diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,26467 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 3775,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013245033112582781,
+      "grad_norm": 2.4951015857107692,
+      "learning_rate": 2.1164021164021165e-07,
+      "loss": 0.9195,
+      "step": 1
+    },
+    {
+      "epoch": 0.0026490066225165563,
+      "grad_norm": 2.5069483858941553,
+      "learning_rate": 4.232804232804233e-07,
+      "loss": 0.9229,
+      "step": 2
+    },
+    {
+      "epoch": 0.003973509933774834,
+      "grad_norm": 2.518336367953602,
+      "learning_rate": 6.34920634920635e-07,
+      "loss": 0.9224,
+      "step": 3
+    },
+    {
+      "epoch": 0.005298013245033113,
+      "grad_norm": 2.475571336113595,
+      "learning_rate": 8.465608465608466e-07,
+      "loss": 0.9216,
+      "step": 4
+    },
+    {
+      "epoch": 0.006622516556291391,
+      "grad_norm": 2.459788556793139,
+      "learning_rate": 1.0582010582010582e-06,
+      "loss": 0.9176,
+      "step": 5
+    },
+    {
+      "epoch": 0.007947019867549669,
+      "grad_norm": 2.3516647579018235,
+      "learning_rate": 1.26984126984127e-06,
+      "loss": 0.9149,
+      "step": 6
+    },
+    {
+      "epoch": 0.009271523178807948,
+      "grad_norm": 2.31788045447258,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.9127,
+      "step": 7
+    },
+    {
+      "epoch": 0.010596026490066225,
+      "grad_norm": 2.012302230416029,
+      "learning_rate": 1.6931216931216932e-06,
+      "loss": 0.8916,
+      "step": 8
+    },
+    {
+      "epoch": 0.011920529801324504,
+      "grad_norm": 1.9637254159298372,
+      "learning_rate": 1.904761904761905e-06,
+      "loss": 0.9037,
+      "step": 9
+    },
+    {
+      "epoch": 0.013245033112582781,
+      "grad_norm": 1.8467595197313083,
+      "learning_rate": 2.1164021164021164e-06,
+      "loss": 0.8913,
+      "step": 10
+    },
+    {
+      "epoch": 0.01456953642384106,
+      "grad_norm": 1.4834000398886105,
+      "learning_rate": 2.328042328042328e-06,
+      "loss": 0.8716,
+      "step": 11
+    },
+    {
+      "epoch": 0.015894039735099338,
+      "grad_norm": 1.4260454981783945,
+      "learning_rate": 2.53968253968254e-06,
+      "loss": 0.8664,
+      "step": 12
+    },
+    {
+      "epoch": 0.017218543046357615,
+      "grad_norm": 1.4062886325161967,
+      "learning_rate": 2.7513227513227516e-06,
+      "loss": 0.8613,
+      "step": 13
+    },
+    {
+      "epoch": 0.018543046357615896,
+      "grad_norm": 1.2355284439204997,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.8305,
+      "step": 14
+    },
+    {
+      "epoch": 0.019867549668874173,
+      "grad_norm": 1.2522577174014358,
+      "learning_rate": 3.1746031746031746e-06,
+      "loss": 0.8229,
+      "step": 15
+    },
+    {
+      "epoch": 0.02119205298013245,
+      "grad_norm": 1.153096423991731,
+      "learning_rate": 3.3862433862433864e-06,
+      "loss": 0.8204,
+      "step": 16
+    },
+    {
+      "epoch": 0.022516556291390728,
+      "grad_norm": 1.0018117962951791,
+      "learning_rate": 3.597883597883598e-06,
+      "loss": 0.8064,
+      "step": 17
+    },
+    {
+      "epoch": 0.02384105960264901,
+      "grad_norm": 0.8931930231330436,
+      "learning_rate": 3.80952380952381e-06,
+      "loss": 0.8022,
+      "step": 18
+    },
+    {
+      "epoch": 0.025165562913907286,
+      "grad_norm": 0.8658685201617282,
+      "learning_rate": 4.0211640211640215e-06,
+      "loss": 0.7632,
+      "step": 19
+    },
+    {
+      "epoch": 0.026490066225165563,
+      "grad_norm": 0.9233081270295175,
+      "learning_rate": 4.232804232804233e-06,
+      "loss": 0.7488,
+      "step": 20
+    },
+    {
+      "epoch": 0.02781456953642384,
+      "grad_norm": 0.8985112870724821,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.7458,
+      "step": 21
+    },
+    {
+      "epoch": 0.02913907284768212,
+      "grad_norm": 0.7238701605047487,
+      "learning_rate": 4.656084656084656e-06,
+      "loss": 0.7273,
+      "step": 22
+    },
+    {
+      "epoch": 0.030463576158940398,
+      "grad_norm": 0.6425753268478377,
+      "learning_rate": 4.867724867724868e-06,
+      "loss": 0.7142,
+      "step": 23
+    },
+    {
+      "epoch": 0.031788079470198675,
+      "grad_norm": 0.6055086418332439,
+      "learning_rate": 5.07936507936508e-06,
+      "loss": 0.7144,
+      "step": 24
+    },
+    {
+      "epoch": 0.033112582781456956,
+      "grad_norm": 0.5641360091623852,
+      "learning_rate": 5.291005291005291e-06,
+      "loss": 0.7003,
+      "step": 25
+    },
+    {
+      "epoch": 0.03443708609271523,
+      "grad_norm": 0.6640440849024278,
+      "learning_rate": 5.502645502645503e-06,
+      "loss": 0.6912,
+      "step": 26
+    },
+    {
+      "epoch": 0.03576158940397351,
+      "grad_norm": 0.6497012951429935,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 0.6801,
+      "step": 27
+    },
+    {
+      "epoch": 0.03708609271523179,
+      "grad_norm": 0.6135597374414027,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.672,
+      "step": 28
+    },
+    {
+      "epoch": 0.038410596026490065,
+      "grad_norm": 0.5937016438929659,
+      "learning_rate": 6.137566137566138e-06,
+      "loss": 0.6679,
+      "step": 29
+    },
+    {
+      "epoch": 0.039735099337748346,
+      "grad_norm": 0.5550677979812818,
+      "learning_rate": 6.349206349206349e-06,
+      "loss": 0.6626,
+      "step": 30
+    },
+    {
+      "epoch": 0.04105960264900662,
+      "grad_norm": 0.48095722086181114,
+      "learning_rate": 6.560846560846561e-06,
+      "loss": 0.6576,
+      "step": 31
+    },
+    {
+      "epoch": 0.0423841059602649,
+      "grad_norm": 0.4487364807310198,
+      "learning_rate": 6.772486772486773e-06,
+      "loss": 0.6512,
+      "step": 32
+    },
+    {
+      "epoch": 0.04370860927152318,
+      "grad_norm": 0.43441954711399206,
+      "learning_rate": 6.984126984126984e-06,
+      "loss": 0.6468,
+      "step": 33
+    },
+    {
+      "epoch": 0.045033112582781455,
+      "grad_norm": 0.37446875714690964,
+      "learning_rate": 7.195767195767196e-06,
+      "loss": 0.6415,
+      "step": 34
+    },
+    {
+      "epoch": 0.046357615894039736,
+      "grad_norm": 0.3581926857745742,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.6356,
+      "step": 35
+    },
+    {
+      "epoch": 0.04768211920529802,
+      "grad_norm": 0.34403316294297476,
+      "learning_rate": 7.61904761904762e-06,
+      "loss": 0.6322,
+      "step": 36
+    },
+    {
+      "epoch": 0.04900662251655629,
+      "grad_norm": 0.3629403824068534,
+      "learning_rate": 7.830687830687831e-06,
+      "loss": 0.63,
+      "step": 37
+    },
+    {
+      "epoch": 0.05033112582781457,
+      "grad_norm": 0.33510233023857255,
+      "learning_rate": 8.042328042328043e-06,
+      "loss": 0.6252,
+      "step": 38
+    },
+    {
+      "epoch": 0.051655629139072845,
+      "grad_norm": 0.30034686053524945,
+      "learning_rate": 8.253968253968254e-06,
+      "loss": 0.6211,
+      "step": 39
+    },
+    {
+      "epoch": 0.052980132450331126,
+      "grad_norm": 0.3147600567715714,
+      "learning_rate": 8.465608465608466e-06,
+      "loss": 0.617,
+      "step": 40
+    },
+    {
+      "epoch": 0.054304635761589407,
+      "grad_norm": 0.3204247453459467,
+      "learning_rate": 8.677248677248678e-06,
+      "loss": 0.6161,
+      "step": 41
+    },
+    {
+      "epoch": 0.05562913907284768,
+      "grad_norm": 0.28540137548908057,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.6097,
+      "step": 42
+    },
+    {
+      "epoch": 0.05695364238410596,
+      "grad_norm": 0.2501127862693104,
+      "learning_rate": 9.1005291005291e-06,
+      "loss": 0.6108,
+      "step": 43
+    },
+    {
+      "epoch": 0.05827814569536424,
+      "grad_norm": 0.25626587701400844,
+      "learning_rate": 9.312169312169313e-06,
+      "loss": 0.6066,
+      "step": 44
+    },
+    {
+      "epoch": 0.059602649006622516,
+      "grad_norm": 0.2564531566916178,
+      "learning_rate": 9.523809523809525e-06,
+      "loss": 0.6011,
+      "step": 45
+    },
+    {
+      "epoch": 0.060927152317880796,
+      "grad_norm": 0.23672021203877813,
+      "learning_rate": 9.735449735449735e-06,
+      "loss": 0.6011,
+      "step": 46
+    },
+    {
+      "epoch": 0.06225165562913907,
+      "grad_norm": 0.21407877504188352,
+      "learning_rate": 9.947089947089947e-06,
+      "loss": 0.6024,
+      "step": 47
+    },
+    {
+      "epoch": 0.06357615894039735,
+      "grad_norm": 0.21489582392964784,
+      "learning_rate": 1.015873015873016e-05,
+      "loss": 0.5944,
+      "step": 48
+    },
+    {
+      "epoch": 0.06490066225165562,
+      "grad_norm": 0.22755959239405907,
+      "learning_rate": 1.037037037037037e-05,
+      "loss": 0.5939,
+      "step": 49
+    },
+    {
+      "epoch": 0.06622516556291391,
+      "grad_norm": 0.22124132910025154,
+      "learning_rate": 1.0582010582010582e-05,
+      "loss": 0.5927,
+      "step": 50
+    },
+    {
+      "epoch": 0.06754966887417219,
+      "grad_norm": 0.21354174922636607,
+      "learning_rate": 1.0793650793650794e-05,
+      "loss": 0.5881,
+      "step": 51
+    },
+    {
+      "epoch": 0.06887417218543046,
+      "grad_norm": 0.1820871304742545,
+      "learning_rate": 1.1005291005291006e-05,
+      "loss": 0.5862,
+      "step": 52
+    },
+    {
+      "epoch": 0.07019867549668875,
+      "grad_norm": 0.2042253586725618,
+      "learning_rate": 1.1216931216931217e-05,
+      "loss": 0.586,
+      "step": 53
+    },
+    {
+      "epoch": 0.07152317880794702,
+      "grad_norm": 0.22901914956821787,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 0.5802,
+      "step": 54
+    },
+    {
+      "epoch": 0.0728476821192053,
+      "grad_norm": 0.19770260210224297,
+      "learning_rate": 1.1640211640211641e-05,
+      "loss": 0.5842,
+      "step": 55
+    },
+    {
+      "epoch": 0.07417218543046358,
+      "grad_norm": 0.1677900830228023,
+      "learning_rate": 1.1851851851851852e-05,
+      "loss": 0.5813,
+      "step": 56
+    },
+    {
+      "epoch": 0.07549668874172186,
+      "grad_norm": 0.18709489143601807,
+      "learning_rate": 1.2063492063492064e-05,
+      "loss": 0.579,
+      "step": 57
+    },
+    {
+      "epoch": 0.07682119205298013,
+      "grad_norm": 0.1539489852394959,
+      "learning_rate": 1.2275132275132276e-05,
+      "loss": 0.581,
+      "step": 58
+    },
+    {
+      "epoch": 0.0781456953642384,
+      "grad_norm": 0.14689119168384712,
+      "learning_rate": 1.2486772486772486e-05,
+      "loss": 0.5761,
+      "step": 59
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 0.15682727782159764,
+      "learning_rate": 1.2698412698412699e-05,
+      "loss": 0.577,
+      "step": 60
+    },
+    {
+      "epoch": 0.08079470198675497,
+      "grad_norm": 0.1470743836923586,
+      "learning_rate": 1.291005291005291e-05,
+      "loss": 0.5733,
+      "step": 61
+    },
+    {
+      "epoch": 0.08211920529801324,
+      "grad_norm": 0.14988498882029283,
+      "learning_rate": 1.3121693121693123e-05,
+      "loss": 0.5711,
+      "step": 62
+    },
+    {
+      "epoch": 0.08344370860927153,
+      "grad_norm": 0.1267802410569993,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.5688,
+      "step": 63
+    },
+    {
+      "epoch": 0.0847682119205298,
+      "grad_norm": 0.17077377281036502,
+      "learning_rate": 1.3544973544973545e-05,
+      "loss": 0.573,
+      "step": 64
+    },
+    {
+      "epoch": 0.08609271523178808,
+      "grad_norm": 0.14805309187719867,
+      "learning_rate": 1.3756613756613758e-05,
+      "loss": 0.5668,
+      "step": 65
+    },
+    {
+      "epoch": 0.08741721854304636,
+      "grad_norm": 0.15313738248674905,
+      "learning_rate": 1.3968253968253968e-05,
+      "loss": 0.5714,
+      "step": 66
+    },
+    {
+      "epoch": 0.08874172185430464,
+      "grad_norm": 0.15111482322695463,
+      "learning_rate": 1.417989417989418e-05,
+      "loss": 0.5694,
+      "step": 67
+    },
+    {
+      "epoch": 0.09006622516556291,
+      "grad_norm": 0.1379424462241911,
+      "learning_rate": 1.4391534391534392e-05,
+      "loss": 0.5673,
+      "step": 68
+    },
+    {
+      "epoch": 0.0913907284768212,
+      "grad_norm": 0.1541278363047302,
+      "learning_rate": 1.4603174603174603e-05,
+      "loss": 0.5672,
+      "step": 69
+    },
+    {
+      "epoch": 0.09271523178807947,
+      "grad_norm": 0.14348643924494467,
+      "learning_rate": 1.4814814814814815e-05,
+      "loss": 0.5671,
+      "step": 70
+    },
+    {
+      "epoch": 0.09403973509933775,
+      "grad_norm": 0.14676026060150274,
+      "learning_rate": 1.5026455026455027e-05,
+      "loss": 0.5611,
+      "step": 71
+    },
+    {
+      "epoch": 0.09536423841059603,
+      "grad_norm": 0.168566862517487,
+      "learning_rate": 1.523809523809524e-05,
+      "loss": 0.5622,
+      "step": 72
+    },
+    {
+      "epoch": 0.09668874172185431,
+      "grad_norm": 0.1364659680563039,
+      "learning_rate": 1.544973544973545e-05,
+      "loss": 0.5651,
+      "step": 73
+    },
+    {
+      "epoch": 0.09801324503311258,
+      "grad_norm": 0.14126633334347158,
+      "learning_rate": 1.5661375661375662e-05,
+      "loss": 0.5585,
+      "step": 74
+    },
+    {
+      "epoch": 0.09933774834437085,
+      "grad_norm": 0.13534419833894473,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.5583,
+      "step": 75
+    },
+    {
+      "epoch": 0.10066225165562914,
+      "grad_norm": 0.13215368189128213,
+      "learning_rate": 1.6084656084656086e-05,
+      "loss": 0.5596,
+      "step": 76
+    },
+    {
+      "epoch": 0.10198675496688742,
+      "grad_norm": 0.11835068920200319,
+      "learning_rate": 1.6296296296296297e-05,
+      "loss": 0.5588,
+      "step": 77
+    },
+    {
+      "epoch": 0.10331125827814569,
+      "grad_norm": 0.11800341005964819,
+      "learning_rate": 1.6507936507936507e-05,
+      "loss": 0.5557,
+      "step": 78
+    },
+    {
+      "epoch": 0.10463576158940398,
+      "grad_norm": 0.13131022535026002,
+      "learning_rate": 1.671957671957672e-05,
+      "loss": 0.5534,
+      "step": 79
+    },
+    {
+      "epoch": 0.10596026490066225,
+      "grad_norm": 0.11818517100842556,
+      "learning_rate": 1.693121693121693e-05,
+      "loss": 0.5554,
+      "step": 80
+    },
+    {
+      "epoch": 0.10728476821192053,
+      "grad_norm": 0.15674290688841333,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.5545,
+      "step": 81
+    },
+    {
+      "epoch": 0.10860927152317881,
+      "grad_norm": 0.17319801049607186,
+      "learning_rate": 1.7354497354497356e-05,
+      "loss": 0.5547,
+      "step": 82
+    },
+    {
+      "epoch": 0.10993377483443709,
+      "grad_norm": 0.23904818846326115,
+      "learning_rate": 1.7566137566137566e-05,
+      "loss": 0.5541,
+      "step": 83
+    },
+    {
+      "epoch": 0.11125827814569536,
+      "grad_norm": 0.38006986168276163,
+      "learning_rate": 1.7777777777777777e-05,
+      "loss": 0.5527,
+      "step": 84
+    },
+    {
+      "epoch": 0.11258278145695365,
+      "grad_norm": 0.5232994492471573,
+      "learning_rate": 1.798941798941799e-05,
+      "loss": 0.5565,
+      "step": 85
+    },
+    {
+      "epoch": 0.11390728476821192,
+      "grad_norm": 0.5009679424462643,
+      "learning_rate": 1.82010582010582e-05,
+      "loss": 0.5547,
+      "step": 86
+    },
+    {
+      "epoch": 0.1152317880794702,
+      "grad_norm": 0.3242253809849485,
+      "learning_rate": 1.8412698412698415e-05,
+      "loss": 0.5543,
+      "step": 87
+    },
+    {
+      "epoch": 0.11655629139072848,
+      "grad_norm": 0.2877990294173592,
+      "learning_rate": 1.8624338624338625e-05,
+      "loss": 0.5516,
+      "step": 88
+    },
+    {
+      "epoch": 0.11788079470198676,
+      "grad_norm": 0.44157834065597373,
+      "learning_rate": 1.8835978835978836e-05,
+      "loss": 0.549,
+      "step": 89
+    },
+    {
+      "epoch": 0.11920529801324503,
+      "grad_norm": 0.3960121955066786,
+      "learning_rate": 1.904761904761905e-05,
+      "loss": 0.5497,
+      "step": 90
+    },
+    {
+      "epoch": 0.1205298013245033,
+      "grad_norm": 0.5154668049227998,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 0.5504,
+      "step": 91
+    },
+    {
+      "epoch": 0.12185430463576159,
+      "grad_norm": 0.6310183888100668,
+      "learning_rate": 1.947089947089947e-05,
+      "loss": 0.5497,
+      "step": 92
+    },
+    {
+      "epoch": 0.12317880794701987,
+      "grad_norm": 0.3296007300740119,
+      "learning_rate": 1.9682539682539684e-05,
+      "loss": 0.5484,
+      "step": 93
+    },
+    {
+      "epoch": 0.12450331125827814,
+      "grad_norm": 0.3917349339041057,
+      "learning_rate": 1.9894179894179895e-05,
+      "loss": 0.5473,
+      "step": 94
+    },
+    {
+      "epoch": 0.12582781456953643,
+      "grad_norm": 0.5950029500495336,
+      "learning_rate": 2.010582010582011e-05,
+      "loss": 0.5465,
+      "step": 95
+    },
+    {
+      "epoch": 0.1271523178807947,
+      "grad_norm": 0.4469629552243088,
+      "learning_rate": 2.031746031746032e-05,
+      "loss": 0.5468,
+      "step": 96
+    },
+    {
+      "epoch": 0.12847682119205298,
+      "grad_norm": 0.32437212012705247,
+      "learning_rate": 2.0529100529100533e-05,
+      "loss": 0.5462,
+      "step": 97
+    },
+    {
+      "epoch": 0.12980132450331125,
+      "grad_norm": 0.28808130188943304,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 0.5479,
+      "step": 98
+    },
+    {
+      "epoch": 0.13112582781456952,
+      "grad_norm": 0.26046646010804964,
+      "learning_rate": 2.0952380952380954e-05,
+      "loss": 0.5415,
+      "step": 99
+    },
+    {
+      "epoch": 0.13245033112582782,
+      "grad_norm": 0.26801252012723414,
+      "learning_rate": 2.1164021164021164e-05,
+      "loss": 0.5449,
+      "step": 100
+    },
+    {
+      "epoch": 0.1337748344370861,
+      "grad_norm": 0.28749454494733195,
+      "learning_rate": 2.1375661375661378e-05,
+      "loss": 0.5477,
+      "step": 101
+    },
+    {
+      "epoch": 0.13509933774834437,
+      "grad_norm": 0.3023162459869953,
+      "learning_rate": 2.158730158730159e-05,
+      "loss": 0.5466,
+      "step": 102
+    },
+    {
+      "epoch": 0.13642384105960265,
+      "grad_norm": 0.28713666045021785,
+      "learning_rate": 2.1798941798941802e-05,
+      "loss": 0.5428,
+      "step": 103
+    },
+    {
+      "epoch": 0.13774834437086092,
+      "grad_norm": 0.29763936850785344,
+      "learning_rate": 2.2010582010582013e-05,
+      "loss": 0.5444,
+      "step": 104
+    },
+    {
+      "epoch": 0.1390728476821192,
+      "grad_norm": 0.3370952424876196,
+      "learning_rate": 2.2222222222222227e-05,
+      "loss": 0.5406,
+      "step": 105
+    },
+    {
+      "epoch": 0.1403973509933775,
+      "grad_norm": 0.38096486991621586,
+      "learning_rate": 2.2433862433862434e-05,
+      "loss": 0.5451,
+      "step": 106
+    },
+    {
+      "epoch": 0.14172185430463577,
+      "grad_norm": 0.4804108036815486,
+      "learning_rate": 2.2645502645502648e-05,
+      "loss": 0.5413,
+      "step": 107
+    },
+    {
+      "epoch": 0.14304635761589404,
+      "grad_norm": 0.5307305400970737,
+      "learning_rate": 2.2857142857142858e-05,
+      "loss": 0.5389,
+      "step": 108
+    },
+    {
+      "epoch": 0.14437086092715232,
+      "grad_norm": 0.521366966131432,
+      "learning_rate": 2.3068783068783072e-05,
+      "loss": 0.5382,
+      "step": 109
+    },
+    {
+      "epoch": 0.1456953642384106,
+      "grad_norm": 0.5062885801773422,
+      "learning_rate": 2.3280423280423282e-05,
+      "loss": 0.5426,
+      "step": 110
+    },
+    {
+      "epoch": 0.14701986754966886,
+      "grad_norm": 0.5545096218913367,
+      "learning_rate": 2.3492063492063496e-05,
+      "loss": 0.543,
+      "step": 111
+    },
+    {
+      "epoch": 0.14834437086092717,
+      "grad_norm": 0.5822561904863042,
+      "learning_rate": 2.3703703703703703e-05,
+      "loss": 0.5417,
+      "step": 112
+    },
+    {
+      "epoch": 0.14966887417218544,
+      "grad_norm": 0.5997406893203101,
+      "learning_rate": 2.3915343915343917e-05,
+      "loss": 0.5437,
+      "step": 113
+    },
+    {
+      "epoch": 0.1509933774834437,
+      "grad_norm": 0.6084307897923639,
+      "learning_rate": 2.4126984126984128e-05,
+      "loss": 0.5429,
+      "step": 114
+    },
+    {
+      "epoch": 0.152317880794702,
+      "grad_norm": 0.5592334685869622,
+      "learning_rate": 2.433862433862434e-05,
+      "loss": 0.5378,
+      "step": 115
+    },
+    {
+      "epoch": 0.15364238410596026,
+      "grad_norm": 0.5162144258308952,
+      "learning_rate": 2.4550264550264552e-05,
+      "loss": 0.5437,
+      "step": 116
+    },
+    {
+      "epoch": 0.15496688741721854,
+      "grad_norm": 0.6043308199507917,
+      "learning_rate": 2.4761904761904766e-05,
+      "loss": 0.5416,
+      "step": 117
+    },
+    {
+      "epoch": 0.1562913907284768,
+      "grad_norm": 0.5054948322658758,
+      "learning_rate": 2.4973544973544973e-05,
+      "loss": 0.5387,
+      "step": 118
+    },
+    {
+      "epoch": 0.1576158940397351,
+      "grad_norm": 0.41822740007028225,
+      "learning_rate": 2.5185185185185187e-05,
+      "loss": 0.5397,
+      "step": 119
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 0.531000938638128,
+      "learning_rate": 2.5396825396825397e-05,
+      "loss": 0.5445,
+      "step": 120
+    },
+    {
+      "epoch": 0.16026490066225166,
+      "grad_norm": 0.5112731803547682,
+      "learning_rate": 2.560846560846561e-05,
+      "loss": 0.5374,
+      "step": 121
+    },
+    {
+      "epoch": 0.16158940397350993,
+      "grad_norm": 0.3297989822232066,
+      "learning_rate": 2.582010582010582e-05,
+      "loss": 0.534,
+      "step": 122
+    },
+    {
+      "epoch": 0.1629139072847682,
+      "grad_norm": 0.5942123628877751,
+      "learning_rate": 2.6031746031746035e-05,
+      "loss": 0.534,
+      "step": 123
+    },
+    {
+      "epoch": 0.16423841059602648,
+      "grad_norm": 0.6816722903922403,
+      "learning_rate": 2.6243386243386246e-05,
+      "loss": 0.5369,
+      "step": 124
+    },
+    {
+      "epoch": 0.16556291390728478,
+      "grad_norm": 0.5726197467260425,
+      "learning_rate": 2.645502645502646e-05,
+      "loss": 0.5376,
+      "step": 125
+    },
+    {
+      "epoch": 0.16688741721854305,
+      "grad_norm": 0.41002656620042977,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.5351,
+      "step": 126
+    },
+    {
+      "epoch": 0.16821192052980133,
+      "grad_norm": 0.43308368982033063,
+      "learning_rate": 2.687830687830688e-05,
+      "loss": 0.5375,
+      "step": 127
+    },
+    {
+      "epoch": 0.1695364238410596,
+      "grad_norm": 0.39077302812321885,
+      "learning_rate": 2.708994708994709e-05,
+      "loss": 0.5349,
+      "step": 128
+    },
+    {
+      "epoch": 0.17086092715231788,
+      "grad_norm": 0.38431424612673454,
+      "learning_rate": 2.7301587301587305e-05,
+      "loss": 0.5342,
+      "step": 129
+    },
+    {
+      "epoch": 0.17218543046357615,
+      "grad_norm": 0.6049711391545987,
+      "learning_rate": 2.7513227513227515e-05,
+      "loss": 0.5346,
+      "step": 130
+    },
+    {
+      "epoch": 0.17350993377483442,
+      "grad_norm": 0.7093529661691825,
+      "learning_rate": 2.772486772486773e-05,
+      "loss": 0.5371,
+      "step": 131
+    },
+    {
+      "epoch": 0.17483443708609273,
+      "grad_norm": 0.5968611583124322,
+      "learning_rate": 2.7936507936507936e-05,
+      "loss": 0.5355,
+      "step": 132
+    },
+    {
+      "epoch": 0.176158940397351,
+      "grad_norm": 0.5641176043819496,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.5336,
+      "step": 133
+    },
+    {
+      "epoch": 0.17748344370860927,
+      "grad_norm": 0.5092016000963328,
+      "learning_rate": 2.835978835978836e-05,
+      "loss": 0.5319,
+      "step": 134
+    },
+    {
+      "epoch": 0.17880794701986755,
+      "grad_norm": 0.30200833594842424,
+      "learning_rate": 2.8571428571428574e-05,
+      "loss": 0.5312,
+      "step": 135
+    },
+    {
+      "epoch": 0.18013245033112582,
+      "grad_norm": 0.36848053219613014,
+      "learning_rate": 2.8783068783068785e-05,
+      "loss": 0.5318,
+      "step": 136
+    },
+    {
+      "epoch": 0.1814569536423841,
+      "grad_norm": 0.5333717234632629,
+      "learning_rate": 2.8994708994709e-05,
+      "loss": 0.5326,
+      "step": 137
+    },
+    {
+      "epoch": 0.1827814569536424,
+      "grad_norm": 0.5430183407529882,
+      "learning_rate": 2.9206349206349206e-05,
+      "loss": 0.5296,
+      "step": 138
+    },
+    {
+      "epoch": 0.18410596026490067,
+      "grad_norm": 0.5157803454823627,
+      "learning_rate": 2.941798941798942e-05,
+      "loss": 0.5335,
+      "step": 139
+    },
+    {
+      "epoch": 0.18543046357615894,
+      "grad_norm": 0.5557988284496658,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 0.5341,
+      "step": 140
+    },
+    {
+      "epoch": 0.18675496688741722,
+      "grad_norm": 0.5003165782088166,
+      "learning_rate": 2.9841269841269844e-05,
+      "loss": 0.5339,
+      "step": 141
+    },
+    {
+      "epoch": 0.1880794701986755,
+      "grad_norm": 0.6245467837162767,
+      "learning_rate": 3.0052910052910054e-05,
+      "loss": 0.5342,
+      "step": 142
+    },
+    {
+      "epoch": 0.18940397350993377,
+      "grad_norm": 0.7326360953053956,
+      "learning_rate": 3.0264550264550268e-05,
+      "loss": 0.5306,
+      "step": 143
+    },
+    {
+      "epoch": 0.19072847682119207,
+      "grad_norm": 0.7205957886738492,
+      "learning_rate": 3.047619047619048e-05,
+      "loss": 0.5284,
+      "step": 144
+    },
+    {
+      "epoch": 0.19205298013245034,
+      "grad_norm": 0.7422963502229467,
+      "learning_rate": 3.068783068783069e-05,
+      "loss": 0.5329,
+      "step": 145
+    },
+    {
+      "epoch": 0.19337748344370861,
+      "grad_norm": 0.7675618217801546,
+      "learning_rate": 3.08994708994709e-05,
+      "loss": 0.5277,
+      "step": 146
+    },
+    {
+      "epoch": 0.1947019867549669,
+      "grad_norm": 0.5843961310832951,
+      "learning_rate": 3.111111111111112e-05,
+      "loss": 0.5291,
+      "step": 147
+    },
+    {
+      "epoch": 0.19602649006622516,
+      "grad_norm": 0.4599827105468913,
+      "learning_rate": 3.1322751322751324e-05,
+      "loss": 0.531,
+      "step": 148
+    },
+    {
+      "epoch": 0.19735099337748344,
+      "grad_norm": 0.566549671919622,
+      "learning_rate": 3.153439153439154e-05,
+      "loss": 0.531,
+      "step": 149
+    },
+    {
+      "epoch": 0.1986754966887417,
+      "grad_norm": 0.6057219228713795,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.5277,
+      "step": 150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4850461982888782,
+      "learning_rate": 3.195767195767196e-05,
+      "loss": 0.5272,
+      "step": 151
+    },
+    {
+      "epoch": 0.20132450331125828,
+      "grad_norm": 0.33034864088900084,
+      "learning_rate": 3.216931216931217e-05,
+      "loss": 0.5279,
+      "step": 152
+    },
+    {
+      "epoch": 0.20264900662251656,
+      "grad_norm": 0.43703680977953874,
+      "learning_rate": 3.2380952380952386e-05,
+      "loss": 0.5302,
+      "step": 153
+    },
+    {
+      "epoch": 0.20397350993377483,
+      "grad_norm": 0.45420379377913606,
+      "learning_rate": 3.259259259259259e-05,
+      "loss": 0.5307,
+      "step": 154
+    },
+    {
+      "epoch": 0.2052980132450331,
+      "grad_norm": 0.3534183513966628,
+      "learning_rate": 3.280423280423281e-05,
+      "loss": 0.5273,
+      "step": 155
+    },
+    {
+      "epoch": 0.20662251655629138,
+      "grad_norm": 0.43054231045159597,
+      "learning_rate": 3.3015873015873014e-05,
+      "loss": 0.5301,
+      "step": 156
+    },
+    {
+      "epoch": 0.20794701986754968,
+      "grad_norm": 0.4663130750367461,
+      "learning_rate": 3.322751322751323e-05,
+      "loss": 0.5284,
+      "step": 157
+    },
+    {
+      "epoch": 0.20927152317880796,
+      "grad_norm": 0.5190054988609634,
+      "learning_rate": 3.343915343915344e-05,
+      "loss": 0.5241,
+      "step": 158
+    },
+    {
+      "epoch": 0.21059602649006623,
+      "grad_norm": 0.6381365474313193,
+      "learning_rate": 3.3650793650793656e-05,
+      "loss": 0.5273,
+      "step": 159
+    },
+    {
+      "epoch": 0.2119205298013245,
+      "grad_norm": 0.8026126053063981,
+      "learning_rate": 3.386243386243386e-05,
+      "loss": 0.5282,
+      "step": 160
+    },
+    {
+      "epoch": 0.21324503311258278,
+      "grad_norm": 0.8529320648657448,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 0.5349,
+      "step": 161
+    },
+    {
+      "epoch": 0.21456953642384105,
+      "grad_norm": 0.7555655815587011,
+      "learning_rate": 3.4285714285714284e-05,
+      "loss": 0.5279,
+      "step": 162
+    },
+    {
+      "epoch": 0.21589403973509932,
+      "grad_norm": 0.649240778449811,
+      "learning_rate": 3.44973544973545e-05,
+      "loss": 0.5279,
+      "step": 163
+    },
+    {
+      "epoch": 0.21721854304635763,
+      "grad_norm": 0.49431590295877914,
+      "learning_rate": 3.470899470899471e-05,
+      "loss": 0.5249,
+      "step": 164
+    },
+    {
+      "epoch": 0.2185430463576159,
+      "grad_norm": 0.4918149129710849,
+      "learning_rate": 3.4920634920634925e-05,
+      "loss": 0.5256,
+      "step": 165
+    },
+    {
+      "epoch": 0.21986754966887417,
+      "grad_norm": 0.8238233139033437,
+      "learning_rate": 3.513227513227513e-05,
+      "loss": 0.5305,
+      "step": 166
+    },
+    {
+      "epoch": 0.22119205298013245,
+      "grad_norm": 0.8374290929884594,
+      "learning_rate": 3.5343915343915346e-05,
+      "loss": 0.5276,
+      "step": 167
+    },
+    {
+      "epoch": 0.22251655629139072,
+      "grad_norm": 0.5344879324706897,
+      "learning_rate": 3.555555555555555e-05,
+      "loss": 0.5257,
+      "step": 168
+    },
+    {
+      "epoch": 0.223841059602649,
+      "grad_norm": 0.5216943178821802,
+      "learning_rate": 3.576719576719577e-05,
+      "loss": 0.5253,
+      "step": 169
+    },
+    {
+      "epoch": 0.2251655629139073,
+      "grad_norm": 0.7373643203697191,
+      "learning_rate": 3.597883597883598e-05,
+      "loss": 0.5255,
+      "step": 170
+    },
+    {
+      "epoch": 0.22649006622516557,
+      "grad_norm": 0.6911423465639086,
+      "learning_rate": 3.6190476190476195e-05,
+      "loss": 0.5256,
+      "step": 171
+    },
+    {
+      "epoch": 0.22781456953642384,
+      "grad_norm": 0.43309721989686334,
+      "learning_rate": 3.64021164021164e-05,
+      "loss": 0.5211,
+      "step": 172
+    },
+    {
+      "epoch": 0.22913907284768212,
+      "grad_norm": 0.45286823063223963,
+      "learning_rate": 3.6613756613756616e-05,
+      "loss": 0.5242,
+      "step": 173
+    },
+    {
+      "epoch": 0.2304635761589404,
+      "grad_norm": 0.6565187425827942,
+      "learning_rate": 3.682539682539683e-05,
+      "loss": 0.522,
+      "step": 174
+    },
+    {
+      "epoch": 0.23178807947019867,
+      "grad_norm": 0.5276312437075916,
+      "learning_rate": 3.703703703703704e-05,
+      "loss": 0.5165,
+      "step": 175
+    },
+    {
+      "epoch": 0.23311258278145697,
+      "grad_norm": 0.4235669514378739,
+      "learning_rate": 3.724867724867725e-05,
+      "loss": 0.5179,
+      "step": 176
+    },
+    {
+      "epoch": 0.23443708609271524,
+      "grad_norm": 0.5491233673978111,
+      "learning_rate": 3.7460317460317464e-05,
+      "loss": 0.5209,
+      "step": 177
+    },
+    {
+      "epoch": 0.23576158940397351,
+      "grad_norm": 0.4678788571666783,
+      "learning_rate": 3.767195767195767e-05,
+      "loss": 0.5212,
+      "step": 178
+    },
+    {
+      "epoch": 0.2370860927152318,
+      "grad_norm": 0.42624592684131124,
+      "learning_rate": 3.7883597883597885e-05,
+      "loss": 0.5228,
+      "step": 179
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 0.507039394931601,
+      "learning_rate": 3.80952380952381e-05,
+      "loss": 0.5198,
+      "step": 180
+    },
+    {
+      "epoch": 0.23973509933774834,
+      "grad_norm": 0.5488715487867573,
+      "learning_rate": 3.830687830687831e-05,
+      "loss": 0.521,
+      "step": 181
+    },
+    {
+      "epoch": 0.2410596026490066,
+      "grad_norm": 0.5870361827225258,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.5237,
+      "step": 182
+    },
+    {
+      "epoch": 0.2423841059602649,
+      "grad_norm": 0.5499041816534843,
+      "learning_rate": 3.8730158730158734e-05,
+      "loss": 0.525,
+      "step": 183
+    },
+    {
+      "epoch": 0.24370860927152319,
+      "grad_norm": 0.431689288227396,
+      "learning_rate": 3.894179894179894e-05,
+      "loss": 0.5181,
+      "step": 184
+    },
+    {
+      "epoch": 0.24503311258278146,
+      "grad_norm": 0.48881854124680885,
+      "learning_rate": 3.9153439153439155e-05,
+      "loss": 0.5217,
+      "step": 185
+    },
+    {
+      "epoch": 0.24635761589403973,
+      "grad_norm": 0.6795329291334721,
+      "learning_rate": 3.936507936507937e-05,
+      "loss": 0.5192,
+      "step": 186
+    },
+    {
+      "epoch": 0.247682119205298,
+      "grad_norm": 0.8564478221260102,
+      "learning_rate": 3.957671957671958e-05,
+      "loss": 0.5272,
+      "step": 187
+    },
+    {
+      "epoch": 0.24900662251655628,
+      "grad_norm": 0.9948299039443181,
+      "learning_rate": 3.978835978835979e-05,
+      "loss": 0.5303,
+      "step": 188
+    },
+    {
+      "epoch": 0.2503311258278146,
+      "grad_norm": 0.8948661294554884,
+      "learning_rate": 4e-05,
+      "loss": 0.5303,
+      "step": 189
+    },
+    {
+      "epoch": 0.25165562913907286,
+      "grad_norm": 0.744583344643671,
+      "learning_rate": 4.021164021164022e-05,
+      "loss": 0.5241,
+      "step": 190
+    },
+    {
+      "epoch": 0.25298013245033113,
+      "grad_norm": 0.9278100456575558,
+      "learning_rate": 4.042328042328043e-05,
+      "loss": 0.5261,
+      "step": 191
+    },
+    {
+      "epoch": 0.2543046357615894,
+      "grad_norm": 0.9149706588772626,
+      "learning_rate": 4.063492063492064e-05,
+      "loss": 0.526,
+      "step": 192
+    },
+    {
+      "epoch": 0.2556291390728477,
+      "grad_norm": 0.6918544730017542,
+      "learning_rate": 4.084656084656085e-05,
+      "loss": 0.5252,
+      "step": 193
+    },
+    {
+      "epoch": 0.25695364238410595,
+      "grad_norm": 0.939250594032553,
+      "learning_rate": 4.1058201058201066e-05,
+      "loss": 0.5286,
+      "step": 194
+    },
+    {
+      "epoch": 0.2582781456953642,
+      "grad_norm": 0.7417347411195065,
+      "learning_rate": 4.126984126984128e-05,
+      "loss": 0.5191,
+      "step": 195
+    },
+    {
+      "epoch": 0.2596026490066225,
+      "grad_norm": 0.6302056810316703,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 0.5164,
+      "step": 196
+    },
+    {
+      "epoch": 0.2609271523178808,
+      "grad_norm": 0.7880212869396347,
+      "learning_rate": 4.1693121693121694e-05,
+      "loss": 0.5202,
+      "step": 197
+    },
+    {
+      "epoch": 0.26225165562913905,
+      "grad_norm": 0.4274719218605281,
+      "learning_rate": 4.190476190476191e-05,
+      "loss": 0.5214,
+      "step": 198
+    },
+    {
+      "epoch": 0.2635761589403974,
+      "grad_norm": 0.6601008524310147,
+      "learning_rate": 4.211640211640212e-05,
+      "loss": 0.5208,
+      "step": 199
+    },
+    {
+      "epoch": 0.26490066225165565,
+      "grad_norm": 0.610190171337781,
+      "learning_rate": 4.232804232804233e-05,
+      "loss": 0.5227,
+      "step": 200
+    },
+    {
+      "epoch": 0.2662251655629139,
+      "grad_norm": 0.3954392715673898,
+      "learning_rate": 4.253968253968254e-05,
+      "loss": 0.5198,
+      "step": 201
+    },
+    {
+      "epoch": 0.2675496688741722,
+      "grad_norm": 0.6932597487914673,
+      "learning_rate": 4.2751322751322756e-05,
+      "loss": 0.5194,
+      "step": 202
+    },
+    {
+      "epoch": 0.26887417218543047,
+      "grad_norm": 0.3861371290176906,
+      "learning_rate": 4.296296296296297e-05,
+      "loss": 0.518,
+      "step": 203
+    },
+    {
+      "epoch": 0.27019867549668874,
+      "grad_norm": 0.4659365868278326,
+      "learning_rate": 4.317460317460318e-05,
+      "loss": 0.5155,
+      "step": 204
+    },
+    {
+      "epoch": 0.271523178807947,
+      "grad_norm": 0.4383182855395808,
+      "learning_rate": 4.338624338624339e-05,
+      "loss": 0.5172,
+      "step": 205
+    },
+    {
+      "epoch": 0.2728476821192053,
+      "grad_norm": 0.34668399593076316,
+      "learning_rate": 4.3597883597883605e-05,
+      "loss": 0.5191,
+      "step": 206
+    },
+    {
+      "epoch": 0.27417218543046357,
+      "grad_norm": 0.402673182663992,
+      "learning_rate": 4.380952380952382e-05,
+      "loss": 0.5201,
+      "step": 207
+    },
+    {
+      "epoch": 0.27549668874172184,
+      "grad_norm": 0.38959233113190883,
+      "learning_rate": 4.4021164021164026e-05,
+      "loss": 0.5155,
+      "step": 208
+    },
+    {
+      "epoch": 0.2768211920529801,
+      "grad_norm": 0.3846885163904279,
+      "learning_rate": 4.423280423280424e-05,
+      "loss": 0.5201,
+      "step": 209
+    },
+    {
+      "epoch": 0.2781456953642384,
+      "grad_norm": 0.4389927887356917,
+      "learning_rate": 4.444444444444445e-05,
+      "loss": 0.5147,
+      "step": 210
+    },
+    {
+      "epoch": 0.27947019867549666,
+      "grad_norm": 0.5035182386923328,
+      "learning_rate": 4.465608465608467e-05,
+      "loss": 0.5155,
+      "step": 211
+    },
+    {
+      "epoch": 0.280794701986755,
+      "grad_norm": 0.5473158648406236,
+      "learning_rate": 4.486772486772487e-05,
+      "loss": 0.5192,
+      "step": 212
+    },
+    {
+      "epoch": 0.28211920529801326,
+      "grad_norm": 0.5234006659031107,
+      "learning_rate": 4.507936507936508e-05,
+      "loss": 0.5185,
+      "step": 213
+    },
+    {
+      "epoch": 0.28344370860927154,
+      "grad_norm": 0.605602743841271,
+      "learning_rate": 4.5291005291005295e-05,
+      "loss": 0.5198,
+      "step": 214
+    },
+    {
+      "epoch": 0.2847682119205298,
+      "grad_norm": 0.7406372968164502,
+      "learning_rate": 4.550264550264551e-05,
+      "loss": 0.5195,
+      "step": 215
+    },
+    {
+      "epoch": 0.2860927152317881,
+      "grad_norm": 0.8451588266749478,
+      "learning_rate": 4.5714285714285716e-05,
+      "loss": 0.521,
+      "step": 216
+    },
+    {
+      "epoch": 0.28741721854304636,
+      "grad_norm": 0.8608502983994097,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 0.5248,
+      "step": 217
+    },
+    {
+      "epoch": 0.28874172185430463,
+      "grad_norm": 0.7305767072027236,
+      "learning_rate": 4.6137566137566144e-05,
+      "loss": 0.5252,
+      "step": 218
+    },
+    {
+      "epoch": 0.2900662251655629,
+      "grad_norm": 0.5219867359737332,
+      "learning_rate": 4.634920634920636e-05,
+      "loss": 0.5186,
+      "step": 219
+    },
+    {
+      "epoch": 0.2913907284768212,
+      "grad_norm": 0.5128142888154852,
+      "learning_rate": 4.6560846560846565e-05,
+      "loss": 0.5172,
+      "step": 220
+    },
+    {
+      "epoch": 0.29271523178807946,
+      "grad_norm": 0.621755363641961,
+      "learning_rate": 4.677248677248678e-05,
+      "loss": 0.52,
+      "step": 221
+    },
+    {
+      "epoch": 0.29403973509933773,
+      "grad_norm": 0.4982632580075054,
+      "learning_rate": 4.698412698412699e-05,
+      "loss": 0.5182,
+      "step": 222
+    },
+    {
+      "epoch": 0.295364238410596,
+      "grad_norm": 0.4607373894585796,
+      "learning_rate": 4.7195767195767206e-05,
+      "loss": 0.5165,
+      "step": 223
+    },
+    {
+      "epoch": 0.29668874172185433,
+      "grad_norm": 0.5954643573534928,
+      "learning_rate": 4.7407407407407407e-05,
+      "loss": 0.5195,
+      "step": 224
+    },
+    {
+      "epoch": 0.2980132450331126,
+      "grad_norm": 0.47143652497116906,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.5149,
+      "step": 225
+    },
+    {
+      "epoch": 0.2993377483443709,
+      "grad_norm": 0.3614314553371073,
+      "learning_rate": 4.7830687830687834e-05,
+      "loss": 0.5169,
+      "step": 226
+    },
+    {
+      "epoch": 0.30066225165562915,
+      "grad_norm": 0.45160917557695274,
+      "learning_rate": 4.804232804232805e-05,
+      "loss": 0.5158,
+      "step": 227
+    },
+    {
+      "epoch": 0.3019867549668874,
+      "grad_norm": 0.4724241358780626,
+      "learning_rate": 4.8253968253968255e-05,
+      "loss": 0.5131,
+      "step": 228
+    },
+    {
+      "epoch": 0.3033112582781457,
+      "grad_norm": 0.3415899028519892,
+      "learning_rate": 4.846560846560847e-05,
+      "loss": 0.514,
+      "step": 229
+    },
+    {
+      "epoch": 0.304635761589404,
+      "grad_norm": 0.36493864326463016,
+      "learning_rate": 4.867724867724868e-05,
+      "loss": 0.5148,
+      "step": 230
+    },
+    {
+      "epoch": 0.30596026490066225,
+      "grad_norm": 0.3417222133508779,
+      "learning_rate": 4.88888888888889e-05,
+      "loss": 0.5142,
+      "step": 231
+    },
+    {
+      "epoch": 0.3072847682119205,
+      "grad_norm": 0.32908204107947103,
+      "learning_rate": 4.9100529100529104e-05,
+      "loss": 0.5165,
+      "step": 232
+    },
+    {
+      "epoch": 0.3086092715231788,
+      "grad_norm": 0.3870873732167372,
+      "learning_rate": 4.931216931216932e-05,
+      "loss": 0.5141,
+      "step": 233
+    },
+    {
+      "epoch": 0.30993377483443707,
+      "grad_norm": 0.34940866355409467,
+      "learning_rate": 4.952380952380953e-05,
+      "loss": 0.5147,
+      "step": 234
+    },
+    {
+      "epoch": 0.31125827814569534,
+      "grad_norm": 1.0655406546035853,
+      "learning_rate": 4.9735449735449745e-05,
+      "loss": 0.5137,
+      "step": 235
+    },
+    {
+      "epoch": 0.3125827814569536,
+      "grad_norm": 0.41728752546953346,
+      "learning_rate": 4.9947089947089946e-05,
+      "loss": 0.512,
+      "step": 236
+    },
+    {
+      "epoch": 0.31390728476821195,
+      "grad_norm": 0.7732229384625128,
+      "learning_rate": 5.015873015873016e-05,
+      "loss": 0.519,
+      "step": 237
+    },
+    {
+      "epoch": 0.3152317880794702,
+      "grad_norm": 1.2144498110566946,
+      "learning_rate": 5.037037037037037e-05,
+      "loss": 0.5254,
+      "step": 238
+    },
+    {
+      "epoch": 0.3165562913907285,
+      "grad_norm": 0.6852978548004667,
+      "learning_rate": 5.058201058201059e-05,
+      "loss": 0.5188,
+      "step": 239
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 0.7252922085852391,
+      "learning_rate": 5.0793650793650794e-05,
+      "loss": 0.5187,
+      "step": 240
+    },
+    {
+      "epoch": 0.31920529801324504,
+      "grad_norm": 0.9773163131615488,
+      "learning_rate": 5.100529100529101e-05,
+      "loss": 0.5275,
+      "step": 241
+    },
+    {
+      "epoch": 0.3205298013245033,
+      "grad_norm": 0.8427914929138572,
+      "learning_rate": 5.121693121693122e-05,
+      "loss": 0.5206,
+      "step": 242
+    },
+    {
+      "epoch": 0.3218543046357616,
+      "grad_norm": 0.8418769090770184,
+      "learning_rate": 5.1428571428571436e-05,
+      "loss": 0.5234,
+      "step": 243
+    },
+    {
+      "epoch": 0.32317880794701986,
+      "grad_norm": 0.903465991474773,
+      "learning_rate": 5.164021164021164e-05,
+      "loss": 0.5266,
+      "step": 244
+    },
+    {
+      "epoch": 0.32450331125827814,
+      "grad_norm": 0.5358343173603256,
+      "learning_rate": 5.1851851851851857e-05,
+      "loss": 0.5189,
+      "step": 245
+    },
+    {
+      "epoch": 0.3258278145695364,
+      "grad_norm": 0.9884981589024716,
+      "learning_rate": 5.206349206349207e-05,
+      "loss": 0.5248,
+      "step": 246
+    },
+    {
+      "epoch": 0.3271523178807947,
+      "grad_norm": 0.7616033634991722,
+      "learning_rate": 5.2275132275132284e-05,
+      "loss": 0.5285,
+      "step": 247
+    },
+    {
+      "epoch": 0.32847682119205296,
+      "grad_norm": 0.7512615684692353,
+      "learning_rate": 5.248677248677249e-05,
+      "loss": 0.5304,
+      "step": 248
+    },
+    {
+      "epoch": 0.32980132450331123,
+      "grad_norm": 0.6469267071344874,
+      "learning_rate": 5.2698412698412705e-05,
+      "loss": 0.5215,
+      "step": 249
+    },
+    {
+      "epoch": 0.33112582781456956,
+      "grad_norm": 1.0026232371558015,
+      "learning_rate": 5.291005291005292e-05,
+      "loss": 0.5337,
+      "step": 250
+    },
+    {
+      "epoch": 0.33245033112582784,
+      "grad_norm": 0.6303816738181705,
+      "learning_rate": 5.312169312169313e-05,
+      "loss": 0.5213,
+      "step": 251
+    },
+    {
+      "epoch": 0.3337748344370861,
+      "grad_norm": 0.6191683779703835,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.5228,
+      "step": 252
+    },
+    {
+      "epoch": 0.3350993377483444,
+      "grad_norm": 0.6646112030333788,
+      "learning_rate": 5.354497354497355e-05,
+      "loss": 0.5263,
+      "step": 253
+    },
+    {
+      "epoch": 0.33642384105960266,
+      "grad_norm": 0.4648142135408598,
+      "learning_rate": 5.375661375661376e-05,
+      "loss": 0.5207,
+      "step": 254
+    },
+    {
+      "epoch": 0.33774834437086093,
+      "grad_norm": 0.5116906707589919,
+      "learning_rate": 5.3968253968253975e-05,
+      "loss": 0.5188,
+      "step": 255
+    },
+    {
+      "epoch": 0.3390728476821192,
+      "grad_norm": 0.3844813139916701,
+      "learning_rate": 5.417989417989418e-05,
+      "loss": 0.5223,
+      "step": 256
+    },
+    {
+      "epoch": 0.3403973509933775,
+      "grad_norm": 0.46591098713720946,
+      "learning_rate": 5.4391534391534396e-05,
+      "loss": 0.5195,
+      "step": 257
+    },
+    {
+      "epoch": 0.34172185430463575,
+      "grad_norm": 0.3912414445330882,
+      "learning_rate": 5.460317460317461e-05,
+      "loss": 0.5169,
+      "step": 258
+    },
+    {
+      "epoch": 0.343046357615894,
+      "grad_norm": 0.3906332189138735,
+      "learning_rate": 5.481481481481482e-05,
+      "loss": 0.516,
+      "step": 259
+    },
+    {
+      "epoch": 0.3443708609271523,
+      "grad_norm": 0.47185887245698194,
+      "learning_rate": 5.502645502645503e-05,
+      "loss": 0.5163,
+      "step": 260
+    },
+    {
+      "epoch": 0.3456953642384106,
+      "grad_norm": 0.38191993716562667,
+      "learning_rate": 5.5238095238095244e-05,
+      "loss": 0.513,
+      "step": 261
+    },
+    {
+      "epoch": 0.34701986754966885,
+      "grad_norm": 0.3256383705843388,
+      "learning_rate": 5.544973544973546e-05,
+      "loss": 0.5128,
+      "step": 262
+    },
+    {
+      "epoch": 0.3483443708609272,
+      "grad_norm": 0.35465922664481025,
+      "learning_rate": 5.566137566137567e-05,
+      "loss": 0.514,
+      "step": 263
+    },
+    {
+      "epoch": 0.34966887417218545,
+      "grad_norm": 0.2907795660721273,
+      "learning_rate": 5.587301587301587e-05,
+      "loss": 0.5145,
+      "step": 264
+    },
+    {
+      "epoch": 0.3509933774834437,
+      "grad_norm": 0.3012852796626736,
+      "learning_rate": 5.6084656084656086e-05,
+      "loss": 0.5115,
+      "step": 265
+    },
+    {
+      "epoch": 0.352317880794702,
+      "grad_norm": 0.3035726997194574,
+      "learning_rate": 5.62962962962963e-05,
+      "loss": 0.5086,
+      "step": 266
+    },
+    {
+      "epoch": 0.3536423841059603,
+      "grad_norm": 0.2959368397779791,
+      "learning_rate": 5.6507936507936514e-05,
+      "loss": 0.5125,
+      "step": 267
+    },
+    {
+      "epoch": 0.35496688741721855,
+      "grad_norm": 0.3330216315798089,
+      "learning_rate": 5.671957671957672e-05,
+      "loss": 0.5113,
+      "step": 268
+    },
+    {
+      "epoch": 0.3562913907284768,
+      "grad_norm": 0.3389399390447598,
+      "learning_rate": 5.6931216931216935e-05,
+      "loss": 0.5121,
+      "step": 269
+    },
+    {
+      "epoch": 0.3576158940397351,
+      "grad_norm": 0.2651097591160793,
+      "learning_rate": 5.714285714285715e-05,
+      "loss": 0.5096,
+      "step": 270
+    },
+    {
+      "epoch": 0.35894039735099337,
+      "grad_norm": 0.2784493380439868,
+      "learning_rate": 5.735449735449736e-05,
