 0.3695,
+      "step": 2515
+    },
+    {
+      "epoch": 4.943025540275049,
+      "grad_norm": 0.04987183065951885,
+      "learning_rate": 3.16517326053134e-08,
+      "loss": 0.3645,
+      "step": 2516
+    },
+    {
+      "epoch": 4.944990176817289,
+      "grad_norm": 0.051619588507551196,
+      "learning_rate": 2.9506754110832926e-08,
+      "loss": 0.3701,
+      "step": 2517
+    },
+    {
+      "epoch": 4.946954813359529,
+      "grad_norm": 0.051390907799838924,
+      "learning_rate": 2.7437001659751385e-08,
+      "loss": 0.3763,
+      "step": 2518
+    },
+    {
+      "epoch": 4.948919449901768,
+      "grad_norm": 0.05138230035104072,
+      "learning_rate": 2.5442479147423925e-08,
+      "loss": 0.3617,
+      "step": 2519
+    },
+    {
+      "epoch": 4.950884086444008,
+      "grad_norm": 0.05004133287119141,
+      "learning_rate": 2.352319032761674e-08,
+      "loss": 0.3642,
+      "step": 2520
+    },
+    {
+      "epoch": 4.952848722986247,
+      "grad_norm": 0.049838525827492416,
+      "learning_rate": 2.167913881251593e-08,
+      "loss": 0.3675,
+      "step": 2521
+    },
+    {
+      "epoch": 4.954813359528488,
+      "grad_norm": 0.04943836045475183,
+      "learning_rate": 1.9910328072687556e-08,
+      "loss": 0.3585,
+      "step": 2522
+    },
+    {
+      "epoch": 4.956777996070727,
+      "grad_norm": 0.049932987604008626,
+      "learning_rate": 1.821676143711315e-08,
+      "loss": 0.3688,
+      "step": 2523
+    },
+    {
+      "epoch": 4.958742632612966,
+      "grad_norm": 0.050640230411207805,
+      "learning_rate": 1.6598442093145317e-08,
+      "loss": 0.3767,
+      "step": 2524
+    },
+    {
+      "epoch": 4.960707269155206,
+      "grad_norm": 0.05108818777837528,
+      "learning_rate": 1.5055373086525492e-08,
+      "loss": 0.3632,
+      "step": 2525
+    },
+    {
+      "epoch": 4.962671905697446,
+      "grad_norm": 0.05298813423582169,
+      "learning_rate": 1.3587557321375067e-08,
+      "loss": 0.3695,
+      "step": 2526
+    },
+    {
+      "epoch": 4.964636542239686,
+      "grad_norm": 0.05025535918519285,
+      "learning_rate": 1.2194997560177613e-08,
+      "loss": 0.3575,
+      "step": 2527
+    },
+    {
+      "epoch": 4.966601178781925,
+      "grad_norm": 0.05081190961633483,
+      "learning_rate": 1.0877696423787776e-08,
+      "loss": 0.3596,
+      "step": 2528
+    },
+    {
+      "epoch": 4.968565815324165,
+      "grad_norm": 0.05196737425639162,
+      "learning_rate": 9.635656391413505e-09,
+      "loss": 0.3553,
+      "step": 2529
+    },
+    {
+      "epoch": 4.970530451866405,
+      "grad_norm": 0.05002319093671988,
+      "learning_rate": 8.468879800629381e-09,
+      "loss": 0.3598,
+      "step": 2530
+    },
+    {
+      "epoch": 4.972495088408644,
+      "grad_norm": 0.051850996078254635,
+      "learning_rate": 7.377368847345523e-09,
+      "loss": 0.3627,
+      "step": 2531
+    },
+    {
+      "epoch": 4.974459724950884,
+      "grad_norm": 0.052270740806430994,
+      "learning_rate": 6.361125585838679e-09,
+      "loss": 0.3692,
+      "step": 2532
+    },
+    {
+      "epoch": 4.976424361493124,
+      "grad_norm": 0.05024893194853551,
+      "learning_rate": 5.4201519287166994e-09,
+      "loss": 0.3674,
+      "step": 2533
+    },
+    {
+      "epoch": 4.9783889980353635,
+      "grad_norm": 0.049179460299697136,
+      "learning_rate": 4.554449646922976e-09,
+      "loss": 0.3605,
+      "step": 2534
+    },
+    {
+      "epoch": 4.980353634577603,
+      "grad_norm": 0.051407116504284686,
+      "learning_rate": 3.764020369749766e-09,
+      "loss": 0.3602,
+      "step": 2535
+    },
+    {
+      "epoch": 4.982318271119842,
+      "grad_norm": 0.05127617375830702,
+      "learning_rate": 3.0488655848115444e-09,
+      "loss": 0.3721,
+      "step": 2536
+    },
+    {
+      "epoch": 4.984282907662083,
+      "grad_norm": 0.05009067782621677,
+      "learning_rate": 2.4089866380627714e-09,
+      "loss": 0.3705,
+      "step": 2537
+    },
+    {
+      "epoch": 4.986247544204322,
+      "grad_norm": 0.05315912290748018,
+      "learning_rate": 1.8443847337756837e-09,
+      "loss": 0.3692,
+      "step": 2538
+    },
+    {
+      "epoch": 4.988212180746562,
+      "grad_norm": 0.04986801419032231,
+      "learning_rate": 1.3550609345536203e-09,
+      "loss": 0.3656,
+      "step": 2539
+    },
+    {
+      "epoch": 4.990176817288802,
+      "grad_norm": 0.05060284099467721,
+      "learning_rate": 9.410161613310209e-10,
+      "loss": 0.3655,
+      "step": 2540
+    },
+    {
+      "epoch": 4.992141453831041,
+      "grad_norm": 0.05383007202901866,
+      "learning_rate": 6.022511933467811e-10,
+      "loss": 0.3707,
+      "step": 2541
+    },
+    {
+      "epoch": 4.994106090373281,
+      "grad_norm": 0.04976401962160572,
+      "learning_rate": 3.3876666817533874e-10,
+      "loss": 0.369,
+      "step": 2542
+    },
+    {
+      "epoch": 4.996070726915521,
+      "grad_norm": 0.05407358799073762,
+      "learning_rate": 1.505630817000281e-10,
+      "loss": 0.366,
+      "step": 2543
+    },
+    {
+      "epoch": 4.998035363457761,
+      "grad_norm": 0.04937005305366321,
+      "learning_rate": 3.764078813528471e-11,
+      "loss": 0.3609,
+      "step": 2544
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.051460983349980845,
+      "learning_rate": 0.0,
+      "loss": 0.3611,
+      "step": 2545
+    },
+    {
+      "epoch": 5.0,
+      "step": 2545,
+      "total_flos": 1.4336695323328512e+17,
+      "train_loss": 0.24341244895472275,
+      "train_runtime": 65867.4089,
+      "train_samples_per_second": 19.78,
+      "train_steps_per_second": 0.039
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2545,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4336695323328512e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}