+      "loss": 0.5089,
+      "step": 271
+    },
+    {
+      "epoch": 0.36026490066225164,
+      "grad_norm": 0.2498082137858785,
+      "learning_rate": 5.756613756613757e-05,
+      "loss": 0.5089,
+      "step": 272
+    },
+    {
+      "epoch": 0.3615894039735099,
+      "grad_norm": 0.339742361458311,
+      "learning_rate": 5.777777777777778e-05,
+      "loss": 0.5091,
+      "step": 273
+    },
+    {
+      "epoch": 0.3629139072847682,
+      "grad_norm": 0.38127802570027813,
+      "learning_rate": 5.7989417989418e-05,
+      "loss": 0.5092,
+      "step": 274
+    },
+    {
+      "epoch": 0.36423841059602646,
+      "grad_norm": 0.37835432083820525,
+      "learning_rate": 5.820105820105821e-05,
+      "loss": 0.5095,
+      "step": 275
+    },
+    {
+      "epoch": 0.3655629139072848,
+      "grad_norm": 0.4866102237236228,
+      "learning_rate": 5.841269841269841e-05,
+      "loss": 0.5111,
+      "step": 276
+    },
+    {
+      "epoch": 0.36688741721854307,
+      "grad_norm": 0.545230019387493,
+      "learning_rate": 5.8624338624338625e-05,
+      "loss": 0.5104,
+      "step": 277
+    },
+    {
+      "epoch": 0.36821192052980134,
+      "grad_norm": 0.7166326547515052,
+      "learning_rate": 5.883597883597884e-05,
+      "loss": 0.5113,
+      "step": 278
+    },
+    {
+      "epoch": 0.3695364238410596,
+      "grad_norm": 0.8870882718425024,
+      "learning_rate": 5.904761904761905e-05,
+      "loss": 0.5142,
+      "step": 279
+    },
+    {
+      "epoch": 0.3708609271523179,
+      "grad_norm": 0.888437933804128,
+      "learning_rate": 5.925925925925926e-05,
+      "loss": 0.5145,
+      "step": 280
+    },
+    {
+      "epoch": 0.37218543046357616,
+      "grad_norm": 0.672767225377524,
+      "learning_rate": 5.9470899470899474e-05,
+      "loss": 0.5098,
+      "step": 281
+    },
+    {
+      "epoch": 0.37350993377483444,
+      "grad_norm": 0.5139833553764804,
+      "learning_rate": 5.968253968253969e-05,
+      "loss": 0.513,
+      "step": 282
+    },
+    {
+      "epoch": 0.3748344370860927,
+      "grad_norm": 0.5724485628892855,
+      "learning_rate": 5.98941798941799e-05,
+      "loss": 0.514,
+      "step": 283
+    },
+    {
+      "epoch": 0.376158940397351,
+      "grad_norm": 0.4523675060129603,
+      "learning_rate": 6.010582010582011e-05,
+      "loss": 0.5073,
+      "step": 284
+    },
+    {
+      "epoch": 0.37748344370860926,
+      "grad_norm": 0.37474590094327487,
+      "learning_rate": 6.031746031746032e-05,
+      "loss": 0.5101,
+      "step": 285
+    },
+    {
+      "epoch": 0.37880794701986753,
+      "grad_norm": 0.5455934283566426,
+      "learning_rate": 6.0529100529100536e-05,
+      "loss": 0.5111,
+      "step": 286
+    },
+    {
+      "epoch": 0.3801324503311258,
+      "grad_norm": 0.45059389765447394,
+      "learning_rate": 6.074074074074075e-05,
+      "loss": 0.5032,
+      "step": 287
+    },
+    {
+      "epoch": 0.38145695364238413,
+      "grad_norm": 0.3230700979247049,
+      "learning_rate": 6.095238095238096e-05,
+      "loss": 0.5082,
+      "step": 288
+    },
+    {
+      "epoch": 0.3827814569536424,
+      "grad_norm": 0.5080167498929623,
+      "learning_rate": 6.116402116402116e-05,
+      "loss": 0.5081,
+      "step": 289
+    },
+    {
+      "epoch": 0.3841059602649007,
+      "grad_norm": 0.5228541909457616,
+      "learning_rate": 6.137566137566138e-05,
+      "loss": 0.5092,
+      "step": 290
+    },
+    {
+      "epoch": 0.38543046357615895,
+      "grad_norm": 0.4439239728716328,
+      "learning_rate": 6.158730158730159e-05,
+      "loss": 0.5059,
+      "step": 291
+    },
+    {
+      "epoch": 0.38675496688741723,
+      "grad_norm": 0.442126516899406,
+      "learning_rate": 6.17989417989418e-05,
+      "loss": 0.5066,
+      "step": 292
+    },
+    {
+      "epoch": 0.3880794701986755,
+      "grad_norm": 0.4508294243389612,
+      "learning_rate": 6.201058201058202e-05,
+      "loss": 0.5093,
+      "step": 293
+    },
+    {
+      "epoch": 0.3894039735099338,
+      "grad_norm": 0.435269288616384,
+      "learning_rate": 6.222222222222223e-05,
+      "loss": 0.5055,
+      "step": 294
+    },
+    {
+      "epoch": 0.39072847682119205,
+      "grad_norm": 0.44869249196988187,
+      "learning_rate": 6.243386243386245e-05,
+      "loss": 0.5062,
+      "step": 295
+    },
+    {
+      "epoch": 0.3920529801324503,
+      "grad_norm": 0.4556040194150797,
+      "learning_rate": 6.264550264550265e-05,
+      "loss": 0.5086,
+      "step": 296
+    },
+    {
+      "epoch": 0.3933774834437086,
+      "grad_norm": 0.6217940114015377,
+      "learning_rate": 6.285714285714286e-05,
+      "loss": 0.5065,
+      "step": 297
+    },
+    {
+      "epoch": 0.39470198675496687,
+      "grad_norm": 0.6995120649356269,
+      "learning_rate": 6.306878306878308e-05,
+      "loss": 0.5056,
+      "step": 298
+    },
+    {
+      "epoch": 0.39602649006622515,
+      "grad_norm": 0.6971135186955677,
+      "learning_rate": 6.328042328042329e-05,
+      "loss": 0.5082,
+      "step": 299
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 0.6851325389052091,
+      "learning_rate": 6.349206349206349e-05,
+      "loss": 0.5113,
+      "step": 300
+    },
+    {
+      "epoch": 0.39867549668874175,
+      "grad_norm": 0.6566779948080963,
+      "learning_rate": 6.37037037037037e-05,
+      "loss": 0.5073,
+      "step": 301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.518759749795198,
+      "learning_rate": 6.391534391534392e-05,
+      "loss": 0.5062,
+      "step": 302
+    },
+    {
+      "epoch": 0.4013245033112583,
+      "grad_norm": 0.4367713450500494,
+      "learning_rate": 6.412698412698413e-05,
+      "loss": 0.5104,
+      "step": 303
+    },
+    {
+      "epoch": 0.40264900662251657,
+      "grad_norm": 0.4795521519506873,
+      "learning_rate": 6.433862433862434e-05,
+      "loss": 0.5139,
+      "step": 304
+    },
+    {
+      "epoch": 0.40397350993377484,
+      "grad_norm": 0.544116171375792,
+      "learning_rate": 6.455026455026456e-05,
+      "loss": 0.506,
+      "step": 305
+    },
+    {
+      "epoch": 0.4052980132450331,
+      "grad_norm": 0.47216403988926897,
+      "learning_rate": 6.476190476190477e-05,
+      "loss": 0.5051,
+      "step": 306
+    },
+    {
+      "epoch": 0.4066225165562914,
+      "grad_norm": 0.3591470229326716,
+      "learning_rate": 6.497354497354499e-05,
+      "loss": 0.5055,
+      "step": 307
+    },
+    {
+      "epoch": 0.40794701986754967,
+      "grad_norm": 0.33667631873732506,
+      "learning_rate": 6.518518518518519e-05,
+      "loss": 0.5056,
+      "step": 308
+    },
+    {
+      "epoch": 0.40927152317880794,
+      "grad_norm": 0.40713826264477077,
+      "learning_rate": 6.53968253968254e-05,
+      "loss": 0.5026,
+      "step": 309
+    },
+    {
+      "epoch": 0.4105960264900662,
+      "grad_norm": 0.37079416307656116,
+      "learning_rate": 6.560846560846561e-05,
+      "loss": 0.5046,
+      "step": 310
+    },
+    {
+      "epoch": 0.4119205298013245,
+      "grad_norm": 0.32167079299906054,
+      "learning_rate": 6.582010582010583e-05,
+      "loss": 0.5011,
+      "step": 311
+    },
+    {
+      "epoch": 0.41324503311258276,
+      "grad_norm": 0.4472773731902467,
+      "learning_rate": 6.603174603174603e-05,
+      "loss": 0.5017,
+      "step": 312
+    },
+    {
+      "epoch": 0.41456953642384103,
+      "grad_norm": 0.6254574310034611,
+      "learning_rate": 6.624338624338624e-05,
+      "loss": 0.5058,
+      "step": 313
+    },
+    {
+      "epoch": 0.41589403973509936,
+      "grad_norm": 0.6801700917980404,
+      "learning_rate": 6.645502645502646e-05,
+      "loss": 0.5032,
+      "step": 314
+    },
+    {
+      "epoch": 0.41721854304635764,
+      "grad_norm": 0.7147095103016862,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.511,
+      "step": 315
+    },
+    {
+      "epoch": 0.4185430463576159,
+      "grad_norm": 0.7733912932918299,
+      "learning_rate": 6.687830687830688e-05,
+      "loss": 0.5068,
+      "step": 316
+    },
+    {
+      "epoch": 0.4198675496688742,
+      "grad_norm": 0.7542893498446464,
+      "learning_rate": 6.70899470899471e-05,
+      "loss": 0.5079,
+      "step": 317
+    },
+    {
+      "epoch": 0.42119205298013246,
+      "grad_norm": 0.5804327943229926,
+      "learning_rate": 6.730158730158731e-05,
+      "loss": 0.5035,
+      "step": 318
+    },
+    {
+      "epoch": 0.42251655629139073,
+      "grad_norm": 0.3607882637348594,
+      "learning_rate": 6.751322751322753e-05,
+      "loss": 0.5042,
+      "step": 319
+    },
+    {
+      "epoch": 0.423841059602649,
+      "grad_norm": 0.5057984489574802,
+      "learning_rate": 6.772486772486773e-05,
+      "loss": 0.5086,
+      "step": 320
+    },
+    {
+      "epoch": 0.4251655629139073,
+      "grad_norm": 0.536235946743846,
+      "learning_rate": 6.793650793650794e-05,
+      "loss": 0.5056,
+      "step": 321
+    },
+    {
+      "epoch": 0.42649006622516555,
+      "grad_norm": 0.30237586465760763,
+      "learning_rate": 6.814814814814815e-05,
+      "loss": 0.5046,
+      "step": 322
+    },
+    {
+      "epoch": 0.4278145695364238,
+      "grad_norm": 0.4476954154125454,
+      "learning_rate": 6.835978835978837e-05,
+      "loss": 0.5023,
+      "step": 323
+    },
+    {
+      "epoch": 0.4291390728476821,
+      "grad_norm": 0.4277627583942506,
+      "learning_rate": 6.857142857142857e-05,
+      "loss": 0.5051,
+      "step": 324
+    },
+    {
+      "epoch": 0.4304635761589404,
+      "grad_norm": 0.30253858180041804,
+      "learning_rate": 6.878306878306878e-05,
+      "loss": 0.5015,
+      "step": 325
+    },
+    {
+      "epoch": 0.43178807947019865,
+      "grad_norm": 0.4491283353899929,
+      "learning_rate": 6.8994708994709e-05,
+      "loss": 0.5053,
+      "step": 326
+    },
+    {
+      "epoch": 0.433112582781457,
+      "grad_norm": 0.4719953236712367,
+      "learning_rate": 6.920634920634921e-05,
+      "loss": 0.5062,
+      "step": 327
+    },
+    {
+      "epoch": 0.43443708609271525,
+      "grad_norm": 0.4832026786281093,
+      "learning_rate": 6.941798941798942e-05,
+      "loss": 0.5051,
+      "step": 328
+    },
+    {
+      "epoch": 0.4357615894039735,
+      "grad_norm": 0.5150329401976371,
+      "learning_rate": 6.962962962962964e-05,
+      "loss": 0.5042,
+      "step": 329
+    },
+    {
+      "epoch": 0.4370860927152318,
+      "grad_norm": 0.6312368437654068,
+      "learning_rate": 6.984126984126985e-05,
+      "loss": 0.5118,
+      "step": 330
+    },
+    {
+      "epoch": 0.4384105960264901,
+      "grad_norm": 0.785752368369056,
+      "learning_rate": 7.005291005291006e-05,
+      "loss": 0.5105,
+      "step": 331
+    },
+    {
+      "epoch": 0.43973509933774835,
+      "grad_norm": 0.7078840070207824,
+      "learning_rate": 7.026455026455026e-05,
+      "loss": 0.5109,
+      "step": 332
+    },
+    {
+      "epoch": 0.4410596026490066,
+      "grad_norm": 0.5151313519024922,
+      "learning_rate": 7.047619047619048e-05,
+      "loss": 0.5046,
+      "step": 333
+    },
+    {
+      "epoch": 0.4423841059602649,
+      "grad_norm": 0.5608709489073054,
+      "learning_rate": 7.068783068783069e-05,
+      "loss": 0.5076,
+      "step": 334
+    },
+    {
+      "epoch": 0.44370860927152317,
+      "grad_norm": 0.6565043823482688,
+      "learning_rate": 7.08994708994709e-05,
+      "loss": 0.5049,
+      "step": 335
+    },
+    {
+      "epoch": 0.44503311258278144,
+      "grad_norm": 0.5674549562908847,
+      "learning_rate": 7.11111111111111e-05,
+      "loss": 0.5035,
+      "step": 336
+    },
+    {
+      "epoch": 0.4463576158940397,
+      "grad_norm": 0.43841238098351115,
+      "learning_rate": 7.132275132275132e-05,
+      "loss": 0.5059,
+      "step": 337
+    },
+    {
+      "epoch": 0.447682119205298,
+      "grad_norm": 0.4432039768635806,
+      "learning_rate": 7.153439153439153e-05,
+      "loss": 0.5054,
+      "step": 338
+    },
+    {
+      "epoch": 0.44900662251655626,
+      "grad_norm": 0.4533836431126359,
+      "learning_rate": 7.174603174603175e-05,
+      "loss": 0.5009,
+      "step": 339
+    },
+    {
+      "epoch": 0.4503311258278146,
+      "grad_norm": 0.4977573105299979,
+      "learning_rate": 7.195767195767196e-05,
+      "loss": 0.5081,
+      "step": 340
+    },
+    {
+      "epoch": 0.45165562913907287,
+      "grad_norm": 0.6529301433154492,
+      "learning_rate": 7.216931216931218e-05,
+      "loss": 0.507,
+      "step": 341
+    },
+    {
+      "epoch": 0.45298013245033114,
+      "grad_norm": 0.7022925001002475,
+      "learning_rate": 7.238095238095239e-05,
+      "loss": 0.5054,
+      "step": 342
+    },
+    {
+      "epoch": 0.4543046357615894,
+      "grad_norm": 0.6056227146468415,
+      "learning_rate": 7.25925925925926e-05,
+      "loss": 0.5062,
+      "step": 343
+    },
+    {
+      "epoch": 0.4556291390728477,
+      "grad_norm": 0.5327503047210377,
+      "learning_rate": 7.28042328042328e-05,
+      "loss": 0.5035,
+      "step": 344
+    },
+    {
+      "epoch": 0.45695364238410596,
+      "grad_norm": 0.48745684108137133,
+      "learning_rate": 7.301587301587302e-05,
+      "loss": 0.5031,
+      "step": 345
+    },
+    {
+      "epoch": 0.45827814569536424,
+      "grad_norm": 0.3940793656347632,
+      "learning_rate": 7.322751322751323e-05,
+      "loss": 0.504,
+      "step": 346
+    },
+    {
+      "epoch": 0.4596026490066225,
+      "grad_norm": 0.4212626635310372,
+      "learning_rate": 7.343915343915345e-05,
+      "loss": 0.5061,
+      "step": 347
+    },
+    {
+      "epoch": 0.4609271523178808,
+      "grad_norm": 0.41018063734951116,
+      "learning_rate": 7.365079365079366e-05,
+      "loss": 0.5066,
+      "step": 348
+    },
+    {
+      "epoch": 0.46225165562913906,
+      "grad_norm": 0.37658624720053224,
+      "learning_rate": 7.386243386243387e-05,
+      "loss": 0.504,
+      "step": 349
+    },
+    {
+      "epoch": 0.46357615894039733,
+      "grad_norm": 0.4445430627969012,
+      "learning_rate": 7.407407407407409e-05,
+      "loss": 0.5004,
+      "step": 350
+    },
+    {
+      "epoch": 0.4649006622516556,
+      "grad_norm": 0.4185383976437788,
+      "learning_rate": 7.42857142857143e-05,
+      "loss": 0.5009,
+      "step": 351
+    },
+    {
+      "epoch": 0.46622516556291393,
+      "grad_norm": 0.2964127118828609,
+      "learning_rate": 7.44973544973545e-05,
+      "loss": 0.5023,
+      "step": 352
+    },
+    {
+      "epoch": 0.4675496688741722,
+      "grad_norm": 0.30814673837876816,
+      "learning_rate": 7.470899470899471e-05,
+      "loss": 0.5034,
+      "step": 353
+    },
+    {
+      "epoch": 0.4688741721854305,
+      "grad_norm": 0.32122310301415125,
+      "learning_rate": 7.492063492063493e-05,
+      "loss": 0.5028,
+      "step": 354
+    },
+    {
+      "epoch": 0.47019867549668876,
+      "grad_norm": 0.38137228592452643,
+      "learning_rate": 7.513227513227514e-05,
+      "loss": 0.5033,
+      "step": 355
+    },
+    {
+      "epoch": 0.47152317880794703,
+      "grad_norm": 0.5002872812284009,
+      "learning_rate": 7.534391534391534e-05,
+      "loss": 0.5008,
+      "step": 356
+    },
+    {
+      "epoch": 0.4728476821192053,
+      "grad_norm": 0.5959526980771447,
+      "learning_rate": 7.555555555555556e-05,
+      "loss": 0.5003,
+      "step": 357
+    },
+    {
+      "epoch": 0.4741721854304636,
+      "grad_norm": 0.6760648306744391,
+      "learning_rate": 7.576719576719577e-05,
+      "loss": 0.5036,
+      "step": 358
+    },
+    {
+      "epoch": 0.47549668874172185,
+      "grad_norm": 0.7309286158818808,
+      "learning_rate": 7.597883597883598e-05,
+      "loss": 0.5095,
+      "step": 359
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 0.7680989242671499,
+      "learning_rate": 7.61904761904762e-05,
+      "loss": 0.5099,
+      "step": 360
+    },
+    {
+      "epoch": 0.4781456953642384,
+      "grad_norm": 0.7400650742913737,
+      "learning_rate": 7.640211640211641e-05,
+      "loss": 0.5051,
+      "step": 361
+    },
+    {
+      "epoch": 0.4794701986754967,
+      "grad_norm": 0.6241168901560405,
+      "learning_rate": 7.661375661375663e-05,
+      "loss": 0.5041,
+      "step": 362
+    },
+    {
+      "epoch": 0.48079470198675495,
+      "grad_norm": 0.5681956255488533,
+      "learning_rate": 7.682539682539684e-05,
+      "loss": 0.5089,
+      "step": 363
+    },
+    {
+      "epoch": 0.4821192052980132,
+      "grad_norm": 0.47090240359075985,
+      "learning_rate": 7.703703703703704e-05,
+      "loss": 0.5026,
+      "step": 364
+    },
+    {
+      "epoch": 0.48344370860927155,
+      "grad_norm": 0.360569154240506,
+      "learning_rate": 7.724867724867725e-05,
+      "loss": 0.5006,
+      "step": 365
+    },
+    {
+      "epoch": 0.4847682119205298,
+      "grad_norm": 0.4224931656095163,
+      "learning_rate": 7.746031746031747e-05,
+      "loss": 0.5066,
+      "step": 366
+    },
+    {
+      "epoch": 0.4860927152317881,
+      "grad_norm": 0.46306143341023753,
+      "learning_rate": 7.767195767195768e-05,
+      "loss": 0.4981,
+      "step": 367
+    },
+    {
+      "epoch": 0.48741721854304637,
+      "grad_norm": 0.4695318334227369,
+      "learning_rate": 7.788359788359788e-05,
+      "loss": 0.4984,
+      "step": 368
+    },
+    {
+      "epoch": 0.48874172185430464,
+      "grad_norm": 0.45900498464532535,
+      "learning_rate": 7.80952380952381e-05,
+      "loss": 0.5037,
+      "step": 369
+    },
+    {
+      "epoch": 0.4900662251655629,
+      "grad_norm": 0.3985807198091737,
+      "learning_rate": 7.830687830687831e-05,
+      "loss": 0.4977,
+      "step": 370
+    },
+    {
+      "epoch": 0.4913907284768212,
+      "grad_norm": 0.3996571986231609,
+      "learning_rate": 7.851851851851852e-05,
+      "loss": 0.4997,
+      "step": 371
+    },
+    {
+      "epoch": 0.49271523178807947,
+      "grad_norm": 0.3951161429228177,
+      "learning_rate": 7.873015873015874e-05,
+      "loss": 0.4998,
+      "step": 372
+    },
+    {
+      "epoch": 0.49403973509933774,
+      "grad_norm": 0.40028195672492733,
+      "learning_rate": 7.894179894179895e-05,
+      "loss": 0.5,
+      "step": 373
+    },
+    {
+      "epoch": 0.495364238410596,
+      "grad_norm": 0.4652092105480949,
+      "learning_rate": 7.915343915343916e-05,
+      "loss": 0.4952,
+      "step": 374
+    },
+    {
+      "epoch": 0.4966887417218543,
+      "grad_norm": 0.47248481411223076,
+      "learning_rate": 7.936507936507938e-05,
+      "loss": 0.4977,
+      "step": 375
+    },
+    {
+      "epoch": 0.49801324503311256,
+      "grad_norm": 0.47280504777523125,
+      "learning_rate": 7.957671957671958e-05,
+      "loss": 0.5041,
+      "step": 376
+    },
+    {
+      "epoch": 0.49933774834437084,
+      "grad_norm": 0.5355218930413733,
+      "learning_rate": 7.978835978835979e-05,
+      "loss": 0.5008,
+      "step": 377
+    },
+    {
+      "epoch": 0.5006622516556292,
+      "grad_norm": 0.5758243072784106,
+      "learning_rate": 8e-05,
+      "loss": 0.4976,
+      "step": 378
+    },
+    {
+      "epoch": 0.5019867549668874,
+      "grad_norm": 0.5599460929386382,
+      "learning_rate": 7.999998289438802e-05,
+      "loss": 0.5058,
+      "step": 379
+    },
+    {
+      "epoch": 0.5033112582781457,
+      "grad_norm": 0.5113331459011213,
+      "learning_rate": 7.999993157756672e-05,
+      "loss": 0.5002,
+      "step": 380
+    },
+    {
+      "epoch": 0.5046357615894039,
+      "grad_norm": 0.5437063876012365,
+      "learning_rate": 7.999984604957997e-05,
+      "loss": 0.5008,
+      "step": 381
+    },
+    {
+      "epoch": 0.5059602649006623,
+      "grad_norm": 0.5666340978735532,
+      "learning_rate": 7.999972631050092e-05,
+      "loss": 0.4992,
+      "step": 382
+    },
+    {
+      "epoch": 0.5072847682119205,
+      "grad_norm": 0.568351147059187,
+      "learning_rate": 7.999957236043202e-05,
+      "loss": 0.4986,
+      "step": 383
+    },
+    {
+      "epoch": 0.5086092715231788,
+      "grad_norm": 0.5767225260476471,
+      "learning_rate": 7.99993841995049e-05,
+      "loss": 0.5053,
+      "step": 384
+    },
+    {
+      "epoch": 0.5099337748344371,
+      "grad_norm": 0.5548557931339406,
+      "learning_rate": 7.999916182788049e-05,
+      "loss": 0.5005,
+      "step": 385
+    },
+    {
+      "epoch": 0.5112582781456954,
+      "grad_norm": 0.4951465769561094,
+      "learning_rate": 7.999890524574901e-05,
+      "loss": 0.5047,
+      "step": 386
+    },
+    {
+      "epoch": 0.5125827814569537,
+      "grad_norm": 0.4003749160877621,
+      "learning_rate": 7.999861445332988e-05,
+      "loss": 0.4999,
+      "step": 387
+    },
+    {
+      "epoch": 0.5139072847682119,
+      "grad_norm": 0.4417790405429633,
+      "learning_rate": 7.999828945087182e-05,
+      "loss": 0.4993,
+      "step": 388
+    },
+    {
+      "epoch": 0.5152317880794702,
+      "grad_norm": 0.422963728081838,
+      "learning_rate": 7.99979302386528e-05,
+      "loss": 0.4988,
+      "step": 389
+    },
+    {
+      "epoch": 0.5165562913907285,
+      "grad_norm": 0.3449239339862523,
+      "learning_rate": 7.999753681698006e-05,
+      "loss": 0.4957,
+      "step": 390
+    },
+    {
+      "epoch": 0.5178807947019868,
+      "grad_norm": 0.35609189708353833,
+      "learning_rate": 7.999710918619006e-05,
+      "loss": 0.4938,
+      "step": 391
+    },
+    {
+      "epoch": 0.519205298013245,
+      "grad_norm": 0.36330343619003014,
+      "learning_rate": 7.999664734664857e-05,
+      "loss": 0.4955,
+      "step": 392
+    },
+    {
+      "epoch": 0.5205298013245033,
+      "grad_norm": 0.2939792369324543,
+      "learning_rate": 7.999615129875057e-05,
+      "loss": 0.4976,
+      "step": 393
+    },
+    {
+      "epoch": 0.5218543046357615,
+      "grad_norm": 0.26966484025953474,
+      "learning_rate": 7.999562104292033e-05,
+      "loss": 0.4971,
+      "step": 394
+    },
+    {
+      "epoch": 0.5231788079470199,
+      "grad_norm": 0.3338156891825692,
+      "learning_rate": 7.999505657961137e-05,
+      "loss": 0.5004,
+      "step": 395
+    },
+    {
+      "epoch": 0.5245033112582781,
+      "grad_norm": 0.36209020756528865,
+      "learning_rate": 7.999445790930645e-05,
+      "loss": 0.4955,
+      "step": 396
+    },
+    {
+      "epoch": 0.5258278145695364,
+      "grad_norm": 0.39252259520406285,
+      "learning_rate": 7.999382503251764e-05,
+      "loss": 0.4994,
+      "step": 397
+    },
+    {
+      "epoch": 0.5271523178807948,
+      "grad_norm": 0.37008003273962986,
+      "learning_rate": 7.999315794978618e-05,
+      "loss": 0.4998,
+      "step": 398
+    },
+    {
+      "epoch": 0.528476821192053,
+      "grad_norm": 0.3830770643300346,
+      "learning_rate": 7.999245666168265e-05,
+      "loss": 0.4977,
+      "step": 399
+    },
+    {
+      "epoch": 0.5298013245033113,
+      "grad_norm": 0.5394619114110106,
+      "learning_rate": 7.99917211688068e-05,
+      "loss": 0.4969,
+      "step": 400
+    },
+    {
+      "epoch": 0.5311258278145695,
+      "grad_norm": 0.6958439970092543,
+      "learning_rate": 7.999095147178774e-05,
+      "loss": 0.4966,
+      "step": 401
+    },
+    {
+      "epoch": 0.5324503311258278,
+      "grad_norm": 0.8523881511439382,
+      "learning_rate": 7.999014757128374e-05,
+      "loss": 0.5007,
+      "step": 402
+    },
+    {
+      "epoch": 0.5337748344370861,
+      "grad_norm": 0.6673807186503627,
+      "learning_rate": 7.998930946798238e-05,
+      "loss": 0.495,
+      "step": 403
+    },
+    {
+      "epoch": 0.5350993377483444,
+      "grad_norm": 0.3914729152093163,
+      "learning_rate": 7.998843716260047e-05,
+      "loss": 0.4959,
+      "step": 404
+    },
+    {
+      "epoch": 0.5364238410596026,
+      "grad_norm": 0.6033660561570864,
+      "learning_rate": 7.998753065588405e-05,
+      "loss": 0.4978,
+      "step": 405
+    },
+    {
+      "epoch": 0.5377483443708609,
+      "grad_norm": 0.7659583145407987,
+      "learning_rate": 7.998658994860848e-05,
+      "loss": 0.5012,
+      "step": 406
+    },
+    {
+      "epoch": 0.5390728476821192,
+      "grad_norm": 0.5849288906436081,
+      "learning_rate": 7.998561504157828e-05,
+      "loss": 0.4969,
+      "step": 407
+    },
+    {
+      "epoch": 0.5403973509933775,
+      "grad_norm": 0.39592099887297716,
+      "learning_rate": 7.998460593562732e-05,
+      "loss": 0.5022,
+      "step": 408
+    },
+    {
+      "epoch": 0.5417218543046357,
+      "grad_norm": 0.5282940083414958,
+      "learning_rate": 7.998356263161863e-05,
+      "loss": 0.497,
+      "step": 409
+    },
+    {
+      "epoch": 0.543046357615894,
+      "grad_norm": 0.5757130702489279,
+      "learning_rate": 7.998248513044456e-05,
+      "loss": 0.4957,
+      "step": 410
+    },
+    {
+      "epoch": 0.5443708609271524,
+      "grad_norm": 0.48447098561749147,
+      "learning_rate": 7.998137343302665e-05,
+      "loss": 0.5008,
+      "step": 411
+    },
+    {
+      "epoch": 0.5456953642384106,
+      "grad_norm": 0.4129861473523075,
+      "learning_rate": 7.998022754031572e-05,
+      "loss": 0.4982,
+      "step": 412
+    },
+    {
+      "epoch": 0.5470198675496689,
+      "grad_norm": 0.3907525787992232,
+      "learning_rate": 7.997904745329183e-05,
+      "loss": 0.5015,
+      "step": 413
+    },
+    {
+      "epoch": 0.5483443708609271,
+      "grad_norm": 0.3893516053534045,
+      "learning_rate": 7.997783317296429e-05,
+      "loss": 0.4994,
+      "step": 414
+    },
+    {
+      "epoch": 0.5496688741721855,
+      "grad_norm": 0.38448859535898816,
+      "learning_rate": 7.997658470037165e-05,
+      "loss": 0.4958,
+      "step": 415
+    },
+    {
+      "epoch": 0.5509933774834437,
+      "grad_norm": 0.3385521636986559,
+      "learning_rate": 7.997530203658169e-05,
+      "loss": 0.4976,
+      "step": 416
+    },
+    {
+      "epoch": 0.552317880794702,
+      "grad_norm": 0.35422930654658397,
+      "learning_rate": 7.997398518269148e-05,
+      "loss": 0.4993,
+      "step": 417
+    },
+    {
+      "epoch": 0.5536423841059602,
+      "grad_norm": 0.32944486339936824,
+      "learning_rate": 7.997263413982726e-05,
+      "loss": 0.4971,
+      "step": 418
+    },
+    {
+      "epoch": 0.5549668874172186,
+      "grad_norm": 0.2615349851701184,
+      "learning_rate": 7.997124890914458e-05,
+      "loss": 0.4916,
+      "step": 419
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 0.2780751111622494,
+      "learning_rate": 7.996982949182818e-05,
+      "loss": 0.4975,
+      "step": 420
+    },
+    {
+      "epoch": 0.5576158940397351,
+      "grad_norm": 0.28593478791451615,
+      "learning_rate": 7.996837588909207e-05,
+      "loss": 0.4973,
+      "step": 421
+    },
+    {
+      "epoch": 0.5589403973509933,
+      "grad_norm": 0.2643741194291814,
+      "learning_rate": 7.996688810217951e-05,
+      "loss": 0.4945,
+      "step": 422
+    },
+    {
+      "epoch": 0.5602649006622517,
+      "grad_norm": 0.26956051645184437,
+      "learning_rate": 7.996536613236294e-05,
+      "loss": 0.494,
+      "step": 423
+    },
+    {
+      "epoch": 0.56158940397351,
+      "grad_norm": 0.3143605283448494,
+      "learning_rate": 7.996380998094409e-05,
+      "loss": 0.4918,
+      "step": 424
+    },
+    {
+      "epoch": 0.5629139072847682,
+      "grad_norm": 0.4065675237246672,
+      "learning_rate": 7.996221964925388e-05,
+      "loss": 0.4956,
+      "step": 425
+    },
+    {
+      "epoch": 0.5642384105960265,
+      "grad_norm": 0.44375027366269765,
+      "learning_rate": 7.996059513865254e-05,
+      "loss": 0.4958,
+      "step": 426
+    },
+    {
+      "epoch": 0.5655629139072847,
+      "grad_norm": 0.4409376905831458,
+      "learning_rate": 7.995893645052943e-05,
+      "loss": 0.4983,
+      "step": 427
+    },
+    {
+      "epoch": 0.5668874172185431,
+      "grad_norm": 0.41011031638078005,
+      "learning_rate": 7.995724358630325e-05,
+      "loss": 0.4998,
+      "step": 428
+    },
+    {
+      "epoch": 0.5682119205298013,
+      "grad_norm": 0.4059765432616618,
+      "learning_rate": 7.995551654742182e-05,
+      "loss": 0.4971,
+      "step": 429
+    },
+    {
+      "epoch": 0.5695364238410596,
+      "grad_norm": 0.35029597061990575,
+      "learning_rate": 7.995375533536226e-05,
+      "loss": 0.4924,
+      "step": 430
+    },
+    {
+      "epoch": 0.5708609271523178,
+      "grad_norm": 0.2761262420621809,
+      "learning_rate": 7.995195995163092e-05,
+      "loss": 0.4969,
+      "step": 431
+    },
+    {
+      "epoch": 0.5721854304635762,
+      "grad_norm": 0.26888530329987675,
+      "learning_rate": 7.995013039776333e-05,
+      "loss": 0.492,
+      "step": 432
+    },
+    {
+      "epoch": 0.5735099337748344,
+      "grad_norm": 0.3018334521642464,
+      "learning_rate": 7.994826667532428e-05,
+      "loss": 0.4973,
+      "step": 433
+    },
+    {
+      "epoch": 0.5748344370860927,
+      "grad_norm": 0.3402039421622009,
+      "learning_rate": 7.994636878590779e-05,
+      "loss": 0.4944,
+      "step": 434
+    },
+    {
+      "epoch": 0.5761589403973509,
+      "grad_norm": 0.38649933104507933,
+      "learning_rate": 7.994443673113706e-05,
+      "loss": 0.4965,
+      "step": 435
+    },
+    {
+      "epoch": 0.5774834437086093,
+      "grad_norm": 0.4259953682568037,
+      "learning_rate": 7.994247051266457e-05,
+      "loss": 0.4975,
+      "step": 436
+    },
+    {
+      "epoch": 0.5788079470198676,
+      "grad_norm": 0.5341579750182491,
+      "learning_rate": 7.994047013217197e-05,
+      "loss": 0.4927,
+      "step": 437
+    },
+    {
+      "epoch": 0.5801324503311258,
+      "grad_norm": 0.6090751107999659,
+      "learning_rate": 7.993843559137016e-05,
+      "loss": 0.4965,
+      "step": 438
+    },
+    {
+      "epoch": 0.5814569536423841,
+      "grad_norm": 0.5476728119143996,
+      "learning_rate": 7.993636689199923e-05,
+      "loss": 0.4988,
+      "step": 439
+    },
+    {
+      "epoch": 0.5827814569536424,
+      "grad_norm": 0.36528909096358364,
+      "learning_rate": 7.99342640358285e-05,
+      "loss": 0.4963,
+      "step": 440
+    },
+    {
+      "epoch": 0.5841059602649007,
+      "grad_norm": 0.2744469205120959,
+      "learning_rate": 7.99321270246565e-05,
+      "loss": 0.4951,
+      "step": 441
+    },
+    {
+      "epoch": 0.5854304635761589,
+      "grad_norm": 0.37776723137966445,
+      "learning_rate": 7.9929955860311e-05,
+      "loss": 0.4935,
+      "step": 442
+    },
+    {
+      "epoch": 0.5867549668874172,
+      "grad_norm": 0.38917474277875125,
+      "learning_rate": 7.992775054464891e-05,
+      "loss": 0.4944,
+      "step": 443
+    },
+    {
+      "epoch": 0.5880794701986755,
+      "grad_norm": 0.30241322350455957,
+      "learning_rate": 7.992551107955643e-05,
+      "loss": 0.4922,
+      "step": 444
+    },
+    {
+      "epoch": 0.5894039735099338,
+      "grad_norm": 0.3443134241678532,
+      "learning_rate": 7.992323746694892e-05,
+      "loss": 0.4968,
+      "step": 445
+    },
+    {
+      "epoch": 0.590728476821192,
+      "grad_norm": 0.38433534435193767,
+      "learning_rate": 7.992092970877097e-05,
+      "loss": 0.4934,
+      "step": 446
+    },
+    {
+      "epoch": 0.5920529801324503,
+      "grad_norm": 0.27501359481782994,
+      "learning_rate": 7.991858780699632e-05,
+      "loss": 0.4919,
+      "step": 447
+    },
+    {
+      "epoch": 0.5933774834437087,
+      "grad_norm": 0.3080892104362082,
+      "learning_rate": 7.9916211763628e-05,
+      "loss": 0.4915,
+      "step": 448
+    },
+    {
+      "epoch": 0.5947019867549669,
+      "grad_norm": 0.3439283733668184,
+      "learning_rate": 7.991380158069814e-05,
+      "loss": 0.4955,
+      "step": 449
+    },
+    {
+      "epoch": 0.5960264900662252,
+      "grad_norm": 0.28097320249192514,
+      "learning_rate": 7.991135726026819e-05,
+      "loss": 0.4902,
+      "step": 450
+    },
+    {
+      "epoch": 0.5973509933774834,
+      "grad_norm": 0.2748409879499819,
+      "learning_rate": 7.990887880442869e-05,
+      "loss": 0.4928,
+      "step": 451
+    },
+    {
+      "epoch": 0.5986754966887418,
+      "grad_norm": 0.2821085410941838,
+      "learning_rate": 7.990636621529941e-05,
+      "loss": 0.4931,
+      "step": 452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.22858396103817458,
+      "learning_rate": 7.990381949502933e-05,
+      "loss": 0.4913,
+      "step": 453
+    },
+    {
+      "epoch": 0.6013245033112583,
+      "grad_norm": 0.23801299179832028,
+      "learning_rate": 7.990123864579661e-05,
+      "loss": 0.4904,
+      "step": 454
+    },
+    {
+      "epoch": 0.6026490066225165,
+      "grad_norm": 0.24218572784751297,
+      "learning_rate": 7.989862366980861e-05,
+      "loss": 0.4922,
+      "step": 455
+    },
+    {
+      "epoch": 0.6039735099337749,
+      "grad_norm": 0.23628404791114782,
+      "learning_rate": 7.989597456930185e-05,
+      "loss": 0.4946,
+      "step": 456
+    },
+    {
+      "epoch": 0.6052980132450331,
+      "grad_norm": 0.24244485287124384,
+      "learning_rate": 7.989329134654207e-05,
+      "loss": 0.4906,
+      "step": 457
+    },
+    {
+      "epoch": 0.6066225165562914,
+      "grad_norm": 0.2735303695114769,
+      "learning_rate": 7.989057400382416e-05,
+      "loss": 0.4938,
+      "step": 458
+    },
+    {
+      "epoch": 0.6079470198675496,
+      "grad_norm": 0.29058449945976517,
+      "learning_rate": 7.988782254347225e-05,
+      "loss": 0.494,
+      "step": 459
+    },
+    {
+      "epoch": 0.609271523178808,
+      "grad_norm": 0.34328453413707194,
+      "learning_rate": 7.988503696783955e-05,
+      "loss": 0.4935,
+      "step": 460
+    },
+    {
+      "epoch": 0.6105960264900663,
+      "grad_norm": 0.39492792546870614,
+      "learning_rate": 7.988221727930856e-05,
+      "loss": 0.496,
+      "step": 461
+    },
+    {
+      "epoch": 0.6119205298013245,
+      "grad_norm": 0.45689190355679954,
+      "learning_rate": 7.987936348029088e-05,
+      "loss": 0.4923,
+      "step": 462
+    },
+    {
+      "epoch": 0.6132450331125828,
+      "grad_norm": 0.5158431960832088,
+      "learning_rate": 7.987647557322732e-05,
+      "loss": 0.4941,
+      "step": 463
+    },
+    {
+      "epoch": 0.614569536423841,
+      "grad_norm": 0.5370274683516859,
+      "learning_rate": 7.987355356058784e-05,
+      "loss": 0.4925,
+      "step": 464
+    },
+    {
+      "epoch": 0.6158940397350994,
+      "grad_norm": 0.48487361163563913,
+      "learning_rate": 7.987059744487159e-05,
+      "loss": 0.4965,
+      "step": 465
+    },
+    {
+      "epoch": 0.6172185430463576,
+      "grad_norm": 0.3634979575207017,
+      "learning_rate": 7.986760722860687e-05,
+      "loss": 0.4917,
+      "step": 466
+    },
+    {
+      "epoch": 0.6185430463576159,
+      "grad_norm": 0.26822680265541626,
+      "learning_rate": 7.986458291435117e-05,
+      "loss": 0.4896,
+      "step": 467
+    },
+    {
+      "epoch": 0.6198675496688741,
+      "grad_norm": 0.32634382854835614,
+      "learning_rate": 7.98615245046911e-05,
+      "loss": 0.4868,
+      "step": 468
+    },
+    {
+      "epoch": 0.6211920529801325,
+      "grad_norm": 0.32856939176638467,
+      "learning_rate": 7.98584320022425e-05,
+      "loss": 0.4937,
+      "step": 469
+    },
+    {
+      "epoch": 0.6225165562913907,
+      "grad_norm": 0.30050632835736657,
+      "learning_rate": 7.985530540965027e-05,
+      "loss": 0.4863,
+      "step": 470
+    },
+    {
+      "epoch": 0.623841059602649,
+      "grad_norm": 0.29235230332255246,
+      "learning_rate": 7.985214472958858e-05,
+      "loss": 0.4948,
+      "step": 471
+    },
+    {
+      "epoch": 0.6251655629139072,
+      "grad_norm": 0.2866094295627474,
+      "learning_rate": 7.984894996476068e-05,
+      "loss": 0.4957,
+      "step": 472
+    },
+    {
+      "epoch": 0.6264900662251656,
+      "grad_norm": 0.27235551363429733,
+      "learning_rate": 7.984572111789896e-05,
+      "loss": 0.4923,
+      "step": 473
+    },
+    {
+      "epoch": 0.6278145695364239,
+      "grad_norm": 0.2589467868770216,
+      "learning_rate": 7.984245819176503e-05,
+      "loss": 0.4921,
+      "step": 474
+    },
+    {
+      "epoch": 0.6291390728476821,
+      "grad_norm": 0.3137404783399244,
+      "learning_rate": 7.98391611891496e-05,
+      "loss": 0.4896,
+      "step": 475
+    },
+    {
+      "epoch": 0.6304635761589404,
+      "grad_norm": 0.380178447363786,
+      "learning_rate": 7.983583011287251e-05,
+      "loss": 0.4935,
+      "step": 476
+    },
+    {
+      "epoch": 0.6317880794701987,
+      "grad_norm": 0.40236818094166893,
+      "learning_rate": 7.983246496578279e-05,
+      "loss": 0.4898,
+      "step": 477
+    },
+    {
+      "epoch": 0.633112582781457,
+      "grad_norm": 0.3451737141162106,
+      "learning_rate": 7.982906575075855e-05,
+      "loss": 0.4912,
+      "step": 478
+    },
+    {
+      "epoch": 0.6344370860927152,
+      "grad_norm": 0.3503661916541925,
+      "learning_rate": 7.982563247070712e-05,
+      "loss": 0.4962,
+      "step": 479
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 0.4020976711576733,
+      "learning_rate": 7.98221651285649e-05,
+      "loss": 0.4933,
+      "step": 480
+    },
+    {
+      "epoch": 0.6370860927152318,
+      "grad_norm": 0.48922215914022216,
+      "learning_rate": 7.981866372729743e-05,
+      "loss": 0.4946,
+      "step": 481
+    },
+    {
+      "epoch": 0.6384105960264901,
+      "grad_norm": 0.5252053881879244,
+      "learning_rate": 7.981512826989939e-05,
+      "loss": 0.4969,
+      "step": 482
+    },
+    {
+      "epoch": 0.6397350993377483,
+      "grad_norm": 0.5109210690968093,
+      "learning_rate": 7.98115587593946e-05,
+      "loss": 0.493,
+      "step": 483
+    },
+    {
+      "epoch": 0.6410596026490066,
+      "grad_norm": 0.46811045053694195,
+      "learning_rate": 7.980795519883599e-05,
+      "loss": 0.4976,
+      "step": 484
+    },
+    {
+      "epoch": 0.6423841059602649,
+      "grad_norm": 0.42649531363750715,
+      "learning_rate": 7.98043175913056e-05,
+      "loss": 0.4932,
+      "step": 485
+    },
+    {
+      "epoch": 0.6437086092715232,
+      "grad_norm": 0.4457925635706888,
+      "learning_rate": 7.980064593991464e-05,
+      "loss": 0.4887,
+      "step": 486
+    },
+    {
+      "epoch": 0.6450331125827815,
+      "grad_norm": 0.45330852233129665,
+      "learning_rate": 7.979694024780336e-05,
+      "loss": 0.4948,
+      "step": 487
+    },
+    {
+      "epoch": 0.6463576158940397,
+      "grad_norm": 0.38494570447816584,
+      "learning_rate": 7.97932005181412e-05,
+      "loss": 0.4883,
+      "step": 488
+    },
+    {
+      "epoch": 0.6476821192052981,
+      "grad_norm": 0.32678498047602356,
+      "learning_rate": 7.978942675412666e-05,
+      "loss": 0.4918,
+      "step": 489
+    },
+    {
+      "epoch": 0.6490066225165563,
+      "grad_norm": 0.4021135385651809,
+      "learning_rate": 7.978561895898736e-05,
+      "loss": 0.4895,
+      "step": 490
+    },
+    {
+      "epoch": 0.6503311258278146,
+      "grad_norm": 0.3776595562941202,
+      "learning_rate": 7.978177713598005e-05,
+      "loss": 0.4946,
+      "step": 491
+    },
+    {
+      "epoch": 0.6516556291390728,
+      "grad_norm": 0.3081169563352137,
+      "learning_rate": 7.977790128839058e-05,
+      "loss": 0.491,
+      "step": 492
+    },
+    {
+      "epoch": 0.6529801324503312,
+      "grad_norm": 0.31573366485798376,
+      "learning_rate": 7.977399141953385e-05,
+      "loss": 0.4901,
+      "step": 493
+    },
+    {
+      "epoch": 0.6543046357615894,
+      "grad_norm": 0.2868330800903801,
+      "learning_rate": 7.977004753275393e-05,
+      "loss": 0.4879,
+      "step": 494
+    },
+    {
+      "epoch": 0.6556291390728477,
+      "grad_norm": 0.2931923997347364,
+      "learning_rate": 7.976606963142393e-05,
+      "loss": 0.4865,
+      "step": 495
+    },
+    {
+      "epoch": 0.6569536423841059,
+      "grad_norm": 0.33237660641813527,
+      "learning_rate": 7.976205771894606e-05,
+      "loss": 0.4878,
+      "step": 496
+    },
+    {
+      "epoch": 0.6582781456953642,
+      "grad_norm": 0.2660067370879585,
+      "learning_rate": 7.975801179875166e-05,
+      "loss": 0.4887,
+      "step": 497
+    },
+    {
+      "epoch": 0.6596026490066225,
+      "grad_norm": 0.25438665331833193,
+      "learning_rate": 7.975393187430112e-05,
+      "loss": 0.4905,
+      "step": 498
+    },
+    {
+      "epoch": 0.6609271523178808,
+      "grad_norm": 0.2386914489425516,
+      "learning_rate": 7.974981794908391e-05,
+      "loss": 0.491,
+      "step": 499
+    },
+    {
+      "epoch": 0.6622516556291391,
+      "grad_norm": 0.22404410748083084,
+      "learning_rate": 7.97456700266186e-05,
+      "loss": 0.4871,
+      "step": 500
+    },
+    {
+      "epoch": 0.6635761589403973,
+      "grad_norm": 0.22378811671818158,
+      "learning_rate": 7.974148811045281e-05,
+      "loss": 0.4901,
+      "step": 501
+    },
+    {
+      "epoch": 0.6649006622516557,
+      "grad_norm": 0.2673176628690433,
+      "learning_rate": 7.973727220416328e-05,
+      "loss": 0.4864,
+      "step": 502
+    },
+    {
+      "epoch": 0.6662251655629139,
+      "grad_norm": 0.28894729593995955,
+      "learning_rate": 7.973302231135576e-05,
+      "loss": 0.4874,
+      "step": 503
+    },
+    {
+      "epoch": 0.6675496688741722,
+      "grad_norm": 0.25028114561307746,
+      "learning_rate": 7.972873843566512e-05,
+      "loss": 0.4855,
+      "step": 504
+    },
+    {
+      "epoch": 0.6688741721854304,
+      "grad_norm": 0.24880809842096896,
+      "learning_rate": 7.972442058075528e-05,
+      "loss": 0.4885,
+      "step": 505
+    },
+    {
+      "epoch": 0.6701986754966888,
+      "grad_norm": 0.25626108315831025,
+      "learning_rate": 7.972006875031922e-05,
+      "loss": 0.487,
+      "step": 506
+    },
+    {
+      "epoch": 0.671523178807947,
+      "grad_norm": 0.3027788598512414,
+      "learning_rate": 7.971568294807894e-05,
+      "loss": 0.487,
+      "step": 507
+    },
+    {
+      "epoch": 0.6728476821192053,
+      "grad_norm": 0.3550308025727053,
+      "learning_rate": 7.971126317778558e-05,
+      "loss": 0.4846,
+      "step": 508
+    },
+    {
+      "epoch": 0.6741721854304635,
+      "grad_norm": 0.3829935907196281,
+      "learning_rate": 7.970680944321926e-05,
+      "loss": 0.4852,
+      "step": 509
+    },
+    {
+      "epoch": 0.6754966887417219,
+      "grad_norm": 0.4186170801788798,
+      "learning_rate": 7.970232174818917e-05,
+      "loss": 0.4915,
+      "step": 510
+    },
+    {
+      "epoch": 0.6768211920529801,
+      "grad_norm": 0.47109126926234396,
+      "learning_rate": 7.969780009653356e-05,
+      "loss": 0.4888,
+      "step": 511
+    },
+    {
+      "epoch": 0.6781456953642384,
+      "grad_norm": 0.5821030553775024,
+      "learning_rate": 7.96932444921197e-05,
+      "loss": 0.4877,
+      "step": 512
+    },
+    {
+      "epoch": 0.6794701986754967,
+      "grad_norm": 0.6897817145723125,
+      "learning_rate": 7.968865493884391e-05,
+      "loss": 0.4928,
+      "step": 513
+    },
+    {
+      "epoch": 0.680794701986755,
+      "grad_norm": 0.7386361764057519,
+      "learning_rate": 7.968403144063156e-05,
+      "loss": 0.4891,
+      "step": 514
+    },
+    {
+      "epoch": 0.6821192052980133,
+      "grad_norm": 0.578138164455798,
+      "learning_rate": 7.967937400143703e-05,
+      "loss": 0.4883,
+      "step": 515
+    },
+    {
+      "epoch": 0.6834437086092715,
+      "grad_norm": 0.3434135832640883,
+      "learning_rate": 7.967468262524374e-05,
+      "loss": 0.4909,
+      "step": 516
+    },
+    {
+      "epoch": 0.6847682119205298,
+      "grad_norm": 0.3935841029823858,
+      "learning_rate": 7.966995731606411e-05,
+      "loss": 0.4902,
+      "step": 517
+    },
+    {
+      "epoch": 0.686092715231788,
+      "grad_norm": 0.4346875870543704,
+      "learning_rate": 7.966519807793963e-05,
+      "loss": 0.4945,
+      "step": 518
+    },
+    {
+      "epoch": 0.6874172185430464,
+      "grad_norm": 0.37089213269275306,
+      "learning_rate": 7.966040491494079e-05,
+      "loss": 0.4898,
+      "step": 519
+    },
+    {
+      "epoch": 0.6887417218543046,
+      "grad_norm": 0.3617652493340738,
+      "learning_rate": 7.965557783116708e-05,
+      "loss": 0.4879,
+      "step": 520
+    },
+    {
+      "epoch": 0.6900662251655629,
+      "grad_norm": 0.3696667922370984,
+      "learning_rate": 7.9650716830747e-05,
+      "loss": 0.4886,
+      "step": 521
+    },
+    {
+      "epoch": 0.6913907284768211,
+      "grad_norm": 0.3312282081892088,
+      "learning_rate": 7.964582191783808e-05,
+      "loss": 0.4913,
+      "step": 522
+    },
+    {
+      "epoch": 0.6927152317880795,
+      "grad_norm": 0.24327864637441138,
+      "learning_rate": 7.964089309662685e-05,
+      "loss": 0.4877,
+      "step": 523
+    },
+    {
+      "epoch": 0.6940397350993377,
+      "grad_norm": 0.25527509713482777,
+      "learning_rate": 7.963593037132881e-05,
+      "loss": 0.4882,
+      "step": 524
+    },
+    {
+      "epoch": 0.695364238410596,
+      "grad_norm": 0.25986113303710506,
+      "learning_rate": 7.963093374618852e-05,
+      "loss": 0.4873,
+      "step": 525
+    },
+    {
+      "epoch": 0.6966887417218544,
+      "grad_norm": 0.22894913334186004,
+      "learning_rate": 7.962590322547947e-05,
+      "loss": 0.486,
+      "step": 526
+    },
+    {
+      "epoch": 0.6980132450331126,
+      "grad_norm": 0.3098463477431134,
+      "learning_rate": 7.962083881350417e-05,
+      "loss": 0.4889,
+      "step": 527
+    },
+    {
+      "epoch": 0.6993377483443709,
+      "grad_norm": 0.2963693087911011,
+      "learning_rate": 7.961574051459412e-05,
+      "loss": 0.4862,
+      "step": 528
+    },
+    {
+      "epoch": 0.7006622516556291,
+      "grad_norm": 0.2014388572559124,
+      "learning_rate": 7.96106083331098e-05,
+      "loss": 0.4905,
+      "step": 529
+    },
+    {
+      "epoch": 0.7019867549668874,
+      "grad_norm": 0.28925101664455793,
+      "learning_rate": 7.960544227344064e-05,
+      "loss": 0.49,
+      "step": 530
+    },
+    {
+      "epoch": 0.7033112582781457,
+      "grad_norm": 0.34190680033364335,
+      "learning_rate": 7.960024234000512e-05,
+      "loss": 0.4893,
+      "step": 531
+    },
+    {
+      "epoch": 0.704635761589404,
+      "grad_norm": 0.3016898586337942,
+      "learning_rate": 7.959500853725058e-05,
+      "loss": 0.4881,
+      "step": 532
+    },
+    {
+      "epoch": 0.7059602649006622,
+      "grad_norm": 0.25394983146297545,
+      "learning_rate": 7.958974086965344e-05,
+      "loss": 0.4905,
+      "step": 533
+    },
+    {
+      "epoch": 0.7072847682119205,
+      "grad_norm": 0.3225568014029614,
+      "learning_rate": 7.9584439341719e-05,
+      "loss": 0.4887,
+      "step": 534
+    },
+    {
+      "epoch": 0.7086092715231788,
+      "grad_norm": 0.36057603215255396,
+      "learning_rate": 7.957910395798158e-05,
+      "loss": 0.4895,
+      "step": 535
+    },
+    {
+      "epoch": 0.7099337748344371,
+      "grad_norm": 0.39291050337209693,
+      "learning_rate": 7.957373472300442e-05,
+      "loss": 0.4927,
+      "step": 536
+    },
+    {
+      "epoch": 0.7112582781456953,
+      "grad_norm": 0.44851489943888484,
+      "learning_rate": 7.95683316413797e-05,
+      "loss": 0.4901,
+      "step": 537
+    },
+    {
+      "epoch": 0.7125827814569536,
+      "grad_norm": 0.4956906699990852,
+      "learning_rate": 7.956289471772862e-05,
+      "loss": 0.4942,
+      "step": 538
+    },
+    {
+      "epoch": 0.713907284768212,
+      "grad_norm": 0.4919071732967479,
+      "learning_rate": 7.955742395670125e-05,
+      "loss": 0.4926,
+      "step": 539
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 0.42799938237634877,
+      "learning_rate": 7.95519193629766e-05,
+      "loss": 0.4891,
+      "step": 540
+    },
+    {
+      "epoch": 0.7165562913907285,
+      "grad_norm": 0.479143617674278,
+      "learning_rate": 7.954638094126268e-05,
+      "loss": 0.4907,
+      "step": 541
+    },
+    {
+      "epoch": 0.7178807947019867,
+      "grad_norm": 0.5287942000083011,
+      "learning_rate": 7.954080869629637e-05,
+      "loss": 0.491,
+      "step": 542
+    },
+    {
+      "epoch": 0.7192052980132451,
+      "grad_norm": 0.4968817807777099,
+      "learning_rate": 7.953520263284353e-05,
+      "loss": 0.491,
+      "step": 543
+    },
+    {
+      "epoch": 0.7205298013245033,
+      "grad_norm": 0.435560066057734,
+      "learning_rate": 7.95295627556989e-05,
+      "loss": 0.4915,
+      "step": 544
+    },
+    {
+      "epoch": 0.7218543046357616,
+      "grad_norm": 0.3842524837977631,
+      "learning_rate": 7.952388906968616e-05,
+      "loss": 0.4899,
+      "step": 545
+    },
+    {
+      "epoch": 0.7231788079470198,
+      "grad_norm": 0.33526439341394304,
+      "learning_rate": 7.951818157965788e-05,
+      "loss": 0.4871,
+      "step": 546
+    },
+    {
+      "epoch": 0.7245033112582782,
+      "grad_norm": 0.31852576143799965,
+      "learning_rate": 7.951244029049561e-05,
+      "loss": 0.4887,
+      "step": 547
+    },
+    {
+      "epoch": 0.7258278145695364,
+      "grad_norm": 0.4149505120988515,
+      "learning_rate": 7.950666520710973e-05,
+      "loss": 0.4915,
+      "step": 548
+    },
+    {
+      "epoch": 0.7271523178807947,
+      "grad_norm": 0.4175572838386353,
+      "learning_rate": 7.950085633443958e-05,
+      "loss": 0.4897,
+      "step": 549
+    },
+    {
+      "epoch": 0.7284768211920529,
+      "grad_norm": 0.30674662183925594,
+      "learning_rate": 7.949501367745335e-05,
+      "loss": 0.4878,
+      "step": 550
+    },
+    {
+      "epoch": 0.7298013245033113,
+      "grad_norm": 0.2678119803358583,
+      "learning_rate": 7.948913724114816e-05,
+      "loss": 0.488,
+      "step": 551
+    },
+    {
+      "epoch": 0.7311258278145696,
+      "grad_norm": 0.4062728291877582,
+      "learning_rate": 7.948322703055002e-05,
+      "loss": 0.4875,
+      "step": 552
+    },
+    {
+      "epoch": 0.7324503311258278,
+      "grad_norm": 0.397205996181469,
+      "learning_rate": 7.947728305071383e-05,
+      "loss": 0.4878,
+      "step": 553
+    },
+    {
+      "epoch": 0.7337748344370861,
+      "grad_norm": 0.25740744980455793,
+      "learning_rate": 7.947130530672333e-05,
+      "loss": 0.4883,
+      "step": 554
+    },
+    {
+      "epoch": 0.7350993377483444,
+      "grad_norm": 0.19482473310852072,
+      "learning_rate": 7.94652938036912e-05,
+      "loss": 0.4889,
+      "step": 555
+    },
+    {
+      "epoch": 0.7364238410596027,
+      "grad_norm": 0.21816763251181084,
+      "learning_rate": 7.945924854675892e-05,
+      "loss": 0.4823,
+      "step": 556
+    },
+    {
+      "epoch": 0.7377483443708609,
+      "grad_norm": 0.2173393576543725,
+      "learning_rate": 7.945316954109693e-05,
+      "loss": 0.4856,
+      "step": 557
+    },
+    {
+      "epoch": 0.7390728476821192,
+      "grad_norm": 0.2474376240334169,
+      "learning_rate": 7.944705679190446e-05,
+      "loss": 0.4874,
+      "step": 558
+    },
+    {
+      "epoch": 0.7403973509933774,
+      "grad_norm": 0.24169568491259857,
+      "learning_rate": 7.944091030440962e-05,
+      "loss": 0.4896,
+      "step": 559
+    },
+    {
+      "epoch": 0.7417218543046358,
+      "grad_norm": 0.23168861162357549,
+      "learning_rate": 7.94347300838694e-05,
+      "loss": 0.487,
+      "step": 560
+    },
+    {
+      "epoch": 0.743046357615894,
+      "grad_norm": 0.2367274215856281,
+      "learning_rate": 7.942851613556961e-05,
+      "loss": 0.4879,
+      "step": 561
+    },
+    {
+      "epoch": 0.7443708609271523,
+      "grad_norm": 0.24186600206866915,
+      "learning_rate": 7.942226846482491e-05,
+      "loss": 0.4841,
+      "step": 562
+    },
+    {
+      "epoch": 0.7456953642384105,
+      "grad_norm": 0.23046180993770768,
+      "learning_rate": 7.941598707697884e-05,
+      "loss": 0.49,
+      "step": 563
+    },
+    {
+      "epoch": 0.7470198675496689,
+      "grad_norm": 0.22905805725330533,
+      "learning_rate": 7.940967197740373e-05,
+      "loss": 0.4889,
+      "step": 564
+    },
+    {
+      "epoch": 0.7483443708609272,
+      "grad_norm": 0.24044171304604883,
+      "learning_rate": 7.940332317150075e-05,
+      "loss": 0.4847,
+      "step": 565
+    },
+    {
+      "epoch": 0.7496688741721854,
+      "grad_norm": 0.25311948741020424,
+      "learning_rate": 7.939694066469995e-05,
+      "loss": 0.4862,
+      "step": 566
+    },
+    {
+      "epoch": 0.7509933774834437,
+      "grad_norm": 0.26438249647737994,
+      "learning_rate": 7.939052446246013e-05,
+      "loss": 0.4846,
+      "step": 567
+    },
+    {
+      "epoch": 0.752317880794702,
+      "grad_norm": 0.3291303361981443,
+      "learning_rate": 7.938407457026894e-05,
+      "loss": 0.4899,
+      "step": 568
+    },
+    {
+      "epoch": 0.7536423841059603,
+      "grad_norm": 0.4106629429443559,
+      "learning_rate": 7.937759099364288e-05,
+      "loss": 0.4864,
+      "step": 569
+    },
+    {
+      "epoch": 0.7549668874172185,
+      "grad_norm": 0.4651594826449493,
+      "learning_rate": 7.93710737381272e-05,
+      "loss": 0.4883,
+      "step": 570
+    },
+    {
+      "epoch": 0.7562913907284768,
+      "grad_norm": 0.45563542939486323,
+      "learning_rate": 7.936452280929599e-05,
+      "loss": 0.4886,
+      "step": 571
+    },
+    {
+      "epoch": 0.7576158940397351,
+      "grad_norm": 0.37378397548019787,
+      "learning_rate": 7.935793821275212e-05,
+      "loss": 0.4833,
+      "step": 572
+    },
+    {
+      "epoch": 0.7589403973509934,
+      "grad_norm": 0.3177955228744069,
+      "learning_rate": 7.93513199541273e-05,
+      "loss": 0.4856,
+      "step": 573
+    },
+    {
+      "epoch": 0.7602649006622516,
+      "grad_norm": 0.2953268549901922,
+      "learning_rate": 7.934466803908197e-05,
+      "loss": 0.4916,
+      "step": 574
+    },
+    {
+      "epoch": 0.7615894039735099,
+      "grad_norm": 0.3200111091609561,
+      "learning_rate": 7.93379824733054e-05,
+      "loss": 0.4903,
+      "step": 575
+    },
+    {
+      "epoch": 0.7629139072847683,
+      "grad_norm": 0.33332371250512793,
+      "learning_rate": 7.933126326251562e-05,
+      "loss": 0.4878,
+      "step": 576
+    },
+    {
+      "epoch": 0.7642384105960265,
+      "grad_norm": 0.2847232747856637,
+      "learning_rate": 7.932451041245941e-05,
+      "loss": 0.4856,
+      "step": 577
+    },
+    {
+      "epoch": 0.7655629139072848,
+      "grad_norm": 0.2599630242930639,
+      "learning_rate": 7.931772392891241e-05,
+      "loss": 0.4867,
+      "step": 578
+    },
+    {
+      "epoch": 0.766887417218543,
+      "grad_norm": 0.22874782088896542,
+      "learning_rate": 7.931090381767891e-05,
+      "loss": 0.4835,
+      "step": 579
+    },
+    {
+      "epoch": 0.7682119205298014,
+      "grad_norm": 0.25710170674585636,
+      "learning_rate": 7.930405008459207e-05,
+      "loss": 0.4851,
+      "step": 580
+    },
+    {
+      "epoch": 0.7695364238410596,
+      "grad_norm": 0.24255108257272834,
+      "learning_rate": 7.929716273551371e-05,
+      "loss": 0.4845,
+      "step": 581
+    },
+    {
+      "epoch": 0.7708609271523179,
+      "grad_norm": 0.21798424207468572,
+      "learning_rate": 7.929024177633448e-05,
+      "loss": 0.4893,
+      "step": 582
+    },
+    {
+      "epoch": 0.7721854304635761,
+      "grad_norm": 0.25313450841838214,
+      "learning_rate": 7.928328721297373e-05,
+      "loss": 0.4861,
+      "step": 583
+    },
+    {
+      "epoch": 0.7735099337748345,
+      "grad_norm": 0.2861386950532355,
+      "learning_rate": 7.927629905137957e-05,
+      "loss": 0.4857,
+      "step": 584
+    },
+    {
+      "epoch": 0.7748344370860927,
+      "grad_norm": 0.2582810738494978,
+      "learning_rate": 7.926927729752881e-05,
+      "loss": 0.4845,
+      "step": 585
+    },
+    {
+      "epoch": 0.776158940397351,
+      "grad_norm": 0.21070213995527853,
+      "learning_rate": 7.926222195742705e-05,
+      "loss": 0.4844,
+      "step": 586
+    },
+    {
+      "epoch": 0.7774834437086092,
+      "grad_norm": 0.22811581490015165,
+      "learning_rate": 7.925513303710859e-05,
+      "loss": 0.4817,
+      "step": 587
+    },
+    {
+      "epoch": 0.7788079470198676,
+      "grad_norm": 0.2506660566616494,
+      "learning_rate": 7.924801054263642e-05,
+      "loss": 0.4853,
+      "step": 588
+    },
+    {
+      "epoch": 0.7801324503311259,
+      "grad_norm": 0.2554417432981022,
+      "learning_rate": 7.924085448010228e-05,
+      "loss": 0.4874,
+      "step": 589
+    },
+    {
+      "epoch": 0.7814569536423841,
+      "grad_norm": 0.3069836512287508,
+      "learning_rate": 7.923366485562663e-05,
+      "loss": 0.4855,
+      "step": 590
+    },
+    {
+      "epoch": 0.7827814569536424,
+      "grad_norm": 0.35095786737160733,
+      "learning_rate": 7.922644167535859e-05,
+      "loss": 0.4862,
+      "step": 591
+    },
+    {
+      "epoch": 0.7841059602649006,
+      "grad_norm": 0.31709598277471535,
+      "learning_rate": 7.921918494547601e-05,
+      "loss": 0.4828,
+      "step": 592
+    },
+    {
+      "epoch": 0.785430463576159,
+      "grad_norm": 0.25022118160117557,
+      "learning_rate": 7.921189467218546e-05,
+      "loss": 0.4895,
+      "step": 593
+    },
+    {
+      "epoch": 0.7867549668874172,
+      "grad_norm": 0.2601919458412492,
+      "learning_rate": 7.920457086172212e-05,
+      "loss": 0.4848,
+      "step": 594
+    },
+    {
+      "epoch": 0.7880794701986755,
+      "grad_norm": 0.2807824808289332,
+      "learning_rate": 7.919721352034994e-05,
+      "loss": 0.4856,
+      "step": 595
+    },
+    {
+      "epoch": 0.7894039735099337,
+      "grad_norm": 0.2972072871799282,
+      "learning_rate": 7.918982265436152e-05,
+      "loss": 0.4839,
+      "step": 596
+    },
+    {
+      "epoch": 0.7907284768211921,
+      "grad_norm": 0.3093859737508261,
+      "learning_rate": 7.918239827007808e-05,
+      "loss": 0.4875,
+      "step": 597
+    },
+    {
+      "epoch": 0.7920529801324503,
+      "grad_norm": 0.30615266613389974,
+      "learning_rate": 7.917494037384959e-05,
+      "loss": 0.4863,
+      "step": 598
+    },
+    {
+      "epoch": 0.7933774834437086,
+      "grad_norm": 0.29386499154184187,
+      "learning_rate": 7.916744897205464e-05,
+      "loss": 0.4866,
+      "step": 599
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 0.2882918531306945,
+      "learning_rate": 7.915992407110047e-05,
+      "loss": 0.4834,
+      "step": 600
+    },
+    {
+      "epoch": 0.7960264900662252,
+      "grad_norm": 0.23333876849398316,
+      "learning_rate": 7.915236567742297e-05,
+      "loss": 0.4855,
+      "step": 601
+    },
+    {
+      "epoch": 0.7973509933774835,
+      "grad_norm": 0.20112653203954525,
+      "learning_rate": 7.914477379748671e-05,
+      "loss": 0.4826,
+      "step": 602
+    },
+    {
+      "epoch": 0.7986754966887417,
+      "grad_norm": 0.24376756727043822,
+      "learning_rate": 7.913714843778486e-05,
+      "loss": 0.4866,
+      "step": 603
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32120081266986006,
+      "learning_rate": 7.912948960483926e-05,
+      "loss": 0.4849,
+      "step": 604
+    },
+    {
+      "epoch": 0.8013245033112583,
+      "grad_norm": 0.3379512556311551,
+      "learning_rate": 7.912179730520037e-05,
+      "loss": 0.4823,
+      "step": 605
+    },
+    {
+      "epoch": 0.8026490066225166,
+      "grad_norm": 0.286496221451673,
+      "learning_rate": 7.911407154544721e-05,
+      "loss": 0.4831,
+      "step": 606
+    },
+    {
+      "epoch": 0.8039735099337748,
+      "grad_norm": 0.23384225296287295,
+      "learning_rate": 7.910631233218754e-05,
+      "loss": 0.4809,
+      "step": 607
+    },
+    {
+      "epoch": 0.8052980132450331,
+      "grad_norm": 0.25724493505693996,
+      "learning_rate": 7.909851967205762e-05,
+      "loss": 0.4829,
+      "step": 608
+    },
+    {
+      "epoch": 0.8066225165562914,
+      "grad_norm": 0.3192957079336321,
+      "learning_rate": 7.909069357172238e-05,
+      "loss": 0.4819,
+      "step": 609
+    },
+    {
+      "epoch": 0.8079470198675497,
+      "grad_norm": 0.34920284690919595,
+      "learning_rate": 7.908283403787532e-05,
+      "loss": 0.4838,
+      "step": 610
+    },
+    {
+      "epoch": 0.8092715231788079,
+      "grad_norm": 0.3523853085711709,
+      "learning_rate": 7.907494107723857e-05,
+      "loss": 0.4848,
+      "step": 611
+    },
+    {
+      "epoch": 0.8105960264900662,
+      "grad_norm": 0.3156758192112876,
+      "learning_rate": 7.90670146965628e-05,
+      "loss": 0.4874,
+      "step": 612
+    },
+    {
+      "epoch": 0.8119205298013245,
+      "grad_norm": 0.2684545819961005,
+      "learning_rate": 7.905905490262729e-05,
+      "loss": 0.4825,
+      "step": 613
+    },
+    {
+      "epoch": 0.8132450331125828,
+      "grad_norm": 0.31329881085374195,
+      "learning_rate": 7.905106170223991e-05,
+      "loss": 0.4837,
+      "step": 614
+    },
+    {
+      "epoch": 0.8145695364238411,
+      "grad_norm": 0.40677676188351664,
+      "learning_rate": 7.90430351022371e-05,
+      "loss": 0.4838,
+      "step": 615
+    },
+    {
+      "epoch": 0.8158940397350993,
+      "grad_norm": 0.4143482055649819,
+      "learning_rate": 7.903497510948383e-05,
+      "loss": 0.4848,
+      "step": 616
+    },
+    {
+      "epoch": 0.8172185430463577,
+      "grad_norm": 0.40314138412951245,
+      "learning_rate": 7.902688173087367e-05,
+      "loss": 0.4882,
+      "step": 617
+    },
+    {
+      "epoch": 0.8185430463576159,
+      "grad_norm": 0.36238760429483446,
+      "learning_rate": 7.901875497332872e-05,
+      "loss": 0.4847,
+      "step": 618
+    },
+    {
+      "epoch": 0.8198675496688742,
+      "grad_norm": 0.3329244377028913,
+      "learning_rate": 7.901059484379964e-05,
+      "loss": 0.4837,
+      "step": 619
+    },
+    {
+      "epoch": 0.8211920529801324,
+      "grad_norm": 0.350165826400946,
+      "learning_rate": 7.900240134926564e-05,
+      "loss": 0.4823,
+      "step": 620
+    },
+    {
+      "epoch": 0.8225165562913908,
+      "grad_norm": 0.34464226132253617,
+      "learning_rate": 7.899417449673447e-05,
+      "loss": 0.482,
+      "step": 621
+    },
+    {
+      "epoch": 0.823841059602649,
+      "grad_norm": 0.34637949987033556,
+      "learning_rate": 7.898591429324237e-05,
+      "loss": 0.4856,
+      "step": 622
+    },
+    {
+      "epoch": 0.8251655629139073,
+      "grad_norm": 0.3313364241464332,
+      "learning_rate": 7.897762074585413e-05,
+      "loss": 0.4843,
+      "step": 623
+    },
+    {
+      "epoch": 0.8264900662251655,
+      "grad_norm": 0.28872493570032826,
+      "learning_rate": 7.896929386166308e-05,
+      "loss": 0.4835,
+      "step": 624
+    },
+    {
+      "epoch": 0.8278145695364238,
+      "grad_norm": 0.2273226704743246,
+      "learning_rate": 7.896093364779104e-05,
+      "loss": 0.4803,
+      "step": 625
+    },
+    {
+      "epoch": 0.8291390728476821,
+      "grad_norm": 0.23321118174261696,
+      "learning_rate": 7.895254011138833e-05,
+      "loss": 0.4842,
+      "step": 626
+    },
+    {
+      "epoch": 0.8304635761589404,
+      "grad_norm": 0.2854337722092818,
+      "learning_rate": 7.894411325963376e-05,
+      "loss": 0.4836,
+      "step": 627
+    },
+    {
+      "epoch": 0.8317880794701987,
+      "grad_norm": 0.27197198671058387,
+      "learning_rate": 7.89356530997347e-05,
+      "loss": 0.4781,
+      "step": 628
+    },
+    {
+      "epoch": 0.833112582781457,
+      "grad_norm": 0.2272148956347377,
+      "learning_rate": 7.892715963892691e-05,
+      "loss": 0.4825,
+      "step": 629
+    },
+    {
+      "epoch": 0.8344370860927153,
+      "grad_norm": 0.2089042827781077,
+      "learning_rate": 7.891863288447473e-05,
+      "loss": 0.4831,
+      "step": 630
+    },
+    {
+      "epoch": 0.8357615894039735,
+      "grad_norm": 0.24233813179628674,
+      "learning_rate": 7.89100728436709e-05,
+      "loss": 0.48,
+      "step": 631
+    },
+    {
+      "epoch": 0.8370860927152318,
+      "grad_norm": 0.23457880217868535,
+      "learning_rate": 7.890147952383666e-05,
+      "loss": 0.4844,
+      "step": 632
+    },
+    {
+      "epoch": 0.83841059602649,
+      "grad_norm": 0.24132985632537876,
+      "learning_rate": 7.889285293232171e-05,
+      "loss": 0.4802,
+      "step": 633
+    },
+    {
+      "epoch": 0.8397350993377484,
+      "grad_norm": 0.24537278375956395,
+      "learning_rate": 7.88841930765042e-05,
+      "loss": 0.4831,
+      "step": 634
+    },
+    {
+      "epoch": 0.8410596026490066,
+      "grad_norm": 0.20648375506024633,
+      "learning_rate": 7.887549996379075e-05,
+      "loss": 0.4817,
+      "step": 635
+    },
+    {
+      "epoch": 0.8423841059602649,
+      "grad_norm": 0.21167680716535964,
+      "learning_rate": 7.88667736016164e-05,
+      "loss": 0.4813,
+      "step": 636
+    },
+    {
+      "epoch": 0.8437086092715231,
+      "grad_norm": 0.2691453070520354,
+      "learning_rate": 7.885801399744465e-05,
+      "loss": 0.4831,
+      "step": 637
+    },
+    {
+      "epoch": 0.8450331125827815,
+      "grad_norm": 0.28834909699721756,
+      "learning_rate": 7.884922115876741e-05,
+      "loss": 0.4813,
+      "step": 638
+    },
+    {
+      "epoch": 0.8463576158940397,
+      "grad_norm": 0.23052988621110965,
+      "learning_rate": 7.884039509310504e-05,
+      "loss": 0.4803,
+      "step": 639
+    },
+    {
+      "epoch": 0.847682119205298,
+      "grad_norm": 0.21388828192768736,
+      "learning_rate": 7.883153580800629e-05,
+      "loss": 0.48,
+      "step": 640
+    },
+    {
+      "epoch": 0.8490066225165563,
+      "grad_norm": 0.2904358915815279,
+      "learning_rate": 7.882264331104831e-05,
+      "loss": 0.4862,
+      "step": 641
+    },
+    {
+      "epoch": 0.8503311258278146,
+      "grad_norm": 0.3175041265229936,
+      "learning_rate": 7.881371760983671e-05,
+      "loss": 0.4853,
+      "step": 642
+    },
+    {
+      "epoch": 0.8516556291390729,
+      "grad_norm": 0.31752691059181426,
+      "learning_rate": 7.880475871200548e-05,
+      "loss": 0.4825,
+      "step": 643
+    },
+    {
+      "epoch": 0.8529801324503311,
+      "grad_norm": 0.30795770709704146,
+      "learning_rate": 7.879576662521696e-05,
+      "loss": 0.4849,
+      "step": 644
+    },
+    {
+      "epoch": 0.8543046357615894,
+      "grad_norm": 0.32844449648489926,
+      "learning_rate": 7.878674135716193e-05,
+      "loss": 0.4823,
+      "step": 645
+    },
+    {
+      "epoch": 0.8556291390728477,
+      "grad_norm": 0.36376460264031807,
+      "learning_rate": 7.877768291555952e-05,
+      "loss": 0.4827,
+      "step": 646
+    },
+    {
+      "epoch": 0.856953642384106,
+      "grad_norm": 0.35912636986291424,
+      "learning_rate": 7.876859130815723e-05,
+      "loss": 0.486,
+      "step": 647
+    },
+    {
+      "epoch": 0.8582781456953642,
+      "grad_norm": 0.3238635030876165,
+      "learning_rate": 7.875946654273094e-05,
+      "loss": 0.4842,
+      "step": 648
+    },
+    {
+      "epoch": 0.8596026490066225,
+      "grad_norm": 0.3138827657343674,
+      "learning_rate": 7.875030862708491e-05,
+      "loss": 0.4837,
+      "step": 649
+    },
+    {
+      "epoch": 0.8609271523178808,
+      "grad_norm": 0.3252050168455758,
+      "learning_rate": 7.87411175690517e-05,
+      "loss": 0.48,
+      "step": 650
+    },
+    {
+      "epoch": 0.8622516556291391,
+      "grad_norm": 0.349182100080766,
+      "learning_rate": 7.873189337649222e-05,
+      "loss": 0.4811,
+      "step": 651
+    },
+    {
+      "epoch": 0.8635761589403973,
+      "grad_norm": 0.39011588714570816,
+      "learning_rate": 7.87226360572958e-05,
+      "loss": 0.4809,
+      "step": 652
+    },
+    {
+      "epoch": 0.8649006622516556,
+      "grad_norm": 0.36213154703285366,
+      "learning_rate": 7.871334561938003e-05,
+      "loss": 0.4838,
+      "step": 653
+    },
+    {
+      "epoch": 0.866225165562914,
+      "grad_norm": 0.2724263249491812,
+      "learning_rate": 7.870402207069081e-05,
+      "loss": 0.4814,
+      "step": 654
+    },
+    {
+      "epoch": 0.8675496688741722,
+      "grad_norm": 0.20118308690104308,
+      "learning_rate": 7.869466541920242e-05,
+      "loss": 0.4821,
+      "step": 655
+    },
+    {
+      "epoch": 0.8688741721854305,
+      "grad_norm": 0.2527983215321377,
+      "learning_rate": 7.86852756729174e-05,
+      "loss": 0.4819,
+      "step": 656
+    },
+    {
+      "epoch": 0.8701986754966887,
+      "grad_norm": 0.2987379598814082,
+      "learning_rate": 7.867585283986664e-05,
+      "loss": 0.4842,
+      "step": 657
+    },
+    {
+      "epoch": 0.871523178807947,
+      "grad_norm": 0.31134849376976437,
+      "learning_rate": 7.86663969281093e-05,
+      "loss": 0.4851,
+      "step": 658
+    },
+    {
+      "epoch": 0.8728476821192053,
+      "grad_norm": 0.299617338657101,
+      "learning_rate": 7.865690794573283e-05,
+      "loss": 0.4818,
+      "step": 659
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 0.3313965959996047,
+      "learning_rate": 7.864738590085297e-05,
+      "loss": 0.4882,
+      "step": 660
+    },
+    {
+      "epoch": 0.8754966887417218,
+      "grad_norm": 0.3432631895467345,
+      "learning_rate": 7.863783080161376e-05,
+      "loss": 0.4813,
+      "step": 661
+    },
+    {
+      "epoch": 0.8768211920529801,
+      "grad_norm": 0.30717653253432015,
+      "learning_rate": 7.862824265618747e-05,
+      "loss": 0.4811,
+      "step": 662
+    },
+    {
+      "epoch": 0.8781456953642384,
+      "grad_norm": 0.25874881971951896,
+      "learning_rate": 7.861862147277466e-05,
+      "loss": 0.4811,
+      "step": 663
+    },
+    {
+      "epoch": 0.8794701986754967,
+      "grad_norm": 0.2913731627623871,
+      "learning_rate": 7.860896725960415e-05,
+      "loss": 0.4833,
+      "step": 664
+    },
+    {
+      "epoch": 0.8807947019867549,
+      "grad_norm": 0.35756305698161034,
+      "learning_rate": 7.859928002493301e-05,
+      "loss": 0.4814,
+      "step": 665
+    },
+    {
+      "epoch": 0.8821192052980132,
+      "grad_norm": 0.3516816572992563,
+      "learning_rate": 7.85895597770465e-05,
+      "loss": 0.481,
+      "step": 666
+    },
+    {
+      "epoch": 0.8834437086092716,
+      "grad_norm": 0.3331370222088617,
+      "learning_rate": 7.857980652425821e-05,
+      "loss": 0.4828,
+      "step": 667
+    },
+    {
+      "epoch": 0.8847682119205298,
+      "grad_norm": 0.3481608528081449,
+      "learning_rate": 7.857002027490987e-05,
+      "loss": 0.4826,
+      "step": 668
+    },
+    {
+      "epoch": 0.8860927152317881,
+      "grad_norm": 0.3392037873959969,
+      "learning_rate": 7.85602010373715e-05,
+      "loss": 0.4847,
+      "step": 669
+    },
+    {
+      "epoch": 0.8874172185430463,
+      "grad_norm": 0.308269061256026,
+      "learning_rate": 7.855034882004129e-05,
+      "loss": 0.4786,
+      "step": 670
+    },
+    {
+      "epoch": 0.8887417218543047,
+      "grad_norm": 0.3334801655936555,
+      "learning_rate": 7.854046363134565e-05,
+      "loss": 0.481,
+      "step": 671
+    },
+    {
+      "epoch": 0.8900662251655629,
+      "grad_norm": 0.3992379184264764,
+      "learning_rate": 7.853054547973917e-05,
+      "loss": 0.4854,
+      "step": 672
+    },
+    {
+      "epoch": 0.8913907284768212,
+      "grad_norm": 0.33908550696365136,
+      "learning_rate": 7.85205943737047e-05,
+      "loss": 0.4813,
+      "step": 673
+    },
+    {
+      "epoch": 0.8927152317880794,
+      "grad_norm": 0.3047711010431091,
+      "learning_rate": 7.851061032175318e-05,
+      "loss": 0.4822,
+      "step": 674
+    },
+    {
+      "epoch": 0.8940397350993378,
+      "grad_norm": 0.40168595466288554,
+      "learning_rate": 7.850059333242381e-05,
+      "loss": 0.4813,
+      "step": 675
+    },
+    {
+      "epoch": 0.895364238410596,
+      "grad_norm": 0.48446403715321423,
+      "learning_rate": 7.849054341428392e-05,
+      "loss": 0.4868,
+      "step": 676
+    },
+    {
+      "epoch": 0.8966887417218543,
+      "grad_norm": 0.3917937553453123,
+      "learning_rate": 7.848046057592898e-05,
+      "loss": 0.4813,
+      "step": 677
+    },
+    {
+      "epoch": 0.8980132450331125,
+      "grad_norm": 0.23644352106920238,
+      "learning_rate": 7.847034482598268e-05,
+      "loss": 0.4829,
+      "step": 678
+    },
+    {
+      "epoch": 0.8993377483443709,
+      "grad_norm": 0.2637214027904466,
+      "learning_rate": 7.846019617309683e-05,
+      "loss": 0.4804,
+      "step": 679
+    },
+    {
+      "epoch": 0.9006622516556292,
+      "grad_norm": 0.279908505779954,
+      "learning_rate": 7.845001462595134e-05,
+      "loss": 0.4838,
+      "step": 680
+    },
+    {
+      "epoch": 0.9019867549668874,
+      "grad_norm": 0.20961128177589614,
+      "learning_rate": 7.843980019325431e-05,
+      "loss": 0.4802,
+      "step": 681
+    },
+    {
+      "epoch": 0.9033112582781457,
+      "grad_norm": 0.2481424384616096,
+      "learning_rate": 7.842955288374195e-05,
+      "loss": 0.4834,
+      "step": 682
+    },
+    {
+      "epoch": 0.904635761589404,
+      "grad_norm": 0.3816502461969246,
+      "learning_rate": 7.841927270617857e-05,
+      "loss": 0.4816,
+      "step": 683
+    },
+    {
+      "epoch": 0.9059602649006623,
+      "grad_norm": 0.4372527580773711,
+      "learning_rate": 7.840895966935665e-05,
+      "loss": 0.4797,
+      "step": 684
+    },
+    {
+      "epoch": 0.9072847682119205,
+      "grad_norm": 0.29547492593962593,
+      "learning_rate": 7.839861378209666e-05,
+      "loss": 0.4846,
+      "step": 685
+    },
+    {
+      "epoch": 0.9086092715231788,
+      "grad_norm": 0.2749375588295858,
+      "learning_rate": 7.83882350532473e-05,
+      "loss": 0.4817,
+      "step": 686
+    },
+    {
+      "epoch": 0.909933774834437,
+      "grad_norm": 0.38275878712261874,
+      "learning_rate": 7.837782349168524e-05,
+      "loss": 0.4823,
+      "step": 687
+    },
+    {
+      "epoch": 0.9112582781456954,
+      "grad_norm": 0.33104961470737865,
+      "learning_rate": 7.836737910631534e-05,
+      "loss": 0.482,
+      "step": 688
+    },
+    {
+      "epoch": 0.9125827814569536,
+      "grad_norm": 0.2377563856877746,
+      "learning_rate": 7.835690190607046e-05,
+      "loss": 0.4824,
+      "step": 689
+    },
+    {
+      "epoch": 0.9139072847682119,
+      "grad_norm": 0.3010875277415439,
+      "learning_rate": 7.834639189991153e-05,
+      "loss": 0.4807,
+      "step": 690
+    },
+    {
+      "epoch": 0.9152317880794701,
+      "grad_norm": 0.27459747828347714,
+      "learning_rate": 7.833584909682756e-05,
+      "loss": 0.4765,
+      "step": 691
+    },
+    {
+      "epoch": 0.9165562913907285,
+      "grad_norm": 0.2515488491182144,
+      "learning_rate": 7.832527350583563e-05,
+      "loss": 0.4814,
+      "step": 692
+    },
+    {
+      "epoch": 0.9178807947019868,
+      "grad_norm": 0.21719190344570832,
+      "learning_rate": 7.831466513598081e-05,
+      "loss": 0.4791,
+      "step": 693
+    },
+    {
+      "epoch": 0.919205298013245,
+      "grad_norm": 0.23182356502794255,
+      "learning_rate": 7.830402399633624e-05,
+      "loss": 0.4862,
+      "step": 694
+    },
+    {
+      "epoch": 0.9205298013245033,
+      "grad_norm": 0.2689850595331572,
+      "learning_rate": 7.82933500960031e-05,
+      "loss": 0.481,
+      "step": 695
+    },
+    {
+      "epoch": 0.9218543046357616,
+      "grad_norm": 0.2823307350570989,
+      "learning_rate": 7.828264344411053e-05,
+      "loss": 0.4816,
+      "step": 696
+    },
+    {
+      "epoch": 0.9231788079470199,
+      "grad_norm": 0.27990834741655785,
+      "learning_rate": 7.827190404981575e-05,
+      "loss": 0.4784,
+      "step": 697
+    },
+    {
+      "epoch": 0.9245033112582781,
+      "grad_norm": 0.2492731503871584,
+      "learning_rate": 7.826113192230395e-05,
+      "loss": 0.4836,
+      "step": 698
+    },
+    {
+      "epoch": 0.9258278145695364,
+      "grad_norm": 0.20529132507609182,
+      "learning_rate": 7.825032707078832e-05,
+      "loss": 0.4815,
+      "step": 699
+    },
+    {
+      "epoch": 0.9271523178807947,
+      "grad_norm": 0.2087048107147062,
+      "learning_rate": 7.823948950451005e-05,
+      "loss": 0.4804,
+      "step": 700
+    },
+    {
+      "epoch": 0.928476821192053,
+      "grad_norm": 0.2620437706439073,
+      "learning_rate": 7.82286192327383e-05,
+      "loss": 0.478,
+      "step": 701
+    },
+    {
+      "epoch": 0.9298013245033112,
+      "grad_norm": 0.32925088172767214,
+      "learning_rate": 7.821771626477017e-05,
+      "loss": 0.4814,
+      "step": 702
+    },
+    {
+      "epoch": 0.9311258278145695,
+      "grad_norm": 0.37310606702027327,
+      "learning_rate": 7.820678060993078e-05,
+      "loss": 0.482,
+      "step": 703
+    },
+    {
+      "epoch": 0.9324503311258279,
+      "grad_norm": 0.3281609752012142,
+      "learning_rate": 7.819581227757321e-05,
+      "loss": 0.4795,
+      "step": 704
+    },
+    {
+      "epoch": 0.9337748344370861,
+      "grad_norm": 0.2638706491772062,
+      "learning_rate": 7.818481127707842e-05,
+      "loss": 0.4862,
+      "step": 705
+    },
+    {
+      "epoch": 0.9350993377483444,
+      "grad_norm": 0.2783208359135352,
+      "learning_rate": 7.817377761785536e-05,
+      "loss": 0.4786,
+      "step": 706
+    },
+    {
+      "epoch": 0.9364238410596026,
+      "grad_norm": 0.3249237459860902,
+      "learning_rate": 7.816271130934092e-05,
+      "loss": 0.4789,
+      "step": 707
+    },
+    {
+      "epoch": 0.937748344370861,
+      "grad_norm": 0.4250076609216278,
+      "learning_rate": 7.81516123609999e-05,
+      "loss": 0.4791,
+      "step": 708
+    },
+    {
+      "epoch": 0.9390728476821192,
+      "grad_norm": 0.5465774781999677,
+      "learning_rate": 7.8140480782325e-05,
+      "loss": 0.4748,
+      "step": 709
+    },
+    {
+      "epoch": 0.9403973509933775,
+      "grad_norm": 0.551071876939898,
+      "learning_rate": 7.812931658283686e-05,
+      "loss": 0.4836,
+      "step": 710
+    },
+    {
+      "epoch": 0.9417218543046357,
+      "grad_norm": 0.4853450339248303,
+      "learning_rate": 7.811811977208399e-05,
+      "loss": 0.4821,
+      "step": 711
+    },
+    {
+      "epoch": 0.9430463576158941,
+      "grad_norm": 0.416291567363337,
+      "learning_rate": 7.81068903596428e-05,
+      "loss": 0.4853,
+      "step": 712
+    },
+    {
+      "epoch": 0.9443708609271523,
+      "grad_norm": 0.40584990524999903,
+      "learning_rate": 7.80956283551176e-05,
+      "loss": 0.4833,
+      "step": 713
+    },
+    {
+      "epoch": 0.9456953642384106,
+      "grad_norm": 0.3421893558441057,
+      "learning_rate": 7.808433376814057e-05,
+      "loss": 0.4792,
+      "step": 714
+    },
+    {
+      "epoch": 0.9470198675496688,
+      "grad_norm": 0.26192378408362266,
+      "learning_rate": 7.807300660837174e-05,
+      "loss": 0.4771,
+      "step": 715
+    },
+    {
+      "epoch": 0.9483443708609272,
+      "grad_norm": 0.3307013997948393,
+      "learning_rate": 7.806164688549901e-05,
+      "loss": 0.4782,
+      "step": 716
+    },
+    {
+      "epoch": 0.9496688741721855,
+      "grad_norm": 0.37098894990171905,
+      "learning_rate": 7.805025460923814e-05,
+      "loss": 0.4807,
+      "step": 717
+    },
+    {
+      "epoch": 0.9509933774834437,
+      "grad_norm": 0.31418671423072825,
+      "learning_rate": 7.80388297893327e-05,
+      "loss": 0.4813,
+      "step": 718
+    },
+    {
+      "epoch": 0.952317880794702,
+      "grad_norm": 0.2288167145145745,
+      "learning_rate": 7.802737243555415e-05,
+      "loss": 0.4825,
+      "step": 719
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 0.22553498975705524,
+      "learning_rate": 7.801588255770172e-05,
+      "loss": 0.4798,
+      "step": 720
+    },
+    {
+      "epoch": 0.9549668874172186,
+      "grad_norm": 0.3086376109173349,
+      "learning_rate": 7.800436016560248e-05,
+      "loss": 0.4798,
+      "step": 721
+    },
+    {
+      "epoch": 0.9562913907284768,
+      "grad_norm": 0.3116396291619925,
+      "learning_rate": 7.799280526911132e-05,
+      "loss": 0.4811,
+      "step": 722
+    },
+    {
+      "epoch": 0.9576158940397351,
+      "grad_norm": 0.22990942051526322,
+      "learning_rate": 7.79812178781109e-05,
+      "loss": 0.4799,
+      "step": 723
+    },
+    {
+      "epoch": 0.9589403973509933,
+      "grad_norm": 0.2227839042918503,
+      "learning_rate": 7.796959800251171e-05,
+      "loss": 0.4825,
+      "step": 724
+    },
+    {
+      "epoch": 0.9602649006622517,
+      "grad_norm": 0.2858407623269414,
+      "learning_rate": 7.795794565225199e-05,
+      "loss": 0.4773,
+      "step": 725
+    },
+    {
+      "epoch": 0.9615894039735099,
+      "grad_norm": 0.25822521728761383,
+      "learning_rate": 7.79462608372978e-05,
+      "loss": 0.4796,
+      "step": 726
+    },
+    {
+      "epoch": 0.9629139072847682,
+      "grad_norm": 0.2129808588513104,
+      "learning_rate": 7.793454356764289e-05,
+      "loss": 0.4772,
+      "step": 727
+    },
+    {
+      "epoch": 0.9642384105960264,
+      "grad_norm": 0.22199646848608,
+      "learning_rate": 7.792279385330884e-05,
+      "loss": 0.4797,
+      "step": 728
+    },
+    {
+      "epoch": 0.9655629139072848,
+      "grad_norm": 0.2122461945927918,
+      "learning_rate": 7.791101170434495e-05,
+      "loss": 0.4788,
+      "step": 729
+    },
+    {
+      "epoch": 0.9668874172185431,
+      "grad_norm": 0.21497031200802819,
+      "learning_rate": 7.789919713082826e-05,
+      "loss": 0.4786,
+      "step": 730
+    },
+    {
+      "epoch": 0.9682119205298013,
+      "grad_norm": 0.22826103454737598,
+      "learning_rate": 7.788735014286354e-05,
+      "loss": 0.4786,
+      "step": 731
+    },
+    {
+      "epoch": 0.9695364238410596,
+      "grad_norm": 0.24424654894084258,
+      "learning_rate": 7.78754707505833e-05,
+      "loss": 0.4792,
+      "step": 732
+    },
+    {
+      "epoch": 0.9708609271523179,
+      "grad_norm": 0.22797501623996758,
+      "learning_rate": 7.786355896414775e-05,
+      "loss": 0.4805,
+      "step": 733
+    },
+    {
+      "epoch": 0.9721854304635762,
+      "grad_norm": 0.20042848442174116,
+      "learning_rate": 7.785161479374481e-05,
+      "loss": 0.4765,
+      "step": 734
+    },
+    {
+      "epoch": 0.9735099337748344,
+      "grad_norm": 0.1708682562526585,
+      "learning_rate": 7.78396382495901e-05,
+      "loss": 0.482,
+      "step": 735
+    },
+    {
+      "epoch": 0.9748344370860927,
+      "grad_norm": 0.18625605366316933,
+      "learning_rate": 7.782762934192692e-05,
+      "loss": 0.4783,
+      "step": 736
+    },
+    {
+      "epoch": 0.976158940397351,
+      "grad_norm": 0.23441836249013337,
+      "learning_rate": 7.781558808102626e-05,
+      "loss": 0.4801,
+      "step": 737
+    },
+    {
+      "epoch": 0.9774834437086093,
+      "grad_norm": 0.23861586275339544,
+      "learning_rate": 7.780351447718676e-05,
+      "loss": 0.4772,
+      "step": 738
+    },
+    {
+      "epoch": 0.9788079470198675,
+      "grad_norm": 0.2481946271154018,
+      "learning_rate": 7.779140854073477e-05,
+      "loss": 0.4818,
+      "step": 739
+    },
+    {
+      "epoch": 0.9801324503311258,
+      "grad_norm": 0.28521797646834773,
+      "learning_rate": 7.777927028202425e-05,
+      "loss": 0.4769,
+      "step": 740
+    },
+    {
+      "epoch": 0.9814569536423841,
+      "grad_norm": 0.37290541277431394,
+      "learning_rate": 7.77670997114368e-05,
+      "loss": 0.4771,
+      "step": 741
+    },
+    {
+      "epoch": 0.9827814569536424,
+      "grad_norm": 0.405505212332374,
+      "learning_rate": 7.775489683938169e-05,
+      "loss": 0.481,
+      "step": 742
+    },
+    {
+      "epoch": 0.9841059602649007,
+      "grad_norm": 0.3975640233328856,
+      "learning_rate": 7.77426616762958e-05,
+      "loss": 0.4813,
+      "step": 743
+    },
+    {
+      "epoch": 0.9854304635761589,
+      "grad_norm": 0.3423929950698897,
+      "learning_rate": 7.773039423264362e-05,
+      "loss": 0.4796,
+      "step": 744
+    },
+    {
+      "epoch": 0.9867549668874173,
+      "grad_norm": 0.26404635732713716,
+      "learning_rate": 7.771809451891726e-05,
+      "loss": 0.481,
+      "step": 745
+    },
+    {
+      "epoch": 0.9880794701986755,
+      "grad_norm": 0.22570895737744853,
+      "learning_rate": 7.770576254563643e-05,
+      "loss": 0.4803,
+      "step": 746
+    },
+    {
+      "epoch": 0.9894039735099338,
+      "grad_norm": 0.2575767643047987,
+      "learning_rate": 7.769339832334843e-05,
+      "loss": 0.4812,
+      "step": 747
+    },
+    {
+      "epoch": 0.990728476821192,
+      "grad_norm": 0.2920721372133632,
+      "learning_rate": 7.768100186262812e-05,
+      "loss": 0.478,
+      "step": 748
+    },
+    {
+      "epoch": 0.9920529801324504,
+      "grad_norm": 0.28969990427597975,
+      "learning_rate": 7.766857317407799e-05,
+      "loss": 0.4795,
+      "step": 749
+    },
+    {
+      "epoch": 0.9933774834437086,
+      "grad_norm": 0.2966368362281624,
+      "learning_rate": 7.765611226832802e-05,
+      "loss": 0.4793,
+      "step": 750
+    },
+    {
+      "epoch": 0.9947019867549669,
+      "grad_norm": 0.33740161692349524,
+      "learning_rate": 7.764361915603578e-05,
+      "loss": 0.4794,
+      "step": 751
+    },
+    {
+      "epoch": 0.9960264900662251,
+      "grad_norm": 0.35333120573462484,
+      "learning_rate": 7.763109384788641e-05,
+      "loss": 0.4829,
+      "step": 752
+    },
+    {
+      "epoch": 0.9973509933774835,
+      "grad_norm": 0.3599567804900422,
+      "learning_rate": 7.761853635459256e-05,
+      "loss": 0.48,
+      "step": 753
+    },
+    {
+      "epoch": 0.9986754966887417,
+      "grad_norm": 0.342135217600496,
+      "learning_rate": 7.76059466868944e-05,
+      "loss": 0.4816,
+      "step": 754
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29846343152057386,
+      "learning_rate": 7.759332485555964e-05,
+      "loss": 0.4756,
+      "step": 755
+    },
+    {
+      "epoch": 1.0013245033112583,
+      "grad_norm": 0.2920664725458839,
+      "learning_rate": 7.758067087138346e-05,
+      "loss": 0.4642,
+      "step": 756
+    },
+    {
+      "epoch": 1.0026490066225167,
+      "grad_norm": 0.3079898777228831,
+      "learning_rate": 7.756798474518859e-05,
+      "loss": 0.4671,
+      "step": 757
+    },
+    {
+      "epoch": 1.0039735099337748,
+      "grad_norm": 0.33499366397020724,
+      "learning_rate": 7.755526648782523e-05,
+      "loss": 0.4655,
+      "step": 758
+    },
+    {
+      "epoch": 1.005298013245033,
+      "grad_norm": 0.3406767958077454,
+      "learning_rate": 7.754251611017105e-05,
+      "loss": 0.4672,
+      "step": 759
+    },
+    {
+      "epoch": 1.0066225165562914,
+      "grad_norm": 0.32091481317655224,
+      "learning_rate": 7.752973362313122e-05,
+      "loss": 0.4665,
+      "step": 760
+    },
+    {
+      "epoch": 1.0079470198675498,
+      "grad_norm": 0.3067019383572125,
+      "learning_rate": 7.751691903763832e-05,
+      "loss": 0.4683,
+      "step": 761
+    },
+    {
+      "epoch": 1.0092715231788079,
+      "grad_norm": 0.3054505230397844,
+      "learning_rate": 7.750407236465243e-05,
+      "loss": 0.466,
+      "step": 762
+    },
+    {
+      "epoch": 1.0105960264900662,
+      "grad_norm": 0.2794059751200471,
+      "learning_rate": 7.749119361516107e-05,
+      "loss": 0.4656,
+      "step": 763
+    },
+    {
+      "epoch": 1.0119205298013245,
+      "grad_norm": 0.24752659970523377,
+      "learning_rate": 7.747828280017917e-05,
+      "loss": 0.4654,
+      "step": 764
+    },
+    {
+      "epoch": 1.0132450331125828,
+      "grad_norm": 0.2544624150174001,
+      "learning_rate": 7.746533993074912e-05,
+      "loss": 0.4644,
+      "step": 765
+    },
+    {
+      "epoch": 1.014569536423841,
+      "grad_norm": 0.25548196077297153,
+      "learning_rate": 7.745236501794068e-05,
+      "loss": 0.4684,
+      "step": 766
+    },
+    {
+      "epoch": 1.0158940397350993,
+      "grad_norm": 0.2822319305420211,
+      "learning_rate": 7.743935807285108e-05,
+      "loss": 0.4657,
+      "step": 767
+    },
+    {
+      "epoch": 1.0172185430463576,
+      "grad_norm": 0.28723007069567386,
+      "learning_rate": 7.742631910660486e-05,
+      "loss": 0.4691,
+      "step": 768
+    },
+    {
+      "epoch": 1.018543046357616,
+      "grad_norm": 0.24247411255980603,
+      "learning_rate": 7.741324813035402e-05,
+      "loss": 0.4628,
+      "step": 769
+    },
+    {
+      "epoch": 1.0198675496688743,
+      "grad_norm": 0.22880484739726545,
+      "learning_rate": 7.74001451552779e-05,
+      "loss": 0.4668,
+      "step": 770
+    },
+    {
+      "epoch": 1.0211920529801324,
+      "grad_norm": 0.2553450750954043,
+      "learning_rate": 7.738701019258324e-05,
+      "loss": 0.4664,
+      "step": 771
+    },
+    {
+      "epoch": 1.0225165562913907,
+      "grad_norm": 0.2801003033681704,
+      "learning_rate": 7.737384325350411e-05,
+      "loss": 0.4668,
+      "step": 772
+    },
+    {
+      "epoch": 1.023841059602649,
+      "grad_norm": 0.26832971058523225,
+      "learning_rate": 7.736064434930193e-05,
+      "loss": 0.4663,
+      "step": 773
+    },
+    {
+      "epoch": 1.0251655629139074,
+      "grad_norm": 0.24450272099660988,
+      "learning_rate": 7.734741349126547e-05,
+      "loss": 0.4652,
+      "step": 774
+    },
+    {
+      "epoch": 1.0264900662251655,
+      "grad_norm": 0.2864157484459465,
+      "learning_rate": 7.733415069071083e-05,
+      "loss": 0.4682,
+      "step": 775
+    },
+    {
+      "epoch": 1.0278145695364238,
+      "grad_norm": 0.3091302960126214,
+      "learning_rate": 7.732085595898143e-05,
+      "loss": 0.4684,
+      "step": 776
+    },
+    {
+      "epoch": 1.0291390728476821,
+      "grad_norm": 0.2855131610310513,
+      "learning_rate": 7.730752930744798e-05,
+      "loss": 0.4647,
+      "step": 777
+    },
+    {
+      "epoch": 1.0304635761589405,
+      "grad_norm": 0.28825301621492705,
+      "learning_rate": 7.729417074750853e-05,
+      "loss": 0.4643,
+      "step": 778
+    },
+    {
+      "epoch": 1.0317880794701986,
+      "grad_norm": 0.30165431193632924,
+      "learning_rate": 7.728078029058837e-05,
+      "loss": 0.4678,
+      "step": 779
+    },
+    {
+      "epoch": 1.033112582781457,
+      "grad_norm": 0.36513890585897735,
+      "learning_rate": 7.726735794814013e-05,
+      "loss": 0.4676,
+      "step": 780
+    },
+    {
+      "epoch": 1.0344370860927152,
+      "grad_norm": 0.332467784835273,
+      "learning_rate": 7.725390373164365e-05,
+      "loss": 0.465,
+      "step": 781
+    },
+    {
+      "epoch": 1.0357615894039736,
+      "grad_norm": 0.272305051221066,
+      "learning_rate": 7.724041765260607e-05,
+      "loss": 0.466,
+      "step": 782
+    },
+    {
+      "epoch": 1.0370860927152319,
+      "grad_norm": 0.2371494701709062,
+      "learning_rate": 7.722689972256178e-05,
+      "loss": 0.4655,
+      "step": 783
+    },
+    {
+      "epoch": 1.03841059602649,
+      "grad_norm": 0.23570599124393343,
+      "learning_rate": 7.721334995307239e-05,
+      "loss": 0.4634,
+      "step": 784
+    },
+    {
+      "epoch": 1.0397350993377483,
+      "grad_norm": 0.22060205870197463,
+      "learning_rate": 7.719976835572677e-05,
+      "loss": 0.4661,
+      "step": 785
+    },
+    {
+      "epoch": 1.0410596026490067,
+      "grad_norm": 0.21617564094458827,
+      "learning_rate": 7.718615494214098e-05,
+      "loss": 0.4666,
+      "step": 786
+    },
+    {
+      "epoch": 1.042384105960265,
+      "grad_norm": 0.23977442565016757,
+      "learning_rate": 7.717250972395833e-05,
+      "loss": 0.4673,
+      "step": 787
+    },
+    {
+      "epoch": 1.043708609271523,
+      "grad_norm": 0.21398065428880023,
+      "learning_rate": 7.715883271284927e-05,
+      "loss": 0.4705,
+      "step": 788
+    },
+    {
+      "epoch": 1.0450331125827814,
+      "grad_norm": 0.20304220608237689,
+      "learning_rate": 7.714512392051153e-05,
+      "loss": 0.4633,
+      "step": 789
+    },
+    {
+      "epoch": 1.0463576158940397,
+      "grad_norm": 0.24222695154173474,
+      "learning_rate": 7.713138335866995e-05,
+      "loss": 0.467,
+      "step": 790
+    },
+    {
+      "epoch": 1.047682119205298,
+      "grad_norm": 0.2544460070269825,
+      "learning_rate": 7.711761103907657e-05,
+      "loss": 0.465,
+      "step": 791
+    },
+    {
+      "epoch": 1.0490066225165562,
+      "grad_norm": 0.23361487179674312,
+      "learning_rate": 7.71038069735106e-05,
+      "loss": 0.4671,
+      "step": 792
+    },
+    {
+      "epoch": 1.0503311258278145,
+      "grad_norm": 0.24282395686138458,
+      "learning_rate": 7.708997117377835e-05,
+      "loss": 0.465,
+      "step": 793
+    },
+    {
+      "epoch": 1.0516556291390728,
+      "grad_norm": 0.24799669527857493,
+      "learning_rate": 7.707610365171336e-05,
+      "loss": 0.4637,
+      "step": 794
+    },
+    {
+      "epoch": 1.0529801324503312,
+      "grad_norm": 0.2826400941937532,
+      "learning_rate": 7.706220441917621e-05,
+      "loss": 0.4701,
+      "step": 795
+    },
+    {
+      "epoch": 1.0543046357615895,
+      "grad_norm": 0.29178340660854385,
+      "learning_rate": 7.704827348805468e-05,
+      "loss": 0.4648,
+      "step": 796
+    },
+    {
+      "epoch": 1.0556291390728476,
+      "grad_norm": 0.22751585892522452,
+      "learning_rate": 7.703431087026361e-05,
+      "loss": 0.4673,
+      "step": 797
+    },
+    {
+      "epoch": 1.056953642384106,
+      "grad_norm": 0.15568161607960904,
+      "learning_rate": 7.702031657774495e-05,
+      "loss": 0.4681,
+      "step": 798
+    },
+    {
+      "epoch": 1.0582781456953643,
+      "grad_norm": 0.15475983387346162,
+      "learning_rate": 7.700629062246777e-05,
+      "loss": 0.4647,
+      "step": 799
+    },
+    {
+      "epoch": 1.0596026490066226,
+      "grad_norm": 0.2144024036837161,
+      "learning_rate": 7.699223301642816e-05,
+      "loss": 0.4639,
+      "step": 800
+    },
+    {
+      "epoch": 1.0609271523178807,
+      "grad_norm": 0.25672647862967335,
+      "learning_rate": 7.697814377164934e-05,
+      "loss": 0.4659,
+      "step": 801
+    },
+    {
+      "epoch": 1.062251655629139,
+      "grad_norm": 0.2639927208494949,
+      "learning_rate": 7.696402290018156e-05,
+      "loss": 0.4624,
+      "step": 802
+    },
+    {
+      "epoch": 1.0635761589403974,
+      "grad_norm": 0.265745242756065,
+      "learning_rate": 7.694987041410213e-05,
+      "loss": 0.4658,
+      "step": 803
+    },
+    {
+      "epoch": 1.0649006622516557,
+      "grad_norm": 0.2683789443166138,
+      "learning_rate": 7.693568632551543e-05,
+      "loss": 0.4657,
+      "step": 804
+    },
+    {
+      "epoch": 1.0662251655629138,
+      "grad_norm": 0.2783893203841122,
+      "learning_rate": 7.692147064655278e-05,
+      "loss": 0.4663,
+      "step": 805
+    },
+    {
+      "epoch": 1.0675496688741721,
+      "grad_norm": 0.2564155394000131,
+      "learning_rate": 7.690722338937261e-05,
+      "loss": 0.4612,
+      "step": 806
+    },
+    {
+      "epoch": 1.0688741721854305,
+      "grad_norm": 0.21833661967385223,
+      "learning_rate": 7.689294456616031e-05,
+      "loss": 0.469,
+      "step": 807
+    },
+    {
+      "epoch": 1.0701986754966888,
+      "grad_norm": 0.1728388380273878,
+      "learning_rate": 7.687863418912827e-05,
+      "loss": 0.4639,
+      "step": 808
+    },
+    {
+      "epoch": 1.0715231788079471,
+      "grad_norm": 0.18218967654995266,
+      "learning_rate": 7.686429227051591e-05,
+      "loss": 0.465,
+      "step": 809
+    },
+    {
+      "epoch": 1.0728476821192052,
+      "grad_norm": 0.19928231994899645,
+      "learning_rate": 7.684991882258957e-05,
+      "loss": 0.466,
+      "step": 810
+    },
+    {
+      "epoch": 1.0741721854304636,
+      "grad_norm": 0.23238843209999333,
+      "learning_rate": 7.683551385764259e-05,
+      "loss": 0.4627,
+      "step": 811
+    },
+    {
+      "epoch": 1.0754966887417219,
+      "grad_norm": 0.2815597537914883,
+      "learning_rate": 7.682107738799524e-05,
+      "loss": 0.4665,
+      "step": 812
+    },
+    {
+      "epoch": 1.0768211920529802,
+      "grad_norm": 0.3337657315603249,
+      "learning_rate": 7.680660942599476e-05,
+      "loss": 0.4667,
+      "step": 813
+    },
+    {
+      "epoch": 1.0781456953642383,
+      "grad_norm": 0.3777897075017307,
+      "learning_rate": 7.679210998401534e-05,
+      "loss": 0.4686,
+      "step": 814
+    },
+    {
+      "epoch": 1.0794701986754967,
+      "grad_norm": 0.3891871217250742,
+      "learning_rate": 7.677757907445805e-05,
+      "loss": 0.4664,
+      "step": 815
+    },
+    {
+      "epoch": 1.080794701986755,
+      "grad_norm": 0.3198865389374094,
+      "learning_rate": 7.67630167097509e-05,
+      "loss": 0.466,
+      "step": 816
+    },
+    {
+      "epoch": 1.0821192052980133,
+      "grad_norm": 0.2858348869484707,
+      "learning_rate": 7.67484229023488e-05,
+      "loss": 0.4676,
+      "step": 817
+    },
+    {
+      "epoch": 1.0834437086092716,
+      "grad_norm": 0.37771195019935544,
+      "learning_rate": 7.673379766473355e-05,
+      "loss": 0.4638,
+      "step": 818
+    },
+    {
+      "epoch": 1.0847682119205297,
+      "grad_norm": 0.41871720097165804,
+      "learning_rate": 7.671914100941382e-05,
+      "loss": 0.4683,
+      "step": 819
+    },
+    {
+      "epoch": 1.086092715231788,
+      "grad_norm": 0.3965414188799175,
+      "learning_rate": 7.67044529489252e-05,
+      "loss": 0.4687,
+      "step": 820
+    },
+    {
+      "epoch": 1.0874172185430464,
+      "grad_norm": 0.362666771906542,
+      "learning_rate": 7.668973349583004e-05,
+      "loss": 0.4649,
+      "step": 821
+    },
+    {
+      "epoch": 1.0887417218543047,
+      "grad_norm": 0.3151779056033609,
+      "learning_rate": 7.667498266271765e-05,
+      "loss": 0.4666,
+      "step": 822
+    },
+    {
+      "epoch": 1.0900662251655628,
+      "grad_norm": 0.25311950379727277,
+      "learning_rate": 7.666020046220413e-05,
+      "loss": 0.4648,
+      "step": 823
+    },
+    {
+      "epoch": 1.0913907284768212,
+      "grad_norm": 0.253369681771004,
+      "learning_rate": 7.664538690693239e-05,
+      "loss": 0.4661,
+      "step": 824
+    },
+    {
+      "epoch": 1.0927152317880795,
+      "grad_norm": 0.31592741788320766,
+      "learning_rate": 7.66305420095722e-05,
+      "loss": 0.4668,
+      "step": 825
+    },
+    {
+      "epoch": 1.0940397350993378,
+      "grad_norm": 0.34512629787623816,
+      "learning_rate": 7.661566578282008e-05,
+      "loss": 0.4673,
+      "step": 826
+    },
+    {
+      "epoch": 1.095364238410596,
+      "grad_norm": 0.28046958705515346,
+      "learning_rate": 7.660075823939941e-05,
+      "loss": 0.4655,
+      "step": 827
+    },
+    {
+      "epoch": 1.0966887417218543,
+      "grad_norm": 0.25468713198972986,
+      "learning_rate": 7.65858193920603e-05,
+      "loss": 0.4684,
+      "step": 828
+    },
+    {
+      "epoch": 1.0980132450331126,
+      "grad_norm": 0.3153291395316737,
+      "learning_rate": 7.657084925357969e-05,
+      "loss": 0.4602,
+      "step": 829
+    },
+    {
+      "epoch": 1.099337748344371,
+      "grad_norm": 0.33358172519991275,
+      "learning_rate": 7.65558478367612e-05,
+      "loss": 0.4639,
+      "step": 830
+    },
+    {
+      "epoch": 1.100662251655629,
+      "grad_norm": 0.29492663842063044,
+      "learning_rate": 7.654081515443527e-05,
+      "loss": 0.4682,
+      "step": 831
+    },
+    {
+      "epoch": 1.1019867549668874,
+      "grad_norm": 0.203804711012212,
+      "learning_rate": 7.652575121945909e-05,
+      "loss": 0.4656,
+      "step": 832
+    },
+    {
+      "epoch": 1.1033112582781457,
+      "grad_norm": 0.16044268575777845,
+      "learning_rate": 7.651065604471651e-05,
+      "loss": 0.4655,
+      "step": 833
+    },
+    {
+      "epoch": 1.104635761589404,
+      "grad_norm": 0.16854387475463642,
+      "learning_rate": 7.649552964311817e-05,
+      "loss": 0.4644,
+      "step": 834
+    },
+    {
+      "epoch": 1.1059602649006623,
+      "grad_norm": 0.18965914423225044,
+      "learning_rate": 7.648037202760137e-05,
+      "loss": 0.4673,
+      "step": 835
+    },
+    {
+      "epoch": 1.1072847682119205,
+      "grad_norm": 0.2494294793657683,
+      "learning_rate": 7.646518321113013e-05,
+      "loss": 0.4659,
+      "step": 836
+    },
+    {
+      "epoch": 1.1086092715231788,
+      "grad_norm": 0.25623203438887143,
+      "learning_rate": 7.644996320669515e-05,
+      "loss": 0.4636,
+      "step": 837
+    },
+    {
+      "epoch": 1.1099337748344371,
+      "grad_norm": 0.26259009042708653,
+      "learning_rate": 7.643471202731381e-05,
+      "loss": 0.4671,
+      "step": 838
+    },
+    {
+      "epoch": 1.1112582781456954,
+      "grad_norm": 0.31370203796185736,
+      "learning_rate": 7.641942968603014e-05,
+      "loss": 0.4694,
+      "step": 839
+    },
+    {
+      "epoch": 1.1125827814569536,
+      "grad_norm": 0.29479607221538773,
+      "learning_rate": 7.640411619591484e-05,
+      "loss": 0.4648,
+      "step": 840
+    },
+    {
+      "epoch": 1.1139072847682119,
+      "grad_norm": 0.22388843778074605,
+      "learning_rate": 7.638877157006523e-05,
+      "loss": 0.4658,
+      "step": 841
+    },
+    {
+      "epoch": 1.1152317880794702,
+      "grad_norm": 0.19485539020131404,
+      "learning_rate": 7.637339582160528e-05,
+      "loss": 0.4667,
+      "step": 842
+    },
+    {
+      "epoch": 1.1165562913907285,
+      "grad_norm": 0.1970175068749773,
+      "learning_rate": 7.635798896368555e-05,
+      "loss": 0.4665,
+      "step": 843
+    },
+    {
+      "epoch": 1.1178807947019869,
+      "grad_norm": 0.2032788203007247,
+      "learning_rate": 7.634255100948325e-05,
+      "loss": 0.4666,
+      "step": 844
+    },
+    {
+      "epoch": 1.119205298013245,
+      "grad_norm": 0.23826875146293003,
+      "learning_rate": 7.632708197220215e-05,
+      "loss": 0.4613,
+      "step": 845
+    },
+    {
+      "epoch": 1.1205298013245033,
+      "grad_norm": 0.2916576358774355,
+      "learning_rate": 7.631158186507263e-05,
+      "loss": 0.4663,
+      "step": 846
+    },
+    {
+      "epoch": 1.1218543046357616,
+      "grad_norm": 0.3113753618036428,
+      "learning_rate": 7.629605070135162e-05,
+      "loss": 0.466,
+      "step": 847
+    },
+    {
+      "epoch": 1.12317880794702,
+      "grad_norm": 0.31452966957665085,
+      "learning_rate": 7.628048849432262e-05,
+      "loss": 0.4624,
+      "step": 848
+    },
+    {
+      "epoch": 1.124503311258278,
+      "grad_norm": 0.2766729457040173,
+      "learning_rate": 7.626489525729568e-05,
+      "loss": 0.4632,
+      "step": 849
+    },
+    {
+      "epoch": 1.1258278145695364,
+      "grad_norm": 0.23639099963713267,
+      "learning_rate": 7.624927100360742e-05,
+      "loss": 0.4648,
+      "step": 850
+    },
+    {
+      "epoch": 1.1271523178807947,
+      "grad_norm": 0.2543543416615765,
+      "learning_rate": 7.623361574662094e-05,
+      "loss": 0.4644,
+      "step": 851
+    },
+    {
+      "epoch": 1.128476821192053,
+      "grad_norm": 0.26302680239388593,
+      "learning_rate": 7.621792949972588e-05,
+      "loss": 0.4659,
+      "step": 852
+    },
+    {
+      "epoch": 1.1298013245033112,
+      "grad_norm": 0.2812169310193035,
+      "learning_rate": 7.620221227633837e-05,
+      "loss": 0.4669,
+      "step": 853
+    },
+    {
+      "epoch": 1.1311258278145695,
+      "grad_norm": 0.3093678668935426,
+      "learning_rate": 7.618646408990106e-05,
+      "loss": 0.4644,
+      "step": 854
+    },
+    {
+      "epoch": 1.1324503311258278,
+      "grad_norm": 0.3251171302100022,
+      "learning_rate": 7.617068495388308e-05,
+      "loss": 0.4679,
+      "step": 855
+    },
+    {
+      "epoch": 1.1337748344370862,
+      "grad_norm": 0.2790180323371915,
+      "learning_rate": 7.615487488178002e-05,
+      "loss": 0.4659,
+      "step": 856
+    },
+    {
+      "epoch": 1.1350993377483443,
+      "grad_norm": 0.2401769033505207,
+      "learning_rate": 7.613903388711388e-05,
+      "loss": 0.4681,
+      "step": 857
+    },
+    {
+      "epoch": 1.1364238410596026,
+      "grad_norm": 0.23823696754404458,
+      "learning_rate": 7.612316198343321e-05,
+      "loss": 0.4633,
+      "step": 858
+    },
+    {
+      "epoch": 1.137748344370861,
+      "grad_norm": 0.2268348719737744,
+      "learning_rate": 7.610725918431292e-05,
+      "loss": 0.4622,
+      "step": 859
+    },
+    {
+      "epoch": 1.1390728476821192,
+      "grad_norm": 0.20981703643070393,
+      "learning_rate": 7.609132550335437e-05,
+      "loss": 0.4646,
+      "step": 860
+    },
+    {
+      "epoch": 1.1403973509933776,
+      "grad_norm": 0.20208310879369645,
+      "learning_rate": 7.607536095418532e-05,
+      "loss": 0.4635,
+      "step": 861
+    },
+    {
+      "epoch": 1.1417218543046357,
+      "grad_norm": 0.2162568259277357,
+      "learning_rate": 7.605936555045995e-05,
+      "loss": 0.4627,
+      "step": 862
+    },
+    {
+      "epoch": 1.143046357615894,
+      "grad_norm": 0.23057998758158343,
+      "learning_rate": 7.60433393058588e-05,
+      "loss": 0.4639,
+      "step": 863
+    },
+    {
+      "epoch": 1.1443708609271523,
+      "grad_norm": 0.24583425029643566,
+      "learning_rate": 7.602728223408884e-05,
+      "loss": 0.4688,
+      "step": 864
+    },
+    {
+      "epoch": 1.1456953642384107,
+      "grad_norm": 0.2756709687344738,
+      "learning_rate": 7.601119434888333e-05,
+      "loss": 0.4644,
+      "step": 865
+    },
+    {
+      "epoch": 1.1470198675496688,
+      "grad_norm": 0.32159628294784376,
+      "learning_rate": 7.599507566400195e-05,
+      "loss": 0.4652,
+      "step": 866
+    },
+    {
+      "epoch": 1.148344370860927,
+      "grad_norm": 0.3266316332271428,
+      "learning_rate": 7.597892619323071e-05,
+      "loss": 0.4685,
+      "step": 867
+    },
+    {
+      "epoch": 1.1496688741721854,
+      "grad_norm": 0.32000623785998406,
+      "learning_rate": 7.596274595038191e-05,
+      "loss": 0.4674,
+      "step": 868
+    },
+    {
+      "epoch": 1.1509933774834438,
+      "grad_norm": 0.2846199188396625,
+      "learning_rate": 7.594653494929422e-05,
+      "loss": 0.4627,
+      "step": 869
+    },
+    {
+      "epoch": 1.152317880794702,
+      "grad_norm": 0.240868251540691,
+      "learning_rate": 7.593029320383258e-05,
+      "loss": 0.4619,
+      "step": 870
+    },
+    {
+      "epoch": 1.1536423841059602,
+      "grad_norm": 0.25247584949919727,
+      "learning_rate": 7.591402072788824e-05,
+      "loss": 0.4639,
+      "step": 871
+    },
+    {
+      "epoch": 1.1549668874172185,
+      "grad_norm": 0.2768341519621789,
+      "learning_rate": 7.589771753537877e-05,
+      "loss": 0.4654,
+      "step": 872
+    },
+    {
+      "epoch": 1.1562913907284769,
+      "grad_norm": 0.3017285904733528,
+      "learning_rate": 7.588138364024792e-05,
+      "loss": 0.4675,
+      "step": 873
+    },
+    {
+      "epoch": 1.1576158940397352,
+      "grad_norm": 0.2752304779920257,
+      "learning_rate": 7.586501905646578e-05,
+      "loss": 0.4645,
+      "step": 874
+    },
+    {
+      "epoch": 1.1589403973509933,
+      "grad_norm": 0.2698436758845606,
+      "learning_rate": 7.584862379802866e-05,
+      "loss": 0.4635,
+      "step": 875
+    },
+    {
+      "epoch": 1.1602649006622516,
+      "grad_norm": 0.28533746450402453,
+      "learning_rate": 7.583219787895911e-05,
+      "loss": 0.4685,
+      "step": 876
+    },
+    {
+      "epoch": 1.16158940397351,
+      "grad_norm": 0.32440597110067604,
+      "learning_rate": 7.581574131330588e-05,
+      "loss": 0.4649,
+      "step": 877
+    },
+    {
+      "epoch": 1.1629139072847683,
+      "grad_norm": 0.359720121190048,
+      "learning_rate": 7.579925411514398e-05,
+      "loss": 0.4628,
+      "step": 878
+    },
+    {
+      "epoch": 1.1642384105960264,
+      "grad_norm": 0.3024942425554212,
+      "learning_rate": 7.578273629857457e-05,
+      "loss": 0.4649,
+      "step": 879
+    },
+    {
+      "epoch": 1.1655629139072847,
+      "grad_norm": 0.21390189383622835,
+      "learning_rate": 7.576618787772503e-05,
+      "loss": 0.4666,
+      "step": 880
+    },
+    {
+      "epoch": 1.166887417218543,
+      "grad_norm": 0.21777736989996926,
+      "learning_rate": 7.57496088667489e-05,
+      "loss": 0.4668,
+      "step": 881
+    },
+    {
+      "epoch": 1.1682119205298014,
+      "grad_norm": 0.2293079532536812,
+      "learning_rate": 7.573299927982588e-05,
+      "loss": 0.4612,
+      "step": 882
+    },
+    {
+      "epoch": 1.1695364238410595,
+      "grad_norm": 0.2495645705441454,
+      "learning_rate": 7.571635913116183e-05,
+      "loss": 0.4619,
+      "step": 883
+    },
+    {
+      "epoch": 1.1708609271523178,
+      "grad_norm": 0.1987718899453046,
+      "learning_rate": 7.569968843498877e-05,
+      "loss": 0.4609,
+      "step": 884
+    },
+    {
+      "epoch": 1.1721854304635762,
+      "grad_norm": 0.2009785530214835,
+      "learning_rate": 7.568298720556477e-05,
+      "loss": 0.4639,
+      "step": 885
+    },
+    {
+      "epoch": 1.1735099337748345,
+      "grad_norm": 0.18172935186102276,
+      "learning_rate": 7.566625545717412e-05,
+      "loss": 0.4658,
+      "step": 886
+    },
+    {
+      "epoch": 1.1748344370860928,
+      "grad_norm": 0.1518119933300951,
+      "learning_rate": 7.564949320412712e-05,
+      "loss": 0.4606,
+      "step": 887
+    },
+    {
+      "epoch": 1.176158940397351,
+      "grad_norm": 0.20879662053701278,
+      "learning_rate": 7.563270046076023e-05,
+      "loss": 0.4657,
+      "step": 888
+    },
+    {
+      "epoch": 1.1774834437086092,
+      "grad_norm": 0.22772022740607786,
+      "learning_rate": 7.561587724143595e-05,
+      "loss": 0.4638,
+      "step": 889
+    },
+    {
+      "epoch": 1.1788079470198676,
+      "grad_norm": 0.2516785118952263,
+      "learning_rate": 7.559902356054285e-05,
+      "loss": 0.4637,
+      "step": 890
+    },
+    {
+      "epoch": 1.180132450331126,
+      "grad_norm": 0.27044708702097775,
+      "learning_rate": 7.558213943249553e-05,
+      "loss": 0.4663,
+      "step": 891
+    },
+    {
+      "epoch": 1.181456953642384,
+      "grad_norm": 0.21527808020046452,
+      "learning_rate": 7.556522487173472e-05,
+      "loss": 0.4632,
+      "step": 892
+    },
+    {
+      "epoch": 1.1827814569536423,
+      "grad_norm": 0.16769258926770036,
+      "learning_rate": 7.554827989272707e-05,
+      "loss": 0.4644,
+      "step": 893
+    },
+    {
+      "epoch": 1.1841059602649007,
+      "grad_norm": 0.19606658215927392,
+      "learning_rate": 7.553130450996528e-05,
+      "loss": 0.462,
+      "step": 894
+    },
+    {
+      "epoch": 1.185430463576159,
+      "grad_norm": 0.19801783763234718,
+      "learning_rate": 7.55142987379681e-05,
+      "loss": 0.4645,
+      "step": 895
+    },
+    {
+      "epoch": 1.1867549668874173,
+      "grad_norm": 0.180348553834159,
+      "learning_rate": 7.549726259128022e-05,
+      "loss": 0.4614,
+      "step": 896
+    },
+    {
+      "epoch": 1.1880794701986754,
+      "grad_norm": 0.19738159469172495,
+      "learning_rate": 7.548019608447232e-05,
+      "loss": 0.4601,
+      "step": 897
+    },
+    {
+      "epoch": 1.1894039735099338,
+      "grad_norm": 0.22886714382315074,
+      "learning_rate": 7.546309923214108e-05,
+      "loss": 0.4655,
+      "step": 898
+    },
+    {
+      "epoch": 1.190728476821192,
+      "grad_norm": 0.2571297118175685,
+      "learning_rate": 7.544597204890908e-05,
+      "loss": 0.4643,
+      "step": 899
+    },
+    {
+      "epoch": 1.1920529801324504,
+      "grad_norm": 0.26296978571427754,
+      "learning_rate": 7.542881454942487e-05,
+      "loss": 0.4672,
+      "step": 900
+    },
+    {
+      "epoch": 1.1933774834437085,
+      "grad_norm": 0.2799140476468322,
+      "learning_rate": 7.541162674836293e-05,
+      "loss": 0.4633,
+      "step": 901
+    },
+    {
+      "epoch": 1.1947019867549669,
+      "grad_norm": 0.2687108086544877,
+      "learning_rate": 7.539440866042365e-05,
+      "loss": 0.4645,
+      "step": 902
+    },
+    {
+      "epoch": 1.1960264900662252,
+      "grad_norm": 0.2401976678612489,
+      "learning_rate": 7.537716030033333e-05,
+      "loss": 0.4655,
+      "step": 903
+    },
+    {
+      "epoch": 1.1973509933774835,
+      "grad_norm": 0.23220708140140578,
+      "learning_rate": 7.535988168284417e-05,
+      "loss": 0.4627,
+      "step": 904
+    },
+    {
+      "epoch": 1.1986754966887416,
+      "grad_norm": 0.3045235232239418,
+      "learning_rate": 7.534257282273422e-05,
+      "loss": 0.4679,
+      "step": 905
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.3068677773175233,
+      "learning_rate": 7.532523373480742e-05,
+      "loss": 0.4681,
+      "step": 906
+    },
+    {
+      "epoch": 1.2013245033112583,
+      "grad_norm": 0.27866319665734024,
+      "learning_rate": 7.530786443389353e-05,
+      "loss": 0.4657,
+      "step": 907
+    },
+    {
+      "epoch": 1.2026490066225166,
+      "grad_norm": 0.2990425442824105,
+      "learning_rate": 7.529046493484821e-05,
+      "loss": 0.4634,
+      "step": 908
+    },
+    {
+      "epoch": 1.2039735099337747,
+      "grad_norm": 0.3084569593219229,
+      "learning_rate": 7.527303525255291e-05,
+      "loss": 0.4656,
+      "step": 909
+    },
+    {
+      "epoch": 1.205298013245033,
+      "grad_norm": 0.3121978259642484,
+      "learning_rate": 7.525557540191487e-05,
+      "loss": 0.4625,
+      "step": 910
+    },
+    {
+      "epoch": 1.2066225165562914,
+      "grad_norm": 0.2853724337095738,
+      "learning_rate": 7.52380853978672e-05,
+      "loss": 0.467,
+      "step": 911
+    },
+    {
+      "epoch": 1.2079470198675497,
+      "grad_norm": 0.2644390170940376,
+      "learning_rate": 7.522056525536874e-05,
+      "loss": 0.4627,
+      "step": 912
+    },
+    {
+      "epoch": 1.209271523178808,
+      "grad_norm": 0.24503078864693983,
+      "learning_rate": 7.520301498940415e-05,
+      "loss": 0.4633,
+      "step": 913
+    },
+    {
+      "epoch": 1.2105960264900661,
+      "grad_norm": 0.23591016125250766,
+      "learning_rate": 7.518543461498379e-05,
+      "loss": 0.4662,
+      "step": 914
+    },
+    {
+      "epoch": 1.2119205298013245,
+      "grad_norm": 0.23800324989272584,
+      "learning_rate": 7.516782414714383e-05,
+      "loss": 0.4638,
+      "step": 915
+    },
+    {
+      "epoch": 1.2132450331125828,
+      "grad_norm": 0.2578342875280947,
+      "learning_rate": 7.51501836009462e-05,
+      "loss": 0.4647,
+      "step": 916
+    },
+    {
+      "epoch": 1.2145695364238411,
+      "grad_norm": 0.26250379161379356,
+      "learning_rate": 7.513251299147846e-05,
+      "loss": 0.4626,
+      "step": 917
+    },
+    {
+      "epoch": 1.2158940397350992,
+      "grad_norm": 0.24704600433474733,
+      "learning_rate": 7.511481233385396e-05,
+      "loss": 0.4644,
+      "step": 918
+    },
+    {
+      "epoch": 1.2172185430463576,
+      "grad_norm": 0.2502574498351733,
+      "learning_rate": 7.509708164321175e-05,
+      "loss": 0.4663,
+      "step": 919
+    },
+    {
+      "epoch": 1.218543046357616,
+      "grad_norm": 0.24489889690164898,
+      "learning_rate": 7.507932093471652e-05,
+      "loss": 0.4603,
+      "step": 920
+    },
+    {
+      "epoch": 1.2198675496688742,
+      "grad_norm": 0.2116413510260588,
+      "learning_rate": 7.506153022355868e-05,
+      "loss": 0.464,
+      "step": 921
+    },
+    {
+      "epoch": 1.2211920529801326,
+      "grad_norm": 0.20399157862268613,
+      "learning_rate": 7.504370952495426e-05,
+      "loss": 0.4642,
+      "step": 922
+    },
+    {
+      "epoch": 1.2225165562913907,
+      "grad_norm": 0.25419134821667866,
+      "learning_rate": 7.502585885414497e-05,
+      "loss": 0.461,
+      "step": 923
+    },
+    {
+      "epoch": 1.223841059602649,
+      "grad_norm": 0.28509421237179666,
+      "learning_rate": 7.500797822639813e-05,
+      "loss": 0.4669,
+      "step": 924
+    },
+    {
+      "epoch": 1.2251655629139073,
+      "grad_norm": 0.2425171912549894,
+      "learning_rate": 7.49900676570067e-05,
+      "loss": 0.4639,
+      "step": 925
+    },
+    {
+      "epoch": 1.2264900662251657,
+      "grad_norm": 0.20248086019575606,
+      "learning_rate": 7.497212716128927e-05,
+      "loss": 0.4624,
+      "step": 926
+    },
+    {
+      "epoch": 1.2278145695364238,
+      "grad_norm": 0.23055073536701057,
+      "learning_rate": 7.495415675458995e-05,
+      "loss": 0.4641,
+      "step": 927
+    },
+    {
+      "epoch": 1.229139072847682,
+      "grad_norm": 0.2025889517888478,
+      "learning_rate": 7.493615645227853e-05,
+      "loss": 0.4616,
+      "step": 928
+    },
+    {
+      "epoch": 1.2304635761589404,
+      "grad_norm": 0.1991242115229505,
+      "learning_rate": 7.491812626975026e-05,
+      "loss": 0.464,
+      "step": 929
+    },
+    {
+      "epoch": 1.2317880794701987,
+      "grad_norm": 0.2802379143634475,
+      "learning_rate": 7.490006622242605e-05,
+      "loss": 0.463,
+      "step": 930
+    },
+    {
+      "epoch": 1.233112582781457,
+      "grad_norm": 0.3808923694898013,
+      "learning_rate": 7.488197632575232e-05,
+      "loss": 0.4655,
+      "step": 931
+    },
+    {
+      "epoch": 1.2344370860927152,
+      "grad_norm": 0.34638411517106926,
+      "learning_rate": 7.486385659520096e-05,
+      "loss": 0.4649,
+      "step": 932
+    },
+    {
+      "epoch": 1.2357615894039735,
+      "grad_norm": 0.22434127911916757,
+      "learning_rate": 7.484570704626945e-05,
+      "loss": 0.4613,
+      "step": 933
+    },
+    {
+      "epoch": 1.2370860927152318,
+      "grad_norm": 0.20271490068475856,
+      "learning_rate": 7.482752769448074e-05,
+      "loss": 0.4629,
+      "step": 934
+    },
+    {
+      "epoch": 1.23841059602649,
+      "grad_norm": 0.2243224276212201,
+      "learning_rate": 7.480931855538329e-05,
+      "loss": 0.4587,
+      "step": 935
+    },
+    {
+      "epoch": 1.2397350993377483,
+      "grad_norm": 0.25323004479598543,
+      "learning_rate": 7.479107964455103e-05,
+      "loss": 0.4652,
+      "step": 936
+    },
+    {
+      "epoch": 1.2410596026490066,
+      "grad_norm": 0.27344186803016246,
+      "learning_rate": 7.477281097758331e-05,
+      "loss": 0.4659,
+      "step": 937
+    },
+    {
+      "epoch": 1.242384105960265,
+      "grad_norm": 0.3043584376229346,
+      "learning_rate": 7.4754512570105e-05,
+      "loss": 0.4601,
+      "step": 938
+    },
+    {
+      "epoch": 1.2437086092715233,
+      "grad_norm": 0.3236551498671955,
+      "learning_rate": 7.473618443776635e-05,
+      "loss": 0.4641,
+      "step": 939
+    },
+    {
+      "epoch": 1.2450331125827814,
+      "grad_norm": 0.3410325297418939,
+      "learning_rate": 7.471782659624307e-05,
+      "loss": 0.4676,
+      "step": 940
+    },
+    {
+      "epoch": 1.2463576158940397,
+      "grad_norm": 0.3972224962830227,
+      "learning_rate": 7.469943906123627e-05,
+      "loss": 0.4694,
+      "step": 941
+    },
+    {
+      "epoch": 1.247682119205298,
+      "grad_norm": 0.42420740276338037,
+      "learning_rate": 7.468102184847244e-05,
+      "loss": 0.4647,
+      "step": 942
+    },
+    {
+      "epoch": 1.2490066225165564,
+      "grad_norm": 0.37696580264250895,
+      "learning_rate": 7.466257497370348e-05,
+      "loss": 0.4631,
+      "step": 943
+    },
+    {
+      "epoch": 1.2503311258278145,
+      "grad_norm": 0.25047218953999495,
+      "learning_rate": 7.464409845270663e-05,
+      "loss": 0.4623,
+      "step": 944
+    },
+    {
+      "epoch": 1.2516556291390728,
+      "grad_norm": 0.22209059008872112,
+      "learning_rate": 7.462559230128449e-05,
+      "loss": 0.4656,
+      "step": 945
+    },
+    {
+      "epoch": 1.2529801324503311,
+      "grad_norm": 0.2514973981745101,
+      "learning_rate": 7.460705653526503e-05,
+      "loss": 0.4631,
+      "step": 946
+    },
+    {
+      "epoch": 1.2543046357615895,
+      "grad_norm": 0.24132412494267116,
+      "learning_rate": 7.458849117050154e-05,
+      "loss": 0.4596,
+      "step": 947
+    },
+    {
+      "epoch": 1.2556291390728478,
+      "grad_norm": 0.23347035116489362,
+      "learning_rate": 7.45698962228726e-05,
+      "loss": 0.463,
+      "step": 948
+    },
+    {
+      "epoch": 1.256953642384106,
+      "grad_norm": 0.19259928423946424,
+      "learning_rate": 7.45512717082821e-05,
+      "loss": 0.4643,
+      "step": 949
+    },
+    {
+      "epoch": 1.2582781456953642,
+      "grad_norm": 0.1918238015252186,
+      "learning_rate": 7.453261764265925e-05,
+      "loss": 0.467,
+      "step": 950
+    },
+    {
+      "epoch": 1.2596026490066226,
+      "grad_norm": 0.19337845153479535,
+      "learning_rate": 7.45139340419585e-05,
+      "loss": 0.4613,
+      "step": 951
+    },
+    {
+      "epoch": 1.2609271523178807,
+      "grad_norm": 0.19173539315567018,
+      "learning_rate": 7.449522092215956e-05,
+      "loss": 0.4651,
+      "step": 952
+    },
+    {
+      "epoch": 1.262251655629139,
+      "grad_norm": 0.18880269516066997,
+      "learning_rate": 7.447647829926742e-05,
+      "loss": 0.4638,
+      "step": 953
+    },
+    {
+      "epoch": 1.2635761589403973,
+      "grad_norm": 0.1979626294650203,
+      "learning_rate": 7.445770618931225e-05,
+      "loss": 0.4612,
+      "step": 954
+    },
+    {
+      "epoch": 1.2649006622516556,
+      "grad_norm": 0.2041573552808363,
+      "learning_rate": 7.44389046083495e-05,
+      "loss": 0.4628,
+      "step": 955
+    },
+    {
+      "epoch": 1.266225165562914,
+      "grad_norm": 0.19716410140542828,
+      "learning_rate": 7.442007357245978e-05,
+      "loss": 0.4641,
+      "step": 956
+    },
+    {
+      "epoch": 1.2675496688741723,
+      "grad_norm": 0.17157484408126042,
+      "learning_rate": 7.440121309774894e-05,
+      "loss": 0.4645,
+      "step": 957
+    },
+    {
+      "epoch": 1.2688741721854304,
+      "grad_norm": 0.2073716306132351,
+      "learning_rate": 7.438232320034793e-05,
+      "loss": 0.4632,
+      "step": 958
+    },
+    {
+      "epoch": 1.2701986754966887,
+      "grad_norm": 0.29462492445087907,
+      "learning_rate": 7.436340389641296e-05,
+      "loss": 0.464,
+      "step": 959
+    },
+    {
+      "epoch": 1.271523178807947,
+      "grad_norm": 0.3225360782571023,
+      "learning_rate": 7.43444552021253e-05,
+      "loss": 0.4628,
+      "step": 960
+    },
+    {
+      "epoch": 1.2728476821192052,
+      "grad_norm": 0.27896474411730754,
+      "learning_rate": 7.432547713369143e-05,
+      "loss": 0.4654,
+      "step": 961
+    },
+    {
+      "epoch": 1.2741721854304635,
+      "grad_norm": 0.2489699215636332,
+      "learning_rate": 7.430646970734293e-05,
+      "loss": 0.4645,
+      "step": 962
+    },
+    {
+      "epoch": 1.2754966887417218,
+      "grad_norm": 0.22333562273730462,
+      "learning_rate": 7.428743293933646e-05,
+      "loss": 0.4625,
+      "step": 963
+    },
+    {
+      "epoch": 1.2768211920529802,
+      "grad_norm": 0.23446256070403693,
+      "learning_rate": 7.42683668459538e-05,
+      "loss": 0.4614,
+      "step": 964
+    },
+    {
+      "epoch": 1.2781456953642385,
+      "grad_norm": 0.23881028772011084,
+      "learning_rate": 7.424927144350183e-05,
+      "loss": 0.4633,
+      "step": 965
+    },
+    {
+      "epoch": 1.2794701986754966,
+      "grad_norm": 0.2340964928043272,
+      "learning_rate": 7.423014674831246e-05,
+      "loss": 0.465,
+      "step": 966
+    },
+    {
+      "epoch": 1.280794701986755,
+      "grad_norm": 0.23566490466688517,
+      "learning_rate": 7.421099277674269e-05,
+      "loss": 0.466,
+      "step": 967
+    },
+    {
+      "epoch": 1.2821192052980133,
+      "grad_norm": 0.2458264292477425,
+      "learning_rate": 7.419180954517449e-05,
+      "loss": 0.4656,
+      "step": 968
+    },
+    {
+      "epoch": 1.2834437086092716,
+      "grad_norm": 0.2801678535576697,
+      "learning_rate": 7.417259707001498e-05,
+      "loss": 0.4611,
+      "step": 969
+    },
+    {
+      "epoch": 1.2847682119205297,
+      "grad_norm": 0.30935588164087774,
+      "learning_rate": 7.415335536769615e-05,
+      "loss": 0.4634,
+      "step": 970
+    },
+    {
+      "epoch": 1.286092715231788,
+      "grad_norm": 0.2910695571114939,
+      "learning_rate": 7.41340844546751e-05,
+      "loss": 0.4614,
+      "step": 971
+    },
+    {
+      "epoch": 1.2874172185430464,
+      "grad_norm": 0.24602199317501514,
+      "learning_rate": 7.411478434743385e-05,
+      "loss": 0.463,
+      "step": 972
+    },
+    {
+      "epoch": 1.2887417218543047,
+      "grad_norm": 0.24438182860630314,
+      "learning_rate": 7.40954550624794e-05,
+      "loss": 0.4632,
+      "step": 973
+    },
+    {
+      "epoch": 1.290066225165563,
+      "grad_norm": 0.2397679209385836,
+      "learning_rate": 7.407609661634372e-05,
+      "loss": 0.4613,
+      "step": 974
+    },
+    {
+      "epoch": 1.2913907284768211,
+      "grad_norm": 0.22489954937935938,
+      "learning_rate": 7.405670902558374e-05,
+      "loss": 0.4639,
+      "step": 975
+    },
+    {
+      "epoch": 1.2927152317880795,
+      "grad_norm": 0.21796831663065422,
+      "learning_rate": 7.403729230678123e-05,
+      "loss": 0.4633,
+      "step": 976
+    },
+    {
+      "epoch": 1.2940397350993378,
+      "grad_norm": 0.23302293272749802,
+      "learning_rate": 7.4017846476543e-05,
+      "loss": 0.4644,
+      "step": 977
+    },
+    {
+      "epoch": 1.295364238410596,
+      "grad_norm": 0.2277700157518215,
+      "learning_rate": 7.399837155150064e-05,
+      "loss": 0.4661,
+      "step": 978
+    },
+    {
+      "epoch": 1.2966887417218542,
+      "grad_norm": 0.22212718252040603,
+      "learning_rate": 7.397886754831069e-05,
+      "loss": 0.4612,
+      "step": 979
+    },
+    {
+      "epoch": 1.2980132450331126,
+      "grad_norm": 0.23568974797796793,
+      "learning_rate": 7.395933448365457e-05,
+      "loss": 0.4607,
+      "step": 980
+    },
+    {
+      "epoch": 1.2993377483443709,
+      "grad_norm": 0.22649768729110417,
+      "learning_rate": 7.39397723742385e-05,
+      "loss": 0.4644,
+      "step": 981
+    },
+    {
+      "epoch": 1.3006622516556292,
+      "grad_norm": 0.20461506813543495,
+      "learning_rate": 7.392018123679357e-05,
+      "loss": 0.465,
+      "step": 982
+    },
+    {
+      "epoch": 1.3019867549668875,
+      "grad_norm": 0.17281035325525482,
+      "learning_rate": 7.390056108807573e-05,
+      "loss": 0.4653,
+      "step": 983
+    },
+    {
+      "epoch": 1.3033112582781456,
+      "grad_norm": 0.1992957068978901,
+      "learning_rate": 7.38809119448657e-05,
+      "loss": 0.4611,
+      "step": 984
+    },
+    {
+      "epoch": 1.304635761589404,
+      "grad_norm": 0.2181271265053973,
+      "learning_rate": 7.3861233823969e-05,
+      "loss": 0.465,
+      "step": 985
+    },
+    {
+      "epoch": 1.3059602649006623,
+      "grad_norm": 0.21781302938026725,
+      "learning_rate": 7.384152674221595e-05,
+      "loss": 0.4602,
+      "step": 986
+    },
+    {
+      "epoch": 1.3072847682119204,
+      "grad_norm": 0.19748275361903553,
+      "learning_rate": 7.382179071646164e-05,
+      "loss": 0.4644,
+      "step": 987
+    },
+    {
+      "epoch": 1.3086092715231787,
+      "grad_norm": 0.16574348729594404,
+      "learning_rate": 7.380202576358591e-05,
+      "loss": 0.4592,
+      "step": 988
+    },
+    {
+      "epoch": 1.309933774834437,
+      "grad_norm": 0.20731708930740908,
+      "learning_rate": 7.378223190049335e-05,
+      "loss": 0.4625,
+      "step": 989
+    },
+    {
+      "epoch": 1.3112582781456954,
+      "grad_norm": 0.21342552635310938,
+      "learning_rate": 7.376240914411323e-05,
+      "loss": 0.4633,
+      "step": 990
+    },
+    {
+      "epoch": 1.3125827814569537,
+      "grad_norm": 0.238232771844375,
+      "learning_rate": 7.374255751139961e-05,
+      "loss": 0.4621,
+      "step": 991
+    },
+    {
+      "epoch": 1.313907284768212,
+      "grad_norm": 0.22923696027603951,
+      "learning_rate": 7.372267701933118e-05,
+      "loss": 0.4655,
+      "step": 992
+    },
+    {
+      "epoch": 1.3152317880794702,
+      "grad_norm": 0.1885592705653636,
+      "learning_rate": 7.370276768491136e-05,
+      "loss": 0.4624,
+      "step": 993
+    },
+    {
+      "epoch": 1.3165562913907285,
+      "grad_norm": 0.21081518203726748,
+      "learning_rate": 7.36828295251682e-05,
+      "loss": 0.4621,
+      "step": 994
+    },
+    {
+      "epoch": 1.3178807947019868,
+      "grad_norm": 0.21813005335039617,
+      "learning_rate": 7.366286255715443e-05,
+      "loss": 0.4625,
+      "step": 995
+    },
+    {
+      "epoch": 1.319205298013245,
+      "grad_norm": 0.24722887080102995,
+      "learning_rate": 7.36428667979474e-05,
+      "loss": 0.4629,
+      "step": 996
+    },
+    {
+      "epoch": 1.3205298013245033,
+      "grad_norm": 0.26720412665097254,
+      "learning_rate": 7.362284226464912e-05,
+      "loss": 0.464,
+      "step": 997
+    },
+    {
+      "epoch": 1.3218543046357616,
+      "grad_norm": 0.271182922994299,
+      "learning_rate": 7.360278897438615e-05,
+      "loss": 0.4612,
+      "step": 998
+    },
+    {
+      "epoch": 1.32317880794702,
+      "grad_norm": 0.23866729762525563,
+      "learning_rate": 7.35827069443097e-05,
+      "loss": 0.4578,
+      "step": 999
+    },
+    {
+      "epoch": 1.3245033112582782,
+      "grad_norm": 0.21014831628424668,
+      "learning_rate": 7.356259619159556e-05,
+      "loss": 0.4625,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3258278145695364,
+      "grad_norm": 0.2190013024453031,
+      "learning_rate": 7.354245673344403e-05,
+      "loss": 0.463,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3271523178807947,
+      "grad_norm": 0.21232805868004515,
+      "learning_rate": 7.352228858708002e-05,
+      "loss": 0.4615,
+      "step": 1002
+    },
+    {
+      "epoch": 1.328476821192053,
+      "grad_norm": 0.18294397524632722,
+      "learning_rate": 7.350209176975293e-05,
+      "loss": 0.4606,
+      "step": 1003
+    },
+    {
+      "epoch": 1.3298013245033111,
+      "grad_norm": 0.2130272099507456,
+      "learning_rate": 7.348186629873674e-05,
+      "loss": 0.4603,
+      "step": 1004
+    },
+    {
+      "epoch": 1.3311258278145695,
+      "grad_norm": 0.23037537634841326,
+      "learning_rate": 7.346161219132988e-05,
+      "loss": 0.4627,
+      "step": 1005
+    },
+    {
+      "epoch": 1.3324503311258278,
+      "grad_norm": 0.2238367653949882,
+      "learning_rate": 7.34413294648553e-05,
+      "loss": 0.4616,
+      "step": 1006
+    },
+    {
+      "epoch": 1.333774834437086,
+      "grad_norm": 0.23190292586153338,
+      "learning_rate": 7.342101813666044e-05,
+      "loss": 0.4646,
+      "step": 1007
+    },
+    {
+      "epoch": 1.3350993377483444,
+      "grad_norm": 0.2268187398112637,
+      "learning_rate": 7.340067822411716e-05,
+      "loss": 0.461,
+      "step": 1008
+    },
+    {
+      "epoch": 1.3364238410596028,
+      "grad_norm": 0.2092069221315842,
+      "learning_rate": 7.33803097446218e-05,
+      "loss": 0.462,
+      "step": 1009
+    },
+    {
+      "epoch": 1.3377483443708609,
+      "grad_norm": 0.2160363252070283,
+      "learning_rate": 7.335991271559512e-05,
+      "loss": 0.4615,
+      "step": 1010
+    },
+    {
+      "epoch": 1.3390728476821192,
+      "grad_norm": 0.2095520694315062,
+      "learning_rate": 7.333948715448232e-05,
+      "loss": 0.4638,
+      "step": 1011
+    },
+    {
+      "epoch": 1.3403973509933775,
+      "grad_norm": 0.22699075921731868,
+      "learning_rate": 7.331903307875297e-05,
+      "loss": 0.4641,
+      "step": 1012
+    },
+    {
+      "epoch": 1.3417218543046356,
+      "grad_norm": 0.21410810726277307,
+      "learning_rate": 7.329855050590105e-05,
+      "loss": 0.4619,
+      "step": 1013
+    },
+    {
+      "epoch": 1.343046357615894,
+      "grad_norm": 0.21049278122678208,
+      "learning_rate": 7.327803945344492e-05,
+      "loss": 0.4629,
+      "step": 1014
+    },
+    {
+      "epoch": 1.3443708609271523,
+      "grad_norm": 0.20447963913685124,
+      "learning_rate": 7.325749993892727e-05,
+      "loss": 0.4637,
+      "step": 1015
+    },
+    {
+      "epoch": 1.3456953642384106,
+      "grad_norm": 0.2116119119141276,
+      "learning_rate": 7.323693197991515e-05,
+      "loss": 0.4632,
+      "step": 1016
+    },
+    {
+      "epoch": 1.347019867549669,
+      "grad_norm": 0.2105859347575885,
+      "learning_rate": 7.321633559399992e-05,
+      "loss": 0.4609,
+      "step": 1017
+    },
+    {
+      "epoch": 1.3483443708609273,
+      "grad_norm": 0.2083858327200109,
+      "learning_rate": 7.319571079879731e-05,
+      "loss": 0.4603,
+      "step": 1018
+    },
+    {
+      "epoch": 1.3496688741721854,
+      "grad_norm": 0.22895220596532986,
+      "learning_rate": 7.317505761194727e-05,
+      "loss": 0.4631,
+      "step": 1019
+    },
+    {
+      "epoch": 1.3509933774834437,
+      "grad_norm": 0.2598369722107028,
+      "learning_rate": 7.31543760511141e-05,
+      "loss": 0.4604,
+      "step": 1020
+    },
+    {
+      "epoch": 1.352317880794702,
+      "grad_norm": 0.23851218299611354,
+      "learning_rate": 7.31336661339863e-05,
+      "loss": 0.4623,
+      "step": 1021
+    },
+    {
+      "epoch": 1.3536423841059602,
+      "grad_norm": 0.20436321994082038,
+      "learning_rate": 7.31129278782767e-05,
+      "loss": 0.4615,
+      "step": 1022
+    },
+    {
+      "epoch": 1.3549668874172185,
+      "grad_norm": 0.18248218538202485,
+      "learning_rate": 7.30921613017223e-05,
+      "loss": 0.4596,
+      "step": 1023
+    },
+    {
+      "epoch": 1.3562913907284768,
+      "grad_norm": 0.24349771108785107,
+      "learning_rate": 7.307136642208436e-05,
+      "loss": 0.4622,
+      "step": 1024
+    },
+    {
+      "epoch": 1.3576158940397351,
+      "grad_norm": 0.3065692165247032,
+      "learning_rate": 7.305054325714833e-05,
+      "loss": 0.4605,
+      "step": 1025
+    },
+    {
+      "epoch": 1.3589403973509935,
+      "grad_norm": 0.3136641756536974,
+      "learning_rate": 7.302969182472389e-05,
+      "loss": 0.4589,
+      "step": 1026
+    },
+    {
+      "epoch": 1.3602649006622516,
+      "grad_norm": 0.3016975574547123,
+      "learning_rate": 7.300881214264482e-05,
+      "loss": 0.4628,
+      "step": 1027
+    },
+    {
+      "epoch": 1.36158940397351,
+      "grad_norm": 0.3198935150121867,
+      "learning_rate": 7.298790422876913e-05,
+      "loss": 0.4651,
+      "step": 1028
+    },
+    {
+      "epoch": 1.3629139072847682,
+      "grad_norm": 0.3610679602167911,
+      "learning_rate": 7.296696810097896e-05,
+      "loss": 0.4648,
+      "step": 1029
+    },
+    {
+      "epoch": 1.3642384105960264,
+      "grad_norm": 0.38272676006355794,
+      "learning_rate": 7.294600377718057e-05,
+      "loss": 0.4633,
+      "step": 1030
+    },
+    {
+      "epoch": 1.3655629139072847,
+      "grad_norm": 0.3358539867000369,
+      "learning_rate": 7.292501127530432e-05,
+      "loss": 0.4643,
+      "step": 1031
+    },
+    {
+      "epoch": 1.366887417218543,
+      "grad_norm": 0.2964943653009151,
+      "learning_rate": 7.290399061330473e-05,
+      "loss": 0.4622,
+      "step": 1032
+    },
+    {
+      "epoch": 1.3682119205298013,
+      "grad_norm": 0.2769477634569621,
+      "learning_rate": 7.288294180916032e-05,
+      "loss": 0.4622,
+      "step": 1033
+    },
+    {
+      "epoch": 1.3695364238410597,
+      "grad_norm": 0.253072083998996,
+      "learning_rate": 7.286186488087374e-05,
+      "loss": 0.4662,
+      "step": 1034
+    },
+    {
+      "epoch": 1.370860927152318,
+      "grad_norm": 0.2619157454927392,
+      "learning_rate": 7.284075984647169e-05,
+      "loss": 0.4641,
+      "step": 1035
+    },
+    {
+      "epoch": 1.372185430463576,
+      "grad_norm": 0.2809983021256224,
+      "learning_rate": 7.281962672400488e-05,
+      "loss": 0.4641,
+      "step": 1036
+    },
+    {
+      "epoch": 1.3735099337748344,
+      "grad_norm": 0.26998743824576665,
+      "learning_rate": 7.279846553154807e-05,
+      "loss": 0.4621,
+      "step": 1037
+    },
+    {
+      "epoch": 1.3748344370860928,
+      "grad_norm": 0.20830351456899088,
+      "learning_rate": 7.277727628720001e-05,
+      "loss": 0.461,
+      "step": 1038
+    },
+    {
+      "epoch": 1.3761589403973509,
+      "grad_norm": 0.2148127045760067,
+      "learning_rate": 7.275605900908346e-05,
+      "loss": 0.4619,
+      "step": 1039
+    },
+    {
+      "epoch": 1.3774834437086092,
+      "grad_norm": 0.26944995560704643,
+      "learning_rate": 7.273481371534513e-05,
+      "loss": 0.4656,
+      "step": 1040
+    },
+    {
+      "epoch": 1.3788079470198675,
+      "grad_norm": 0.23818441243543717,
+      "learning_rate": 7.271354042415572e-05,
+      "loss": 0.46,
+      "step": 1041
+    },
+    {
+      "epoch": 1.3801324503311259,
+      "grad_norm": 0.20403299194358496,
+      "learning_rate": 7.269223915370987e-05,
+      "loss": 0.4609,
+      "step": 1042
+    },
+    {
+      "epoch": 1.3814569536423842,
+      "grad_norm": 0.20581442177167955,
+      "learning_rate": 7.267090992222614e-05,
+      "loss": 0.4572,
+      "step": 1043
+    },
+    {
+      "epoch": 1.3827814569536425,
+      "grad_norm": 0.1982493489067554,
+      "learning_rate": 7.264955274794698e-05,
+      "loss": 0.4607,
+      "step": 1044
+    },
+    {
+      "epoch": 1.3841059602649006,
+      "grad_norm": 0.20112010247964382,
+      "learning_rate": 7.262816764913882e-05,
+      "loss": 0.4598,
+      "step": 1045
+    },
+    {
+      "epoch": 1.385430463576159,
+      "grad_norm": 0.22316049220702247,
+      "learning_rate": 7.260675464409186e-05,
+      "loss": 0.4623,
+      "step": 1046
+    },
+    {
+      "epoch": 1.3867549668874173,
+      "grad_norm": 0.2123397257256748,
+      "learning_rate": 7.258531375112027e-05,
+      "loss": 0.4635,
+      "step": 1047
+    },
+    {
+      "epoch": 1.3880794701986754,
+      "grad_norm": 0.1727616814032157,
+      "learning_rate": 7.256384498856201e-05,
+      "loss": 0.4585,
+      "step": 1048
+    },
+    {
+      "epoch": 1.3894039735099337,
+      "grad_norm": 0.17933566658703173,
+      "learning_rate": 7.25423483747789e-05,
+      "loss": 0.4641,
+      "step": 1049
+    },
+    {
+      "epoch": 1.390728476821192,
+      "grad_norm": 0.20848334040758829,
+      "learning_rate": 7.25208239281566e-05,
+      "loss": 0.4608,
+      "step": 1050
+    },
+    {
+      "epoch": 1.3920529801324504,
+      "grad_norm": 0.20870192391982353,
+      "learning_rate": 7.249927166710451e-05,
+      "loss": 0.4618,
+      "step": 1051
+    },
+    {
+      "epoch": 1.3933774834437087,
+      "grad_norm": 0.21309147911116313,
+      "learning_rate": 7.247769161005589e-05,
+      "loss": 0.4624,
+      "step": 1052
+    },
+    {
+      "epoch": 1.3947019867549668,
+      "grad_norm": 0.1930581125036177,
+      "learning_rate": 7.245608377546774e-05,
+      "loss": 0.461,
+      "step": 1053
+    },
+    {
+      "epoch": 1.3960264900662251,
+      "grad_norm": 0.2057470872476602,
+      "learning_rate": 7.243444818182082e-05,
+      "loss": 0.4588,
+      "step": 1054
+    },
+    {
+      "epoch": 1.3973509933774835,
+      "grad_norm": 0.20280666019453578,
+      "learning_rate": 7.241278484761962e-05,
+      "loss": 0.4628,
+      "step": 1055
+    },
+    {
+      "epoch": 1.3986754966887418,
+      "grad_norm": 0.2073331415352651,
+      "learning_rate": 7.239109379139237e-05,
+      "loss": 0.4643,
+      "step": 1056
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.21138347972150553,
+      "learning_rate": 7.236937503169104e-05,
+      "loss": 0.4643,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4013245033112582,
+      "grad_norm": 0.2135785248608845,
+      "learning_rate": 7.234762858709124e-05,
+      "loss": 0.4622,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4026490066225166,
+      "grad_norm": 0.2263059865912043,
+      "learning_rate": 7.232585447619227e-05,
+      "loss": 0.4596,
+      "step": 1059
+    },
+    {
+      "epoch": 1.403973509933775,
+      "grad_norm": 0.26016842175403604,
+      "learning_rate": 7.230405271761715e-05,
+      "loss": 0.4611,
+      "step": 1060
+    },
+    {
+      "epoch": 1.4052980132450332,
+      "grad_norm": 0.2610904460597662,
+      "learning_rate": 7.228222333001245e-05,
+      "loss": 0.4619,
+      "step": 1061
+    },
+    {
+      "epoch": 1.4066225165562913,
+      "grad_norm": 0.26110924555616566,
+      "learning_rate": 7.226036633204845e-05,
+      "loss": 0.4624,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4079470198675497,
+      "grad_norm": 0.2595067486799945,
+      "learning_rate": 7.223848174241902e-05,
+      "loss": 0.4581,
+      "step": 1063
+    },
+    {
+      "epoch": 1.409271523178808,
+      "grad_norm": 0.2261800314468133,
+      "learning_rate": 7.22165695798416e-05,
+      "loss": 0.4627,
+      "step": 1064
+    },
+    {
+      "epoch": 1.410596026490066,
+      "grad_norm": 0.2374941141236378,
+      "learning_rate": 7.219462986305725e-05,
+      "loss": 0.4632,
+      "step": 1065
+    },
+    {
+      "epoch": 1.4119205298013244,
+      "grad_norm": 0.24562264407198894,
+      "learning_rate": 7.21726626108306e-05,
+      "loss": 0.4579,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4132450331125828,
+      "grad_norm": 0.21476933522260547,
+      "learning_rate": 7.21506678419498e-05,
+      "loss": 0.4634,
+      "step": 1067
+    },
+    {
+      "epoch": 1.414569536423841,
+      "grad_norm": 0.2468416977148505,
+      "learning_rate": 7.212864557522653e-05,
+      "loss": 0.4603,
+      "step": 1068
+    },
+    {
+      "epoch": 1.4158940397350994,
+      "grad_norm": 0.23051184718690906,
+      "learning_rate": 7.210659582949606e-05,
+      "loss": 0.465,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4172185430463577,
+      "grad_norm": 0.22499118440187768,
+      "learning_rate": 7.208451862361704e-05,
+      "loss": 0.4604,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4185430463576159,
+      "grad_norm": 0.23628929761276637,
+      "learning_rate": 7.206241397647174e-05,
+      "loss": 0.465,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4198675496688742,
+      "grad_norm": 0.2372597022116873,
+      "learning_rate": 7.20402819069658e-05,
+      "loss": 0.4604,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4211920529801325,
+      "grad_norm": 0.20800843716274772,
+      "learning_rate": 7.201812243402837e-05,
+      "loss": 0.4561,
+      "step": 1073
+    },
+    {
+      "epoch": 1.4225165562913906,
+      "grad_norm": 0.1986126046904411,
+      "learning_rate": 7.1995935576612e-05,
+      "loss": 0.4605,
+      "step": 1074
+    },
+    {
+      "epoch": 1.423841059602649,
+      "grad_norm": 0.1896396027428134,
+      "learning_rate": 7.197372135369268e-05,
+      "loss": 0.4614,
+      "step": 1075
+    },
+    {
+      "epoch": 1.4251655629139073,
+      "grad_norm": 0.22347074084553045,
+      "learning_rate": 7.19514797842698e-05,
+      "loss": 0.4579,
+      "step": 1076
+    },
+    {
+      "epoch": 1.4264900662251656,
+      "grad_norm": 0.22996451450248026,
+      "learning_rate": 7.192921088736617e-05,
+      "loss": 0.461,
+      "step": 1077
+    },
+    {
+      "epoch": 1.427814569536424,
+      "grad_norm": 0.1818612431335228,
+      "learning_rate": 7.190691468202791e-05,
+      "loss": 0.4596,
+      "step": 1078
+    },
+    {
+      "epoch": 1.429139072847682,
+      "grad_norm": 0.17844673292520666,
+      "learning_rate": 7.188459118732455e-05,
+      "loss": 0.4635,
+      "step": 1079
+    },
+    {
+      "epoch": 1.4304635761589404,
+      "grad_norm": 0.2374573742786211,
+      "learning_rate": 7.186224042234896e-05,
+      "loss": 0.4634,
+      "step": 1080
+    },
+    {
+      "epoch": 1.4317880794701987,
+      "grad_norm": 0.26673854218216225,
+      "learning_rate": 7.183986240621728e-05,
+      "loss": 0.4618,
+      "step": 1081
+    },
+    {
+      "epoch": 1.433112582781457,
+      "grad_norm": 0.23774544927620384,
+      "learning_rate": 7.181745715806902e-05,
+      "loss": 0.4589,
+      "step": 1082
+    },
+    {
+      "epoch": 1.4344370860927151,
+      "grad_norm": 0.213756255253391,
+      "learning_rate": 7.179502469706693e-05,
+      "loss": 0.4649,
+      "step": 1083
+    },
+    {
+      "epoch": 1.4357615894039735,
+      "grad_norm": 0.2656133870154096,
+      "learning_rate": 7.177256504239708e-05,
+      "loss": 0.4617,
+      "step": 1084
+    },
+    {
+      "epoch": 1.4370860927152318,
+      "grad_norm": 0.2860937801880211,
+      "learning_rate": 7.175007821326877e-05,
+      "loss": 0.4563,
+      "step": 1085
+    },
+    {
+      "epoch": 1.4384105960264901,
+      "grad_norm": 0.28804169964722004,
+      "learning_rate": 7.172756422891455e-05,
+      "loss": 0.4639,
+      "step": 1086
+    },
+    {
+      "epoch": 1.4397350993377485,
+      "grad_norm": 0.2424712471853641,
+      "learning_rate": 7.170502310859019e-05,
+      "loss": 0.4627,
+      "step": 1087
+    },
+    {
+      "epoch": 1.4410596026490066,
+      "grad_norm": 0.17755982334511505,
+      "learning_rate": 7.168245487157466e-05,
+      "loss": 0.4577,
+      "step": 1088
+    },
+    {
+      "epoch": 1.442384105960265,
+      "grad_norm": 0.15277792289869552,
+      "learning_rate": 7.165985953717017e-05,
+      "loss": 0.4592,
+      "step": 1089
+    },
+    {
+      "epoch": 1.4437086092715232,
+      "grad_norm": 0.1437031451263441,
+      "learning_rate": 7.163723712470204e-05,
+      "loss": 0.4619,
+      "step": 1090
+    },
+    {
+      "epoch": 1.4450331125827813,
+      "grad_norm": 0.1698977093510239,
+      "learning_rate": 7.16145876535188e-05,
+      "loss": 0.4638,
+      "step": 1091
+    },
+    {
+      "epoch": 1.4463576158940397,
+      "grad_norm": 0.20927357496042656,
+      "learning_rate": 7.159191114299208e-05,
+      "loss": 0.462,
+      "step": 1092
+    },
+    {
+      "epoch": 1.447682119205298,
+      "grad_norm": 0.237156415336418,
+      "learning_rate": 7.15692076125167e-05,
+      "loss": 0.4633,
+      "step": 1093
+    },
+    {
+      "epoch": 1.4490066225165563,
+      "grad_norm": 0.20398451405812942,
+      "learning_rate": 7.15464770815105e-05,
+      "loss": 0.4598,
+      "step": 1094
+    },
+    {
+      "epoch": 1.4503311258278146,
+      "grad_norm": 0.17005905379409286,
+      "learning_rate": 7.15237195694145e-05,
+      "loss": 0.4626,
+      "step": 1095
+    },
+    {
+      "epoch": 1.451655629139073,
+      "grad_norm": 0.16578815479149542,
+      "learning_rate": 7.150093509569272e-05,
+      "loss": 0.4624,
+      "step": 1096
+    },
+    {
+      "epoch": 1.452980132450331,
+      "grad_norm": 0.1793486727182083,
+      "learning_rate": 7.147812367983234e-05,
+      "loss": 0.4607,
+      "step": 1097
+    },
+    {
+      "epoch": 1.4543046357615894,
+      "grad_norm": 0.205895039489702,
+      "learning_rate": 7.145528534134345e-05,
+      "loss": 0.4596,
+      "step": 1098
+    },
+    {
+      "epoch": 1.4556291390728477,
+      "grad_norm": 0.18195260711723962,
+      "learning_rate": 7.143242009975929e-05,
+      "loss": 0.4626,
+      "step": 1099
+    },
+    {
+      "epoch": 1.4569536423841059,
+      "grad_norm": 0.17086724031933276,
+      "learning_rate": 7.140952797463604e-05,
+      "loss": 0.462,
+      "step": 1100
+    },
+    {
+      "epoch": 1.4582781456953642,
+      "grad_norm": 0.15944307987371498,
+      "learning_rate": 7.138660898555289e-05,
+      "loss": 0.4604,
+      "step": 1101
+    },
+    {
+      "epoch": 1.4596026490066225,
+      "grad_norm": 0.1435493465096214,
+      "learning_rate": 7.136366315211199e-05,
+      "loss": 0.4634,
+      "step": 1102
+    },
+    {
+      "epoch": 1.4609271523178808,
+      "grad_norm": 0.15722799412439806,
+      "learning_rate": 7.13406904939385e-05,
+      "loss": 0.4586,
+      "step": 1103
+    },
+    {
+      "epoch": 1.4622516556291392,
+      "grad_norm": 0.1804732794812046,
+      "learning_rate": 7.131769103068045e-05,
+      "loss": 0.4614,
+      "step": 1104
+    },
+    {
+      "epoch": 1.4635761589403973,
+      "grad_norm": 0.1518905521087175,
+      "learning_rate": 7.129466478200889e-05,
+      "loss": 0.4598,
+      "step": 1105
+    },
+    {
+      "epoch": 1.4649006622516556,
+      "grad_norm": 0.15532387270319342,
+      "learning_rate": 7.127161176761767e-05,
+      "loss": 0.4594,
+      "step": 1106
+    },
+    {
+      "epoch": 1.466225165562914,
+      "grad_norm": 0.17734033507929392,
+      "learning_rate": 7.12485320072236e-05,
+      "loss": 0.4603,
+      "step": 1107
+    },
+    {
+      "epoch": 1.4675496688741723,
+      "grad_norm": 0.23319006315126187,
+      "learning_rate": 7.122542552056637e-05,
+      "loss": 0.4592,
+      "step": 1108
+    },
+    {
+      "epoch": 1.4688741721854304,
+      "grad_norm": 0.25677714222646264,
+      "learning_rate": 7.12022923274085e-05,
+      "loss": 0.4577,
+      "step": 1109
+    },
+    {
+      "epoch": 1.4701986754966887,
+      "grad_norm": 0.2732688831138733,
+      "learning_rate": 7.117913244753535e-05,
+      "loss": 0.4637,
+      "step": 1110
+    },
+    {
+      "epoch": 1.471523178807947,
+      "grad_norm": 0.27179408802374827,
+      "learning_rate": 7.115594590075512e-05,
+      "loss": 0.4594,
+      "step": 1111
+    },
+    {
+      "epoch": 1.4728476821192054,
+      "grad_norm": 0.26382363873749914,
+      "learning_rate": 7.113273270689883e-05,
+      "loss": 0.4628,
+      "step": 1112
+    },
+    {
+      "epoch": 1.4741721854304637,
+      "grad_norm": 0.2920235367155696,
+      "learning_rate": 7.110949288582024e-05,
+      "loss": 0.4631,
+      "step": 1113
+    },
+    {
+      "epoch": 1.4754966887417218,
+      "grad_norm": 0.30451445184720155,
+      "learning_rate": 7.108622645739596e-05,
+      "loss": 0.4621,
+      "step": 1114
+    },
+    {
+      "epoch": 1.4768211920529801,
+      "grad_norm": 0.34548796150704525,
+      "learning_rate": 7.106293344152528e-05,
+      "loss": 0.4621,
+      "step": 1115
+    },
+    {
+      "epoch": 1.4781456953642385,
+      "grad_norm": 0.3870552741469224,
+      "learning_rate": 7.10396138581303e-05,
+      "loss": 0.4642,
+      "step": 1116
+    },
+    {
+      "epoch": 1.4794701986754966,
+      "grad_norm": 0.3529218569215964,
+      "learning_rate": 7.101626772715578e-05,
+      "loss": 0.4625,
+      "step": 1117
+    },
+    {
+      "epoch": 1.480794701986755,
+      "grad_norm": 0.28534173797844775,
+      "learning_rate": 7.099289506856922e-05,
+      "loss": 0.4651,
+      "step": 1118
+    },
+    {
+      "epoch": 1.4821192052980132,
+      "grad_norm": 0.17475438818358124,
+      "learning_rate": 7.096949590236079e-05,
+      "loss": 0.459,
+      "step": 1119
+    },
+    {
+      "epoch": 1.4834437086092715,
+      "grad_norm": 0.18423648123110545,
+      "learning_rate": 7.094607024854336e-05,
+      "loss": 0.4595,
+      "step": 1120
+    },
+    {
+      "epoch": 1.4847682119205299,
+      "grad_norm": 0.29097949780340987,
+      "learning_rate": 7.092261812715243e-05,
+      "loss": 0.4603,
+      "step": 1121
+    },
+    {
+      "epoch": 1.4860927152317882,
+      "grad_norm": 0.34480539357307416,
+      "learning_rate": 7.089913955824616e-05,
+      "loss": 0.4652,
+      "step": 1122
+    },
+    {
+      "epoch": 1.4874172185430463,
+      "grad_norm": 0.29925271871869896,
+      "learning_rate": 7.087563456190529e-05,
+      "loss": 0.4617,
+      "step": 1123
+    },
+    {
+      "epoch": 1.4887417218543046,
+      "grad_norm": 0.21470433698537697,
+      "learning_rate": 7.08521031582332e-05,
+      "loss": 0.4618,
+      "step": 1124
+    },
+    {
+      "epoch": 1.490066225165563,
+      "grad_norm": 0.1752998156689193,
+      "learning_rate": 7.082854536735583e-05,
+      "loss": 0.462,
+      "step": 1125
+    },
+    {
+      "epoch": 1.491390728476821,
+      "grad_norm": 0.21993271626476413,
+      "learning_rate": 7.080496120942173e-05,
+      "loss": 0.4582,
+      "step": 1126
+    },
+    {
+      "epoch": 1.4927152317880794,
+      "grad_norm": 0.2827226719326882,
+      "learning_rate": 7.078135070460194e-05,
+      "loss": 0.4654,
+      "step": 1127
+    },
+    {
+      "epoch": 1.4940397350993377,
+      "grad_norm": 0.2774991268783928,
+      "learning_rate": 7.075771387309008e-05,
+      "loss": 0.4594,
+      "step": 1128
+    },
+    {
+      "epoch": 1.495364238410596,
+      "grad_norm": 0.21028424241072677,
+      "learning_rate": 7.07340507351023e-05,
+      "loss": 0.4615,
+      "step": 1129
+    },
+    {
+      "epoch": 1.4966887417218544,
+      "grad_norm": 0.1400742560926179,
+      "learning_rate": 7.071036131087716e-05,
+      "loss": 0.4607,
+      "step": 1130
+    },
+    {
+      "epoch": 1.4980132450331125,
+      "grad_norm": 0.20237496689165338,
+      "learning_rate": 7.068664562067582e-05,
+      "loss": 0.4576,
+      "step": 1131
+    },
+    {
+      "epoch": 1.4993377483443708,
+      "grad_norm": 0.24884297173640077,
+      "learning_rate": 7.066290368478184e-05,
+      "loss": 0.4599,
+      "step": 1132
+    },
+    {
+      "epoch": 1.5006622516556292,
+      "grad_norm": 0.25775273784167696,
+      "learning_rate": 7.063913552350123e-05,
+      "loss": 0.4581,
+      "step": 1133
+    },
+    {
+      "epoch": 1.5019867549668873,
+      "grad_norm": 0.20761106628013062,
+      "learning_rate": 7.061534115716243e-05,
+      "loss": 0.4609,
+      "step": 1134
+    },
+    {
+      "epoch": 1.5033112582781456,
+      "grad_norm": 0.19875959711508612,
+      "learning_rate": 7.05915206061163e-05,
+      "loss": 0.459,
+      "step": 1135
+    },
+    {
+      "epoch": 1.504635761589404,
+      "grad_norm": 0.23581757956173993,
+      "learning_rate": 7.056767389073612e-05,
+      "loss": 0.4624,
+      "step": 1136
+    },
+    {
+      "epoch": 1.5059602649006623,
+      "grad_norm": 0.2189049929215525,
+      "learning_rate": 7.054380103141749e-05,
+      "loss": 0.4613,
+      "step": 1137
+    },
+    {
+      "epoch": 1.5072847682119206,
+      "grad_norm": 0.22606162457347956,
+      "learning_rate": 7.051990204857842e-05,
+      "loss": 0.4586,
+      "step": 1138
+    },
+    {
+      "epoch": 1.508609271523179,
+      "grad_norm": 0.20994731120532403,
+      "learning_rate": 7.049597696265924e-05,
+      "loss": 0.4627,
+      "step": 1139
+    },
+    {
+      "epoch": 1.5099337748344372,
+      "grad_norm": 0.1668366662878674,
+      "learning_rate": 7.047202579412263e-05,
+      "loss": 0.4588,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5112582781456954,
+      "grad_norm": 0.14129908030117797,
+      "learning_rate": 7.044804856345354e-05,
+      "loss": 0.4612,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5125827814569537,
+      "grad_norm": 0.14208874570106672,
+      "learning_rate": 7.042404529115924e-05,
+      "loss": 0.4591,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5139072847682118,
+      "grad_norm": 0.14439546079983848,
+      "learning_rate": 7.040001599776926e-05,
+      "loss": 0.4594,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5152317880794701,
+      "grad_norm": 0.19295445993603727,
+      "learning_rate": 7.037596070383539e-05,
+      "loss": 0.4621,
+      "step": 1144
+    },
+    {
+      "epoch": 1.5165562913907285,
+      "grad_norm": 0.2266912716147751,
+      "learning_rate": 7.035187942993163e-05,
+      "loss": 0.4586,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5178807947019868,
+      "grad_norm": 0.2088670673251553,
+      "learning_rate": 7.03277721966543e-05,
+      "loss": 0.4579,
+      "step": 1146
+    },
+    {
+      "epoch": 1.519205298013245,
+      "grad_norm": 0.20351059976425556,
+      "learning_rate": 7.030363902462176e-05,
+      "loss": 0.4587,
+      "step": 1147
+    },
+    {
+      "epoch": 1.5205298013245034,
+      "grad_norm": 0.22143524895587602,
+      "learning_rate": 7.027947993447469e-05,
+      "loss": 0.4625,
+      "step": 1148
+    },
+    {
+      "epoch": 1.5218543046357615,
+      "grad_norm": 0.22872422882907228,
+      "learning_rate": 7.025529494687588e-05,
+      "loss": 0.4636,
+      "step": 1149
+    },
+    {
+      "epoch": 1.5231788079470199,
+      "grad_norm": 0.20979146652482103,
+      "learning_rate": 7.02310840825103e-05,
+      "loss": 0.4635,
+      "step": 1150
+    },
+    {
+      "epoch": 1.524503311258278,
+      "grad_norm": 0.1663951200307748,
+      "learning_rate": 7.020684736208501e-05,
+      "loss": 0.4614,
+      "step": 1151
+    },
+    {
+      "epoch": 1.5258278145695363,
+      "grad_norm": 0.16142709046275283,
+      "learning_rate": 7.018258480632922e-05,
+      "loss": 0.4608,
+      "step": 1152
+    },
+    {
+      "epoch": 1.5271523178807946,
+      "grad_norm": 0.19060198368360098,
+      "learning_rate": 7.01582964359942e-05,
+      "loss": 0.4615,
+      "step": 1153
+    },
+    {
+      "epoch": 1.528476821192053,
+      "grad_norm": 0.21467260105461144,
+      "learning_rate": 7.013398227185336e-05,
+      "loss": 0.4607,
+      "step": 1154
+    },
+    {
+      "epoch": 1.5298013245033113,
+      "grad_norm": 0.2117774449275097,
+      "learning_rate": 7.01096423347021e-05,
+      "loss": 0.463,
+      "step": 1155
+    },
+    {
+      "epoch": 1.5311258278145696,
+      "grad_norm": 0.19199498761399875,
+      "learning_rate": 7.008527664535791e-05,
+      "loss": 0.4624,
+      "step": 1156
+    },
+    {
+      "epoch": 1.532450331125828,
+      "grad_norm": 0.2291246645744794,
+      "learning_rate": 7.00608852246603e-05,
+      "loss": 0.4591,
+      "step": 1157
+    },
+    {
+      "epoch": 1.533774834437086,
+      "grad_norm": 0.26229873610731447,
+      "learning_rate": 7.003646809347076e-05,
+      "loss": 0.4606,
+      "step": 1158
+    },
+    {
+      "epoch": 1.5350993377483444,
+      "grad_norm": 0.2096784662039535,
+      "learning_rate": 7.001202527267281e-05,
+      "loss": 0.4613,
+      "step": 1159
+    },
+    {
+      "epoch": 1.5364238410596025,
+      "grad_norm": 0.15699423471928547,
+      "learning_rate": 6.998755678317189e-05,
+      "loss": 0.4602,
+      "step": 1160
+    },
+    {
+      "epoch": 1.5377483443708608,
+      "grad_norm": 0.1832366663063814,
+      "learning_rate": 6.996306264589544e-05,
+      "loss": 0.46,
+      "step": 1161
+    },
+    {
+      "epoch": 1.5390728476821192,
+      "grad_norm": 0.18355430297018466,
+      "learning_rate": 6.993854288179284e-05,
+      "loss": 0.4599,
+      "step": 1162
+    },
+    {
+      "epoch": 1.5403973509933775,
+      "grad_norm": 0.1801413115131054,
+      "learning_rate": 6.991399751183535e-05,
+      "loss": 0.461,
+      "step": 1163
+    },
+    {
+      "epoch": 1.5417218543046358,
+      "grad_norm": 0.1936348147588445,
+      "learning_rate": 6.988942655701613e-05,
+      "loss": 0.4624,
+      "step": 1164
+    },
+    {
+      "epoch": 1.5430463576158941,
+      "grad_norm": 0.18922926201389467,
+      "learning_rate": 6.986483003835027e-05,
+      "loss": 0.4603,
+      "step": 1165
+    },
+    {
+      "epoch": 1.5443708609271525,
+      "grad_norm": 0.22181726815280522,
+      "learning_rate": 6.984020797687469e-05,
+      "loss": 0.4564,
+      "step": 1166
+    },
+    {
+      "epoch": 1.5456953642384106,
+      "grad_norm": 0.24110743815635935,
+      "learning_rate": 6.981556039364817e-05,
+      "loss": 0.4593,
+      "step": 1167
+    },
+    {
+      "epoch": 1.547019867549669,
+      "grad_norm": 0.2572368803357817,
+      "learning_rate": 6.979088730975128e-05,
+      "loss": 0.4606,
+      "step": 1168
+    },
+    {
+      "epoch": 1.548344370860927,
+      "grad_norm": 0.27340819943958145,
+      "learning_rate": 6.976618874628646e-05,
+      "loss": 0.4612,
+      "step": 1169
+    },
+    {
+      "epoch": 1.5496688741721854,
+      "grad_norm": 0.27280995175480377,
+      "learning_rate": 6.974146472437789e-05,
+      "loss": 0.4623,
+      "step": 1170
+    },
+    {
+      "epoch": 1.5509933774834437,
+      "grad_norm": 0.2690203751377856,
+      "learning_rate": 6.971671526517155e-05,
+      "loss": 0.4593,
+      "step": 1171
+    },
+    {
+      "epoch": 1.552317880794702,
+      "grad_norm": 0.2574563197572923,
+      "learning_rate": 6.96919403898352e-05,
+      "loss": 0.463,
+      "step": 1172
+    },
+    {
+      "epoch": 1.5536423841059603,
+      "grad_norm": 0.23589670956857772,
+      "learning_rate": 6.966714011955826e-05,
+      "loss": 0.458,
+      "step": 1173
+    },
+    {
+      "epoch": 1.5549668874172187,
+      "grad_norm": 0.2164193193816133,
+      "learning_rate": 6.964231447555196e-05,
+      "loss": 0.4625,
+      "step": 1174
+    },
+    {
+      "epoch": 1.5562913907284768,
+      "grad_norm": 0.22694373412435137,
+      "learning_rate": 6.961746347904917e-05,
+      "loss": 0.4609,
+      "step": 1175
+    },
+    {
+      "epoch": 1.557615894039735,
+      "grad_norm": 0.21208003204019388,
+      "learning_rate": 6.95925871513045e-05,
+      "loss": 0.4623,
+      "step": 1176
+    },
+    {
+      "epoch": 1.5589403973509932,
+      "grad_norm": 0.19719284048458996,
+      "learning_rate": 6.956768551359415e-05,
+      "loss": 0.4625,
+      "step": 1177
+    },
+    {
+      "epoch": 1.5602649006622515,
+      "grad_norm": 0.20652037816107544,
+      "learning_rate": 6.954275858721603e-05,
+      "loss": 0.4609,
+      "step": 1178
+    },
+    {
+      "epoch": 1.5615894039735099,
+      "grad_norm": 0.24290207003577066,
+      "learning_rate": 6.951780639348963e-05,
+      "loss": 0.4596,
+      "step": 1179
+    },
+    {
+      "epoch": 1.5629139072847682,
+      "grad_norm": 0.2667229256771541,
+      "learning_rate": 6.949282895375612e-05,
+      "loss": 0.4615,
+      "step": 1180
+    },
+    {
+      "epoch": 1.5642384105960265,
+      "grad_norm": 0.22813027799551208,
+      "learning_rate": 6.946782628937819e-05,
+      "loss": 0.458,
+      "step": 1181
+    },
+    {
+      "epoch": 1.5655629139072849,
+      "grad_norm": 0.1861439690300803,
+      "learning_rate": 6.944279842174013e-05,
+      "loss": 0.4599,
+      "step": 1182
+    },
+    {
+      "epoch": 1.5668874172185432,
+      "grad_norm": 0.1822432134381999,
+      "learning_rate": 6.941774537224781e-05,
+      "loss": 0.4616,
+      "step": 1183
+    },
+    {
+      "epoch": 1.5682119205298013,
+      "grad_norm": 0.17253357637648026,
+      "learning_rate": 6.93926671623286e-05,
+      "loss": 0.4628,
+      "step": 1184
+    },
+    {
+      "epoch": 1.5695364238410596,
+      "grad_norm": 0.1807157195190629,
+      "learning_rate": 6.936756381343142e-05,
+      "loss": 0.4589,
+      "step": 1185
+    },
+    {
+      "epoch": 1.5708609271523177,
+      "grad_norm": 0.16122340785252048,
+      "learning_rate": 6.934243534702667e-05,
+      "loss": 0.458,
+      "step": 1186
+    },
+    {
+      "epoch": 1.572185430463576,
+      "grad_norm": 0.1862503451811599,
+      "learning_rate": 6.931728178460624e-05,
+      "loss": 0.4605,
+      "step": 1187
+    },
+    {
+      "epoch": 1.5735099337748344,
+      "grad_norm": 0.19415174260141882,
+      "learning_rate": 6.929210314768348e-05,
+      "loss": 0.4607,
+      "step": 1188
+    },
+    {
+      "epoch": 1.5748344370860927,
+      "grad_norm": 0.164851758193945,
+      "learning_rate": 6.92668994577932e-05,
+      "loss": 0.4632,
+      "step": 1189
+    },
+    {
+      "epoch": 1.576158940397351,
+      "grad_norm": 0.1700663123191463,
+      "learning_rate": 6.924167073649163e-05,
+      "loss": 0.46,
+      "step": 1190
+    },
+    {
+      "epoch": 1.5774834437086094,
+      "grad_norm": 0.2015498342125039,
+      "learning_rate": 6.921641700535638e-05,
+      "loss": 0.4594,
+      "step": 1191
+    },
+    {
+      "epoch": 1.5788079470198677,
+      "grad_norm": 0.19991491166394315,
+      "learning_rate": 6.91911382859865e-05,
+      "loss": 0.46,
+      "step": 1192
+    },
+    {
+      "epoch": 1.5801324503311258,
+      "grad_norm": 0.1852689212675992,
+      "learning_rate": 6.916583460000238e-05,
+      "loss": 0.4599,
+      "step": 1193
+    },
+    {
+      "epoch": 1.5814569536423841,
+      "grad_norm": 0.2204416990881589,
+      "learning_rate": 6.914050596904577e-05,
+      "loss": 0.4585,
+      "step": 1194
+    },
+    {
+      "epoch": 1.5827814569536423,
+      "grad_norm": 0.25714725912813485,
+      "learning_rate": 6.911515241477978e-05,
+      "loss": 0.4581,
+      "step": 1195
+    },
+    {
+      "epoch": 1.5841059602649006,
+      "grad_norm": 0.25934623135671764,
+      "learning_rate": 6.908977395888876e-05,
+      "loss": 0.4593,
+      "step": 1196
+    },
+    {
+      "epoch": 1.585430463576159,
+      "grad_norm": 0.2231170119500944,
+      "learning_rate": 6.906437062307847e-05,
+      "loss": 0.4567,
+      "step": 1197
+    },
+    {
+      "epoch": 1.5867549668874172,
+      "grad_norm": 0.146522132913816,
+      "learning_rate": 6.903894242907586e-05,
+      "loss": 0.4616,
+      "step": 1198
+    },
+    {
+      "epoch": 1.5880794701986756,
+      "grad_norm": 0.19768120768488517,
+      "learning_rate": 6.901348939862916e-05,
+      "loss": 0.4599,
+      "step": 1199
+    },
+    {
+      "epoch": 1.589403973509934,
+      "grad_norm": 0.2364082626612461,
+      "learning_rate": 6.898801155350788e-05,
+      "loss": 0.4611,
+      "step": 1200
+    },
+    {
+      "epoch": 1.590728476821192,
+      "grad_norm": 0.23356005948632258,
+      "learning_rate": 6.896250891550271e-05,
+      "loss": 0.4603,
+      "step": 1201
+    },
+    {
+      "epoch": 1.5920529801324503,
+      "grad_norm": 0.1887773419844515,
+      "learning_rate": 6.893698150642558e-05,
+      "loss": 0.461,
+      "step": 1202
+    },
+    {
+      "epoch": 1.5933774834437087,
+      "grad_norm": 0.18868352770781913,
+      "learning_rate": 6.891142934810956e-05,
+      "loss": 0.4623,
+      "step": 1203
+    },
+    {
+      "epoch": 1.5947019867549668,
+      "grad_norm": 0.1972835536569781,
+      "learning_rate": 6.888585246240895e-05,
+      "loss": 0.4589,
+      "step": 1204
+    },
+    {
+      "epoch": 1.596026490066225,
+      "grad_norm": 0.1732569549175012,
+      "learning_rate": 6.886025087119911e-05,
+      "loss": 0.4597,
+      "step": 1205
+    },
+    {
+      "epoch": 1.5973509933774834,
+      "grad_norm": 0.17558604253926785,
+      "learning_rate": 6.883462459637664e-05,
+      "loss": 0.4543,
+      "step": 1206
+    },
+    {
+      "epoch": 1.5986754966887418,
+      "grad_norm": 0.18008887233697357,
+      "learning_rate": 6.880897365985917e-05,
+      "loss": 0.457,
+      "step": 1207
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.18271459981079283,
+      "learning_rate": 6.878329808358544e-05,
+      "loss": 0.4585,
+      "step": 1208
+    },
+    {
+      "epoch": 1.6013245033112584,
+      "grad_norm": 0.22069896930911517,
+      "learning_rate": 6.87575978895153e-05,
+      "loss": 0.4589,
+      "step": 1209
+    },
+    {
+      "epoch": 1.6026490066225165,
+      "grad_norm": 0.24034056563095274,
+      "learning_rate": 6.87318730996296e-05,
+      "loss": 0.4609,
+      "step": 1210
+    },
+    {
+      "epoch": 1.6039735099337749,
+      "grad_norm": 0.20445185933472385,
+      "learning_rate": 6.870612373593028e-05,
+      "loss": 0.4572,
+      "step": 1211
+    },
+    {
+      "epoch": 1.605298013245033,
+      "grad_norm": 0.18842201594896324,
+      "learning_rate": 6.868034982044025e-05,
+      "loss": 0.4611,
+      "step": 1212
+    },
+    {
+      "epoch": 1.6066225165562913,
+      "grad_norm": 0.2063426301044817,
+      "learning_rate": 6.865455137520344e-05,
+      "loss": 0.4592,
+      "step": 1213
+    },
+    {
+      "epoch": 1.6079470198675496,
+      "grad_norm": 0.1860122286057625,
+      "learning_rate": 6.862872842228478e-05,
+      "loss": 0.4578,
+      "step": 1214
+    },
+    {
+      "epoch": 1.609271523178808,
+      "grad_norm": 0.1830152953510653,
+      "learning_rate": 6.860288098377011e-05,
+      "loss": 0.4586,
+      "step": 1215
+    },
+    {
+      "epoch": 1.6105960264900663,
+      "grad_norm": 0.21105960990710787,
+      "learning_rate": 6.857700908176625e-05,
+      "loss": 0.4598,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6119205298013246,
+      "grad_norm": 0.19571939351648607,
+      "learning_rate": 6.855111273840097e-05,
+      "loss": 0.4604,
+      "step": 1217
+    },
+    {
+      "epoch": 1.613245033112583,
+      "grad_norm": 0.17070160009595178,
+      "learning_rate": 6.852519197582288e-05,
+      "loss": 0.4596,
+      "step": 1218
+    },
+    {
+      "epoch": 1.614569536423841,
+      "grad_norm": 0.1791220301718667,
+      "learning_rate": 6.84992468162015e-05,
+      "loss": 0.4606,
+      "step": 1219
+    },
+    {
+      "epoch": 1.6158940397350994,
+      "grad_norm": 0.20384129119953623,
+      "learning_rate": 6.847327728172725e-05,
+      "loss": 0.4627,
+      "step": 1220
+    },
+    {
+      "epoch": 1.6172185430463575,
+      "grad_norm": 0.20458024535613722,
+      "learning_rate": 6.844728339461133e-05,
+      "loss": 0.4587,
+      "step": 1221
+    },
+    {
+      "epoch": 1.6185430463576158,
+      "grad_norm": 0.2027415499260645,
+      "learning_rate": 6.842126517708584e-05,
+      "loss": 0.4559,
+      "step": 1222
+    },
+    {
+      "epoch": 1.6198675496688741,
+      "grad_norm": 0.1805192376721496,
+      "learning_rate": 6.839522265140365e-05,
+      "loss": 0.4585,
+      "step": 1223
+    },
+    {
+      "epoch": 1.6211920529801325,
+      "grad_norm": 0.16443558987947632,
+      "learning_rate": 6.836915583983841e-05,
+      "loss": 0.4575,
+      "step": 1224
+    },
+    {
+      "epoch": 1.6225165562913908,
+      "grad_norm": 0.1935359853360884,
+      "learning_rate": 6.834306476468457e-05,
+      "loss": 0.4598,
+      "step": 1225
+    },
+    {
+      "epoch": 1.6238410596026491,
+      "grad_norm": 0.2075640530356208,
+      "learning_rate": 6.831694944825733e-05,
+      "loss": 0.4575,
+      "step": 1226
+    },
+    {
+      "epoch": 1.6251655629139072,
+      "grad_norm": 0.22740340261359548,
+      "learning_rate": 6.829080991289261e-05,
+      "loss": 0.4568,
+      "step": 1227
+    },
+    {
+      "epoch": 1.6264900662251656,
+      "grad_norm": 0.24822708874657887,
+      "learning_rate": 6.826464618094703e-05,
+      "loss": 0.4639,
+      "step": 1228
+    },
+    {
+      "epoch": 1.627814569536424,
+      "grad_norm": 0.24124029459813678,
+      "learning_rate": 6.823845827479794e-05,
+      "loss": 0.4605,
+      "step": 1229
+    },
+    {
+      "epoch": 1.629139072847682,
+      "grad_norm": 0.21193012852069015,
+      "learning_rate": 6.821224621684334e-05,
+      "loss": 0.462,
+      "step": 1230
+    },
+    {
+      "epoch": 1.6304635761589403,
+      "grad_norm": 0.198682025042783,
+      "learning_rate": 6.818601002950189e-05,
+      "loss": 0.4608,
+      "step": 1231
+    },
+    {
+      "epoch": 1.6317880794701987,
+      "grad_norm": 0.17344437746731664,
+      "learning_rate": 6.815974973521291e-05,
+      "loss": 0.4605,
+      "step": 1232
+    },
+    {
+      "epoch": 1.633112582781457,
+      "grad_norm": 0.19345055554324278,
+      "learning_rate": 6.813346535643631e-05,
+      "loss": 0.4571,
+      "step": 1233
+    },
+    {
+      "epoch": 1.6344370860927153,
+      "grad_norm": 0.23142533371047533,
+      "learning_rate": 6.81071569156526e-05,
+      "loss": 0.4569,
+      "step": 1234
+    },
+    {
+      "epoch": 1.6357615894039736,
+      "grad_norm": 0.26684249832478324,
+      "learning_rate": 6.808082443536289e-05,
+      "loss": 0.4589,
+      "step": 1235
+    },
+    {
+      "epoch": 1.6370860927152318,
+      "grad_norm": 0.24559722812765936,
+      "learning_rate": 6.805446793808884e-05,
+      "loss": 0.4593,
+      "step": 1236
+    },
+    {
+      "epoch": 1.63841059602649,
+      "grad_norm": 0.18036623718320416,
+      "learning_rate": 6.802808744637263e-05,
+      "loss": 0.461,
+      "step": 1237
+    },
+    {
+      "epoch": 1.6397350993377482,
+      "grad_norm": 0.1996660897639992,
+      "learning_rate": 6.800168298277702e-05,
+      "loss": 0.4593,
+      "step": 1238
+    },
+    {
+      "epoch": 1.6410596026490065,
+      "grad_norm": 0.2681490668895438,
+      "learning_rate": 6.79752545698852e-05,
+      "loss": 0.4608,
+      "step": 1239
+    },
+    {
+      "epoch": 1.6423841059602649,
+      "grad_norm": 0.2544290983298035,
+      "learning_rate": 6.794880223030091e-05,
+      "loss": 0.4595,
+      "step": 1240
+    },
+    {
+      "epoch": 1.6437086092715232,
+      "grad_norm": 0.2216076123411865,
+      "learning_rate": 6.792232598664831e-05,
+      "loss": 0.4608,
+      "step": 1241
+    },
+    {
+      "epoch": 1.6450331125827815,
+      "grad_norm": 0.1734189764652301,
+      "learning_rate": 6.789582586157201e-05,
+      "loss": 0.4556,
+      "step": 1242
+    },
+    {
+      "epoch": 1.6463576158940398,
+      "grad_norm": 0.160170391464827,
+      "learning_rate": 6.786930187773705e-05,
+      "loss": 0.4559,
+      "step": 1243
+    },
+    {
+      "epoch": 1.6476821192052982,
+      "grad_norm": 0.21380922552971757,
+      "learning_rate": 6.784275405782889e-05,
+      "loss": 0.4614,
+      "step": 1244
+    },
+    {
+      "epoch": 1.6490066225165563,
+      "grad_norm": 0.21215140039430652,
+      "learning_rate": 6.781618242455336e-05,
+      "loss": 0.4579,
+      "step": 1245
+    },
+    {
+      "epoch": 1.6503311258278146,
+      "grad_norm": 0.15800921664322257,
+      "learning_rate": 6.778958700063668e-05,
+      "loss": 0.4621,
+      "step": 1246
+    },
+    {
+      "epoch": 1.6516556291390727,
+      "grad_norm": 0.14255461879047662,
+      "learning_rate": 6.776296780882537e-05,
+      "loss": 0.456,
+      "step": 1247
+    },
+    {
+      "epoch": 1.652980132450331,
+      "grad_norm": 0.18087417457085486,
+      "learning_rate": 6.773632487188631e-05,
+      "loss": 0.4614,
+      "step": 1248
+    },
+    {
+      "epoch": 1.6543046357615894,
+      "grad_norm": 0.19299185598100338,
+      "learning_rate": 6.77096582126067e-05,
+      "loss": 0.4611,
+      "step": 1249
+    },
+    {
+      "epoch": 1.6556291390728477,
+      "grad_norm": 0.18289776786796577,
+      "learning_rate": 6.768296785379404e-05,
+      "loss": 0.4587,
+      "step": 1250
+    },
+    {
+      "epoch": 1.656953642384106,
+      "grad_norm": 0.18450752778682764,
+      "learning_rate": 6.765625381827605e-05,
+      "loss": 0.4587,
+      "step": 1251
+    },
+    {
+      "epoch": 1.6582781456953644,
+      "grad_norm": 0.16664978854504942,
+      "learning_rate": 6.762951612890073e-05,
+      "loss": 0.4576,
+      "step": 1252
+    },
+    {
+      "epoch": 1.6596026490066225,
+      "grad_norm": 0.1595324176718375,
+      "learning_rate": 6.760275480853629e-05,
+      "loss": 0.4591,
+      "step": 1253
+    },
+    {
+      "epoch": 1.6609271523178808,
+      "grad_norm": 0.18793976302194315,
+      "learning_rate": 6.757596988007118e-05,
+      "loss": 0.4585,
+      "step": 1254
+    },
+    {
+      "epoch": 1.6622516556291391,
+      "grad_norm": 0.17641883372856854,
+      "learning_rate": 6.754916136641405e-05,
+      "loss": 0.4609,
+      "step": 1255
+    },
+    {
+      "epoch": 1.6635761589403972,
+      "grad_norm": 0.20559329600997198,
+      "learning_rate": 6.752232929049367e-05,
+      "loss": 0.4605,
+      "step": 1256
+    },
+    {
+      "epoch": 1.6649006622516556,
+      "grad_norm": 0.20525465518718466,
+      "learning_rate": 6.749547367525903e-05,
+      "loss": 0.4586,
+      "step": 1257
+    },
+    {
+      "epoch": 1.666225165562914,
+      "grad_norm": 0.15213014180033305,
+      "learning_rate": 6.746859454367917e-05,
+      "loss": 0.4624,
+      "step": 1258
+    },
+    {
+      "epoch": 1.6675496688741722,
+      "grad_norm": 0.14521695458624856,
+      "learning_rate": 6.744169191874334e-05,
+      "loss": 0.4586,
+      "step": 1259
+    },
+    {
+      "epoch": 1.6688741721854305,
+      "grad_norm": 0.17095996926205456,
+      "learning_rate": 6.741476582346079e-05,
+      "loss": 0.4591,
+      "step": 1260
+    },
+    {
+      "epoch": 1.6701986754966889,
+      "grad_norm": 0.17910450197397276,
+      "learning_rate": 6.738781628086092e-05,
+      "loss": 0.4604,
+      "step": 1261
+    },
+    {
+      "epoch": 1.671523178807947,
+      "grad_norm": 0.16811543888464683,
+      "learning_rate": 6.736084331399312e-05,
+      "loss": 0.4568,
+      "step": 1262
+    },
+    {
+      "epoch": 1.6728476821192053,
+      "grad_norm": 0.1870040039495568,
+      "learning_rate": 6.733384694592687e-05,
+      "loss": 0.4572,
+      "step": 1263
+    },
+    {
+      "epoch": 1.6741721854304634,
+      "grad_norm": 0.18295354345907902,
+      "learning_rate": 6.730682719975163e-05,
+      "loss": 0.4615,
+      "step": 1264
+    },
+    {
+      "epoch": 1.6754966887417218,
+      "grad_norm": 0.17651386247022202,
+      "learning_rate": 6.727978409857687e-05,
+      "loss": 0.4611,
+      "step": 1265
+    },
+    {
+      "epoch": 1.67682119205298,
+      "grad_norm": 0.20903239865141005,
+      "learning_rate": 6.725271766553201e-05,
+      "loss": 0.4551,
+      "step": 1266
+    },
+    {
+      "epoch": 1.6781456953642384,
+      "grad_norm": 0.20919567969680206,
+      "learning_rate": 6.722562792376647e-05,
+      "loss": 0.4628,
+      "step": 1267
+    },
+    {
+      "epoch": 1.6794701986754967,
+      "grad_norm": 0.1638423861844039,
+      "learning_rate": 6.719851489644958e-05,
+      "loss": 0.4593,
+      "step": 1268
+    },
+    {
+      "epoch": 1.680794701986755,
+      "grad_norm": 0.18695703220317864,
+      "learning_rate": 6.717137860677057e-05,
+      "loss": 0.4616,
+      "step": 1269
+    },
+    {
+      "epoch": 1.6821192052980134,
+      "grad_norm": 0.1761026515510813,
+      "learning_rate": 6.714421907793859e-05,
+      "loss": 0.4565,
+      "step": 1270
+    },
+    {
+      "epoch": 1.6834437086092715,
+      "grad_norm": 0.1618823664075114,
+      "learning_rate": 6.711703633318265e-05,
+      "loss": 0.4591,
+      "step": 1271
+    },
+    {
+      "epoch": 1.6847682119205298,
+      "grad_norm": 0.18071868997001528,
+      "learning_rate": 6.708983039575165e-05,
+      "loss": 0.4594,
+      "step": 1272
+    },
+    {
+      "epoch": 1.686092715231788,
+      "grad_norm": 0.1825567656155801,
+      "learning_rate": 6.706260128891428e-05,
+      "loss": 0.4577,
+      "step": 1273
+    },
+    {
+      "epoch": 1.6874172185430463,
+      "grad_norm": 0.17650769335467226,
+      "learning_rate": 6.703534903595905e-05,
+      "loss": 0.4588,
+      "step": 1274
+    },
+    {
+      "epoch": 1.6887417218543046,
+      "grad_norm": 0.18000288491556438,
+      "learning_rate": 6.700807366019432e-05,
+      "loss": 0.4573,
+      "step": 1275
+    },
+    {
+      "epoch": 1.690066225165563,
+      "grad_norm": 0.1782069988491776,
+      "learning_rate": 6.698077518494817e-05,
+      "loss": 0.4573,
+      "step": 1276
+    },
+    {
+      "epoch": 1.6913907284768213,
+      "grad_norm": 0.19670491878332996,
+      "learning_rate": 6.695345363356846e-05,
+      "loss": 0.459,
+      "step": 1277
+    },
+    {
+      "epoch": 1.6927152317880796,
+      "grad_norm": 0.20382196482499046,
+      "learning_rate": 6.692610902942278e-05,
+      "loss": 0.4592,
+      "step": 1278
+    },
+    {
+      "epoch": 1.6940397350993377,
+      "grad_norm": 0.2363090736103728,
+      "learning_rate": 6.689874139589845e-05,
+      "loss": 0.4554,
+      "step": 1279
+    },
+    {
+      "epoch": 1.695364238410596,
+      "grad_norm": 0.21598957151755055,
+      "learning_rate": 6.687135075640244e-05,
+      "loss": 0.4592,
+      "step": 1280
+    },
+    {
+      "epoch": 1.6966887417218544,
+      "grad_norm": 0.2106909893881921,
+      "learning_rate": 6.684393713436148e-05,
+      "loss": 0.4581,
+      "step": 1281
+    },
+    {
+      "epoch": 1.6980132450331125,
+      "grad_norm": 0.272842150759551,
+      "learning_rate": 6.681650055322188e-05,
+      "loss": 0.4575,
+      "step": 1282
+    },
+    {
+      "epoch": 1.6993377483443708,
+      "grad_norm": 0.3365831460046258,
+      "learning_rate": 6.678904103644963e-05,
+      "loss": 0.4579,
+      "step": 1283
+    },
+    {
+      "epoch": 1.7006622516556291,
+      "grad_norm": 0.36874564401237986,
+      "learning_rate": 6.676155860753032e-05,
+      "loss": 0.4576,
+      "step": 1284
+    },
+    {
+      "epoch": 1.7019867549668874,
+      "grad_norm": 0.4003607118152827,
+      "learning_rate": 6.673405328996912e-05,
+      "loss": 0.4593,
+      "step": 1285
+    },
+    {
+      "epoch": 1.7033112582781458,
+      "grad_norm": 0.3805402588829684,
+      "learning_rate": 6.670652510729081e-05,
+      "loss": 0.4574,
+      "step": 1286
+    },
+    {
+      "epoch": 1.704635761589404,
+      "grad_norm": 0.2782318763500424,
+      "learning_rate": 6.667897408303972e-05,
+      "loss": 0.4606,
+      "step": 1287
+    },
+    {
+      "epoch": 1.7059602649006622,
+      "grad_norm": 0.1529093272984365,
+      "learning_rate": 6.66514002407797e-05,
+      "loss": 0.4588,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7072847682119205,
+      "grad_norm": 0.20632534256754756,
+      "learning_rate": 6.66238036040941e-05,
+      "loss": 0.4613,
+      "step": 1289
+    },
+    {
+      "epoch": 1.7086092715231787,
+      "grad_norm": 0.32963037028284853,
+      "learning_rate": 6.659618419658581e-05,
+      "loss": 0.456,
+      "step": 1290
+    },
+    {
+      "epoch": 1.709933774834437,
+      "grad_norm": 0.4002053716370613,
+      "learning_rate": 6.656854204187718e-05,
+      "loss": 0.4593,
+      "step": 1291
+    },
+    {
+      "epoch": 1.7112582781456953,
+      "grad_norm": 0.3385613016217147,
+      "learning_rate": 6.654087716361e-05,
+      "loss": 0.4591,
+      "step": 1292
+    },
+    {
+      "epoch": 1.7125827814569536,
+      "grad_norm": 0.24908069109814882,
+      "learning_rate": 6.651318958544549e-05,
+      "loss": 0.458,
+      "step": 1293
+    },
+    {
+      "epoch": 1.713907284768212,
+      "grad_norm": 0.18063167005899655,
+      "learning_rate": 6.648547933106431e-05,
+      "loss": 0.463,
+      "step": 1294
+    },
+    {
+      "epoch": 1.7152317880794703,
+      "grad_norm": 0.17647659460058235,
+      "learning_rate": 6.645774642416651e-05,
+      "loss": 0.4617,
+      "step": 1295
+    },
+    {
+      "epoch": 1.7165562913907286,
+      "grad_norm": 0.21938354552066236,
+      "learning_rate": 6.64299908884715e-05,
+      "loss": 0.4603,
+      "step": 1296
+    },
+    {
+      "epoch": 1.7178807947019867,
+      "grad_norm": 0.2521208368816925,
+      "learning_rate": 6.640221274771805e-05,
+      "loss": 0.4593,
+      "step": 1297
+    },
+    {
+      "epoch": 1.719205298013245,
+      "grad_norm": 0.24396979469429406,
+      "learning_rate": 6.637441202566428e-05,
+      "loss": 0.4581,
+      "step": 1298
+    },
+    {
+      "epoch": 1.7205298013245032,
+      "grad_norm": 0.19114076985698936,
+      "learning_rate": 6.634658874608757e-05,
+      "loss": 0.4584,
+      "step": 1299
+    },
+    {
+      "epoch": 1.7218543046357615,
+      "grad_norm": 0.18174616076846623,
+      "learning_rate": 6.631874293278467e-05,
+      "loss": 0.4583,
+      "step": 1300
+    },
+    {
+      "epoch": 1.7231788079470198,
+      "grad_norm": 0.17799082682616604,
+      "learning_rate": 6.629087460957154e-05,
+      "loss": 0.4564,
+      "step": 1301
+    },
+    {
+      "epoch": 1.7245033112582782,
+      "grad_norm": 0.16766033865971336,
+      "learning_rate": 6.626298380028344e-05,
+      "loss": 0.4579,
+      "step": 1302
+    },
+    {
+      "epoch": 1.7258278145695365,
+      "grad_norm": 0.17692598115328556,
+      "learning_rate": 6.62350705287748e-05,
+      "loss": 0.46,
+      "step": 1303
+    },
+    {
+      "epoch": 1.7271523178807948,
+      "grad_norm": 0.19882535808781227,
+      "learning_rate": 6.620713481891935e-05,
+      "loss": 0.4597,
+      "step": 1304
+    },
+    {
+      "epoch": 1.728476821192053,
+      "grad_norm": 0.19203925081816803,
+      "learning_rate": 6.617917669460993e-05,
+      "loss": 0.4588,
+      "step": 1305
+    },
+    {
+      "epoch": 1.7298013245033113,
+      "grad_norm": 0.2097721172846332,
+      "learning_rate": 6.615119617975858e-05,
+      "loss": 0.4595,
+      "step": 1306
+    },
+    {
+      "epoch": 1.7311258278145696,
+      "grad_norm": 0.22761456574044486,
+      "learning_rate": 6.61231932982965e-05,
+      "loss": 0.4544,
+      "step": 1307
+    },
+    {
+      "epoch": 1.7324503311258277,
+      "grad_norm": 0.1821961354871429,
+      "learning_rate": 6.6095168074174e-05,
+      "loss": 0.4589,
+      "step": 1308
+    },
+    {
+      "epoch": 1.733774834437086,
+      "grad_norm": 0.14490939524643048,
+      "learning_rate": 6.606712053136052e-05,
+      "loss": 0.4567,
+      "step": 1309
+    },
+    {
+      "epoch": 1.7350993377483444,
+      "grad_norm": 0.1608232006913273,
+      "learning_rate": 6.603905069384458e-05,
+      "loss": 0.4594,
+      "step": 1310
+    },
+    {
+      "epoch": 1.7364238410596027,
+      "grad_norm": 0.14193181740307795,
+      "learning_rate": 6.601095858563376e-05,
+      "loss": 0.4612,
+      "step": 1311
+    },
+    {
+      "epoch": 1.737748344370861,
+      "grad_norm": 0.1574268248734957,
+      "learning_rate": 6.598284423075471e-05,
+      "loss": 0.455,
+      "step": 1312
+    },
+    {
+      "epoch": 1.7390728476821193,
+      "grad_norm": 0.14061219104294895,
+      "learning_rate": 6.595470765325309e-05,
+      "loss": 0.459,
+      "step": 1313
+    },
+    {
+      "epoch": 1.7403973509933774,
+      "grad_norm": 0.13692995844968936,
+      "learning_rate": 6.592654887719355e-05,
+      "loss": 0.458,
+      "step": 1314
+    },
+    {
+      "epoch": 1.7417218543046358,
+      "grad_norm": 0.1479187964023412,
+      "learning_rate": 6.589836792665976e-05,
+      "loss": 0.4585,
+      "step": 1315
+    },
+    {
+      "epoch": 1.7430463576158939,
+      "grad_norm": 0.16641084577580537,
+      "learning_rate": 6.587016482575433e-05,
+      "loss": 0.4611,
+      "step": 1316
+    },
+    {
+      "epoch": 1.7443708609271522,
+      "grad_norm": 0.16438050868935902,
+      "learning_rate": 6.584193959859883e-05,
+      "loss": 0.4612,
+      "step": 1317
+    },
+    {
+      "epoch": 1.7456953642384105,
+      "grad_norm": 0.127630920088683,
+      "learning_rate": 6.581369226933376e-05,
+      "loss": 0.457,
+      "step": 1318
+    },
+    {
+      "epoch": 1.7470198675496689,
+      "grad_norm": 0.13968485349009116,
+      "learning_rate": 6.578542286211851e-05,
+      "loss": 0.459,
+      "step": 1319
+    },
+    {
+      "epoch": 1.7483443708609272,
+      "grad_norm": 0.14367616989940696,
+      "learning_rate": 6.575713140113135e-05,
+      "loss": 0.4575,
+      "step": 1320
+    },
+    {
+      "epoch": 1.7496688741721855,
+      "grad_norm": 0.1690681141425111,
+      "learning_rate": 6.57288179105694e-05,
+      "loss": 0.459,
+      "step": 1321
+    },
+    {
+      "epoch": 1.7509933774834439,
+      "grad_norm": 0.15338485412244704,
+      "learning_rate": 6.570048241464866e-05,
+      "loss": 0.4537,
+      "step": 1322
+    },
+    {
+      "epoch": 1.752317880794702,
+      "grad_norm": 0.17661676978790444,
+      "learning_rate": 6.567212493760394e-05,
+      "loss": 0.4583,
+      "step": 1323
+    },
+    {
+      "epoch": 1.7536423841059603,
+      "grad_norm": 0.2135888389556036,
+      "learning_rate": 6.56437455036888e-05,
+      "loss": 0.4582,
+      "step": 1324
+    },
+    {
+      "epoch": 1.7549668874172184,
+      "grad_norm": 0.23013900710764906,
+      "learning_rate": 6.561534413717568e-05,
+      "loss": 0.4585,
+      "step": 1325
+    },
+    {
+      "epoch": 1.7562913907284767,
+      "grad_norm": 0.2339985543717907,
+      "learning_rate": 6.558692086235565e-05,
+      "loss": 0.46,
+      "step": 1326
+    },
+    {
+      "epoch": 1.757615894039735,
+      "grad_norm": 0.24172730859506247,
+      "learning_rate": 6.555847570353864e-05,
+      "loss": 0.4564,
+      "step": 1327
+    },
+    {
+      "epoch": 1.7589403973509934,
+      "grad_norm": 0.24409055724000997,
+      "learning_rate": 6.553000868505321e-05,
+      "loss": 0.4563,
+      "step": 1328
+    },
+    {
+      "epoch": 1.7602649006622517,
+      "grad_norm": 0.17060909042067615,
+      "learning_rate": 6.550151983124666e-05,
+      "loss": 0.4578,
+      "step": 1329
+    },
+    {
+      "epoch": 1.76158940397351,
+      "grad_norm": 0.14066985057943587,
+      "learning_rate": 6.547300916648496e-05,
+      "loss": 0.4577,
+      "step": 1330
+    },
+    {
+      "epoch": 1.7629139072847684,
+      "grad_norm": 0.14858403209274482,
+      "learning_rate": 6.544447671515272e-05,
+      "loss": 0.4558,
+      "step": 1331
+    },
+    {
+      "epoch": 1.7642384105960265,
+      "grad_norm": 0.18141289438193864,
+      "learning_rate": 6.54159225016532e-05,
+      "loss": 0.458,
+      "step": 1332
+    },
+    {
+      "epoch": 1.7655629139072848,
+      "grad_norm": 0.18496716440881464,
+      "learning_rate": 6.538734655040826e-05,
+      "loss": 0.4604,
+      "step": 1333
+    },
+    {
+      "epoch": 1.766887417218543,
+      "grad_norm": 0.16476540017810706,
+      "learning_rate": 6.535874888585833e-05,
+      "loss": 0.4549,
+      "step": 1334
+    },
+    {
+      "epoch": 1.7682119205298013,
+      "grad_norm": 0.15692598562557855,
+      "learning_rate": 6.533012953246249e-05,
+      "loss": 0.4588,
+      "step": 1335
+    },
+    {
+      "epoch": 1.7695364238410596,
+      "grad_norm": 0.17168433793108323,
+      "learning_rate": 6.530148851469827e-05,
+      "loss": 0.4575,
+      "step": 1336
+    },
+    {
+      "epoch": 1.770860927152318,
+      "grad_norm": 0.19113529740453947,
+      "learning_rate": 6.527282585706182e-05,
+      "loss": 0.459,
+      "step": 1337
+    },
+    {
+      "epoch": 1.7721854304635762,
+      "grad_norm": 0.17624126101269205,
+      "learning_rate": 6.524414158406771e-05,
+      "loss": 0.4548,
+      "step": 1338
+    },
+    {
+      "epoch": 1.7735099337748346,
+      "grad_norm": 0.1739402171074709,
+      "learning_rate": 6.52154357202491e-05,
+      "loss": 0.4559,
+      "step": 1339
+    },
+    {
+      "epoch": 1.7748344370860927,
+      "grad_norm": 0.17985271016027438,
+      "learning_rate": 6.518670829015749e-05,
+      "loss": 0.454,
+      "step": 1340
+    },
+    {
+      "epoch": 1.776158940397351,
+      "grad_norm": 0.15672505299693462,
+      "learning_rate": 6.515795931836293e-05,
+      "loss": 0.4563,
+      "step": 1341
+    },
+    {
+      "epoch": 1.7774834437086091,
+      "grad_norm": 0.16534698302581147,
+      "learning_rate": 6.512918882945387e-05,
+      "loss": 0.4566,
+      "step": 1342
+    },
+    {
+      "epoch": 1.7788079470198674,
+      "grad_norm": 0.17069915824770518,
+      "learning_rate": 6.510039684803712e-05,
+      "loss": 0.4558,
+      "step": 1343
+    },
+    {
+      "epoch": 1.7801324503311258,
+      "grad_norm": 0.1742359376310982,
+      "learning_rate": 6.507158339873793e-05,
+      "loss": 0.4559,
+      "step": 1344
+    },
+    {
+      "epoch": 1.781456953642384,
+      "grad_norm": 0.16033139895676476,
+      "learning_rate": 6.504274850619988e-05,
+      "loss": 0.4588,
+      "step": 1345
+    },
+    {
+      "epoch": 1.7827814569536424,
+      "grad_norm": 0.17524964167224855,
+      "learning_rate": 6.501389219508487e-05,
+      "loss": 0.4565,
+      "step": 1346
+    },
+    {
+      "epoch": 1.7841059602649008,
+      "grad_norm": 0.18795689555973796,
+      "learning_rate": 6.498501449007317e-05,
+      "loss": 0.4558,
+      "step": 1347
+    },
+    {
+      "epoch": 1.785430463576159,
+      "grad_norm": 0.1668243934367087,
+      "learning_rate": 6.49561154158633e-05,
+      "loss": 0.4577,
+      "step": 1348
+    },
+    {
+      "epoch": 1.7867549668874172,
+      "grad_norm": 0.16051145430410596,
+      "learning_rate": 6.49271949971721e-05,
+      "loss": 0.4571,
+      "step": 1349
+    },
+    {
+      "epoch": 1.7880794701986755,
+      "grad_norm": 0.12122081524204774,
+      "learning_rate": 6.489825325873464e-05,
+      "loss": 0.4558,
+      "step": 1350
+    },
+    {
+      "epoch": 1.7894039735099336,
+      "grad_norm": 0.13595018419597787,
+      "learning_rate": 6.48692902253042e-05,
+      "loss": 0.4571,
+      "step": 1351
+    },
+    {
+      "epoch": 1.790728476821192,
+      "grad_norm": 0.16290432334191166,
+      "learning_rate": 6.484030592165232e-05,
+      "loss": 0.4558,
+      "step": 1352
+    },
+    {
+      "epoch": 1.7920529801324503,
+      "grad_norm": 0.15723877169598824,
+      "learning_rate": 6.481130037256874e-05,
+      "loss": 0.4515,
+      "step": 1353
+    },
+    {
+      "epoch": 1.7933774834437086,
+      "grad_norm": 0.13238121856363175,
+      "learning_rate": 6.478227360286129e-05,
+      "loss": 0.4585,
+      "step": 1354
+    },
+    {
+      "epoch": 1.794701986754967,
+      "grad_norm": 0.1488955180979618,
+      "learning_rate": 6.475322563735603e-05,
+      "loss": 0.4609,
+      "step": 1355
+    },
+    {
+      "epoch": 1.7960264900662253,
+      "grad_norm": 0.1541736934111582,
+      "learning_rate": 6.472415650089712e-05,
+      "loss": 0.4561,
+      "step": 1356
+    },
+    {
+      "epoch": 1.7973509933774836,
+      "grad_norm": 0.16236983791119028,
+      "learning_rate": 6.469506621834684e-05,
+      "loss": 0.4557,
+      "step": 1357
+    },
+    {
+      "epoch": 1.7986754966887417,
+      "grad_norm": 0.17161598431613118,
+      "learning_rate": 6.466595481458552e-05,
+      "loss": 0.4575,
+      "step": 1358
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.1727646253051581,
+      "learning_rate": 6.463682231451159e-05,
+      "loss": 0.4548,
+      "step": 1359
+    },
+    {
+      "epoch": 1.8013245033112582,
+      "grad_norm": 0.2019313467388488,
+      "learning_rate": 6.460766874304152e-05,
+      "loss": 0.4583,
+      "step": 1360
+    },
+    {
+      "epoch": 1.8026490066225165,
+      "grad_norm": 0.23862365061888258,
+      "learning_rate": 6.457849412510978e-05,
+      "loss": 0.4584,
+      "step": 1361
+    },
+    {
+      "epoch": 1.8039735099337748,
+      "grad_norm": 0.2558595458264096,
+      "learning_rate": 6.454929848566887e-05,
+      "loss": 0.4593,
+      "step": 1362
+    },
+    {
+      "epoch": 1.8052980132450331,
+      "grad_norm": 0.25217483491059983,
+      "learning_rate": 6.452008184968926e-05,
+      "loss": 0.4579,
+      "step": 1363
+    },
+    {
+      "epoch": 1.8066225165562915,
+      "grad_norm": 0.23004749721746393,
+      "learning_rate": 6.449084424215935e-05,
+      "loss": 0.4578,
+      "step": 1364
+    },
+    {
+      "epoch": 1.8079470198675498,
+      "grad_norm": 0.19291368394540187,
+      "learning_rate": 6.446158568808549e-05,
+      "loss": 0.4591,
+      "step": 1365
+    },
+    {
+      "epoch": 1.809271523178808,
+      "grad_norm": 0.17268581418873852,
+      "learning_rate": 6.443230621249198e-05,
+      "loss": 0.4576,
+      "step": 1366
+    },
+    {
+      "epoch": 1.8105960264900662,
+      "grad_norm": 0.16209516036184654,
+      "learning_rate": 6.440300584042097e-05,
+      "loss": 0.4563,
+      "step": 1367
+    },
+    {
+      "epoch": 1.8119205298013243,
+      "grad_norm": 0.19497648673673784,
+      "learning_rate": 6.437368459693252e-05,
+      "loss": 0.4549,
+      "step": 1368
+    },
+    {
+      "epoch": 1.8132450331125827,
+      "grad_norm": 0.17928542382691906,
+      "learning_rate": 6.434434250710451e-05,
+      "loss": 0.4583,
+      "step": 1369
+    },
+    {
+      "epoch": 1.814569536423841,
+      "grad_norm": 0.23449638037960718,
+      "learning_rate": 6.431497959603265e-05,
+      "loss": 0.4554,
+      "step": 1370
+    },
+    {
+      "epoch": 1.8158940397350993,
+      "grad_norm": 0.22793401846561953,
+      "learning_rate": 6.428559588883048e-05,
+      "loss": 0.4609,
+      "step": 1371
+    },
+    {
+      "epoch": 1.8172185430463577,
+      "grad_norm": 0.2003466738133925,
+      "learning_rate": 6.425619141062932e-05,
+      "loss": 0.4583,
+      "step": 1372
+    },
+    {
+      "epoch": 1.818543046357616,
+      "grad_norm": 0.23437410327822442,
+      "learning_rate": 6.422676618657823e-05,
+      "loss": 0.4561,
+      "step": 1373
+    },
+    {
+      "epoch": 1.8198675496688743,
+      "grad_norm": 0.2566537950545314,
+      "learning_rate": 6.419732024184407e-05,
+      "loss": 0.4569,
+      "step": 1374
+    },
+    {
+      "epoch": 1.8211920529801324,
+      "grad_norm": 0.24750385963879834,
+      "learning_rate": 6.416785360161132e-05,
+      "loss": 0.4566,
+      "step": 1375
+    },
+    {
+      "epoch": 1.8225165562913908,
+      "grad_norm": 0.19623937831507052,
+      "learning_rate": 6.413836629108228e-05,
+      "loss": 0.4604,
+      "step": 1376
+    },
+    {
+      "epoch": 1.8238410596026489,
+      "grad_norm": 0.14349582853376272,
+      "learning_rate": 6.410885833547686e-05,
+      "loss": 0.4566,
+      "step": 1377
+    },
+    {
+      "epoch": 1.8251655629139072,
+      "grad_norm": 0.15824690495748867,
+      "learning_rate": 6.407932976003266e-05,
+      "loss": 0.459,
+      "step": 1378
+    },
+    {
+      "epoch": 1.8264900662251655,
+      "grad_norm": 0.16529849570799415,
+      "learning_rate": 6.404978059000485e-05,
+      "loss": 0.4577,
+      "step": 1379
+    },
+    {
+      "epoch": 1.8278145695364238,
+      "grad_norm": 0.16075620086039508,
+      "learning_rate": 6.40202108506663e-05,
+      "loss": 0.4578,
+      "step": 1380
+    },
+    {
+      "epoch": 1.8291390728476822,
+      "grad_norm": 0.16889697572248744,
+      "learning_rate": 6.399062056730744e-05,
+      "loss": 0.4586,
+      "step": 1381
+    },
+    {
+      "epoch": 1.8304635761589405,
+      "grad_norm": 0.15939201677535858,
+      "learning_rate": 6.396100976523623e-05,
+      "loss": 0.4585,
+      "step": 1382
+    },
+    {
+      "epoch": 1.8317880794701988,
+      "grad_norm": 0.13704592759846707,
+      "learning_rate": 6.393137846977824e-05,
+      "loss": 0.4555,
+      "step": 1383
+    },
+    {
+      "epoch": 1.833112582781457,
+      "grad_norm": 0.1584399850527092,
+      "learning_rate": 6.390172670627656e-05,
+      "loss": 0.457,
+      "step": 1384
+    },
+    {
+      "epoch": 1.8344370860927153,
+      "grad_norm": 0.179875818195689,
+      "learning_rate": 6.387205450009172e-05,
+      "loss": 0.4568,
+      "step": 1385
+    },
+    {
+      "epoch": 1.8357615894039734,
+      "grad_norm": 0.17381003736142653,
+      "learning_rate": 6.384236187660182e-05,
+      "loss": 0.4541,
+      "step": 1386
+    },
+    {
+      "epoch": 1.8370860927152317,
+      "grad_norm": 0.16339124387872528,
+      "learning_rate": 6.381264886120238e-05,
+      "loss": 0.4553,
+      "step": 1387
+    },
+    {
+      "epoch": 1.83841059602649,
+      "grad_norm": 0.171198049462701,
+      "learning_rate": 6.378291547930634e-05,
+      "loss": 0.459,
+      "step": 1388
+    },
+    {
+      "epoch": 1.8397350993377484,
+      "grad_norm": 0.15397061060565295,
+      "learning_rate": 6.375316175634412e-05,
+      "loss": 0.4594,
+      "step": 1389
+    },
+    {
+      "epoch": 1.8410596026490067,
+      "grad_norm": 0.14917028686839576,
+      "learning_rate": 6.372338771776348e-05,
+      "loss": 0.4581,
+      "step": 1390
+    },
+    {
+      "epoch": 1.842384105960265,
+      "grad_norm": 0.13452688816517128,
+      "learning_rate": 6.369359338902956e-05,
+      "loss": 0.4582,
+      "step": 1391
+    },
+    {
+      "epoch": 1.8437086092715231,
+      "grad_norm": 0.14758923306275876,
+      "learning_rate": 6.366377879562493e-05,
+      "loss": 0.4581,
+      "step": 1392
+    },
+    {
+      "epoch": 1.8450331125827815,
+      "grad_norm": 0.16810960719073922,
+      "learning_rate": 6.363394396304937e-05,
+      "loss": 0.4554,
+      "step": 1393
+    },
+    {
+      "epoch": 1.8463576158940396,
+      "grad_norm": 0.16506882794936406,
+      "learning_rate": 6.360408891682008e-05,
+      "loss": 0.4573,
+      "step": 1394
+    },
+    {
+      "epoch": 1.847682119205298,
+      "grad_norm": 0.19066275725799983,
+      "learning_rate": 6.357421368247147e-05,
+      "loss": 0.4588,
+      "step": 1395
+    },
+    {
+      "epoch": 1.8490066225165562,
+      "grad_norm": 0.1928033332617369,
+      "learning_rate": 6.354431828555527e-05,
+      "loss": 0.4622,
+      "step": 1396
+    },
+    {
+      "epoch": 1.8503311258278146,
+      "grad_norm": 0.15191966034781854,
+      "learning_rate": 6.351440275164042e-05,
+      "loss": 0.4552,
+      "step": 1397
+    },
+    {
+      "epoch": 1.851655629139073,
+      "grad_norm": 0.15193256445899544,
+      "learning_rate": 6.34844671063131e-05,
+      "loss": 0.4584,
+      "step": 1398
+    },
+    {
+      "epoch": 1.8529801324503312,
+      "grad_norm": 0.14749723495592806,
+      "learning_rate": 6.345451137517668e-05,
+      "loss": 0.4596,
+      "step": 1399
+    },
+    {
+      "epoch": 1.8543046357615895,
+      "grad_norm": 0.18258400823872453,
+      "learning_rate": 6.342453558385174e-05,
+      "loss": 0.4561,
+      "step": 1400
+    },
+    {
+      "epoch": 1.8556291390728477,
+      "grad_norm": 0.1715174210221506,
+      "learning_rate": 6.339453975797595e-05,
+      "loss": 0.4571,
+      "step": 1401
+    },
+    {
+      "epoch": 1.856953642384106,
+      "grad_norm": 0.19566132464923308,
+      "learning_rate": 6.33645239232042e-05,
+      "loss": 0.4554,
+      "step": 1402
+    },
+    {
+      "epoch": 1.858278145695364,
+      "grad_norm": 0.24917708103596473,
+      "learning_rate": 6.333448810520843e-05,
+      "loss": 0.4607,
+      "step": 1403
+    },
+    {
+      "epoch": 1.8596026490066224,
+      "grad_norm": 0.2522111128504153,
+      "learning_rate": 6.33044323296777e-05,
+      "loss": 0.4561,
+      "step": 1404
+    },
+    {
+      "epoch": 1.8609271523178808,
+      "grad_norm": 0.17657870425395444,
+      "learning_rate": 6.327435662231812e-05,
+      "loss": 0.4567,
+      "step": 1405
+    },
+    {
+      "epoch": 1.862251655629139,
+      "grad_norm": 0.14232847863222456,
+      "learning_rate": 6.324426100885287e-05,
+      "loss": 0.4556,
+      "step": 1406
+    },
+    {
+      "epoch": 1.8635761589403974,
+      "grad_norm": 0.16277145402651072,
+      "learning_rate": 6.321414551502215e-05,
+      "loss": 0.4569,
+      "step": 1407
+    },
+    {
+      "epoch": 1.8649006622516557,
+      "grad_norm": 0.1467320403669246,
+      "learning_rate": 6.318401016658313e-05,
+      "loss": 0.4542,
+      "step": 1408
+    },
+    {
+      "epoch": 1.866225165562914,
+      "grad_norm": 0.14051266796917805,
+      "learning_rate": 6.315385498931e-05,
+      "loss": 0.4563,
+      "step": 1409
+    },
+    {
+      "epoch": 1.8675496688741722,
+      "grad_norm": 0.18161996885877996,
+      "learning_rate": 6.312368000899394e-05,
+      "loss": 0.4558,
+      "step": 1410
+    },
+    {
+      "epoch": 1.8688741721854305,
+      "grad_norm": 0.2066708038186849,
+      "learning_rate": 6.309348525144298e-05,
+      "loss": 0.4608,
+      "step": 1411
+    },
+    {
+      "epoch": 1.8701986754966886,
+      "grad_norm": 0.18546840396712855,
+      "learning_rate": 6.30632707424821e-05,
+      "loss": 0.4583,
+      "step": 1412
+    },
+    {
+      "epoch": 1.871523178807947,
+      "grad_norm": 0.17553796807205696,
+      "learning_rate": 6.303303650795323e-05,
+      "loss": 0.4559,
+      "step": 1413
+    },
+    {
+      "epoch": 1.8728476821192053,
+      "grad_norm": 0.15092185287015317,
+      "learning_rate": 6.300278257371509e-05,
+      "loss": 0.4579,
+      "step": 1414
+    },
+    {
+      "epoch": 1.8741721854304636,
+      "grad_norm": 0.1406567635175933,
+      "learning_rate": 6.297250896564329e-05,
+      "loss": 0.4606,
+      "step": 1415
+    },
+    {
+      "epoch": 1.875496688741722,
+      "grad_norm": 0.1907713082074752,
+      "learning_rate": 6.294221570963027e-05,
+      "loss": 0.4556,
+      "step": 1416
+    },
+    {
+      "epoch": 1.8768211920529803,
+      "grad_norm": 0.23536054396998218,
+      "learning_rate": 6.291190283158524e-05,
+      "loss": 0.4576,
+      "step": 1417
+    },
+    {
+      "epoch": 1.8781456953642384,
+      "grad_norm": 0.2402201394222149,
+      "learning_rate": 6.288157035743425e-05,
+      "loss": 0.4588,
+      "step": 1418
+    },
+    {
+      "epoch": 1.8794701986754967,
+      "grad_norm": 0.22047044802420182,
+      "learning_rate": 6.285121831312005e-05,
+      "loss": 0.4567,
+      "step": 1419
+    },
+    {
+      "epoch": 1.8807947019867548,
+      "grad_norm": 0.19967474751943765,
+      "learning_rate": 6.282084672460215e-05,
+      "loss": 0.4603,
+      "step": 1420
+    },
+    {
+      "epoch": 1.8821192052980131,
+      "grad_norm": 0.22623993396772246,
+      "learning_rate": 6.279045561785681e-05,
+      "loss": 0.4577,
+      "step": 1421
+    },
+    {
+      "epoch": 1.8834437086092715,
+      "grad_norm": 0.24181065512433514,
+      "learning_rate": 6.276004501887693e-05,
+      "loss": 0.4545,
+      "step": 1422
+    },
+    {
+      "epoch": 1.8847682119205298,
+      "grad_norm": 0.19237064276347068,
+      "learning_rate": 6.272961495367212e-05,
+      "loss": 0.4538,
+      "step": 1423
+    },
+    {
+      "epoch": 1.8860927152317881,
+      "grad_norm": 0.16812674181061435,
+      "learning_rate": 6.269916544826861e-05,
+      "loss": 0.4568,
+      "step": 1424
+    },
+    {
+      "epoch": 1.8874172185430464,
+      "grad_norm": 0.20280559037852816,
+      "learning_rate": 6.266869652870929e-05,
+      "loss": 0.458,
+      "step": 1425
+    },
+    {
+      "epoch": 1.8887417218543048,
+      "grad_norm": 0.20100849796501596,
+      "learning_rate": 6.263820822105363e-05,
+      "loss": 0.4561,
+      "step": 1426
+    },
+    {
+      "epoch": 1.8900662251655629,
+      "grad_norm": 0.18558946393106066,
+      "learning_rate": 6.260770055137767e-05,
+      "loss": 0.4545,
+      "step": 1427
+    },
+    {
+      "epoch": 1.8913907284768212,
+      "grad_norm": 0.1648670439364104,
+      "learning_rate": 6.257717354577405e-05,
+      "loss": 0.4601,
+      "step": 1428
+    },
+    {
+      "epoch": 1.8927152317880793,
+      "grad_norm": 0.16449168479302959,
+      "learning_rate": 6.254662723035189e-05,
+      "loss": 0.4574,
+      "step": 1429
+    },
+    {
+      "epoch": 1.8940397350993377,
+      "grad_norm": 0.158330099580357,
+      "learning_rate": 6.251606163123691e-05,
+      "loss": 0.4558,
+      "step": 1430
+    },
+    {
+      "epoch": 1.895364238410596,
+      "grad_norm": 0.17658298263052527,
+      "learning_rate": 6.248547677457123e-05,
+      "loss": 0.4558,
+      "step": 1431
+    },
+    {
+      "epoch": 1.8966887417218543,
+      "grad_norm": 0.153338492422261,
+      "learning_rate": 6.245487268651353e-05,
+      "loss": 0.4552,
+      "step": 1432
+    },
+    {
+      "epoch": 1.8980132450331126,
+      "grad_norm": 0.14697677266610717,
+      "learning_rate": 6.242424939323886e-05,
+      "loss": 0.4564,
+      "step": 1433
+    },
+    {
+      "epoch": 1.899337748344371,
+      "grad_norm": 0.15862457258781343,
+      "learning_rate": 6.239360692093876e-05,
+      "loss": 0.4511,
+      "step": 1434
+    },
+    {
+      "epoch": 1.9006622516556293,
+      "grad_norm": 0.13083400849565796,
+      "learning_rate": 6.236294529582108e-05,
+      "loss": 0.4596,
+      "step": 1435
+    },
+    {
+      "epoch": 1.9019867549668874,
+      "grad_norm": 0.13133537032207174,
+      "learning_rate": 6.233226454411017e-05,
+      "loss": 0.4557,
+      "step": 1436
+    },
+    {
+      "epoch": 1.9033112582781457,
+      "grad_norm": 0.15789059856382945,
+      "learning_rate": 6.230156469204666e-05,
+      "loss": 0.456,
+      "step": 1437
+    },
+    {
+      "epoch": 1.9046357615894038,
+      "grad_norm": 0.19336782327124435,
+      "learning_rate": 6.227084576588754e-05,
+      "loss": 0.4595,
+      "step": 1438
+    },
+    {
+      "epoch": 1.9059602649006622,
+      "grad_norm": 0.2022249986048075,
+      "learning_rate": 6.224010779190613e-05,
+      "loss": 0.4572,
+      "step": 1439
+    },
+    {
+      "epoch": 1.9072847682119205,
+      "grad_norm": 0.16903189456838572,
+      "learning_rate": 6.220935079639199e-05,
+      "loss": 0.4571,
+      "step": 1440
+    },
+    {
+      "epoch": 1.9086092715231788,
+      "grad_norm": 0.1480653272826143,
+      "learning_rate": 6.2178574805651e-05,
+      "loss": 0.4575,
+      "step": 1441
+    },
+    {
+      "epoch": 1.9099337748344372,
+      "grad_norm": 0.1567014407930392,
+      "learning_rate": 6.214777984600526e-05,
+      "loss": 0.4567,
+      "step": 1442
+    },
+    {
+      "epoch": 1.9112582781456955,
+      "grad_norm": 0.17297367299101593,
+      "learning_rate": 6.211696594379312e-05,
+      "loss": 0.4544,
+      "step": 1443
+    },
+    {
+      "epoch": 1.9125827814569536,
+      "grad_norm": 0.18853642810325438,
+      "learning_rate": 6.208613312536909e-05,
+      "loss": 0.4586,
+      "step": 1444
+    },
+    {
+      "epoch": 1.913907284768212,
+      "grad_norm": 0.15096583562809193,
+      "learning_rate": 6.205528141710391e-05,
+      "loss": 0.4524,
+      "step": 1445
+    },
+    {
+      "epoch": 1.91523178807947,
+      "grad_norm": 0.15357372886069198,
+      "learning_rate": 6.202441084538441e-05,
+      "loss": 0.4562,
+      "step": 1446
+    },
+    {
+      "epoch": 1.9165562913907284,
+      "grad_norm": 0.19577876465501162,
+      "learning_rate": 6.199352143661361e-05,
+      "loss": 0.4529,
+      "step": 1447
+    },
+    {
+      "epoch": 1.9178807947019867,
+      "grad_norm": 0.1792582430112527,
+      "learning_rate": 6.196261321721063e-05,
+      "loss": 0.4549,
+      "step": 1448
+    },
+    {
+      "epoch": 1.919205298013245,
+      "grad_norm": 0.17863404146922918,
+      "learning_rate": 6.193168621361065e-05,
+      "loss": 0.4585,
+      "step": 1449
+    },
+    {
+      "epoch": 1.9205298013245033,
+      "grad_norm": 0.16904576250337883,
+      "learning_rate": 6.190074045226495e-05,
+      "loss": 0.4549,
+      "step": 1450
+    },
+    {
+      "epoch": 1.9218543046357617,
+      "grad_norm": 0.17284401226257734,
+      "learning_rate": 6.186977595964085e-05,
+      "loss": 0.4536,
+      "step": 1451
+    },
+    {
+      "epoch": 1.92317880794702,
+      "grad_norm": 0.16441425857567624,
+      "learning_rate": 6.183879276222165e-05,
+      "loss": 0.4546,
+      "step": 1452
+    },
+    {
+      "epoch": 1.9245033112582781,
+      "grad_norm": 0.16547028248881768,
+      "learning_rate": 6.18077908865067e-05,
+      "loss": 0.4591,
+      "step": 1453
+    },
+    {
+      "epoch": 1.9258278145695364,
+      "grad_norm": 0.19023935463075914,
+      "learning_rate": 6.177677035901131e-05,
+      "loss": 0.4578,
+      "step": 1454
+    },
+    {
+      "epoch": 1.9271523178807946,
+      "grad_norm": 0.2007986257768923,
+      "learning_rate": 6.174573120626672e-05,
+      "loss": 0.452,
+      "step": 1455
+    },
+    {
+      "epoch": 1.9284768211920529,
+      "grad_norm": 0.15892182167264574,
+      "learning_rate": 6.171467345482011e-05,
+      "loss": 0.4559,
+      "step": 1456
+    },
+    {
+      "epoch": 1.9298013245033112,
+      "grad_norm": 0.15247275653321476,
+      "learning_rate": 6.16835971312346e-05,
+      "loss": 0.4567,
+      "step": 1457
+    },
+    {
+      "epoch": 1.9311258278145695,
+      "grad_norm": 0.15439777718540215,
+      "learning_rate": 6.165250226208912e-05,
+      "loss": 0.4578,
+      "step": 1458
+    },
+    {
+      "epoch": 1.9324503311258279,
+      "grad_norm": 0.152069260761263,
+      "learning_rate": 6.162138887397856e-05,
+      "loss": 0.4556,
+      "step": 1459
+    },
+    {
+      "epoch": 1.9337748344370862,
+      "grad_norm": 0.13902266812828998,
+      "learning_rate": 6.159025699351356e-05,
+      "loss": 0.456,
+      "step": 1460
+    },
+    {
+      "epoch": 1.9350993377483445,
+      "grad_norm": 0.1669947383992905,
+      "learning_rate": 6.155910664732061e-05,
+      "loss": 0.4588,
+      "step": 1461
+    },
+    {
+      "epoch": 1.9364238410596026,
+      "grad_norm": 0.16551916716197446,
+      "learning_rate": 6.152793786204204e-05,
+      "loss": 0.4529,
+      "step": 1462
+    },
+    {
+      "epoch": 1.937748344370861,
+      "grad_norm": 0.14038450440472772,
+      "learning_rate": 6.149675066433586e-05,
+      "loss": 0.4567,
+      "step": 1463
+    },
+    {
+      "epoch": 1.939072847682119,
+      "grad_norm": 0.1528043769076294,
+      "learning_rate": 6.146554508087592e-05,
+      "loss": 0.4568,
+      "step": 1464
+    },
+    {
+      "epoch": 1.9403973509933774,
+      "grad_norm": 0.17693022887620993,
+      "learning_rate": 6.143432113835172e-05,
+      "loss": 0.4563,
+      "step": 1465
+    },
+    {
+      "epoch": 1.9417218543046357,
+      "grad_norm": 0.19672591309939141,
+      "learning_rate": 6.140307886346848e-05,
+      "loss": 0.4583,
+      "step": 1466
+    },
+    {
+      "epoch": 1.943046357615894,
+      "grad_norm": 0.1978271667862739,
+      "learning_rate": 6.137181828294714e-05,
+      "loss": 0.4571,
+      "step": 1467
+    },
+    {
+      "epoch": 1.9443708609271524,
+      "grad_norm": 0.17715644765400318,
+      "learning_rate": 6.134053942352426e-05,
+      "loss": 0.4549,
+      "step": 1468
+    },
+    {
+      "epoch": 1.9456953642384107,
+      "grad_norm": 0.15391178923036633,
+      "learning_rate": 6.130924231195203e-05,
+      "loss": 0.4583,
+      "step": 1469
+    },
+    {
+      "epoch": 1.9470198675496688,
+      "grad_norm": 0.15492097853935494,
+      "learning_rate": 6.127792697499828e-05,
+      "loss": 0.4558,
+      "step": 1470
+    },
+    {
+      "epoch": 1.9483443708609272,
+      "grad_norm": 0.15322792918233222,
+      "learning_rate": 6.12465934394464e-05,
+      "loss": 0.4542,
+      "step": 1471
+    },
+    {
+      "epoch": 1.9496688741721855,
+      "grad_norm": 0.1668354463850271,
+      "learning_rate": 6.121524173209535e-05,
+      "loss": 0.4582,
+      "step": 1472
+    },
+    {
+      "epoch": 1.9509933774834436,
+      "grad_norm": 0.19809487519031535,
+      "learning_rate": 6.118387187975965e-05,
+      "loss": 0.4522,
+      "step": 1473
+    },
+    {
+      "epoch": 1.952317880794702,
+      "grad_norm": 0.184288333718565,
+      "learning_rate": 6.115248390926932e-05,
+      "loss": 0.4547,
+      "step": 1474
+    },
+    {
+      "epoch": 1.9536423841059603,
+      "grad_norm": 0.16357263498301558,
+      "learning_rate": 6.112107784746988e-05,
+      "loss": 0.4581,
+      "step": 1475
+    },
+    {
+      "epoch": 1.9549668874172186,
+      "grad_norm": 0.1882029050506093,
+      "learning_rate": 6.108965372122232e-05,
+      "loss": 0.4548,
+      "step": 1476
+    },
+    {
+      "epoch": 1.956291390728477,
+      "grad_norm": 0.1993506555999991,
+      "learning_rate": 6.10582115574031e-05,
+      "loss": 0.458,
+      "step": 1477
+    },
+    {
+      "epoch": 1.9576158940397352,
+      "grad_norm": 0.1854804277191762,
+      "learning_rate": 6.1026751382904086e-05,
+      "loss": 0.4528,
+      "step": 1478
+    },
+    {
+      "epoch": 1.9589403973509933,
+      "grad_norm": 0.20283866627895117,
+      "learning_rate": 6.0995273224632556e-05,
+      "loss": 0.455,
+      "step": 1479
+    },
+    {
+      "epoch": 1.9602649006622517,
+      "grad_norm": 0.22358222335503145,
+      "learning_rate": 6.0963777109511166e-05,
+      "loss": 0.458,
+      "step": 1480
+    },
+    {
+      "epoch": 1.9615894039735098,
+      "grad_norm": 0.202967931077154,
+      "learning_rate": 6.0932263064477935e-05,
+      "loss": 0.4561,
+      "step": 1481
+    },
+    {
+      "epoch": 1.9629139072847681,
+      "grad_norm": 0.19910942607415777,
+      "learning_rate": 6.09007311164862e-05,
+      "loss": 0.4528,
+      "step": 1482
+    },
+    {
+      "epoch": 1.9642384105960264,
+      "grad_norm": 0.17857325763492593,
+      "learning_rate": 6.0869181292504646e-05,
+      "loss": 0.4564,
+      "step": 1483
+    },
+    {
+      "epoch": 1.9655629139072848,
+      "grad_norm": 0.16864330345474904,
+      "learning_rate": 6.083761361951722e-05,
+      "loss": 0.4595,
+      "step": 1484
+    },
+    {
+      "epoch": 1.966887417218543,
+      "grad_norm": 0.1644936147604867,
+      "learning_rate": 6.080602812452313e-05,
+      "loss": 0.4573,
+      "step": 1485
+    },
+    {
+      "epoch": 1.9682119205298014,
+      "grad_norm": 0.18241181921492491,
+      "learning_rate": 6.0774424834536844e-05,
+      "loss": 0.4566,
+      "step": 1486
+    },
+    {
+      "epoch": 1.9695364238410598,
+      "grad_norm": 0.22951225086819727,
+      "learning_rate": 6.074280377658804e-05,
+      "loss": 0.4529,
+      "step": 1487
+    },
+    {
+      "epoch": 1.9708609271523179,
+      "grad_norm": 0.22476026901710608,
+      "learning_rate": 6.07111649777216e-05,
+      "loss": 0.454,
+      "step": 1488
+    },
+    {
+      "epoch": 1.9721854304635762,
+      "grad_norm": 0.16144556114966288,
+      "learning_rate": 6.067950846499756e-05,
+      "loss": 0.4598,
+      "step": 1489
+    },
+    {
+      "epoch": 1.9735099337748343,
+      "grad_norm": 0.14103941762909064,
+      "learning_rate": 6.064783426549115e-05,
+      "loss": 0.4549,
+      "step": 1490
+    },
+    {
+      "epoch": 1.9748344370860926,
+      "grad_norm": 0.1928550721246238,
+      "learning_rate": 6.0616142406292674e-05,
+      "loss": 0.455,
+      "step": 1491
+    },
+    {
+      "epoch": 1.976158940397351,
+      "grad_norm": 0.19144476701344726,
+      "learning_rate": 6.058443291450757e-05,
+      "loss": 0.4592,
+      "step": 1492
+    },
+    {
+      "epoch": 1.9774834437086093,
+      "grad_norm": 0.15529465228721923,
+      "learning_rate": 6.055270581725635e-05,
+      "loss": 0.4571,
+      "step": 1493
+    },
+    {
+      "epoch": 1.9788079470198676,
+      "grad_norm": 0.14196274008209844,
+      "learning_rate": 6.052096114167461e-05,
+      "loss": 0.4557,
+      "step": 1494
+    },
+    {
+      "epoch": 1.980132450331126,
+      "grad_norm": 0.15085179175853095,
+      "learning_rate": 6.0489198914912905e-05,
+      "loss": 0.4573,
+      "step": 1495
+    },
+    {
+      "epoch": 1.981456953642384,
+      "grad_norm": 0.17333623959994102,
+      "learning_rate": 6.04574191641369e-05,
+      "loss": 0.4553,
+      "step": 1496
+    },
+    {
+      "epoch": 1.9827814569536424,
+      "grad_norm": 0.14405001603526288,
+      "learning_rate": 6.0425621916527166e-05,
+      "loss": 0.4578,
+      "step": 1497
+    },
+    {
+      "epoch": 1.9841059602649007,
+      "grad_norm": 0.16799971036277606,
+      "learning_rate": 6.039380719927929e-05,
+      "loss": 0.4571,
+      "step": 1498
+    },
+    {
+      "epoch": 1.9854304635761588,
+      "grad_norm": 0.20584651378470314,
+      "learning_rate": 6.0361975039603765e-05,
+      "loss": 0.4546,
+      "step": 1499
+    },
+    {
+      "epoch": 1.9867549668874172,
+      "grad_norm": 0.18318399637690272,
+      "learning_rate": 6.033012546472604e-05,
+      "loss": 0.4551,
+      "step": 1500
+    },
+    {
+      "epoch": 1.9880794701986755,
+      "grad_norm": 0.20463070913153947,
+      "learning_rate": 6.029825850188644e-05,
+      "loss": 0.4561,
+      "step": 1501
+    },
+    {
+      "epoch": 1.9894039735099338,
+      "grad_norm": 0.2171380371331036,
+      "learning_rate": 6.026637417834013e-05,
+      "loss": 0.4591,
+      "step": 1502
+    },
+    {
+      "epoch": 1.9907284768211921,
+      "grad_norm": 0.19963283583468963,
+      "learning_rate": 6.023447252135718e-05,
+      "loss": 0.4556,
+      "step": 1503
+    },
+    {
+      "epoch": 1.9920529801324505,
+      "grad_norm": 0.1686910161440413,
+      "learning_rate": 6.0202553558222464e-05,
+      "loss": 0.4581,
+      "step": 1504
+    },
+    {
+      "epoch": 1.9933774834437086,
+      "grad_norm": 0.16433810189055548,
+      "learning_rate": 6.0170617316235634e-05,
+      "loss": 0.4517,
+      "step": 1505
+    },
+    {
+      "epoch": 1.994701986754967,
+      "grad_norm": 0.16459455164773862,
+      "learning_rate": 6.013866382271114e-05,
+      "loss": 0.4525,
+      "step": 1506
+    },
+    {
+      "epoch": 1.996026490066225,
+      "grad_norm": 0.228012064379968,
+      "learning_rate": 6.010669310497819e-05,
+      "loss": 0.4576,
+      "step": 1507
+    },
+    {
+      "epoch": 1.9973509933774833,
+      "grad_norm": 0.2432807121991456,
+      "learning_rate": 6.007470519038072e-05,
+      "loss": 0.4548,
+      "step": 1508
+    },
+    {
+      "epoch": 1.9986754966887417,
+      "grad_norm": 0.22733922692214092,
+      "learning_rate": 6.004270010627735e-05,
+      "loss": 0.4523,
+      "step": 1509
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.24305600360340535,
+      "learning_rate": 6.001067788004146e-05,
+      "loss": 0.4554,
+      "step": 1510
+    },
+    {
+      "epoch": 2.0013245033112583,
+      "grad_norm": 0.2889200550360593,
+      "learning_rate": 5.997863853906099e-05,
+      "loss": 0.4357,
+      "step": 1511
+    },
+    {
+      "epoch": 2.0026490066225167,
+      "grad_norm": 0.49062682374644606,
+      "learning_rate": 5.9946582110738584e-05,
+      "loss": 0.4402,
+      "step": 1512
+    },
+    {
+      "epoch": 2.003973509933775,
+      "grad_norm": 0.31809066821904297,
+      "learning_rate": 5.9914508622491485e-05,
+      "loss": 0.4359,
+      "step": 1513
+    },
+    {
+      "epoch": 2.0052980132450333,
+      "grad_norm": 0.28536509290255063,
+      "learning_rate": 5.988241810175152e-05,
+      "loss": 0.4381,
+      "step": 1514
+    },
+    {
+      "epoch": 2.006622516556291,
+      "grad_norm": 0.3794438807684635,
+      "learning_rate": 5.98503105759651e-05,
+      "loss": 0.4428,
+      "step": 1515
+    },
+    {
+      "epoch": 2.0079470198675495,
+      "grad_norm": 0.4795364360293247,
+      "learning_rate": 5.981818607259315e-05,
+      "loss": 0.4407,
+      "step": 1516
+    },
+    {
+      "epoch": 2.009271523178808,
+      "grad_norm": 0.547255546659395,
+      "learning_rate": 5.978604461911115e-05,
+      "loss": 0.4419,
+      "step": 1517
+    },
+    {
+      "epoch": 2.010596026490066,
+      "grad_norm": 0.6494246687726558,
+      "learning_rate": 5.9753886243009056e-05,
+      "loss": 0.4454,
+      "step": 1518
+    },
+    {
+      "epoch": 2.0119205298013245,
+      "grad_norm": 0.6729533544109833,
+      "learning_rate": 5.9721710971791293e-05,
+      "loss": 0.4429,
+      "step": 1519
+    },
+    {
+      "epoch": 2.013245033112583,
+      "grad_norm": 0.5138673242410033,
+      "learning_rate": 5.968951883297677e-05,
+      "loss": 0.4405,
+      "step": 1520
+    },
+    {
+      "epoch": 2.014569536423841,
+      "grad_norm": 0.4514334386691532,
+      "learning_rate": 5.965730985409878e-05,
+      "loss": 0.4429,
+      "step": 1521
+    },
+    {
+      "epoch": 2.0158940397350995,
+      "grad_norm": 0.5401925059353765,
+      "learning_rate": 5.962508406270505e-05,
+      "loss": 0.438,
+      "step": 1522
+    },
+    {
+      "epoch": 2.017218543046358,
+      "grad_norm": 0.4662513518440839,
+      "learning_rate": 5.959284148635766e-05,
+      "loss": 0.4396,
+      "step": 1523
+    },
+    {
+      "epoch": 2.0185430463576157,
+      "grad_norm": 0.5053798147032519,
+      "learning_rate": 5.956058215263307e-05,
+      "loss": 0.4411,
+      "step": 1524
+    },
+    {
+      "epoch": 2.019867549668874,
+      "grad_norm": 0.6424306111951932,
+      "learning_rate": 5.952830608912206e-05,
+      "loss": 0.4418,
+      "step": 1525
+    },
+    {
+      "epoch": 2.0211920529801324,
+      "grad_norm": 0.5584467814359974,
+      "learning_rate": 5.949601332342973e-05,
+      "loss": 0.4378,
+      "step": 1526
+    },
+    {
+      "epoch": 2.0225165562913907,
+      "grad_norm": 0.31659588502002556,
+      "learning_rate": 5.946370388317543e-05,
+      "loss": 0.4382,
+      "step": 1527
+    },
+    {
+      "epoch": 2.023841059602649,
+      "grad_norm": 0.37337764455424305,
+      "learning_rate": 5.943137779599283e-05,
+      "loss": 0.4369,
+      "step": 1528
+    },
+    {
+      "epoch": 2.0251655629139074,
+      "grad_norm": 0.42414692857815384,
+      "learning_rate": 5.939903508952977e-05,
+      "loss": 0.4372,
+      "step": 1529
+    },
+    {
+      "epoch": 2.0264900662251657,
+      "grad_norm": 0.29628885881246586,
+      "learning_rate": 5.9366675791448375e-05,
+      "loss": 0.4372,
+      "step": 1530
+    },
+    {
+      "epoch": 2.027814569536424,
+      "grad_norm": 0.23047596593457328,
+      "learning_rate": 5.93342999294249e-05,
+      "loss": 0.4395,
+      "step": 1531
+    },
+    {
+      "epoch": 2.029139072847682,
+      "grad_norm": 0.26222380772738524,
+      "learning_rate": 5.9301907531149805e-05,
+      "loss": 0.4338,
+      "step": 1532
+    },
+    {
+      "epoch": 2.0304635761589402,
+      "grad_norm": 0.2420502141203094,
+      "learning_rate": 5.926949862432767e-05,
+      "loss": 0.4396,
+      "step": 1533
+    },
+    {
+      "epoch": 2.0317880794701986,
+      "grad_norm": 0.215176588694931,
+      "learning_rate": 5.9237073236677216e-05,
+      "loss": 0.4379,
+      "step": 1534
+    },
+    {
+      "epoch": 2.033112582781457,
+      "grad_norm": 0.18999325689850474,
+      "learning_rate": 5.9204631395931244e-05,
+      "loss": 0.4355,
+      "step": 1535
+    },
+    {
+      "epoch": 2.0344370860927152,
+      "grad_norm": 0.19777500909916865,
+      "learning_rate": 5.9172173129836614e-05,
+      "loss": 0.4375,
+      "step": 1536
+    },
+    {
+      "epoch": 2.0357615894039736,
+      "grad_norm": 0.21549275815483226,
+      "learning_rate": 5.913969846615429e-05,
+      "loss": 0.4337,
+      "step": 1537
+    },
+    {
+      "epoch": 2.037086092715232,
+      "grad_norm": 0.19885930008442662,
+      "learning_rate": 5.910720743265918e-05,
+      "loss": 0.4351,
+      "step": 1538
+    },
+    {
+      "epoch": 2.03841059602649,
+      "grad_norm": 0.18606620413354488,
+      "learning_rate": 5.907470005714025e-05,
+      "loss": 0.4365,
+      "step": 1539
+    },
+    {
+      "epoch": 2.0397350993377485,
+      "grad_norm": 0.1989940721518127,
+      "learning_rate": 5.9042176367400435e-05,
+      "loss": 0.4349,
+      "step": 1540
+    },
+    {
+      "epoch": 2.0410596026490064,
+      "grad_norm": 0.19831453425765527,
+      "learning_rate": 5.9009636391256604e-05,
+      "loss": 0.4402,
+      "step": 1541
+    },
+    {
+      "epoch": 2.0423841059602648,
+      "grad_norm": 0.1798435644698872,
+      "learning_rate": 5.897708015653958e-05,
+      "loss": 0.4366,
+      "step": 1542
+    },
+    {
+      "epoch": 2.043708609271523,
+      "grad_norm": 0.16100159642279163,
+      "learning_rate": 5.894450769109407e-05,
+      "loss": 0.438,
+      "step": 1543
+    },
+    {
+      "epoch": 2.0450331125827814,
+      "grad_norm": 0.15953490094442624,
+      "learning_rate": 5.8911919022778674e-05,
+      "loss": 0.4348,
+      "step": 1544
+    },
+    {
+      "epoch": 2.0463576158940397,
+      "grad_norm": 0.1896222412980382,
+      "learning_rate": 5.8879314179465847e-05,
+      "loss": 0.4348,
+      "step": 1545
+    },
+    {
+      "epoch": 2.047682119205298,
+      "grad_norm": 0.1562650607570057,
+      "learning_rate": 5.8846693189041885e-05,
+      "loss": 0.4331,
+      "step": 1546
+    },
+    {
+      "epoch": 2.0490066225165564,
+      "grad_norm": 0.18918127747760027,
+      "learning_rate": 5.881405607940688e-05,
+      "loss": 0.4332,
+      "step": 1547
+    },
+    {
+      "epoch": 2.0503311258278147,
+      "grad_norm": 0.14908337969597085,
+      "learning_rate": 5.878140287847473e-05,
+      "loss": 0.4344,
+      "step": 1548
+    },
+    {
+      "epoch": 2.0516556291390726,
+      "grad_norm": 0.14313247559412812,
+      "learning_rate": 5.8748733614173066e-05,
+      "loss": 0.4338,
+      "step": 1549
+    },
+    {
+      "epoch": 2.052980132450331,
+      "grad_norm": 0.16280634381754575,
+      "learning_rate": 5.8716048314443285e-05,
+      "loss": 0.4346,
+      "step": 1550
+    },
+    {
+      "epoch": 2.0543046357615893,
+      "grad_norm": 0.14014742153537318,
+      "learning_rate": 5.86833470072405e-05,
+      "loss": 0.433,
+      "step": 1551
+    },
+    {
+      "epoch": 2.0556291390728476,
+      "grad_norm": 0.1320214140153468,
+      "learning_rate": 5.8650629720533486e-05,
+      "loss": 0.4374,
+      "step": 1552
+    },
+    {
+      "epoch": 2.056953642384106,
+      "grad_norm": 0.13296992930366294,
+      "learning_rate": 5.861789648230472e-05,
+      "loss": 0.4335,
+      "step": 1553
+    },
+    {
+      "epoch": 2.0582781456953643,
+      "grad_norm": 0.13905642856417919,
+      "learning_rate": 5.858514732055029e-05,
+      "loss": 0.4332,
+      "step": 1554
+    },
+    {
+      "epoch": 2.0596026490066226,
+      "grad_norm": 0.11069116515758654,
+      "learning_rate": 5.8552382263279933e-05,
+      "loss": 0.4365,
+      "step": 1555
+    },
+    {
+      "epoch": 2.060927152317881,
+      "grad_norm": 0.14587231074494586,
+      "learning_rate": 5.8519601338516965e-05,
+      "loss": 0.4363,
+      "step": 1556
+    },
+    {
+      "epoch": 2.0622516556291393,
+      "grad_norm": 0.1356961158197318,
+      "learning_rate": 5.8486804574298274e-05,
+      "loss": 0.4307,
+      "step": 1557
+    },
+    {
+      "epoch": 2.063576158940397,
+      "grad_norm": 0.13755915416505474,
+      "learning_rate": 5.845399199867428e-05,
+      "loss": 0.4316,
+      "step": 1558
+    },
+    {
+      "epoch": 2.0649006622516555,
+      "grad_norm": 0.12142098349855084,
+      "learning_rate": 5.8421163639708976e-05,
+      "loss": 0.4322,
+      "step": 1559
+    },
+    {
+      "epoch": 2.066225165562914,
+      "grad_norm": 0.11778999743229701,
+      "learning_rate": 5.8388319525479776e-05,
+      "loss": 0.4332,
+      "step": 1560
+    },
+    {
+      "epoch": 2.067549668874172,
+      "grad_norm": 0.15936411011949986,
+      "learning_rate": 5.835545968407765e-05,
+      "loss": 0.4324,
+      "step": 1561
+    },
+    {
+      "epoch": 2.0688741721854305,
+      "grad_norm": 0.13116272068290705,
+      "learning_rate": 5.832258414360697e-05,
+      "loss": 0.4317,
+      "step": 1562
+    },
+    {
+      "epoch": 2.070198675496689,
+      "grad_norm": 0.13020475656367664,
+      "learning_rate": 5.8289692932185546e-05,
+      "loss": 0.4376,
+      "step": 1563
+    },
+    {
+      "epoch": 2.071523178807947,
+      "grad_norm": 0.14369793752508384,
+      "learning_rate": 5.82567860779446e-05,
+      "loss": 0.435,
+      "step": 1564
+    },
+    {
+      "epoch": 2.0728476821192054,
+      "grad_norm": 0.10893261081833851,
+      "learning_rate": 5.822386360902871e-05,
+      "loss": 0.4352,
+      "step": 1565
+    },
+    {
+      "epoch": 2.0741721854304638,
+      "grad_norm": 0.12259518906345106,
+      "learning_rate": 5.819092555359585e-05,
+      "loss": 0.4332,
+      "step": 1566
+    },
+    {
+      "epoch": 2.0754966887417217,
+      "grad_norm": 0.1479682231584038,
+      "learning_rate": 5.815797193981729e-05,
+      "loss": 0.4352,
+      "step": 1567
+    },
+    {
+      "epoch": 2.07682119205298,
+      "grad_norm": 0.1285501212173335,
+      "learning_rate": 5.8125002795877594e-05,
+      "loss": 0.4349,
+      "step": 1568
+    },
+    {
+      "epoch": 2.0781456953642383,
+      "grad_norm": 0.11204041036917814,
+      "learning_rate": 5.809201814997467e-05,
+      "loss": 0.4356,
+      "step": 1569
+    },
+    {
+      "epoch": 2.0794701986754967,
+      "grad_norm": 0.13787197403459772,
+      "learning_rate": 5.805901803031961e-05,
+      "loss": 0.4342,
+      "step": 1570
+    },
+    {
+      "epoch": 2.080794701986755,
+      "grad_norm": 0.12543009024937735,
+      "learning_rate": 5.80260024651368e-05,
+      "loss": 0.4323,
+      "step": 1571
+    },
+    {
+      "epoch": 2.0821192052980133,
+      "grad_norm": 0.11582030180851238,
+      "learning_rate": 5.79929714826638e-05,
+      "loss": 0.4339,
+      "step": 1572
+    },
+    {
+      "epoch": 2.0834437086092716,
+      "grad_norm": 0.12664064105322856,
+      "learning_rate": 5.7959925111151386e-05,
+      "loss": 0.433,
+      "step": 1573
+    },
+    {
+      "epoch": 2.08476821192053,
+      "grad_norm": 0.09432450937421345,
+      "learning_rate": 5.792686337886344e-05,
+      "loss": 0.4325,
+      "step": 1574
+    },
+    {
+      "epoch": 2.0860927152317883,
+      "grad_norm": 0.11704468268796787,
+      "learning_rate": 5.7893786314077065e-05,
+      "loss": 0.4356,
+      "step": 1575
+    },
+    {
+      "epoch": 2.087417218543046,
+      "grad_norm": 0.13567171186370938,
+      "learning_rate": 5.786069394508241e-05,
+      "loss": 0.4349,
+      "step": 1576
+    },
+    {
+      "epoch": 2.0887417218543045,
+      "grad_norm": 0.1457839042401526,
+      "learning_rate": 5.782758630018274e-05,
+      "loss": 0.4356,
+      "step": 1577
+    },
+    {
+      "epoch": 2.090066225165563,
+      "grad_norm": 0.16522855194591107,
+      "learning_rate": 5.7794463407694375e-05,
+      "loss": 0.435,
+      "step": 1578
+    },
+    {
+      "epoch": 2.091390728476821,
+      "grad_norm": 0.16413562956081745,
+      "learning_rate": 5.77613252959467e-05,
+      "loss": 0.4336,
+      "step": 1579
+    },
+    {
+      "epoch": 2.0927152317880795,
+      "grad_norm": 0.12417364587569958,
+      "learning_rate": 5.7728171993282064e-05,
+      "loss": 0.4342,
+      "step": 1580
+    },
+    {
+      "epoch": 2.094039735099338,
+      "grad_norm": 0.14104793819136102,
+      "learning_rate": 5.7695003528055874e-05,
+      "loss": 0.4342,
+      "step": 1581
+    },
+    {
+      "epoch": 2.095364238410596,
+      "grad_norm": 0.21461526262953803,
+      "learning_rate": 5.7661819928636484e-05,
+      "loss": 0.435,
+      "step": 1582
+    },
+    {
+      "epoch": 2.0966887417218545,
+      "grad_norm": 0.2355366470850295,
+      "learning_rate": 5.762862122340515e-05,
+      "loss": 0.4331,
+      "step": 1583
+    },
+    {
+      "epoch": 2.0980132450331124,
+      "grad_norm": 0.1906914391786148,
+      "learning_rate": 5.75954074407561e-05,
+      "loss": 0.4356,
+      "step": 1584
+    },
+    {
+      "epoch": 2.0993377483443707,
+      "grad_norm": 0.15090563365666954,
+      "learning_rate": 5.756217860909644e-05,
+      "loss": 0.4362,
+      "step": 1585
+    },
+    {
+      "epoch": 2.100662251655629,
+      "grad_norm": 0.12434689713663916,
+      "learning_rate": 5.7528934756846134e-05,
+      "loss": 0.4345,
+      "step": 1586
+    },
+    {
+      "epoch": 2.1019867549668874,
+      "grad_norm": 0.10868743017717654,
+      "learning_rate": 5.749567591243802e-05,
+      "loss": 0.4344,
+      "step": 1587
+    },
+    {
+      "epoch": 2.1033112582781457,
+      "grad_norm": 0.12552950812464744,
+      "learning_rate": 5.7462402104317724e-05,
+      "loss": 0.4343,
+      "step": 1588
+    },
+    {
+      "epoch": 2.104635761589404,
+      "grad_norm": 0.1494775814765201,
+      "learning_rate": 5.7429113360943694e-05,
+      "loss": 0.4376,
+      "step": 1589
+    },
+    {
+      "epoch": 2.1059602649006623,
+      "grad_norm": 0.13200828840931994,
+      "learning_rate": 5.739580971078715e-05,
+      "loss": 0.4367,
+      "step": 1590
+    },
+    {
+      "epoch": 2.1072847682119207,
+      "grad_norm": 0.11307299592530613,
+      "learning_rate": 5.736249118233205e-05,
+      "loss": 0.4347,
+      "step": 1591
+    },
+    {
+      "epoch": 2.108609271523179,
+      "grad_norm": 0.12353253717561186,
+      "learning_rate": 5.73291578040751e-05,
+      "loss": 0.4326,
+      "step": 1592
+    },
+    {
+      "epoch": 2.109933774834437,
+      "grad_norm": 0.1521741113472388,
+      "learning_rate": 5.729580960452568e-05,
+      "loss": 0.4325,
+      "step": 1593
+    },
+    {
+      "epoch": 2.111258278145695,
+      "grad_norm": 0.160983111426557,
+      "learning_rate": 5.726244661220588e-05,
+      "loss": 0.4356,
+      "step": 1594
+    },
+    {
+      "epoch": 2.1125827814569536,
+      "grad_norm": 0.1283840329557453,
+      "learning_rate": 5.722906885565037e-05,
+      "loss": 0.4318,
+      "step": 1595
+    },
+    {
+      "epoch": 2.113907284768212,
+      "grad_norm": 0.11524622238594787,
+      "learning_rate": 5.719567636340655e-05,
+      "loss": 0.4309,
+      "step": 1596
+    },
+    {
+      "epoch": 2.11523178807947,
+      "grad_norm": 0.15608870241866474,
+      "learning_rate": 5.716226916403434e-05,
+      "loss": 0.4366,
+      "step": 1597
+    },
+    {
+      "epoch": 2.1165562913907285,
+      "grad_norm": 0.13085758411991147,
+      "learning_rate": 5.712884728610628e-05,
+      "loss": 0.4363,
+      "step": 1598
+    },
+    {
+      "epoch": 2.117880794701987,
+      "grad_norm": 0.11005315647772583,
+      "learning_rate": 5.7095410758207465e-05,
+      "loss": 0.4358,
+      "step": 1599
+    },
+    {
+      "epoch": 2.119205298013245,
+      "grad_norm": 0.11726173837348314,
+      "learning_rate": 5.706195960893549e-05,
+      "loss": 0.4341,
+      "step": 1600
+    },
+    {
+      "epoch": 2.120529801324503,
+      "grad_norm": 0.1310011976930671,
+      "learning_rate": 5.702849386690047e-05,
+      "loss": 0.431,
+      "step": 1601
+    },
+    {
+      "epoch": 2.1218543046357614,
+      "grad_norm": 0.12249139420537569,
+      "learning_rate": 5.6995013560725024e-05,
+      "loss": 0.4329,
+      "step": 1602
+    },
+    {
+      "epoch": 2.1231788079470197,
+      "grad_norm": 0.12477865849500473,
+      "learning_rate": 5.6961518719044205e-05,
+      "loss": 0.4372,
+      "step": 1603
+    },
+    {
+      "epoch": 2.124503311258278,
+      "grad_norm": 0.13102809531709494,
+      "learning_rate": 5.6928009370505485e-05,
+      "loss": 0.433,
+      "step": 1604
+    },
+    {
+      "epoch": 2.1258278145695364,
+      "grad_norm": 0.1448313895957034,
+      "learning_rate": 5.689448554376878e-05,
+      "loss": 0.435,
+      "step": 1605
+    },
+    {
+      "epoch": 2.1271523178807947,
+      "grad_norm": 0.14536118572238488,
+      "learning_rate": 5.686094726750636e-05,
+      "loss": 0.4343,
+      "step": 1606
+    },
+    {
+      "epoch": 2.128476821192053,
+      "grad_norm": 0.14207194346114713,
+      "learning_rate": 5.682739457040286e-05,
+      "loss": 0.4367,
+      "step": 1607
+    },
+    {
+      "epoch": 2.1298013245033114,
+      "grad_norm": 0.17596798589311444,
+      "learning_rate": 5.679382748115525e-05,
+      "loss": 0.4368,
+      "step": 1608
+    },
+    {
+      "epoch": 2.1311258278145697,
+      "grad_norm": 0.16108740847246547,
+      "learning_rate": 5.676024602847281e-05,
+      "loss": 0.4313,
+      "step": 1609
+    },
+    {
+      "epoch": 2.1324503311258276,
+      "grad_norm": 0.11390281779620849,
+      "learning_rate": 5.6726650241077113e-05,
+      "loss": 0.4326,
+      "step": 1610
+    },
+    {
+      "epoch": 2.133774834437086,
+      "grad_norm": 0.15470280946980208,
+      "learning_rate": 5.669304014770198e-05,
+      "loss": 0.4339,
+      "step": 1611
+    },
+    {
+      "epoch": 2.1350993377483443,
+      "grad_norm": 0.12705971943663616,
+      "learning_rate": 5.665941577709346e-05,
+      "loss": 0.4349,
+      "step": 1612
+    },
+    {
+      "epoch": 2.1364238410596026,
+      "grad_norm": 0.11099793080261494,
+      "learning_rate": 5.662577715800984e-05,
+      "loss": 0.4339,
+      "step": 1613
+    },
+    {
+      "epoch": 2.137748344370861,
+      "grad_norm": 0.11192932847446964,
+      "learning_rate": 5.659212431922158e-05,
+      "loss": 0.4342,
+      "step": 1614
+    },
+    {
+      "epoch": 2.1390728476821192,
+      "grad_norm": 0.14495007943311636,
+      "learning_rate": 5.655845728951129e-05,
+      "loss": 0.431,
+      "step": 1615
+    },
+    {
+      "epoch": 2.1403973509933776,
+      "grad_norm": 0.1620179076586833,
+      "learning_rate": 5.6524776097673726e-05,
+      "loss": 0.4314,
+      "step": 1616
+    },
+    {
+      "epoch": 2.141721854304636,
+      "grad_norm": 0.13487365712351376,
+      "learning_rate": 5.649108077251577e-05,
+      "loss": 0.4339,
+      "step": 1617
+    },
+    {
+      "epoch": 2.1430463576158942,
+      "grad_norm": 0.12636894202224241,
+      "learning_rate": 5.6457371342856365e-05,
+      "loss": 0.4326,
+      "step": 1618
+    },
+    {
+      "epoch": 2.144370860927152,
+      "grad_norm": 0.12671608219335437,
+      "learning_rate": 5.642364783752655e-05,
+      "loss": 0.4325,
+      "step": 1619
+    },
+    {
+      "epoch": 2.1456953642384105,
+      "grad_norm": 0.11620356457225499,
+      "learning_rate": 5.638991028536938e-05,
+      "loss": 0.4311,
+      "step": 1620
+    },
+    {
+      "epoch": 2.147019867549669,
+      "grad_norm": 0.12104568697994726,
+      "learning_rate": 5.6356158715239905e-05,
+      "loss": 0.4324,
+      "step": 1621
+    },
+    {
+      "epoch": 2.148344370860927,
+      "grad_norm": 0.11675168882792084,
+      "learning_rate": 5.6322393156005216e-05,
+      "loss": 0.4342,
+      "step": 1622
+    },
+    {
+      "epoch": 2.1496688741721854,
+      "grad_norm": 0.12715667339976872,
+      "learning_rate": 5.628861363654433e-05,
+      "loss": 0.4311,
+      "step": 1623
+    },
+    {
+      "epoch": 2.1509933774834438,
+      "grad_norm": 0.11189713316580231,
+      "learning_rate": 5.625482018574822e-05,
+      "loss": 0.4367,
+      "step": 1624
+    },
+    {
+      "epoch": 2.152317880794702,
+      "grad_norm": 0.1296301348583053,
+      "learning_rate": 5.622101283251976e-05,
+      "loss": 0.4337,
+      "step": 1625
+    },
+    {
+      "epoch": 2.1536423841059604,
+      "grad_norm": 0.14356003207609858,
+      "learning_rate": 5.618719160577373e-05,
+      "loss": 0.4326,
+      "step": 1626
+    },
+    {
+      "epoch": 2.1549668874172188,
+      "grad_norm": 0.14369738363617432,
+      "learning_rate": 5.615335653443676e-05,
+      "loss": 0.4342,
+      "step": 1627
+    },
+    {
+      "epoch": 2.1562913907284766,
+      "grad_norm": 0.14042107107447105,
+      "learning_rate": 5.6119507647447356e-05,
+      "loss": 0.4347,
+      "step": 1628
+    },
+    {
+      "epoch": 2.157615894039735,
+      "grad_norm": 0.12928991754102834,
+      "learning_rate": 5.608564497375579e-05,
+      "loss": 0.4356,
+      "step": 1629
+    },
+    {
+      "epoch": 2.1589403973509933,
+      "grad_norm": 0.14991129495074407,
+      "learning_rate": 5.605176854232415e-05,
+      "loss": 0.4322,
+      "step": 1630
+    },
+    {
+      "epoch": 2.1602649006622516,
+      "grad_norm": 0.13195292379565143,
+      "learning_rate": 5.60178783821263e-05,
+      "loss": 0.4351,
+      "step": 1631
+    },
+    {
+      "epoch": 2.16158940397351,
+      "grad_norm": 0.12920339096434544,
+      "learning_rate": 5.5983974522147816e-05,
+      "loss": 0.4354,
+      "step": 1632
+    },
+    {
+      "epoch": 2.1629139072847683,
+      "grad_norm": 0.14017786242029165,
+      "learning_rate": 5.5950056991386055e-05,
+      "loss": 0.4325,
+      "step": 1633
+    },
+    {
+      "epoch": 2.1642384105960266,
+      "grad_norm": 0.14310518170424447,
+      "learning_rate": 5.5916125818849986e-05,
+      "loss": 0.4326,
+      "step": 1634
+    },
+    {
+      "epoch": 2.165562913907285,
+      "grad_norm": 0.15032489481609645,
+      "learning_rate": 5.5882181033560285e-05,
+      "loss": 0.4348,
+      "step": 1635
+    },
+    {
+      "epoch": 2.1668874172185433,
+      "grad_norm": 0.14598744201095945,
+      "learning_rate": 5.584822266454928e-05,
+      "loss": 0.4336,
+      "step": 1636
+    },
+    {
+      "epoch": 2.168211920529801,
+      "grad_norm": 0.16769750273800763,
+      "learning_rate": 5.5814250740860906e-05,
+      "loss": 0.4343,
+      "step": 1637
+    },
+    {
+      "epoch": 2.1695364238410595,
+      "grad_norm": 0.20694652710015365,
+      "learning_rate": 5.578026529155069e-05,
+      "loss": 0.4355,
+      "step": 1638
+    },
+    {
+      "epoch": 2.170860927152318,
+      "grad_norm": 0.1896226450427282,
+      "learning_rate": 5.574626634568572e-05,
+      "loss": 0.4343,
+      "step": 1639
+    },
+    {
+      "epoch": 2.172185430463576,
+      "grad_norm": 0.12031134515325016,
+      "learning_rate": 5.571225393234463e-05,
+      "loss": 0.4338,
+      "step": 1640
+    },
+    {
+      "epoch": 2.1735099337748345,
+      "grad_norm": 0.1441553695138287,
+      "learning_rate": 5.5678228080617596e-05,
+      "loss": 0.4351,
+      "step": 1641
+    },
+    {
+      "epoch": 2.174834437086093,
+      "grad_norm": 0.16101976893219944,
+      "learning_rate": 5.564418881960624e-05,
+      "loss": 0.4359,
+      "step": 1642
+    },
+    {
+      "epoch": 2.176158940397351,
+      "grad_norm": 0.14102447938267243,
+      "learning_rate": 5.561013617842371e-05,
+      "loss": 0.4337,
+      "step": 1643
+    },
+    {
+      "epoch": 2.1774834437086095,
+      "grad_norm": 0.1260114768993658,
+      "learning_rate": 5.5576070186194575e-05,
+      "loss": 0.4334,
+      "step": 1644
+    },
+    {
+      "epoch": 2.1788079470198674,
+      "grad_norm": 0.13218888513415805,
+      "learning_rate": 5.5541990872054785e-05,
+      "loss": 0.4339,
+      "step": 1645
+    },
+    {
+      "epoch": 2.1801324503311257,
+      "grad_norm": 0.15063350052696672,
+      "learning_rate": 5.550789826515173e-05,
+      "loss": 0.4343,
+      "step": 1646
+    },
+    {
+      "epoch": 2.181456953642384,
+      "grad_norm": 0.14387422923593118,
+      "learning_rate": 5.547379239464416e-05,
+      "loss": 0.4315,
+      "step": 1647
+    },
+    {
+      "epoch": 2.1827814569536423,
+      "grad_norm": 0.15318402477184806,
+      "learning_rate": 5.543967328970217e-05,
+      "loss": 0.4352,
+      "step": 1648
+    },
+    {
+      "epoch": 2.1841059602649007,
+      "grad_norm": 0.1396642770928862,
+      "learning_rate": 5.540554097950715e-05,
+      "loss": 0.4375,
+      "step": 1649
+    },
+    {
+      "epoch": 2.185430463576159,
+      "grad_norm": 0.12259419466220746,
+      "learning_rate": 5.537139549325183e-05,
+      "loss": 0.4335,
+      "step": 1650
+    },
+    {
+      "epoch": 2.1867549668874173,
+      "grad_norm": 0.14001981824313284,
+      "learning_rate": 5.533723686014015e-05,
+      "loss": 0.4311,
+      "step": 1651
+    },
+    {
+      "epoch": 2.1880794701986757,
+      "grad_norm": 0.1570562453029728,
+      "learning_rate": 5.530306510938734e-05,
+      "loss": 0.4344,
+      "step": 1652
+    },
+    {
+      "epoch": 2.1894039735099335,
+      "grad_norm": 0.16360734803193616,
+      "learning_rate": 5.5268880270219835e-05,
+      "loss": 0.432,
+      "step": 1653
+    },
+    {
+      "epoch": 2.190728476821192,
+      "grad_norm": 0.15475179457615604,
+      "learning_rate": 5.523468237187528e-05,
+      "loss": 0.4336,
+      "step": 1654
+    },
+    {
+      "epoch": 2.19205298013245,
+      "grad_norm": 0.1464842023213899,
+      "learning_rate": 5.520047144360245e-05,
+      "loss": 0.4331,
+      "step": 1655
+    },
+    {
+      "epoch": 2.1933774834437085,
+      "grad_norm": 0.1304407360428309,
+      "learning_rate": 5.51662475146613e-05,
+      "loss": 0.4384,
+      "step": 1656
+    },
+    {
+      "epoch": 2.194701986754967,
+      "grad_norm": 0.11394466010693859,
+      "learning_rate": 5.51320106143229e-05,
+      "loss": 0.433,
+      "step": 1657
+    },
+    {
+      "epoch": 2.196026490066225,
+      "grad_norm": 0.1399760076451764,
+      "learning_rate": 5.509776077186939e-05,
+      "loss": 0.4333,
+      "step": 1658
+    },
+    {
+      "epoch": 2.1973509933774835,
+      "grad_norm": 0.13573264514821448,
+      "learning_rate": 5.5063498016594006e-05,
+      "loss": 0.4342,
+      "step": 1659
+    },
+    {
+      "epoch": 2.198675496688742,
+      "grad_norm": 0.12201277146774503,
+      "learning_rate": 5.502922237780102e-05,
+      "loss": 0.4328,
+      "step": 1660
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.12364361687480331,
+      "learning_rate": 5.499493388480571e-05,
+      "loss": 0.4328,
+      "step": 1661
+    },
+    {
+      "epoch": 2.201324503311258,
+      "grad_norm": 0.13826342388377677,
+      "learning_rate": 5.496063256693436e-05,
+      "loss": 0.4341,
+      "step": 1662
+    },
+    {
+      "epoch": 2.2026490066225164,
+      "grad_norm": 0.13538470433795208,
+      "learning_rate": 5.4926318453524234e-05,
+      "loss": 0.4336,
+      "step": 1663
+    },
+    {
+      "epoch": 2.2039735099337747,
+      "grad_norm": 0.1258832130498806,
+      "learning_rate": 5.489199157392352e-05,
+      "loss": 0.4342,
+      "step": 1664
+    },
+    {
+      "epoch": 2.205298013245033,
+      "grad_norm": 0.1617931844422333,
+      "learning_rate": 5.485765195749134e-05,
+      "loss": 0.4377,
+      "step": 1665
+    },
+    {
+      "epoch": 2.2066225165562914,
+      "grad_norm": 0.14449935043571618,
+      "learning_rate": 5.482329963359768e-05,
+      "loss": 0.4347,
+      "step": 1666
+    },
+    {
+      "epoch": 2.2079470198675497,
+      "grad_norm": 0.1558407556976361,
+      "learning_rate": 5.478893463162343e-05,
+      "loss": 0.4346,
+      "step": 1667
+    },
+    {
+      "epoch": 2.209271523178808,
+      "grad_norm": 0.1539207213490265,
+      "learning_rate": 5.4754556980960314e-05,
+      "loss": 0.4369,
+      "step": 1668
+    },
+    {
+      "epoch": 2.2105960264900664,
+      "grad_norm": 0.14208206228586495,
+      "learning_rate": 5.472016671101087e-05,
+      "loss": 0.4308,
+      "step": 1669
+    },
+    {
+      "epoch": 2.2119205298013247,
+      "grad_norm": 0.13036771128774202,
+      "learning_rate": 5.4685763851188426e-05,
+      "loss": 0.437,
+      "step": 1670
+    },
+    {
+      "epoch": 2.2132450331125826,
+      "grad_norm": 0.15628717453722565,
+      "learning_rate": 5.4651348430917076e-05,
+      "loss": 0.4352,
+      "step": 1671
+    },
+    {
+      "epoch": 2.214569536423841,
+      "grad_norm": 0.17787283380857,
+      "learning_rate": 5.4616920479631665e-05,
+      "loss": 0.4345,
+      "step": 1672
+    },
+    {
+      "epoch": 2.2158940397350992,
+      "grad_norm": 0.1793189978219582,
+      "learning_rate": 5.458248002677774e-05,
+      "loss": 0.4357,
+      "step": 1673
+    },
+    {
+      "epoch": 2.2172185430463576,
+      "grad_norm": 0.16200250095334512,
+      "learning_rate": 5.454802710181158e-05,
+      "loss": 0.4368,
+      "step": 1674
+    },
+    {
+      "epoch": 2.218543046357616,
+      "grad_norm": 0.13900108655233512,
+      "learning_rate": 5.4513561734200074e-05,
+      "loss": 0.433,
+      "step": 1675
+    },
+    {
+      "epoch": 2.2198675496688742,
+      "grad_norm": 0.1510395034896451,
+      "learning_rate": 5.447908395342079e-05,
+      "loss": 0.435,
+      "step": 1676
+    },
+    {
+      "epoch": 2.2211920529801326,
+      "grad_norm": 0.17542674080276205,
+      "learning_rate": 5.444459378896192e-05,
+      "loss": 0.4365,
+      "step": 1677
+    },
+    {
+      "epoch": 2.222516556291391,
+      "grad_norm": 0.14012233875753557,
+      "learning_rate": 5.441009127032222e-05,
+      "loss": 0.4358,
+      "step": 1678
+    },
+    {
+      "epoch": 2.223841059602649,
+      "grad_norm": 0.13693354756854054,
+      "learning_rate": 5.437557642701102e-05,
+      "loss": 0.4342,
+      "step": 1679
+    },
+    {
+      "epoch": 2.225165562913907,
+      "grad_norm": 0.14188014653836178,
+      "learning_rate": 5.434104928854822e-05,
+      "loss": 0.4343,
+      "step": 1680
+    },
+    {
+      "epoch": 2.2264900662251654,
+      "grad_norm": 0.15793300325867773,
+      "learning_rate": 5.4306509884464184e-05,
+      "loss": 0.4371,
+      "step": 1681
+    },
+    {
+      "epoch": 2.2278145695364238,
+      "grad_norm": 0.1585361615718665,
+      "learning_rate": 5.42719582442998e-05,
+      "loss": 0.4391,
+      "step": 1682
+    },
+    {
+      "epoch": 2.229139072847682,
+      "grad_norm": 0.13138096177885705,
+      "learning_rate": 5.423739439760642e-05,
+      "loss": 0.4355,
+      "step": 1683
+    },
+    {
+      "epoch": 2.2304635761589404,
+      "grad_norm": 0.11593545417362722,
+      "learning_rate": 5.4202818373945834e-05,
+      "loss": 0.4371,
+      "step": 1684
+    },
+    {
+      "epoch": 2.2317880794701987,
+      "grad_norm": 0.14169482537922423,
+      "learning_rate": 5.416823020289025e-05,
+      "loss": 0.4355,
+      "step": 1685
+    },
+    {
+      "epoch": 2.233112582781457,
+      "grad_norm": 0.13035096359959272,
+      "learning_rate": 5.413362991402224e-05,
+      "loss": 0.4347,
+      "step": 1686
+    },
+    {
+      "epoch": 2.2344370860927154,
+      "grad_norm": 0.11925502715649695,
+      "learning_rate": 5.409901753693477e-05,
+      "loss": 0.4375,
+      "step": 1687
+    },
+    {
+      "epoch": 2.2357615894039737,
+      "grad_norm": 0.12502647853485654,
+      "learning_rate": 5.406439310123114e-05,
+      "loss": 0.4359,
+      "step": 1688
+    },
+    {
+      "epoch": 2.2370860927152316,
+      "grad_norm": 0.1334657376391245,
+      "learning_rate": 5.402975663652497e-05,
+      "loss": 0.437,
+      "step": 1689
+    },
+    {
+      "epoch": 2.23841059602649,
+      "grad_norm": 0.13481070787719326,
+      "learning_rate": 5.3995108172440115e-05,
+      "loss": 0.4318,
+      "step": 1690
+    },
+    {
+      "epoch": 2.2397350993377483,
+      "grad_norm": 0.12286314264125177,
+      "learning_rate": 5.396044773861078e-05,
+      "loss": 0.4338,
+      "step": 1691
+    },
+    {
+      "epoch": 2.2410596026490066,
+      "grad_norm": 0.1204690975231818,
+      "learning_rate": 5.392577536468133e-05,
+      "loss": 0.436,
+      "step": 1692
+    },
+    {
+      "epoch": 2.242384105960265,
+      "grad_norm": 0.12263620896664064,
+      "learning_rate": 5.389109108030638e-05,
+      "loss": 0.4356,
+      "step": 1693
+    },
+    {
+      "epoch": 2.2437086092715233,
+      "grad_norm": 0.13838995386309705,
+      "learning_rate": 5.385639491515073e-05,
+      "loss": 0.4343,
+      "step": 1694
+    },
+    {
+      "epoch": 2.2450331125827816,
+      "grad_norm": 0.1254052922124742,
+      "learning_rate": 5.382168689888935e-05,
+      "loss": 0.4316,
+      "step": 1695
+    },
+    {
+      "epoch": 2.24635761589404,
+      "grad_norm": 0.13693645647049985,
+      "learning_rate": 5.378696706120731e-05,
+      "loss": 0.4384,
+      "step": 1696
+    },
+    {
+      "epoch": 2.247682119205298,
+      "grad_norm": 0.14262993259226864,
+      "learning_rate": 5.375223543179982e-05,
+      "loss": 0.4327,
+      "step": 1697
+    },
+    {
+      "epoch": 2.249006622516556,
+      "grad_norm": 0.14900039215177394,
+      "learning_rate": 5.371749204037218e-05,
+      "loss": 0.4306,
+      "step": 1698
+    },
+    {
+      "epoch": 2.2503311258278145,
+      "grad_norm": 0.1667430167842563,
+      "learning_rate": 5.368273691663973e-05,
+      "loss": 0.4318,
+      "step": 1699
+    },
+    {
+      "epoch": 2.251655629139073,
+      "grad_norm": 0.11167852610162267,
+      "learning_rate": 5.364797009032786e-05,
+      "loss": 0.4358,
+      "step": 1700
+    },
+    {
+      "epoch": 2.252980132450331,
+      "grad_norm": 0.1537799450801514,
+      "learning_rate": 5.361319159117195e-05,
+      "loss": 0.4317,
+      "step": 1701
+    },
+    {
+      "epoch": 2.2543046357615895,
+      "grad_norm": 0.168514845554482,
+      "learning_rate": 5.3578401448917384e-05,
+      "loss": 0.4329,
+      "step": 1702
+    },
+    {
+      "epoch": 2.255629139072848,
+      "grad_norm": 0.14721515680923172,
+      "learning_rate": 5.3543599693319484e-05,
+      "loss": 0.4362,
+      "step": 1703
+    },
+    {
+      "epoch": 2.256953642384106,
+      "grad_norm": 0.14591040937258806,
+      "learning_rate": 5.3508786354143533e-05,
+      "loss": 0.4361,
+      "step": 1704
+    },
+    {
+      "epoch": 2.258278145695364,
+      "grad_norm": 0.15126949761974146,
+      "learning_rate": 5.3473961461164706e-05,
+      "loss": 0.4342,
+      "step": 1705
+    },
+    {
+      "epoch": 2.2596026490066223,
+      "grad_norm": 0.19493493906166734,
+      "learning_rate": 5.3439125044168035e-05,
+      "loss": 0.4349,
+      "step": 1706
+    },
+    {
+      "epoch": 2.2609271523178807,
+      "grad_norm": 0.18129318569449251,
+      "learning_rate": 5.340427713294844e-05,
+      "loss": 0.4373,
+      "step": 1707
+    },
+    {
+      "epoch": 2.262251655629139,
+      "grad_norm": 0.15053628534685887,
+      "learning_rate": 5.336941775731069e-05,
+      "loss": 0.4381,
+      "step": 1708
+    },
+    {
+      "epoch": 2.2635761589403973,
+      "grad_norm": 0.1332607563852359,
+      "learning_rate": 5.33345469470693e-05,
+      "loss": 0.4328,
+      "step": 1709
+    },
+    {
+      "epoch": 2.2649006622516556,
+      "grad_norm": 0.11203160571750988,
+      "learning_rate": 5.329966473204862e-05,
+      "loss": 0.4386,
+      "step": 1710
+    },
+    {
+      "epoch": 2.266225165562914,
+      "grad_norm": 0.14137358552443474,
+      "learning_rate": 5.326477114208272e-05,
+      "loss": 0.4361,
+      "step": 1711
+    },
+    {
+      "epoch": 2.2675496688741723,
+      "grad_norm": 0.17230318305054027,
+      "learning_rate": 5.3229866207015405e-05,
+      "loss": 0.4351,
+      "step": 1712
+    },
+    {
+      "epoch": 2.2688741721854306,
+      "grad_norm": 0.18438308217220137,
+      "learning_rate": 5.31949499567002e-05,
+      "loss": 0.4318,
+      "step": 1713
+    },
+    {
+      "epoch": 2.2701986754966885,
+      "grad_norm": 0.13977337229119174,
+      "learning_rate": 5.3160022421000284e-05,
+      "loss": 0.4334,
+      "step": 1714
+    },
+    {
+      "epoch": 2.271523178807947,
+      "grad_norm": 0.1190564795273831,
+      "learning_rate": 5.312508362978852e-05,
+      "loss": 0.4352,
+      "step": 1715
+    },
+    {
+      "epoch": 2.272847682119205,
+      "grad_norm": 0.13801914306495694,
+      "learning_rate": 5.309013361294737e-05,
+      "loss": 0.4355,
+      "step": 1716
+    },
+    {
+      "epoch": 2.2741721854304635,
+      "grad_norm": 0.14286007098679387,
+      "learning_rate": 5.305517240036889e-05,
+      "loss": 0.4323,
+      "step": 1717
+    },
+    {
+      "epoch": 2.275496688741722,
+      "grad_norm": 0.17686950045999425,
+      "learning_rate": 5.302020002195474e-05,
+      "loss": 0.4342,
+      "step": 1718
+    },
+    {
+      "epoch": 2.27682119205298,
+      "grad_norm": 0.16412156377736006,
+      "learning_rate": 5.298521650761613e-05,
+      "loss": 0.4379,
+      "step": 1719
+    },
+    {
+      "epoch": 2.2781456953642385,
+      "grad_norm": 0.14710299760413173,
+      "learning_rate": 5.295022188727376e-05,
+      "loss": 0.4306,
+      "step": 1720
+    },
+    {
+      "epoch": 2.279470198675497,
+      "grad_norm": 0.1513752108814609,
+      "learning_rate": 5.291521619085785e-05,
+      "loss": 0.4372,
+      "step": 1721
+    },
+    {
+      "epoch": 2.280794701986755,
+      "grad_norm": 0.1843464444735487,
+      "learning_rate": 5.288019944830812e-05,
+      "loss": 0.435,
+      "step": 1722
+    },
+    {
+      "epoch": 2.282119205298013,
+      "grad_norm": 0.15600156389727562,
+      "learning_rate": 5.284517168957367e-05,
+      "loss": 0.4311,
+      "step": 1723
+    },
+    {
+      "epoch": 2.2834437086092714,
+      "grad_norm": 0.1372117139493852,
+      "learning_rate": 5.2810132944613086e-05,
+      "loss": 0.4327,
+      "step": 1724
+    },
+    {
+      "epoch": 2.2847682119205297,
+      "grad_norm": 0.1477999646229247,
+      "learning_rate": 5.277508324339433e-05,
+      "loss": 0.4366,
+      "step": 1725
+    },
+    {
+      "epoch": 2.286092715231788,
+      "grad_norm": 0.17042487413416033,
+      "learning_rate": 5.2740022615894725e-05,
+      "loss": 0.4351,
+      "step": 1726
+    },
+    {
+      "epoch": 2.2874172185430464,
+      "grad_norm": 0.15026981393923752,
+      "learning_rate": 5.2704951092100945e-05,
+      "loss": 0.4346,
+      "step": 1727
+    },
+    {
+      "epoch": 2.2887417218543047,
+      "grad_norm": 0.13824802948931553,
+      "learning_rate": 5.2669868702008975e-05,
+      "loss": 0.4342,
+      "step": 1728
+    },
+    {
+      "epoch": 2.290066225165563,
+      "grad_norm": 0.11328445177742642,
+      "learning_rate": 5.263477547562413e-05,
+      "loss": 0.4314,
+      "step": 1729
+    },
+    {
+      "epoch": 2.2913907284768213,
+      "grad_norm": 0.16578625677444184,
+      "learning_rate": 5.259967144296094e-05,
+      "loss": 0.4371,
+      "step": 1730
+    },
+    {
+      "epoch": 2.2927152317880797,
+      "grad_norm": 0.17999253771577015,
+      "learning_rate": 5.2564556634043214e-05,
+      "loss": 0.434,
+      "step": 1731
+    },
+    {
+      "epoch": 2.2940397350993376,
+      "grad_norm": 0.1285623477699793,
+      "learning_rate": 5.2529431078903955e-05,
+      "loss": 0.436,
+      "step": 1732
+    },
+    {
+      "epoch": 2.295364238410596,
+      "grad_norm": 0.1526676898601384,
+      "learning_rate": 5.249429480758539e-05,
+      "loss": 0.435,
+      "step": 1733
+    },
+    {
+      "epoch": 2.296688741721854,
+      "grad_norm": 0.20501892227701013,
+      "learning_rate": 5.2459147850138864e-05,
+      "loss": 0.4346,
+      "step": 1734
+    },
+    {
+      "epoch": 2.2980132450331126,
+      "grad_norm": 0.1349169402214536,
+      "learning_rate": 5.242399023662492e-05,
+      "loss": 0.4346,
+      "step": 1735
+    },
+    {
+      "epoch": 2.299337748344371,
+      "grad_norm": 0.10367044825544644,
+      "learning_rate": 5.238882199711315e-05,
+      "loss": 0.4384,
+      "step": 1736
+    },
+    {
+      "epoch": 2.300662251655629,
+      "grad_norm": 0.14589436502938075,
+      "learning_rate": 5.235364316168228e-05,
+      "loss": 0.4343,
+      "step": 1737
+    },
+    {
+      "epoch": 2.3019867549668875,
+      "grad_norm": 0.12625160250231135,
+      "learning_rate": 5.23184537604201e-05,
+      "loss": 0.4363,
+      "step": 1738
+    },
+    {
+      "epoch": 2.303311258278146,
+      "grad_norm": 0.12157475799974742,
+      "learning_rate": 5.2283253823423394e-05,
+      "loss": 0.4345,
+      "step": 1739
+    },
+    {
+      "epoch": 2.304635761589404,
+      "grad_norm": 0.14429126651409022,
+      "learning_rate": 5.2248043380798016e-05,
+      "loss": 0.4387,
+      "step": 1740
+    },
+    {
+      "epoch": 2.305960264900662,
+      "grad_norm": 0.13102126462486444,
+      "learning_rate": 5.221282246265875e-05,
+      "loss": 0.4344,
+      "step": 1741
+    },
+    {
+      "epoch": 2.3072847682119204,
+      "grad_norm": 0.12075067986635224,
+      "learning_rate": 5.217759109912937e-05,
+      "loss": 0.4329,
+      "step": 1742
+    },
+    {
+      "epoch": 2.3086092715231787,
+      "grad_norm": 0.1476638510581227,
+      "learning_rate": 5.214234932034259e-05,
+      "loss": 0.4313,
+      "step": 1743
+    },
+    {
+      "epoch": 2.309933774834437,
+      "grad_norm": 0.1312653366201799,
+      "learning_rate": 5.210709715644e-05,
+      "loss": 0.4337,
+      "step": 1744
+    },
+    {
+      "epoch": 2.3112582781456954,
+      "grad_norm": 0.1205104802220811,
+      "learning_rate": 5.2071834637572106e-05,
+      "loss": 0.4343,
+      "step": 1745
+    },
+    {
+      "epoch": 2.3125827814569537,
+      "grad_norm": 0.11409449596566638,
+      "learning_rate": 5.2036561793898254e-05,
+      "loss": 0.434,
+      "step": 1746
+    },
+    {
+      "epoch": 2.313907284768212,
+      "grad_norm": 0.12993149164364132,
+      "learning_rate": 5.200127865558662e-05,
+      "loss": 0.4357,
+      "step": 1747
+    },
+    {
+      "epoch": 2.3152317880794704,
+      "grad_norm": 0.13607762278536686,
+      "learning_rate": 5.1965985252814184e-05,
+      "loss": 0.4367,
+      "step": 1748
+    },
+    {
+      "epoch": 2.3165562913907287,
+      "grad_norm": 0.11856777296727106,
+      "learning_rate": 5.193068161576672e-05,
+      "loss": 0.4348,
+      "step": 1749
+    },
+    {
+      "epoch": 2.3178807947019866,
+      "grad_norm": 0.11893091583648031,
+      "learning_rate": 5.1895367774638736e-05,
+      "loss": 0.4343,
+      "step": 1750
+    },
+    {
+      "epoch": 2.319205298013245,
+      "grad_norm": 0.13421012083910133,
+      "learning_rate": 5.186004375963348e-05,
+      "loss": 0.4323,
+      "step": 1751
+    },
+    {
+      "epoch": 2.3205298013245033,
+      "grad_norm": 0.12908620612311072,
+      "learning_rate": 5.1824709600962886e-05,
+      "loss": 0.4368,
+      "step": 1752
+    },
+    {
+      "epoch": 2.3218543046357616,
+      "grad_norm": 0.11595701090935862,
+      "learning_rate": 5.178936532884759e-05,
+      "loss": 0.4338,
+      "step": 1753
+    },
+    {
+      "epoch": 2.32317880794702,
+      "grad_norm": 0.131482188340973,
+      "learning_rate": 5.175401097351685e-05,
+      "loss": 0.433,
+      "step": 1754
+    },
+    {
+      "epoch": 2.3245033112582782,
+      "grad_norm": 0.14400932620484277,
+      "learning_rate": 5.1718646565208554e-05,
+      "loss": 0.4331,
+      "step": 1755
+    },
+    {
+      "epoch": 2.3258278145695366,
+      "grad_norm": 0.15672891081393633,
+      "learning_rate": 5.1683272134169216e-05,
+      "loss": 0.4325,
+      "step": 1756
+    },
+    {
+      "epoch": 2.3271523178807945,
+      "grad_norm": 0.12960129123649272,
+      "learning_rate": 5.164788771065389e-05,
+      "loss": 0.4389,
+      "step": 1757
+    },
+    {
+      "epoch": 2.328476821192053,
+      "grad_norm": 0.12464438456925948,
+      "learning_rate": 5.161249332492617e-05,
+      "loss": 0.4327,
+      "step": 1758
+    },
+    {
+      "epoch": 2.329801324503311,
+      "grad_norm": 0.15609011279884452,
+      "learning_rate": 5.157708900725821e-05,
+      "loss": 0.4339,
+      "step": 1759
+    },
+    {
+      "epoch": 2.3311258278145695,
+      "grad_norm": 0.1428351854036793,
+      "learning_rate": 5.154167478793064e-05,
+      "loss": 0.4362,
+      "step": 1760
+    },
+    {
+      "epoch": 2.332450331125828,
+      "grad_norm": 0.13430083691416192,
+      "learning_rate": 5.150625069723254e-05,
+      "loss": 0.4336,
+      "step": 1761
+    },
+    {
+      "epoch": 2.333774834437086,
+      "grad_norm": 0.12172812219241462,
+      "learning_rate": 5.147081676546145e-05,
+      "loss": 0.4337,
+      "step": 1762
+    },
+    {
+      "epoch": 2.3350993377483444,
+      "grad_norm": 0.1115574692289964,
+      "learning_rate": 5.1435373022923326e-05,
+      "loss": 0.4346,
+      "step": 1763
+    },
+    {
+      "epoch": 2.3364238410596028,
+      "grad_norm": 0.14566118098982478,
+      "learning_rate": 5.139991949993252e-05,
+      "loss": 0.4374,
+      "step": 1764
+    },
+    {
+      "epoch": 2.337748344370861,
+      "grad_norm": 0.135038101862186,
+      "learning_rate": 5.136445622681172e-05,
+      "loss": 0.4348,
+      "step": 1765
+    },
+    {
+      "epoch": 2.339072847682119,
+      "grad_norm": 0.11198640532550169,
+      "learning_rate": 5.1328983233892e-05,
+      "loss": 0.4337,
+      "step": 1766
+    },
+    {
+      "epoch": 2.3403973509933773,
+      "grad_norm": 0.11413281793945403,
+      "learning_rate": 5.129350055151271e-05,
+      "loss": 0.4357,
+      "step": 1767
+    },
+    {
+      "epoch": 2.3417218543046356,
+      "grad_norm": 0.14953315075963916,
+      "learning_rate": 5.125800821002151e-05,
+      "loss": 0.4346,
+      "step": 1768
+    },
+    {
+      "epoch": 2.343046357615894,
+      "grad_norm": 0.13511971189895,
+      "learning_rate": 5.1222506239774295e-05,
+      "loss": 0.4325,
+      "step": 1769
+    },
+    {
+      "epoch": 2.3443708609271523,
+      "grad_norm": 0.113581310035483,
+      "learning_rate": 5.1186994671135226e-05,
+      "loss": 0.4358,
+      "step": 1770
+    },
+    {
+      "epoch": 2.3456953642384106,
+      "grad_norm": 0.12178438277696843,
+      "learning_rate": 5.1151473534476664e-05,
+      "loss": 0.4353,
+      "step": 1771
+    },
+    {
+      "epoch": 2.347019867549669,
+      "grad_norm": 0.1071247977438184,
+      "learning_rate": 5.1115942860179134e-05,
+      "loss": 0.4345,
+      "step": 1772
+    },
+    {
+      "epoch": 2.3483443708609273,
+      "grad_norm": 0.11877663348366435,
+      "learning_rate": 5.108040267863133e-05,
+      "loss": 0.4346,
+      "step": 1773
+    },
+    {
+      "epoch": 2.3496688741721856,
+      "grad_norm": 0.12240754290501306,
+      "learning_rate": 5.104485302023008e-05,
+      "loss": 0.4328,
+      "step": 1774
+    },
+    {
+      "epoch": 2.3509933774834435,
+      "grad_norm": 0.11012429258676817,
+      "learning_rate": 5.1009293915380334e-05,
+      "loss": 0.4349,
+      "step": 1775
+    },
+    {
+      "epoch": 2.352317880794702,
+      "grad_norm": 0.12888478680629054,
+      "learning_rate": 5.097372539449508e-05,
+      "loss": 0.435,
+      "step": 1776
+    },
+    {
+      "epoch": 2.35364238410596,
+      "grad_norm": 0.1547483540601825,
+      "learning_rate": 5.093814748799542e-05,
+      "loss": 0.4335,
+      "step": 1777
+    },
+    {
+      "epoch": 2.3549668874172185,
+      "grad_norm": 0.11190369182648183,
+      "learning_rate": 5.0902560226310405e-05,
+      "loss": 0.434,
+      "step": 1778
+    },
+    {
+      "epoch": 2.356291390728477,
+      "grad_norm": 0.12267229623812666,
+      "learning_rate": 5.086696363987715e-05,
+      "loss": 0.433,
+      "step": 1779
+    },
+    {
+      "epoch": 2.357615894039735,
+      "grad_norm": 0.12088864764061134,
+      "learning_rate": 5.0831357759140734e-05,
+      "loss": 0.4333,
+      "step": 1780
+    },
+    {
+      "epoch": 2.3589403973509935,
+      "grad_norm": 0.14730185560947637,
+      "learning_rate": 5.079574261455417e-05,
+      "loss": 0.4341,
+      "step": 1781
+    },
+    {
+      "epoch": 2.360264900662252,
+      "grad_norm": 0.13038878230315598,
+      "learning_rate": 5.076011823657839e-05,
+      "loss": 0.4374,
+      "step": 1782
+    },
+    {
+      "epoch": 2.36158940397351,
+      "grad_norm": 0.15092437258546146,
+      "learning_rate": 5.072448465568223e-05,
+      "loss": 0.437,
+      "step": 1783
+    },
+    {
+      "epoch": 2.362913907284768,
+      "grad_norm": 0.13452502468303415,
+      "learning_rate": 5.068884190234242e-05,
+      "loss": 0.437,
+      "step": 1784
+    },
+    {
+      "epoch": 2.3642384105960264,
+      "grad_norm": 0.11260632723898564,
+      "learning_rate": 5.0653190007043496e-05,
+      "loss": 0.4394,
+      "step": 1785
+    },
+    {
+      "epoch": 2.3655629139072847,
+      "grad_norm": 0.1259784931815095,
+      "learning_rate": 5.061752900027785e-05,
+      "loss": 0.4309,
+      "step": 1786
+    },
+    {
+      "epoch": 2.366887417218543,
+      "grad_norm": 0.14500488852287427,
+      "learning_rate": 5.0581858912545636e-05,
+      "loss": 0.4339,
+      "step": 1787
+    },
+    {
+      "epoch": 2.3682119205298013,
+      "grad_norm": 0.1520969231834478,
+      "learning_rate": 5.05461797743548e-05,
+      "loss": 0.4354,
+      "step": 1788
+    },
+    {
+      "epoch": 2.3695364238410597,
+      "grad_norm": 0.14629893824055312,
+      "learning_rate": 5.0510491616221e-05,
+      "loss": 0.4313,
+      "step": 1789
+    },
+    {
+      "epoch": 2.370860927152318,
+      "grad_norm": 0.13665971046748884,
+      "learning_rate": 5.0474794468667636e-05,
+      "loss": 0.4333,
+      "step": 1790
+    },
+    {
+      "epoch": 2.3721854304635763,
+      "grad_norm": 0.12519847182410923,
+      "learning_rate": 5.04390883622258e-05,
+      "loss": 0.4358,
+      "step": 1791
+    },
+    {
+      "epoch": 2.3735099337748347,
+      "grad_norm": 0.12818216472412058,
+      "learning_rate": 5.04033733274342e-05,
+      "loss": 0.4358,
+      "step": 1792
+    },
+    {
+      "epoch": 2.3748344370860925,
+      "grad_norm": 0.14982375783717294,
+      "learning_rate": 5.036764939483924e-05,
+      "loss": 0.4382,
+      "step": 1793
+    },
+    {
+      "epoch": 2.376158940397351,
+      "grad_norm": 0.15564285725273017,
+      "learning_rate": 5.033191659499489e-05,
+      "loss": 0.4351,
+      "step": 1794
+    },
+    {
+      "epoch": 2.377483443708609,
+      "grad_norm": 0.11722210816250708,
+      "learning_rate": 5.029617495846273e-05,
+      "loss": 0.4347,
+      "step": 1795
+    },
+    {
+      "epoch": 2.3788079470198675,
+      "grad_norm": 0.106971298576703,
+      "learning_rate": 5.026042451581186e-05,
+      "loss": 0.4347,
+      "step": 1796
+    },
+    {
+      "epoch": 2.380132450331126,
+      "grad_norm": 0.1093067358729182,
+      "learning_rate": 5.022466529761899e-05,
+      "loss": 0.4314,
+      "step": 1797
+    },
+    {
+      "epoch": 2.381456953642384,
+      "grad_norm": 0.12544495146006954,
+      "learning_rate": 5.018889733446826e-05,
+      "loss": 0.4311,
+      "step": 1798
+    },
+    {
+      "epoch": 2.3827814569536425,
+      "grad_norm": 0.13120576654940272,
+      "learning_rate": 5.0153120656951294e-05,
+      "loss": 0.4353,
+      "step": 1799
+    },
+    {
+      "epoch": 2.384105960264901,
+      "grad_norm": 0.15273963570542248,
+      "learning_rate": 5.011733529566723e-05,
+      "loss": 0.4364,
+      "step": 1800
+    },
+    {
+      "epoch": 2.385430463576159,
+      "grad_norm": 0.12062473627289505,
+      "learning_rate": 5.008154128122257e-05,
+      "loss": 0.4342,
+      "step": 1801
+    },
+    {
+      "epoch": 2.386754966887417,
+      "grad_norm": 0.16453700316082293,
+      "learning_rate": 5.004573864423124e-05,
+      "loss": 0.4376,
+      "step": 1802
+    },
+    {
+      "epoch": 2.3880794701986754,
+      "grad_norm": 0.20780279741550356,
+      "learning_rate": 5.0009927415314556e-05,
+      "loss": 0.433,
+      "step": 1803
+    },
+    {
+      "epoch": 2.3894039735099337,
+      "grad_norm": 0.1423049997340501,
+      "learning_rate": 4.997410762510115e-05,
+      "loss": 0.4353,
+      "step": 1804
+    },
+    {
+      "epoch": 2.390728476821192,
+      "grad_norm": 0.13370672552819562,
+      "learning_rate": 4.9938279304227006e-05,
+      "loss": 0.4326,
+      "step": 1805
+    },
+    {
+      "epoch": 2.3920529801324504,
+      "grad_norm": 0.16955085892665675,
+      "learning_rate": 4.990244248333537e-05,
+      "loss": 0.4383,
+      "step": 1806
+    },
+    {
+      "epoch": 2.3933774834437087,
+      "grad_norm": 0.14422512279773758,
+      "learning_rate": 4.98665971930768e-05,
+      "loss": 0.4314,
+      "step": 1807
+    },
+    {
+      "epoch": 2.394701986754967,
+      "grad_norm": 0.12122638942652446,
+      "learning_rate": 4.9830743464109086e-05,
+      "loss": 0.4331,
+      "step": 1808
+    },
+    {
+      "epoch": 2.396026490066225,
+      "grad_norm": 0.1086894323156468,
+      "learning_rate": 4.979488132709719e-05,
+      "loss": 0.4347,
+      "step": 1809
+    },
+    {
+      "epoch": 2.3973509933774833,
+      "grad_norm": 0.1162425027795365,
+      "learning_rate": 4.9759010812713344e-05,
+      "loss": 0.437,
+      "step": 1810
+    },
+    {
+      "epoch": 2.3986754966887416,
+      "grad_norm": 0.12023101162875749,
+      "learning_rate": 4.972313195163688e-05,
+      "loss": 0.4333,
+      "step": 1811
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.10199149868165827,
+      "learning_rate": 4.9687244774554304e-05,
+      "loss": 0.4362,
+      "step": 1812
+    },
+    {
+      "epoch": 2.4013245033112582,
+      "grad_norm": 0.12285417150309627,
+      "learning_rate": 4.9651349312159204e-05,
+      "loss": 0.4327,
+      "step": 1813
+    },
+    {
+      "epoch": 2.4026490066225166,
+      "grad_norm": 0.13579010945946668,
+      "learning_rate": 4.9615445595152294e-05,
+      "loss": 0.4335,
+      "step": 1814
+    },
+    {
+      "epoch": 2.403973509933775,
+      "grad_norm": 0.12169553650283792,
+      "learning_rate": 4.95795336542413e-05,
+      "loss": 0.4351,
+      "step": 1815
+    },
+    {
+      "epoch": 2.4052980132450332,
+      "grad_norm": 0.11227496971159953,
+      "learning_rate": 4.954361352014103e-05,
+      "loss": 0.4347,
+      "step": 1816
+    },
+    {
+      "epoch": 2.4066225165562916,
+      "grad_norm": 0.12295812710683345,
+      "learning_rate": 4.9507685223573275e-05,
+      "loss": 0.4313,
+      "step": 1817
+    },
+    {
+      "epoch": 2.4079470198675494,
+      "grad_norm": 0.1216703950399423,
+      "learning_rate": 4.947174879526681e-05,
+      "loss": 0.435,
+      "step": 1818
+    },
+    {
+      "epoch": 2.4092715231788078,
+      "grad_norm": 0.13469591353629776,
+      "learning_rate": 4.943580426595736e-05,
+      "loss": 0.433,
+      "step": 1819
+    },
+    {
+      "epoch": 2.410596026490066,
+      "grad_norm": 0.15564437584521307,
+      "learning_rate": 4.939985166638757e-05,
+      "loss": 0.4364,
+      "step": 1820
+    },
+    {
+      "epoch": 2.4119205298013244,
+      "grad_norm": 0.1283452471811568,
+      "learning_rate": 4.936389102730703e-05,
+      "loss": 0.4322,
+      "step": 1821
+    },
+    {
+      "epoch": 2.4132450331125828,
+      "grad_norm": 0.11326307758035775,
+      "learning_rate": 4.932792237947218e-05,
+      "loss": 0.4331,
+      "step": 1822
+    },
+    {
+      "epoch": 2.414569536423841,
+      "grad_norm": 0.10748096499497292,
+      "learning_rate": 4.929194575364626e-05,
+      "loss": 0.4309,
+      "step": 1823
+    },
+    {
+      "epoch": 2.4158940397350994,
+      "grad_norm": 0.11435328582714152,
+      "learning_rate": 4.925596118059943e-05,
+      "loss": 0.4317,
+      "step": 1824
+    },
+    {
+      "epoch": 2.4172185430463577,
+      "grad_norm": 0.11832638064103104,
+      "learning_rate": 4.921996869110857e-05,
+      "loss": 0.4342,
+      "step": 1825
+    },
+    {
+      "epoch": 2.418543046357616,
+      "grad_norm": 0.11609726971592538,
+      "learning_rate": 4.9183968315957365e-05,
+      "loss": 0.4373,
+      "step": 1826
+    },
+    {
+      "epoch": 2.419867549668874,
+      "grad_norm": 0.10857909579068643,
+      "learning_rate": 4.914796008593624e-05,
+      "loss": 0.438,
+      "step": 1827
+    },
+    {
+      "epoch": 2.4211920529801323,
+      "grad_norm": 0.12283019983578589,
+      "learning_rate": 4.9111944031842334e-05,
+      "loss": 0.4353,
+      "step": 1828
+    },
+    {
+      "epoch": 2.4225165562913906,
+      "grad_norm": 0.10508091994887642,
+      "learning_rate": 4.907592018447947e-05,
+      "loss": 0.4323,
+      "step": 1829
+    },
+    {
+      "epoch": 2.423841059602649,
+      "grad_norm": 0.11186873034342877,
+      "learning_rate": 4.903988857465816e-05,
+      "loss": 0.4352,
+      "step": 1830
+    },
+    {
+      "epoch": 2.4251655629139073,
+      "grad_norm": 0.13838571380496,
+      "learning_rate": 4.900384923319554e-05,
+      "loss": 0.4334,
+      "step": 1831
+    },
+    {
+      "epoch": 2.4264900662251656,
+      "grad_norm": 0.11461333587619857,
+      "learning_rate": 4.896780219091535e-05,
+      "loss": 0.4343,
+      "step": 1832
+    },
+    {
+      "epoch": 2.427814569536424,
+      "grad_norm": 0.1544659906201904,
+      "learning_rate": 4.893174747864793e-05,
+      "loss": 0.4372,
+      "step": 1833
+    },
+    {
+      "epoch": 2.4291390728476823,
+      "grad_norm": 0.16017140391417778,
+      "learning_rate": 4.889568512723017e-05,
+      "loss": 0.4336,
+      "step": 1834
+    },
+    {
+      "epoch": 2.4304635761589406,
+      "grad_norm": 0.10878987355917234,
+      "learning_rate": 4.885961516750552e-05,
+      "loss": 0.4321,
+      "step": 1835
+    },
+    {
+      "epoch": 2.4317880794701985,
+      "grad_norm": 0.1264344173686421,
+      "learning_rate": 4.882353763032389e-05,
+      "loss": 0.4364,
+      "step": 1836
+    },
+    {
+      "epoch": 2.433112582781457,
+      "grad_norm": 0.11999366880468462,
+      "learning_rate": 4.8787452546541716e-05,
+      "loss": 0.4306,
+      "step": 1837
+    },
+    {
+      "epoch": 2.434437086092715,
+      "grad_norm": 0.12523414119649046,
+      "learning_rate": 4.875135994702186e-05,
+      "loss": 0.4366,
+      "step": 1838
+    },
+    {
+      "epoch": 2.4357615894039735,
+      "grad_norm": 0.12495411072629845,
+      "learning_rate": 4.871525986263363e-05,
+      "loss": 0.4349,
+      "step": 1839
+    },
+    {
+      "epoch": 2.437086092715232,
+      "grad_norm": 0.10398812794742524,
+      "learning_rate": 4.8679152324252725e-05,
+      "loss": 0.4347,
+      "step": 1840
+    },
+    {
+      "epoch": 2.43841059602649,
+      "grad_norm": 0.1159948867875482,
+      "learning_rate": 4.8643037362761216e-05,
+      "loss": 0.437,
+      "step": 1841
+    },
+    {
+      "epoch": 2.4397350993377485,
+      "grad_norm": 0.13272829975426037,
+      "learning_rate": 4.860691500904756e-05,
+      "loss": 0.4343,
+      "step": 1842
+    },
+    {
+      "epoch": 2.441059602649007,
+      "grad_norm": 0.12286888178482085,
+      "learning_rate": 4.857078529400644e-05,
+      "loss": 0.4367,
+      "step": 1843
+    },
+    {
+      "epoch": 2.442384105960265,
+      "grad_norm": 0.10774708350168788,
+      "learning_rate": 4.853464824853897e-05,
+      "loss": 0.4381,
+      "step": 1844
+    },
+    {
+      "epoch": 2.443708609271523,
+      "grad_norm": 0.137528373658189,
+      "learning_rate": 4.849850390355242e-05,
+      "loss": 0.4333,
+      "step": 1845
+    },
+    {
+      "epoch": 2.4450331125827813,
+      "grad_norm": 0.13394445464456045,
+      "learning_rate": 4.846235228996037e-05,
+      "loss": 0.435,
+      "step": 1846
+    },
+    {
+      "epoch": 2.4463576158940397,
+      "grad_norm": 0.12262442101703165,
+      "learning_rate": 4.8426193438682566e-05,
+      "loss": 0.4329,
+      "step": 1847
+    },
+    {
+      "epoch": 2.447682119205298,
+      "grad_norm": 0.12612356366898858,
+      "learning_rate": 4.8390027380645e-05,
+      "loss": 0.4321,
+      "step": 1848
+    },
+    {
+      "epoch": 2.4490066225165563,
+      "grad_norm": 0.11768056707401847,
+      "learning_rate": 4.83538541467798e-05,
+      "loss": 0.4321,
+      "step": 1849
+    },
+    {
+      "epoch": 2.4503311258278146,
+      "grad_norm": 0.11611948486027548,
+      "learning_rate": 4.83176737680252e-05,
+      "loss": 0.4394,
+      "step": 1850
+    },
+    {
+      "epoch": 2.451655629139073,
+      "grad_norm": 0.11977744506447154,
+      "learning_rate": 4.828148627532562e-05,
+      "loss": 0.4351,
+      "step": 1851
+    },
+    {
+      "epoch": 2.4529801324503313,
+      "grad_norm": 0.1477714754343818,
+      "learning_rate": 4.824529169963149e-05,
+      "loss": 0.4359,
+      "step": 1852
+    },
+    {
+      "epoch": 2.4543046357615896,
+      "grad_norm": 0.15132114059393634,
+      "learning_rate": 4.820909007189934e-05,
+      "loss": 0.4333,
+      "step": 1853
+    },
+    {
+      "epoch": 2.4556291390728475,
+      "grad_norm": 0.10545727410002335,
+      "learning_rate": 4.817288142309171e-05,
+      "loss": 0.4378,
+      "step": 1854
+    },
+    {
+      "epoch": 2.456953642384106,
+      "grad_norm": 0.11275941471462617,
+      "learning_rate": 4.8136665784177174e-05,
+      "loss": 0.4344,
+      "step": 1855
+    },
+    {
+      "epoch": 2.458278145695364,
+      "grad_norm": 0.10805346412970004,
+      "learning_rate": 4.810044318613024e-05,
+      "loss": 0.4384,
+      "step": 1856
+    },
+    {
+      "epoch": 2.4596026490066225,
+      "grad_norm": 0.11706292180026705,
+      "learning_rate": 4.8064213659931394e-05,
+      "loss": 0.4343,
+      "step": 1857
+    },
+    {
+      "epoch": 2.460927152317881,
+      "grad_norm": 0.12332799917962514,
+      "learning_rate": 4.802797723656707e-05,
+      "loss": 0.434,
+      "step": 1858
+    },
+    {
+      "epoch": 2.462251655629139,
+      "grad_norm": 0.11695837107109343,
+      "learning_rate": 4.799173394702958e-05,
+      "loss": 0.4347,
+      "step": 1859
+    },
+    {
+      "epoch": 2.4635761589403975,
+      "grad_norm": 0.11941200402901435,
+      "learning_rate": 4.795548382231708e-05,
+      "loss": 0.4354,
+      "step": 1860
+    },
+    {
+      "epoch": 2.4649006622516554,
+      "grad_norm": 0.10159563957766254,
+      "learning_rate": 4.79192268934336e-05,
+      "loss": 0.4342,
+      "step": 1861
+    },
+    {
+      "epoch": 2.466225165562914,
+      "grad_norm": 0.10386323498415187,
+      "learning_rate": 4.788296319138901e-05,
+      "loss": 0.433,
+      "step": 1862
+    },
+    {
+      "epoch": 2.467549668874172,
+      "grad_norm": 0.1129276655500794,
+      "learning_rate": 4.784669274719896e-05,
+      "loss": 0.4301,
+      "step": 1863
+    },
+    {
+      "epoch": 2.4688741721854304,
+      "grad_norm": 0.1152912187741091,
+      "learning_rate": 4.781041559188481e-05,
+      "loss": 0.4346,
+      "step": 1864
+    },
+    {
+      "epoch": 2.4701986754966887,
+      "grad_norm": 0.10990129288074314,
+      "learning_rate": 4.7774131756473746e-05,
+      "loss": 0.4359,
+      "step": 1865
+    },
+    {
+      "epoch": 2.471523178807947,
+      "grad_norm": 0.11505039230121866,
+      "learning_rate": 4.773784127199861e-05,
+      "loss": 0.4309,
+      "step": 1866
+    },
+    {
+      "epoch": 2.4728476821192054,
+      "grad_norm": 0.11793544612257888,
+      "learning_rate": 4.770154416949796e-05,
+      "loss": 0.4357,
+      "step": 1867
+    },
+    {
+      "epoch": 2.4741721854304637,
+      "grad_norm": 0.10609822682094047,
+      "learning_rate": 4.766524048001599e-05,
+      "loss": 0.4322,
+      "step": 1868
+    },
+    {
+      "epoch": 2.475496688741722,
+      "grad_norm": 0.1305303098746471,
+      "learning_rate": 4.762893023460256e-05,
+      "loss": 0.4337,
+      "step": 1869
+    },
+    {
+      "epoch": 2.47682119205298,
+      "grad_norm": 0.13734993702443085,
+      "learning_rate": 4.759261346431311e-05,
+      "loss": 0.4322,
+      "step": 1870
+    },
+    {
+      "epoch": 2.4781456953642382,
+      "grad_norm": 0.09599224168265871,
+      "learning_rate": 4.755629020020866e-05,
+      "loss": 0.4335,
+      "step": 1871
+    },
+    {
+      "epoch": 2.4794701986754966,
+      "grad_norm": 0.11292702010883111,
+      "learning_rate": 4.7519960473355814e-05,
+      "loss": 0.4339,
+      "step": 1872
+    },
+    {
+      "epoch": 2.480794701986755,
+      "grad_norm": 0.13221977493772444,
+      "learning_rate": 4.748362431482667e-05,
+      "loss": 0.4344,
+      "step": 1873
+    },
+    {
+      "epoch": 2.482119205298013,
+      "grad_norm": 0.11540748964376425,
+      "learning_rate": 4.744728175569883e-05,
+      "loss": 0.4367,
+      "step": 1874
+    },
+    {
+      "epoch": 2.4834437086092715,
+      "grad_norm": 0.09143288405335911,
+      "learning_rate": 4.7410932827055395e-05,
+      "loss": 0.4341,
+      "step": 1875
+    },
+    {
+      "epoch": 2.48476821192053,
+      "grad_norm": 0.13219100412610219,
+      "learning_rate": 4.737457755998489e-05,
+      "loss": 0.4339,
+      "step": 1876
+    },
+    {
+      "epoch": 2.486092715231788,
+      "grad_norm": 0.1363247253876154,
+      "learning_rate": 4.733821598558127e-05,
+      "loss": 0.4379,
+      "step": 1877
+    },
+    {
+      "epoch": 2.4874172185430465,
+      "grad_norm": 0.11182817583656034,
+      "learning_rate": 4.730184813494389e-05,
+      "loss": 0.4353,
+      "step": 1878
+    },
+    {
+      "epoch": 2.4887417218543044,
+      "grad_norm": 0.12626388427005103,
+      "learning_rate": 4.726547403917746e-05,
+      "loss": 0.434,
+      "step": 1879
+    },
+    {
+      "epoch": 2.4900662251655628,
+      "grad_norm": 0.109066354938557,
+      "learning_rate": 4.722909372939205e-05,
+      "loss": 0.4345,
+      "step": 1880
+    },
+    {
+      "epoch": 2.491390728476821,
+      "grad_norm": 0.11821801363945975,
+      "learning_rate": 4.719270723670301e-05,
+      "loss": 0.4323,
+      "step": 1881
+    },
+    {
+      "epoch": 2.4927152317880794,
+      "grad_norm": 0.11332679971731265,
+      "learning_rate": 4.715631459223103e-05,
+      "loss": 0.4384,
+      "step": 1882
+    },
+    {
+      "epoch": 2.4940397350993377,
+      "grad_norm": 0.1143425015401509,
+      "learning_rate": 4.711991582710203e-05,
+      "loss": 0.4359,
+      "step": 1883
+    },
+    {
+      "epoch": 2.495364238410596,
+      "grad_norm": 0.10464983337637004,
+      "learning_rate": 4.7083510972447124e-05,
+      "loss": 0.4379,
+      "step": 1884
+    },
+    {
+      "epoch": 2.4966887417218544,
+      "grad_norm": 0.12539835261538504,
+      "learning_rate": 4.704710005940273e-05,
+      "loss": 0.4323,
+      "step": 1885
+    },
+    {
+      "epoch": 2.4980132450331127,
+      "grad_norm": 0.11678337568613202,
+      "learning_rate": 4.701068311911037e-05,
+      "loss": 0.4334,
+      "step": 1886
+    },
+    {
+      "epoch": 2.499337748344371,
+      "grad_norm": 0.13384924833668563,
+      "learning_rate": 4.697426018271676e-05,
+      "loss": 0.4352,
+      "step": 1887
+    },
+    {
+      "epoch": 2.500662251655629,
+      "grad_norm": 0.1377091887628952,
+      "learning_rate": 4.6937831281373714e-05,
+      "loss": 0.4337,
+      "step": 1888
+    },
+    {
+      "epoch": 2.5019867549668873,
+      "grad_norm": 0.15236551824685682,
+      "learning_rate": 4.690139644623818e-05,
+      "loss": 0.4345,
+      "step": 1889
+    },
+    {
+      "epoch": 2.5033112582781456,
+      "grad_norm": 0.10206461615737791,
+      "learning_rate": 4.6864955708472157e-05,
+      "loss": 0.4346,
+      "step": 1890
+    },
+    {
+      "epoch": 2.504635761589404,
+      "grad_norm": 0.13285100753144086,
+      "learning_rate": 4.682850909924271e-05,
+      "loss": 0.4345,
+      "step": 1891
+    },
+    {
+      "epoch": 2.5059602649006623,
+      "grad_norm": 0.14098323068172677,
+      "learning_rate": 4.67920566497219e-05,
+      "loss": 0.4301,
+      "step": 1892
+    },
+    {
+      "epoch": 2.5072847682119206,
+      "grad_norm": 0.11158474679786355,
+      "learning_rate": 4.675559839108683e-05,
+      "loss": 0.437,
+      "step": 1893
+    },
+    {
+      "epoch": 2.508609271523179,
+      "grad_norm": 0.10856946252981409,
+      "learning_rate": 4.6719134354519505e-05,
+      "loss": 0.4315,
+      "step": 1894
+    },
+    {
+      "epoch": 2.5099337748344372,
+      "grad_norm": 0.12006682974460438,
+      "learning_rate": 4.668266457120693e-05,
+      "loss": 0.4369,
+      "step": 1895
+    },
+    {
+      "epoch": 2.5112582781456956,
+      "grad_norm": 0.11095239753882401,
+      "learning_rate": 4.664618907234099e-05,
+      "loss": 0.4359,
+      "step": 1896
+    },
+    {
+      "epoch": 2.5125827814569535,
+      "grad_norm": 0.11746005579830146,
+      "learning_rate": 4.660970788911849e-05,
+      "loss": 0.4302,
+      "step": 1897
+    },
+    {
+      "epoch": 2.513907284768212,
+      "grad_norm": 0.11214616995538933,
+      "learning_rate": 4.657322105274105e-05,
+      "loss": 0.4335,
+      "step": 1898
+    },
+    {
+      "epoch": 2.51523178807947,
+      "grad_norm": 0.10101621151245706,
+      "learning_rate": 4.6536728594415184e-05,
+      "loss": 0.4355,
+      "step": 1899
+    },
+    {
+      "epoch": 2.5165562913907285,
+      "grad_norm": 0.12998105690361067,
+      "learning_rate": 4.650023054535217e-05,
+      "loss": 0.4369,
+      "step": 1900
+    },
+    {
+      "epoch": 2.517880794701987,
+      "grad_norm": 0.1625791938724105,
+      "learning_rate": 4.646372693676808e-05,
+      "loss": 0.4362,
+      "step": 1901
+    },
+    {
+      "epoch": 2.519205298013245,
+      "grad_norm": 0.1198636236570356,
+      "learning_rate": 4.6427217799883734e-05,
+      "loss": 0.4333,
+      "step": 1902
+    },
+    {
+      "epoch": 2.5205298013245034,
+      "grad_norm": 0.15734647094948864,
+      "learning_rate": 4.639070316592472e-05,
+      "loss": 0.4348,
+      "step": 1903
+    },
+    {
+      "epoch": 2.5218543046357613,
+      "grad_norm": 0.13313340522878836,
+      "learning_rate": 4.635418306612126e-05,
+      "loss": 0.4363,
+      "step": 1904
+    },
+    {
+      "epoch": 2.52317880794702,
+      "grad_norm": 0.11384600537757321,
+      "learning_rate": 4.6317657531708305e-05,
+      "loss": 0.4311,
+      "step": 1905
+    },
+    {
+      "epoch": 2.524503311258278,
+      "grad_norm": 0.13932887963132873,
+      "learning_rate": 4.628112659392543e-05,
+      "loss": 0.4347,
+      "step": 1906
+    },
+    {
+      "epoch": 2.5258278145695363,
+      "grad_norm": 0.12787490943337876,
+      "learning_rate": 4.624459028401685e-05,
+      "loss": 0.433,
+      "step": 1907
+    },
+    {
+      "epoch": 2.5271523178807946,
+      "grad_norm": 0.11951405473678842,
+      "learning_rate": 4.620804863323133e-05,
+      "loss": 0.4368,
+      "step": 1908
+    },
+    {
+      "epoch": 2.528476821192053,
+      "grad_norm": 0.12862787068554984,
+      "learning_rate": 4.6171501672822264e-05,
+      "loss": 0.4377,
+      "step": 1909
+    },
+    {
+      "epoch": 2.5298013245033113,
+      "grad_norm": 0.12131054240193938,
+      "learning_rate": 4.613494943404756e-05,
+      "loss": 0.4352,
+      "step": 1910
+    },
+    {
+      "epoch": 2.5311258278145696,
+      "grad_norm": 0.10874700876876323,
+      "learning_rate": 4.6098391948169624e-05,
+      "loss": 0.4358,
+      "step": 1911
+    },
+    {
+      "epoch": 2.532450331125828,
+      "grad_norm": 0.1470602588451638,
+      "learning_rate": 4.606182924645536e-05,
+      "loss": 0.4351,
+      "step": 1912
+    },
+    {
+      "epoch": 2.533774834437086,
+      "grad_norm": 0.13492379444020403,
+      "learning_rate": 4.6025261360176164e-05,
+      "loss": 0.4318,
+      "step": 1913
+    },
+    {
+      "epoch": 2.5350993377483446,
+      "grad_norm": 0.14053942700137423,
+      "learning_rate": 4.598868832060783e-05,
+      "loss": 0.4371,
+      "step": 1914
+    },
+    {
+      "epoch": 2.5364238410596025,
+      "grad_norm": 0.14232069435125846,
+      "learning_rate": 4.5952110159030535e-05,
+      "loss": 0.4355,
+      "step": 1915
+    },
+    {
+      "epoch": 2.537748344370861,
+      "grad_norm": 0.14387157764587402,
+      "learning_rate": 4.591552690672891e-05,
+      "loss": 0.4362,
+      "step": 1916
+    },
+    {
+      "epoch": 2.539072847682119,
+      "grad_norm": 0.14014609365119468,
+      "learning_rate": 4.5878938594991886e-05,
+      "loss": 0.4341,
+      "step": 1917
+    },
+    {
+      "epoch": 2.5403973509933775,
+      "grad_norm": 0.13050799540002536,
+      "learning_rate": 4.584234525511274e-05,
+      "loss": 0.4339,
+      "step": 1918
+    },
+    {
+      "epoch": 2.541721854304636,
+      "grad_norm": 0.16654663703770609,
+      "learning_rate": 4.580574691838905e-05,
+      "loss": 0.4357,
+      "step": 1919
+    },
+    {
+      "epoch": 2.543046357615894,
+      "grad_norm": 0.14326527715849693,
+      "learning_rate": 4.576914361612265e-05,
+      "loss": 0.4338,
+      "step": 1920
+    },
+    {
+      "epoch": 2.5443708609271525,
+      "grad_norm": 0.11546787334009136,
+      "learning_rate": 4.5732535379619654e-05,
+      "loss": 0.435,
+      "step": 1921
+    },
+    {
+      "epoch": 2.5456953642384104,
+      "grad_norm": 0.14426094509225693,
+      "learning_rate": 4.569592224019035e-05,
+      "loss": 0.4344,
+      "step": 1922
+    },
+    {
+      "epoch": 2.547019867549669,
+      "grad_norm": 0.19283368145150837,
+      "learning_rate": 4.5659304229149276e-05,
+      "loss": 0.4322,
+      "step": 1923
+    },
+    {
+      "epoch": 2.548344370860927,
+      "grad_norm": 0.13659633400647758,
+      "learning_rate": 4.562268137781509e-05,
+      "loss": 0.4335,
+      "step": 1924
+    },
+    {
+      "epoch": 2.5496688741721854,
+      "grad_norm": 0.13045056712578024,
+      "learning_rate": 4.55860537175106e-05,
+      "loss": 0.4344,
+      "step": 1925
+    },
+    {
+      "epoch": 2.5509933774834437,
+      "grad_norm": 0.1626213308082377,
+      "learning_rate": 4.554942127956274e-05,
+      "loss": 0.4314,
+      "step": 1926
+    },
+    {
+      "epoch": 2.552317880794702,
+      "grad_norm": 0.13277956170796434,
+      "learning_rate": 4.551278409530253e-05,
+      "loss": 0.4342,
+      "step": 1927
+    },
+    {
+      "epoch": 2.5536423841059603,
+      "grad_norm": 0.10638447726932002,
+      "learning_rate": 4.5476142196065025e-05,
+      "loss": 0.4339,
+      "step": 1928
+    },
+    {
+      "epoch": 2.5549668874172187,
+      "grad_norm": 0.15265091145820697,
+      "learning_rate": 4.543949561318935e-05,
+      "loss": 0.4316,
+      "step": 1929
+    },
+    {
+      "epoch": 2.556291390728477,
+      "grad_norm": 0.16078912932413397,
+      "learning_rate": 4.5402844378018606e-05,
+      "loss": 0.432,
+      "step": 1930
+    },
+    {
+      "epoch": 2.557615894039735,
+      "grad_norm": 0.11394131403300516,
+      "learning_rate": 4.536618852189989e-05,
+      "loss": 0.4339,
+      "step": 1931
+    },
+    {
+      "epoch": 2.558940397350993,
+      "grad_norm": 0.14058829583407814,
+      "learning_rate": 4.532952807618423e-05,
+      "loss": 0.4369,
+      "step": 1932
+    },
+    {
+      "epoch": 2.5602649006622515,
+      "grad_norm": 0.11141290152637823,
+      "learning_rate": 4.529286307222662e-05,
+      "loss": 0.4332,
+      "step": 1933
+    },
+    {
+      "epoch": 2.56158940397351,
+      "grad_norm": 0.15697907717182633,
+      "learning_rate": 4.5256193541385916e-05,
+      "loss": 0.4326,
+      "step": 1934
+    },
+    {
+      "epoch": 2.562913907284768,
+      "grad_norm": 0.15759197318566992,
+      "learning_rate": 4.521951951502483e-05,
+      "loss": 0.4339,
+      "step": 1935
+    },
+    {
+      "epoch": 2.5642384105960265,
+      "grad_norm": 0.11632607039686232,
+      "learning_rate": 4.518284102450998e-05,
+      "loss": 0.4335,
+      "step": 1936
+    },
+    {
+      "epoch": 2.565562913907285,
+      "grad_norm": 0.1240713268974873,
+      "learning_rate": 4.514615810121175e-05,
+      "loss": 0.4326,
+      "step": 1937
+    },
+    {
+      "epoch": 2.566887417218543,
+      "grad_norm": 0.12453701658166023,
+      "learning_rate": 4.510947077650434e-05,
+      "loss": 0.4342,
+      "step": 1938
+    },
+    {
+      "epoch": 2.5682119205298015,
+      "grad_norm": 0.11549167346349168,
+      "learning_rate": 4.507277908176569e-05,
+      "loss": 0.4343,
+      "step": 1939
+    },
+    {
+      "epoch": 2.5695364238410594,
+      "grad_norm": 0.11825828387494593,
+      "learning_rate": 4.503608304837753e-05,
+      "loss": 0.4325,
+      "step": 1940
+    },
+    {
+      "epoch": 2.5708609271523177,
+      "grad_norm": 0.11575297330909413,
+      "learning_rate": 4.499938270772523e-05,
+      "loss": 0.4343,
+      "step": 1941
+    },
+    {
+      "epoch": 2.572185430463576,
+      "grad_norm": 0.11712368529136978,
+      "learning_rate": 4.49626780911979e-05,
+      "loss": 0.4361,
+      "step": 1942
+    },
+    {
+      "epoch": 2.5735099337748344,
+      "grad_norm": 0.11101464803380533,
+      "learning_rate": 4.492596923018827e-05,
+      "loss": 0.4328,
+      "step": 1943
+    },
+    {
+      "epoch": 2.5748344370860927,
+      "grad_norm": 0.12866552446214224,
+      "learning_rate": 4.488925615609274e-05,
+      "loss": 0.431,
+      "step": 1944
+    },
+    {
+      "epoch": 2.576158940397351,
+      "grad_norm": 0.12881345833581448,
+      "learning_rate": 4.485253890031127e-05,
+      "loss": 0.4313,
+      "step": 1945
+    },
+    {
+      "epoch": 2.5774834437086094,
+      "grad_norm": 0.11582365921096163,
+      "learning_rate": 4.4815817494247416e-05,
+      "loss": 0.4339,
+      "step": 1946
+    },
+    {
+      "epoch": 2.5788079470198677,
+      "grad_norm": 0.12185139862576552,
+      "learning_rate": 4.4779091969308294e-05,
+      "loss": 0.4312,
+      "step": 1947
+    },
+    {
+      "epoch": 2.580132450331126,
+      "grad_norm": 0.13187299397590257,
+      "learning_rate": 4.474236235690454e-05,
+      "loss": 0.4346,
+      "step": 1948
+    },
+    {
+      "epoch": 2.581456953642384,
+      "grad_norm": 0.12143961667401575,
+      "learning_rate": 4.470562868845026e-05,
+      "loss": 0.4304,
+      "step": 1949
+    },
+    {
+      "epoch": 2.5827814569536423,
+      "grad_norm": 0.14486879191709548,
+      "learning_rate": 4.466889099536305e-05,
+      "loss": 0.4316,
+      "step": 1950
+    },
+    {
+      "epoch": 2.5841059602649006,
+      "grad_norm": 0.11945450026522224,
+      "learning_rate": 4.463214930906397e-05,
+      "loss": 0.4319,
+      "step": 1951
+    },
+    {
+      "epoch": 2.585430463576159,
+      "grad_norm": 0.13364469531011,
+      "learning_rate": 4.459540366097744e-05,
+      "loss": 0.4344,
+      "step": 1952
+    },
+    {
+      "epoch": 2.5867549668874172,
+      "grad_norm": 0.16249427340971714,
+      "learning_rate": 4.455865408253131e-05,
+      "loss": 0.4343,
+      "step": 1953
+    },
+    {
+      "epoch": 2.5880794701986756,
+      "grad_norm": 0.1289324006443511,
+      "learning_rate": 4.45219006051568e-05,
+      "loss": 0.4343,
+      "step": 1954
+    },
+    {
+      "epoch": 2.589403973509934,
+      "grad_norm": 0.12643159449715646,
+      "learning_rate": 4.448514326028844e-05,
+      "loss": 0.4329,
+      "step": 1955
+    },
+    {
+      "epoch": 2.590728476821192,
+      "grad_norm": 0.1360872170361341,
+      "learning_rate": 4.444838207936404e-05,
+      "loss": 0.4336,
+      "step": 1956
+    },
+    {
+      "epoch": 2.5920529801324506,
+      "grad_norm": 0.14434754155352675,
+      "learning_rate": 4.441161709382477e-05,
+      "loss": 0.4325,
+      "step": 1957
+    },
+    {
+      "epoch": 2.5933774834437084,
+      "grad_norm": 0.13569597920577034,
+      "learning_rate": 4.437484833511499e-05,
+      "loss": 0.4334,
+      "step": 1958
+    },
+    {
+      "epoch": 2.5947019867549668,
+      "grad_norm": 0.12739433333776953,
+      "learning_rate": 4.4338075834682285e-05,
+      "loss": 0.4343,
+      "step": 1959
+    },
+    {
+      "epoch": 2.596026490066225,
+      "grad_norm": 0.11842749284813443,
+      "learning_rate": 4.430129962397749e-05,
+      "loss": 0.4349,
+      "step": 1960
+    },
+    {
+      "epoch": 2.5973509933774834,
+      "grad_norm": 0.11934114139374974,
+      "learning_rate": 4.426451973445459e-05,
+      "loss": 0.4354,
+      "step": 1961
+    },
+    {
+      "epoch": 2.5986754966887418,
+      "grad_norm": 0.1113157766056704,
+      "learning_rate": 4.422773619757069e-05,
+      "loss": 0.433,
+      "step": 1962
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.10624409599693035,
+      "learning_rate": 4.419094904478603e-05,
+      "loss": 0.4391,
+      "step": 1963
+    },
+    {
+      "epoch": 2.6013245033112584,
+      "grad_norm": 0.13110385466987032,
+      "learning_rate": 4.4154158307563975e-05,
+      "loss": 0.43,
+      "step": 1964
+    },
+    {
+      "epoch": 2.6026490066225163,
+      "grad_norm": 0.13334645491733696,
+      "learning_rate": 4.4117364017370925e-05,
+      "loss": 0.4348,
+      "step": 1965
+    },
+    {
+      "epoch": 2.603973509933775,
+      "grad_norm": 0.14907795448203673,
+      "learning_rate": 4.408056620567629e-05,
+      "loss": 0.4314,
+      "step": 1966
+    },
+    {
+      "epoch": 2.605298013245033,
+      "grad_norm": 0.13575422639129428,
+      "learning_rate": 4.404376490395255e-05,
+      "loss": 0.4309,
+      "step": 1967
+    },
+    {
+      "epoch": 2.6066225165562913,
+      "grad_norm": 0.13644153334140763,
+      "learning_rate": 4.400696014367514e-05,
+      "loss": 0.4323,
+      "step": 1968
+    },
+    {
+      "epoch": 2.6079470198675496,
+      "grad_norm": 0.1789094363171331,
+      "learning_rate": 4.397015195632247e-05,
+      "loss": 0.4328,
+      "step": 1969
+    },
+    {
+      "epoch": 2.609271523178808,
+      "grad_norm": 0.15940545684588467,
+      "learning_rate": 4.393334037337584e-05,
+      "loss": 0.4344,
+      "step": 1970
+    },
+    {
+      "epoch": 2.6105960264900663,
+      "grad_norm": 0.11951583123273661,
+      "learning_rate": 4.389652542631952e-05,
+      "loss": 0.4311,
+      "step": 1971
+    },
+    {
+      "epoch": 2.6119205298013246,
+      "grad_norm": 0.13821659394759508,
+      "learning_rate": 4.385970714664058e-05,
+      "loss": 0.4359,
+      "step": 1972
+    },
+    {
+      "epoch": 2.613245033112583,
+      "grad_norm": 0.1393337286324959,
+      "learning_rate": 4.382288556582901e-05,
+      "loss": 0.4346,
+      "step": 1973
+    },
+    {
+      "epoch": 2.614569536423841,
+      "grad_norm": 0.13416542435924803,
+      "learning_rate": 4.378606071537758e-05,
+      "loss": 0.4326,
+      "step": 1974
+    },
+    {
+      "epoch": 2.6158940397350996,
+      "grad_norm": 0.10496140877190129,
+      "learning_rate": 4.374923262678189e-05,
+      "loss": 0.4306,
+      "step": 1975
+    },
+    {
+      "epoch": 2.6172185430463575,
+      "grad_norm": 0.10014681999846631,
+      "learning_rate": 4.371240133154025e-05,
+      "loss": 0.4314,
+      "step": 1976
+    },
+    {
+      "epoch": 2.618543046357616,
+      "grad_norm": 0.1144526996249267,
+      "learning_rate": 4.3675566861153785e-05,
+      "loss": 0.4357,
+      "step": 1977
+    },
+    {
+      "epoch": 2.619867549668874,
+      "grad_norm": 0.11985498337322023,
+      "learning_rate": 4.363872924712629e-05,
+      "loss": 0.4339,
+      "step": 1978
+    },
+    {
+      "epoch": 2.6211920529801325,
+      "grad_norm": 0.12082513962918313,
+      "learning_rate": 4.3601888520964274e-05,
+      "loss": 0.438,
+      "step": 1979
+    },
+    {
+      "epoch": 2.622516556291391,
+      "grad_norm": 0.1200263744895839,
+      "learning_rate": 4.356504471417687e-05,
+      "loss": 0.4288,
+      "step": 1980
+    },
+    {
+      "epoch": 2.623841059602649,
+      "grad_norm": 0.12635689257550423,
+      "learning_rate": 4.35281978582759e-05,
+      "loss": 0.4312,
+      "step": 1981
+    },
+    {
+      "epoch": 2.6251655629139075,
+      "grad_norm": 0.13232907520332413,
+      "learning_rate": 4.3491347984775755e-05,
+      "loss": 0.4335,
+      "step": 1982
+    },
+    {
+      "epoch": 2.6264900662251653,
+      "grad_norm": 0.12030662769422258,
+      "learning_rate": 4.3454495125193396e-05,
+      "loss": 0.4341,
+      "step": 1983
+    },
+    {
+      "epoch": 2.627814569536424,
+      "grad_norm": 0.1491585158985016,
+      "learning_rate": 4.341763931104839e-05,
+      "loss": 0.4292,
+      "step": 1984
+    },
+    {
+      "epoch": 2.629139072847682,
+      "grad_norm": 0.1294517223092044,
+      "learning_rate": 4.33807805738628e-05,
+      "loss": 0.4359,
+      "step": 1985
+    },
+    {
+      "epoch": 2.6304635761589403,
+      "grad_norm": 0.1114723408853529,
+      "learning_rate": 4.334391894516115e-05,
+      "loss": 0.4319,
+      "step": 1986
+    },
+    {
+      "epoch": 2.6317880794701987,
+      "grad_norm": 0.1277279244926657,
+      "learning_rate": 4.330705445647052e-05,
+      "loss": 0.4327,
+      "step": 1987
+    },
+    {
+      "epoch": 2.633112582781457,
+      "grad_norm": 0.11632427424225715,
+      "learning_rate": 4.327018713932036e-05,
+      "loss": 0.4338,
+      "step": 1988
+    },
+    {
+      "epoch": 2.6344370860927153,
+      "grad_norm": 0.11030377303971069,
+      "learning_rate": 4.32333170252426e-05,
+      "loss": 0.4307,
+      "step": 1989
+    },
+    {
+      "epoch": 2.6357615894039736,
+      "grad_norm": 0.1401417734279874,
+      "learning_rate": 4.31964441457715e-05,
+      "loss": 0.4338,
+      "step": 1990
+    },
+    {
+      "epoch": 2.637086092715232,
+      "grad_norm": 0.12435631449106338,
+      "learning_rate": 4.315956853244375e-05,
+      "loss": 0.4377,
+      "step": 1991
+    },
+    {
+      "epoch": 2.63841059602649,
+      "grad_norm": 0.12065233271768533,
+      "learning_rate": 4.312269021679833e-05,
+      "loss": 0.4356,
+      "step": 1992
+    },
+    {
+      "epoch": 2.639735099337748,
+      "grad_norm": 0.13895611998992377,
+      "learning_rate": 4.3085809230376554e-05,
+      "loss": 0.4342,
+      "step": 1993
+    },
+    {
+      "epoch": 2.6410596026490065,
+      "grad_norm": 0.11565110232561115,
+      "learning_rate": 4.3048925604722004e-05,
+      "loss": 0.4351,
+      "step": 1994
+    },
+    {
+      "epoch": 2.642384105960265,
+      "grad_norm": 0.09022486978202521,
+      "learning_rate": 4.3012039371380546e-05,
+      "loss": 0.4339,
+      "step": 1995
+    },
+    {
+      "epoch": 2.643708609271523,
+      "grad_norm": 0.11551427983113624,
+      "learning_rate": 4.297515056190024e-05,
+      "loss": 0.4342,
+      "step": 1996
+    },
+    {
+      "epoch": 2.6450331125827815,
+      "grad_norm": 0.11366993291649961,
+      "learning_rate": 4.293825920783138e-05,
+      "loss": 0.4292,
+      "step": 1997
+    },
+    {
+      "epoch": 2.64635761589404,
+      "grad_norm": 0.10538102018911824,
+      "learning_rate": 4.2901365340726436e-05,
+      "loss": 0.4335,
+      "step": 1998
+    },
+    {
+      "epoch": 2.647682119205298,
+      "grad_norm": 0.11626136461156437,
+      "learning_rate": 4.286446899213999e-05,
+      "loss": 0.4343,
+      "step": 1999
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "grad_norm": 0.11607675326893058,
+      "learning_rate": 4.28275701936288e-05,
+      "loss": 0.4307,
+      "step": 2000
+    },
+    {
+      "epoch": 2.6503311258278144,
+      "grad_norm": 0.11345252243595624,
+      "learning_rate": 4.2790668976751675e-05,
+      "loss": 0.431,
+      "step": 2001
+    },
+    {
+      "epoch": 2.6516556291390727,
+      "grad_norm": 0.1095357211025071,
+      "learning_rate": 4.275376537306952e-05,
+      "loss": 0.4356,
+      "step": 2002
+    },
+    {
+      "epoch": 2.652980132450331,
+      "grad_norm": 0.10836690839278865,
+      "learning_rate": 4.2716859414145266e-05,
+      "loss": 0.4341,
+      "step": 2003
+    },
+    {
+      "epoch": 2.6543046357615894,
+      "grad_norm": 0.09943547165744339,
+      "learning_rate": 4.2679951131543855e-05,
+      "loss": 0.432,
+      "step": 2004
+    },
+    {
+      "epoch": 2.6556291390728477,
+      "grad_norm": 0.11246544582714542,
+      "learning_rate": 4.2643040556832244e-05,
+      "loss": 0.4367,
+      "step": 2005
+    },
+    {
+      "epoch": 2.656953642384106,
+      "grad_norm": 0.11233626888138099,
+      "learning_rate": 4.260612772157934e-05,
+      "loss": 0.4354,
+      "step": 2006
+    },
+    {
+      "epoch": 2.6582781456953644,
+      "grad_norm": 0.10003219287554832,
+      "learning_rate": 4.256921265735593e-05,
+      "loss": 0.4312,
+      "step": 2007
+    },
+    {
+      "epoch": 2.6596026490066222,
+      "grad_norm": 0.10728320090763335,
+      "learning_rate": 4.2532295395734795e-05,
+      "loss": 0.4319,
+      "step": 2008
+    },
+    {
+      "epoch": 2.660927152317881,
+      "grad_norm": 0.10976174377594604,
+      "learning_rate": 4.2495375968290536e-05,
+      "loss": 0.4361,
+      "step": 2009
+    },
+    {
+      "epoch": 2.662251655629139,
+      "grad_norm": 0.12624750984123964,
+      "learning_rate": 4.245845440659961e-05,
+      "loss": 0.4341,
+      "step": 2010
+    },
+    {
+      "epoch": 2.6635761589403972,
+      "grad_norm": 0.11615015183289755,
+      "learning_rate": 4.242153074224034e-05,
+      "loss": 0.4312,
+      "step": 2011
+    },
+    {
+      "epoch": 2.6649006622516556,
+      "grad_norm": 0.10807888497422707,
+      "learning_rate": 4.23846050067928e-05,
+      "loss": 0.4351,
+      "step": 2012
+    },
+    {
+      "epoch": 2.666225165562914,
+      "grad_norm": 0.10704772863770866,
+      "learning_rate": 4.234767723183886e-05,
+      "loss": 0.4334,
+      "step": 2013
+    },
+    {
+      "epoch": 2.667549668874172,
+      "grad_norm": 0.13707055917385425,
+      "learning_rate": 4.231074744896213e-05,
+      "loss": 0.4351,
+      "step": 2014
+    },
+    {
+      "epoch": 2.6688741721854305,
+      "grad_norm": 0.1085512154072027,
+      "learning_rate": 4.227381568974795e-05,
+      "loss": 0.4323,
+      "step": 2015
+    },
+    {
+      "epoch": 2.670198675496689,
+      "grad_norm": 0.1203737771290948,
+      "learning_rate": 4.223688198578333e-05,
+      "loss": 0.4323,
+      "step": 2016
+    },
+    {
+      "epoch": 2.6715231788079468,
+      "grad_norm": 0.13441715878608498,
+      "learning_rate": 4.2199946368656916e-05,
+      "loss": 0.4307,
+      "step": 2017
+    },
+    {
+      "epoch": 2.6728476821192055,
+      "grad_norm": 0.11888301979665765,
+      "learning_rate": 4.216300886995908e-05,
+      "loss": 0.4354,
+      "step": 2018
+    },
+    {
+      "epoch": 2.6741721854304634,
+      "grad_norm": 0.11674750757805716,
+      "learning_rate": 4.212606952128171e-05,
+      "loss": 0.4322,
+      "step": 2019
+    },
+    {
+      "epoch": 2.6754966887417218,
+      "grad_norm": 0.13223290108889976,
+      "learning_rate": 4.208912835421833e-05,
+      "loss": 0.4344,
+      "step": 2020
+    },
+    {
+      "epoch": 2.67682119205298,
+      "grad_norm": 0.1525452942699337,
+      "learning_rate": 4.2052185400363996e-05,
+      "loss": 0.4324,
+      "step": 2021
+    },
+    {
+      "epoch": 2.6781456953642384,
+      "grad_norm": 0.11031228071641518,
+      "learning_rate": 4.201524069131531e-05,
+      "loss": 0.431,
+      "step": 2022
+    },
+    {
+      "epoch": 2.6794701986754967,
+      "grad_norm": 0.12665168908657032,
+      "learning_rate": 4.197829425867036e-05,
+      "loss": 0.4323,
+      "step": 2023
+    },
+    {
+      "epoch": 2.680794701986755,
+      "grad_norm": 0.15611109477984908,
+      "learning_rate": 4.194134613402869e-05,
+      "loss": 0.4309,
+      "step": 2024
+    },
+    {
+      "epoch": 2.6821192052980134,
+      "grad_norm": 0.11997037437118005,
+      "learning_rate": 4.190439634899136e-05,
+      "loss": 0.434,
+      "step": 2025
+    },
+    {
+      "epoch": 2.6834437086092713,
+      "grad_norm": 0.11759248676447576,
+      "learning_rate": 4.186744493516078e-05,
+      "loss": 0.433,
+      "step": 2026
+    },
+    {
+      "epoch": 2.68476821192053,
+      "grad_norm": 0.1119334157357217,
+      "learning_rate": 4.183049192414076e-05,
+      "loss": 0.4347,
+      "step": 2027
+    },
+    {
+      "epoch": 2.686092715231788,
+      "grad_norm": 0.10086222744424429,
+      "learning_rate": 4.179353734753653e-05,
+      "loss": 0.4351,
+      "step": 2028
+    },
+    {
+      "epoch": 2.6874172185430463,
+      "grad_norm": 0.10331945923985877,
+      "learning_rate": 4.175658123695459e-05,
+      "loss": 0.4325,
+      "step": 2029
+    },
+    {
+      "epoch": 2.6887417218543046,
+      "grad_norm": 0.09388057243445709,
+      "learning_rate": 4.1719623624002805e-05,
+      "loss": 0.4301,
+      "step": 2030
+    },
+    {
+      "epoch": 2.690066225165563,
+      "grad_norm": 0.09812639920707002,
+      "learning_rate": 4.168266454029028e-05,
+      "loss": 0.4324,
+      "step": 2031
+    },
+    {
+      "epoch": 2.6913907284768213,
+      "grad_norm": 0.0968423254986005,
+      "learning_rate": 4.1645704017427436e-05,
+      "loss": 0.4385,
+      "step": 2032
+    },
+    {
+      "epoch": 2.6927152317880796,
+      "grad_norm": 0.1030292471057085,
+      "learning_rate": 4.1608742087025874e-05,
+      "loss": 0.4376,
+      "step": 2033
+    },
+    {
+      "epoch": 2.694039735099338,
+      "grad_norm": 0.10198809746429072,
+      "learning_rate": 4.157177878069841e-05,
+      "loss": 0.4291,
+      "step": 2034
+    },
+    {
+      "epoch": 2.695364238410596,
+      "grad_norm": 0.10589719386528834,
+      "learning_rate": 4.153481413005905e-05,
+      "loss": 0.4347,
+      "step": 2035
+    },
+    {
+      "epoch": 2.6966887417218546,
+      "grad_norm": 0.10960363376151648,
+      "learning_rate": 4.149784816672295e-05,
+      "loss": 0.4337,
+      "step": 2036
+    },
+    {
+      "epoch": 2.6980132450331125,
+      "grad_norm": 0.11742435671754742,
+      "learning_rate": 4.1460880922306367e-05,
+      "loss": 0.4326,
+      "step": 2037
+    },
+    {
+      "epoch": 2.699337748344371,
+      "grad_norm": 0.10934346835979068,
+      "learning_rate": 4.1423912428426664e-05,
+      "loss": 0.4354,
+      "step": 2038
+    },
+    {
+      "epoch": 2.700662251655629,
+      "grad_norm": 0.11746722125546423,
+      "learning_rate": 4.13869427167023e-05,
+      "loss": 0.4311,
+      "step": 2039
+    },
+    {
+      "epoch": 2.7019867549668874,
+      "grad_norm": 0.12517434006450562,
+      "learning_rate": 4.134997181875273e-05,
+      "loss": 0.437,
+      "step": 2040
+    },
+    {
+      "epoch": 2.703311258278146,
+      "grad_norm": 0.11555804106345016,
+      "learning_rate": 4.131299976619845e-05,
+      "loss": 0.4343,
+      "step": 2041
+    },
+    {
+      "epoch": 2.704635761589404,
+      "grad_norm": 0.11243068558558453,
+      "learning_rate": 4.127602659066095e-05,
+      "loss": 0.4318,
+      "step": 2042
+    },
+    {
+      "epoch": 2.7059602649006624,
+      "grad_norm": 0.11339874919774044,
+      "learning_rate": 4.123905232376267e-05,
+      "loss": 0.4326,
+      "step": 2043
+    },
+    {
+      "epoch": 2.7072847682119203,
+      "grad_norm": 0.10470179104100809,
+      "learning_rate": 4.1202076997126966e-05,
+      "loss": 0.429,
+      "step": 2044
+    },
+    {
+      "epoch": 2.7086092715231787,
+      "grad_norm": 0.11701520941080885,
+      "learning_rate": 4.116510064237813e-05,
+      "loss": 0.4372,
+      "step": 2045
+    },
+    {
+      "epoch": 2.709933774834437,
+      "grad_norm": 0.11518347454785498,
+      "learning_rate": 4.1128123291141324e-05,
+      "loss": 0.4336,
+      "step": 2046
+    },
+    {
+      "epoch": 2.7112582781456953,
+      "grad_norm": 0.11650050638712772,
+      "learning_rate": 4.109114497504256e-05,
+      "loss": 0.432,
+      "step": 2047
+    },
+    {
+      "epoch": 2.7125827814569536,
+      "grad_norm": 0.10372773867280456,
+      "learning_rate": 4.1054165725708646e-05,
+      "loss": 0.4354,
+      "step": 2048
+    },
+    {
+      "epoch": 2.713907284768212,
+      "grad_norm": 0.10062528095862662,
+      "learning_rate": 4.1017185574767254e-05,
+      "loss": 0.4342,
+      "step": 2049
+    },
+    {
+      "epoch": 2.7152317880794703,
+      "grad_norm": 0.09522274961660188,
+      "learning_rate": 4.098020455384678e-05,
+      "loss": 0.4349,
+      "step": 2050
+    },
+    {
+      "epoch": 2.7165562913907286,
+      "grad_norm": 0.1098719962263817,
+      "learning_rate": 4.0943222694576346e-05,
+      "loss": 0.4311,
+      "step": 2051
+    },
+    {
+      "epoch": 2.717880794701987,
+      "grad_norm": 0.11269246740630508,
+      "learning_rate": 4.090624002858586e-05,
+      "loss": 0.4354,
+      "step": 2052
+    },
+    {
+      "epoch": 2.719205298013245,
+      "grad_norm": 0.1275305346890819,
+      "learning_rate": 4.086925658750585e-05,
+      "loss": 0.4299,
+      "step": 2053
+    },
+    {
+      "epoch": 2.720529801324503,
+      "grad_norm": 0.10693388959422007,
+      "learning_rate": 4.083227240296755e-05,
+      "loss": 0.4357,
+      "step": 2054
+    },
+    {
+      "epoch": 2.7218543046357615,
+      "grad_norm": 0.10918194851020067,
+      "learning_rate": 4.0795287506602806e-05,
+      "loss": 0.4317,
+      "step": 2055
+    },
+    {
+      "epoch": 2.72317880794702,
+      "grad_norm": 0.12608750266333257,
+      "learning_rate": 4.075830193004409e-05,
+      "loss": 0.4332,
+      "step": 2056
+    },
+    {
+      "epoch": 2.724503311258278,
+      "grad_norm": 0.1272921850349737,
+      "learning_rate": 4.072131570492446e-05,
+      "loss": 0.4352,
+      "step": 2057
+    },
+    {
+      "epoch": 2.7258278145695365,
+      "grad_norm": 0.10185855717563613,
+      "learning_rate": 4.068432886287748e-05,
+      "loss": 0.4332,
+      "step": 2058
+    },
+    {
+      "epoch": 2.727152317880795,
+      "grad_norm": 0.10313922235572927,
+      "learning_rate": 4.0647341435537306e-05,
+      "loss": 0.4339,
+      "step": 2059
+    },
+    {
+      "epoch": 2.7284768211920527,
+      "grad_norm": 0.09640284514615353,
+      "learning_rate": 4.0610353454538565e-05,
+      "loss": 0.4333,
+      "step": 2060
+    },
+    {
+      "epoch": 2.7298013245033115,
+      "grad_norm": 0.12218149895690432,
+      "learning_rate": 4.057336495151636e-05,
+      "loss": 0.4331,
+      "step": 2061
+    },
+    {
+      "epoch": 2.7311258278145694,
+      "grad_norm": 0.10876981445204967,
+      "learning_rate": 4.053637595810622e-05,
+      "loss": 0.4363,
+      "step": 2062
+    },
+    {
+      "epoch": 2.7324503311258277,
+      "grad_norm": 0.1046996253470179,
+      "learning_rate": 4.0499386505944145e-05,
+      "loss": 0.4315,
+      "step": 2063
+    },
+    {
+      "epoch": 2.733774834437086,
+      "grad_norm": 0.09634390712139536,
+      "learning_rate": 4.0462396626666474e-05,
+      "loss": 0.4355,
+      "step": 2064
+    },
+    {
+      "epoch": 2.7350993377483444,
+      "grad_norm": 0.11232855527231764,
+      "learning_rate": 4.042540635190993e-05,
+      "loss": 0.434,
+      "step": 2065
+    },
+    {
+      "epoch": 2.7364238410596027,
+      "grad_norm": 0.10472789112174005,
+      "learning_rate": 4.03884157133116e-05,
+      "loss": 0.4332,
+      "step": 2066
+    },
+    {
+      "epoch": 2.737748344370861,
+      "grad_norm": 0.09248681390871626,
+      "learning_rate": 4.0351424742508845e-05,
+      "loss": 0.4331,
+      "step": 2067
+    },
+    {
+      "epoch": 2.7390728476821193,
+      "grad_norm": 0.09100139626703747,
+      "learning_rate": 4.031443347113932e-05,
+      "loss": 0.4363,
+      "step": 2068
+    },
+    {
+      "epoch": 2.7403973509933772,
+      "grad_norm": 0.09842510668289631,
+      "learning_rate": 4.027744193084095e-05,
+      "loss": 0.433,
+      "step": 2069
+    },
+    {
+      "epoch": 2.741721854304636,
+      "grad_norm": 0.10550415133522469,
+      "learning_rate": 4.024045015325189e-05,
+      "loss": 0.4327,
+      "step": 2070
+    },
+    {
+      "epoch": 2.743046357615894,
+      "grad_norm": 0.12776932480584233,
+      "learning_rate": 4.0203458170010464e-05,
+      "loss": 0.4353,
+      "step": 2071
+    },
+    {
+      "epoch": 2.744370860927152,
+      "grad_norm": 0.09500591372768596,
+      "learning_rate": 4.016646601275521e-05,
+      "loss": 0.4353,
+      "step": 2072
+    },
+    {
+      "epoch": 2.7456953642384105,
+      "grad_norm": 0.14721013977475383,
+      "learning_rate": 4.012947371312481e-05,
+      "loss": 0.4316,
+      "step": 2073
+    },
+    {
+      "epoch": 2.747019867549669,
+      "grad_norm": 0.0985809708398492,
+      "learning_rate": 4.009248130275806e-05,
+      "loss": 0.4337,
+      "step": 2074
+    },
+    {
+      "epoch": 2.748344370860927,
+      "grad_norm": 0.13070414888059986,
+      "learning_rate": 4.005548881329384e-05,
+      "loss": 0.4328,
+      "step": 2075
+    },
+    {
+      "epoch": 2.7496688741721855,
+      "grad_norm": 0.13114303392144527,
+      "learning_rate": 4.0018496276371115e-05,
+      "loss": 0.4357,
+      "step": 2076
+    },
+    {
+      "epoch": 2.750993377483444,
+      "grad_norm": 0.1038558394283418,
+      "learning_rate": 3.99815037236289e-05,
+      "loss": 0.4313,
+      "step": 2077
+    },
+    {
+      "epoch": 2.7523178807947017,
+      "grad_norm": 0.10490870371128908,
+      "learning_rate": 3.994451118670617e-05,
+      "loss": 0.4325,
+      "step": 2078
+    },
+    {
+      "epoch": 2.7536423841059605,
+      "grad_norm": 0.10778158820076435,
+      "learning_rate": 3.990751869724195e-05,
+      "loss": 0.435,
+      "step": 2079
+    },
+    {
+      "epoch": 2.7549668874172184,
+      "grad_norm": 0.10616560526299287,
+      "learning_rate": 3.9870526286875194e-05,
+      "loss": 0.4336,
+      "step": 2080
+    },
+    {
+      "epoch": 2.7562913907284767,
+      "grad_norm": 0.09924009922209948,
+      "learning_rate": 3.98335339872448e-05,
+      "loss": 0.4329,
+      "step": 2081
+    },
+    {
+      "epoch": 2.757615894039735,
+      "grad_norm": 0.09638503255800969,
+      "learning_rate": 3.9796541829989556e-05,
+      "loss": 0.4326,
+      "step": 2082
+    },
+    {
+      "epoch": 2.7589403973509934,
+      "grad_norm": 0.09530399902147312,
+      "learning_rate": 3.9759549846748126e-05,
+      "loss": 0.4344,
+      "step": 2083
+    },
+    {
+      "epoch": 2.7602649006622517,
+      "grad_norm": 0.10042165975941354,
+      "learning_rate": 3.972255806915907e-05,
+      "loss": 0.4319,
+      "step": 2084
+    },
+    {
+      "epoch": 2.76158940397351,
+      "grad_norm": 0.1078094213914752,
+      "learning_rate": 3.968556652886069e-05,
+      "loss": 0.4343,
+      "step": 2085
+    },
+    {
+      "epoch": 2.7629139072847684,
+      "grad_norm": 0.10646084161421088,
+      "learning_rate": 3.964857525749117e-05,
+      "loss": 0.4335,
+      "step": 2086
+    },
+    {
+      "epoch": 2.7642384105960263,
+      "grad_norm": 0.11806443804473468,
+      "learning_rate": 3.9611584286688406e-05,
+      "loss": 0.4289,
+      "step": 2087
+    },
+    {
+      "epoch": 2.765562913907285,
+      "grad_norm": 0.09346057710152406,
+      "learning_rate": 3.957459364809008e-05,
+      "loss": 0.4327,
+      "step": 2088
+    },
+    {
+      "epoch": 2.766887417218543,
+      "grad_norm": 0.1047412837249641,
+      "learning_rate": 3.953760337333353e-05,
+      "loss": 0.4315,
+      "step": 2089
+    },
+    {
+      "epoch": 2.7682119205298013,
+      "grad_norm": 0.10756864470715347,
+      "learning_rate": 3.950061349405587e-05,
+      "loss": 0.4333,
+      "step": 2090
+    },
+    {
+      "epoch": 2.7695364238410596,
+      "grad_norm": 0.08824900727147929,
+      "learning_rate": 3.946362404189378e-05,
+      "loss": 0.4325,
+      "step": 2091
+    },
+    {
+      "epoch": 2.770860927152318,
+      "grad_norm": 0.11694952461986799,
+      "learning_rate": 3.942663504848365e-05,
+      "loss": 0.4287,
+      "step": 2092
+    },
+    {
+      "epoch": 2.7721854304635762,
+      "grad_norm": 0.09620490643891208,
+      "learning_rate": 3.938964654546145e-05,
+      "loss": 0.4358,
+      "step": 2093
+    },
+    {
+      "epoch": 2.7735099337748346,
+      "grad_norm": 0.104298360594762,
+      "learning_rate": 3.93526585644627e-05,
+      "loss": 0.4334,
+      "step": 2094
+    },
+    {
+      "epoch": 2.774834437086093,
+      "grad_norm": 0.10929534924079674,
+      "learning_rate": 3.931567113712253e-05,
+      "loss": 0.4325,
+      "step": 2095
+    },
+    {
+      "epoch": 2.776158940397351,
+      "grad_norm": 0.11144633579520988,
+      "learning_rate": 3.927868429507555e-05,
+      "loss": 0.4329,
+      "step": 2096
+    },
+    {
+      "epoch": 2.777483443708609,
+      "grad_norm": 0.1010496371608269,
+      "learning_rate": 3.924169806995592e-05,
+      "loss": 0.4346,
+      "step": 2097
+    },
+    {
+      "epoch": 2.7788079470198674,
+      "grad_norm": 0.11068319103306176,
+      "learning_rate": 3.9204712493397194e-05,
+      "loss": 0.4309,
+      "step": 2098
+    },
+    {
+      "epoch": 2.7801324503311258,
+      "grad_norm": 0.09813789574406488,
+      "learning_rate": 3.916772759703246e-05,
+      "loss": 0.4363,
+      "step": 2099
+    },
+    {
+      "epoch": 2.781456953642384,
+      "grad_norm": 0.1320540227908325,
+      "learning_rate": 3.913074341249415e-05,
+      "loss": 0.4337,
+      "step": 2100
+    },
+    {
+      "epoch": 2.7827814569536424,
+      "grad_norm": 0.11031030273469454,
+      "learning_rate": 3.9093759971414155e-05,
+      "loss": 0.4318,
+      "step": 2101
+    },
+    {
+      "epoch": 2.7841059602649008,
+      "grad_norm": 0.10280958591599554,
+      "learning_rate": 3.9056777305423654e-05,
+      "loss": 0.4367,
+      "step": 2102
+    },
+    {
+      "epoch": 2.785430463576159,
+      "grad_norm": 0.14815477635987104,
+      "learning_rate": 3.9019795446153234e-05,
+      "loss": 0.4328,
+      "step": 2103
+    },
+    {
+      "epoch": 2.7867549668874174,
+      "grad_norm": 0.1005492720422899,
+      "learning_rate": 3.8982814425232766e-05,
+      "loss": 0.4335,
+      "step": 2104
+    },
+    {
+      "epoch": 2.7880794701986753,
+      "grad_norm": 0.11788657669731817,
+      "learning_rate": 3.894583427429136e-05,
+      "loss": 0.4352,
+      "step": 2105
+    },
+    {
+      "epoch": 2.7894039735099336,
+      "grad_norm": 0.12800390273637297,
+      "learning_rate": 3.890885502495746e-05,
+      "loss": 0.4342,
+      "step": 2106
+    },
+    {
+      "epoch": 2.790728476821192,
+      "grad_norm": 0.1020995111522734,
+      "learning_rate": 3.887187670885868e-05,
+      "loss": 0.4331,
+      "step": 2107
+    },
+    {
+      "epoch": 2.7920529801324503,
+      "grad_norm": 0.10148183446259089,
+      "learning_rate": 3.8834899357621885e-05,
+      "loss": 0.4357,
+      "step": 2108
+    },
+    {
+      "epoch": 2.7933774834437086,
+      "grad_norm": 0.0925723739393206,
+      "learning_rate": 3.879792300287304e-05,
+      "loss": 0.4325,
+      "step": 2109
+    },
+    {
+      "epoch": 2.794701986754967,
+      "grad_norm": 0.09782920889971077,
+      "learning_rate": 3.8760947676237344e-05,
+      "loss": 0.4321,
+      "step": 2110
+    },
+    {
+      "epoch": 2.7960264900662253,
+      "grad_norm": 0.10151326705805695,
+      "learning_rate": 3.872397340933905e-05,
+      "loss": 0.4336,
+      "step": 2111
+    },
+    {
+      "epoch": 2.7973509933774836,
+      "grad_norm": 0.11446927684096478,
+      "learning_rate": 3.868700023380156e-05,
+      "loss": 0.4332,
+      "step": 2112
+    },
+    {
+      "epoch": 2.798675496688742,
+      "grad_norm": 0.09106688384856204,
+      "learning_rate": 3.865002818124729e-05,
+      "loss": 0.4283,
+      "step": 2113
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.12034456278184624,
+      "learning_rate": 3.861305728329772e-05,
+      "loss": 0.431,
+      "step": 2114
+    },
+    {
+      "epoch": 2.801324503311258,
+      "grad_norm": 0.09739652981008376,
+      "learning_rate": 3.8576087571573356e-05,
+      "loss": 0.4323,
+      "step": 2115
+    },
+    {
+      "epoch": 2.8026490066225165,
+      "grad_norm": 0.11505202845417953,
+      "learning_rate": 3.853911907769365e-05,
+      "loss": 0.4342,
+      "step": 2116
+    },
+    {
+      "epoch": 2.803973509933775,
+      "grad_norm": 0.10488952648107612,
+      "learning_rate": 3.850215183327707e-05,
+      "loss": 0.4339,
+      "step": 2117
+    },
+    {
+      "epoch": 2.805298013245033,
+      "grad_norm": 0.10101256664225082,
+      "learning_rate": 3.846518586994096e-05,
+      "loss": 0.435,
+      "step": 2118
+    },
+    {
+      "epoch": 2.8066225165562915,
+      "grad_norm": 0.09950608548930501,
+      "learning_rate": 3.84282212193016e-05,
+      "loss": 0.4292,
+      "step": 2119
+    },
+    {
+      "epoch": 2.80794701986755,
+      "grad_norm": 0.08489465641691937,
+      "learning_rate": 3.839125791297413e-05,
+      "loss": 0.4334,
+      "step": 2120
+    },
+    {
+      "epoch": 2.8092715231788077,
+      "grad_norm": 0.11645847511511001,
+      "learning_rate": 3.835429598257257e-05,
+      "loss": 0.4328,
+      "step": 2121
+    },
+    {
+      "epoch": 2.8105960264900665,
+      "grad_norm": 0.08879521814592503,
+      "learning_rate": 3.831733545970972e-05,
+      "loss": 0.4347,
+      "step": 2122
+    },
+    {
+      "epoch": 2.8119205298013243,
+      "grad_norm": 0.09812673691808318,
+      "learning_rate": 3.828037637599721e-05,
+      "loss": 0.4308,
+      "step": 2123
+    },
+    {
+      "epoch": 2.8132450331125827,
+      "grad_norm": 0.10565094745351382,
+      "learning_rate": 3.824341876304543e-05,
+      "loss": 0.4331,
+      "step": 2124
+    },
+    {
+      "epoch": 2.814569536423841,
+      "grad_norm": 0.10236745133424366,
+      "learning_rate": 3.8206462652463486e-05,
+      "loss": 0.4313,
+      "step": 2125
+    },
+    {
+      "epoch": 2.8158940397350993,
+      "grad_norm": 0.10173632869761708,
+      "learning_rate": 3.8169508075859244e-05,
+      "loss": 0.435,
+      "step": 2126
+    },
+    {
+      "epoch": 2.8172185430463577,
+      "grad_norm": 0.10356274409635098,
+      "learning_rate": 3.813255506483923e-05,
+      "loss": 0.4338,
+      "step": 2127
+    },
+    {
+      "epoch": 2.818543046357616,
+      "grad_norm": 0.12235829693501114,
+      "learning_rate": 3.8095603651008656e-05,
+      "loss": 0.4335,
+      "step": 2128
+    },
+    {
+      "epoch": 2.8198675496688743,
+      "grad_norm": 0.11623030213798687,
+      "learning_rate": 3.805865386597131e-05,
+      "loss": 0.435,
+      "step": 2129
+    },
+    {
+      "epoch": 2.821192052980132,
+      "grad_norm": 0.10115408402946707,
+      "learning_rate": 3.8021705741329655e-05,
+      "loss": 0.4332,
+      "step": 2130
+    },
+    {
+      "epoch": 2.822516556291391,
+      "grad_norm": 0.11280599862622588,
+      "learning_rate": 3.7984759308684695e-05,
+      "loss": 0.4319,
+      "step": 2131
+    },
+    {
+      "epoch": 2.823841059602649,
+      "grad_norm": 0.1049601374882309,
+      "learning_rate": 3.794781459963601e-05,
+      "loss": 0.4329,
+      "step": 2132
+    },
+    {
+      "epoch": 2.825165562913907,
+      "grad_norm": 0.10347977798621824,
+      "learning_rate": 3.7910871645781684e-05,
+      "loss": 0.4328,
+      "step": 2133
+    },
+    {
+      "epoch": 2.8264900662251655,
+      "grad_norm": 0.11985251737524,
+      "learning_rate": 3.7873930478718294e-05,
+      "loss": 0.4302,
+      "step": 2134
+    },
+    {
+      "epoch": 2.827814569536424,
+      "grad_norm": 0.11608113528170241,
+      "learning_rate": 3.783699113004094e-05,
+      "loss": 0.4316,
+      "step": 2135
+    },
+    {
+      "epoch": 2.829139072847682,
+      "grad_norm": 0.1147254836937296,
+      "learning_rate": 3.780005363134309e-05,
+      "loss": 0.4356,
+      "step": 2136
+    },
+    {
+      "epoch": 2.8304635761589405,
+      "grad_norm": 0.1269284727656141,
+      "learning_rate": 3.7763118014216696e-05,
+      "loss": 0.434,
+      "step": 2137
+    },
+    {
+      "epoch": 2.831788079470199,
+      "grad_norm": 0.10779757080009938,
+      "learning_rate": 3.772618431025205e-05,
+      "loss": 0.4348,
+      "step": 2138
+    },
+    {
+      "epoch": 2.8331125827814567,
+      "grad_norm": 0.12151164206811907,
+      "learning_rate": 3.7689252551037874e-05,
+      "loss": 0.4304,
+      "step": 2139
+    },
+    {
+      "epoch": 2.8344370860927155,
+      "grad_norm": 0.10998308018124178,
+      "learning_rate": 3.765232276816115e-05,
+      "loss": 0.4329,
+      "step": 2140
+    },
+    {
+      "epoch": 2.8357615894039734,
+      "grad_norm": 0.10586557030105485,
+      "learning_rate": 3.761539499320721e-05,
+      "loss": 0.4335,
+      "step": 2141
+    },
+    {
+      "epoch": 2.8370860927152317,
+      "grad_norm": 0.11312567884172808,
+      "learning_rate": 3.757846925775967e-05,
+      "loss": 0.4339,
+      "step": 2142
+    },
+    {
+      "epoch": 2.83841059602649,
+      "grad_norm": 0.12462539429796252,
+      "learning_rate": 3.75415455934004e-05,
+      "loss": 0.433,
+      "step": 2143
+    },
+    {
+      "epoch": 2.8397350993377484,
+      "grad_norm": 0.12193021522328842,
+      "learning_rate": 3.7504624031709484e-05,
+      "loss": 0.4306,
+      "step": 2144
+    },
+    {
+      "epoch": 2.8410596026490067,
+      "grad_norm": 0.10472294680505106,
+      "learning_rate": 3.746770460426522e-05,
+      "loss": 0.433,
+      "step": 2145
+    },
+    {
+      "epoch": 2.842384105960265,
+      "grad_norm": 0.091934546262177,
+      "learning_rate": 3.743078734264408e-05,
+      "loss": 0.4306,
+      "step": 2146
+    },
+    {
+      "epoch": 2.8437086092715234,
+      "grad_norm": 0.10681061052615858,
+      "learning_rate": 3.739387227842067e-05,
+      "loss": 0.4302,
+      "step": 2147
+    },
+    {
+      "epoch": 2.8450331125827812,
+      "grad_norm": 0.10794604058849286,
+      "learning_rate": 3.735695944316776e-05,
+      "loss": 0.4331,
+      "step": 2148
+    },
+    {
+      "epoch": 2.8463576158940396,
+      "grad_norm": 0.10080470883626316,
+      "learning_rate": 3.7320048868456145e-05,
+      "loss": 0.4325,
+      "step": 2149
+    },
+    {
+      "epoch": 2.847682119205298,
+      "grad_norm": 0.12382313251859059,
+      "learning_rate": 3.728314058585474e-05,
+      "loss": 0.4307,
+      "step": 2150
+    },
+    {
+      "epoch": 2.8490066225165562,
+      "grad_norm": 0.10498762340646255,
+      "learning_rate": 3.724623462693048e-05,
+      "loss": 0.4346,
+      "step": 2151
+    },
+    {
+      "epoch": 2.8503311258278146,
+      "grad_norm": 0.13391440852675648,
+      "learning_rate": 3.720933102324834e-05,
+      "loss": 0.4326,
+      "step": 2152
+    },
+    {
+      "epoch": 2.851655629139073,
+      "grad_norm": 0.11754881597956075,
+      "learning_rate": 3.717242980637121e-05,
+      "loss": 0.4336,
+      "step": 2153
+    },
+    {
+      "epoch": 2.852980132450331,
+      "grad_norm": 0.11123142257528643,
+      "learning_rate": 3.7135531007860014e-05,
+      "loss": 0.4322,
+      "step": 2154
+    },
+    {
+      "epoch": 2.8543046357615895,
+      "grad_norm": 0.09958487817682031,
+      "learning_rate": 3.7098634659273584e-05,
+      "loss": 0.433,
+      "step": 2155
+    },
+    {
+      "epoch": 2.855629139072848,
+      "grad_norm": 0.1263714629701585,
+      "learning_rate": 3.7061740792168625e-05,
+      "loss": 0.4319,
+      "step": 2156
+    },
+    {
+      "epoch": 2.8569536423841058,
+      "grad_norm": 0.09715828324748574,
+      "learning_rate": 3.702484943809977e-05,
+      "loss": 0.4305,
+      "step": 2157
+    },
+    {
+      "epoch": 2.858278145695364,
+      "grad_norm": 0.11820363925777144,
+      "learning_rate": 3.698796062861947e-05,
+      "loss": 0.4301,
+      "step": 2158
+    },
+    {
+      "epoch": 2.8596026490066224,
+      "grad_norm": 0.09575030249809552,
+      "learning_rate": 3.695107439527801e-05,
+      "loss": 0.4321,
+      "step": 2159
+    },
+    {
+      "epoch": 2.8609271523178808,
+      "grad_norm": 0.0993138334917628,
+      "learning_rate": 3.691419076962345e-05,
+      "loss": 0.4299,
+      "step": 2160
+    },
+    {
+      "epoch": 2.862251655629139,
+      "grad_norm": 0.093043045817249,
+      "learning_rate": 3.687730978320168e-05,
+      "loss": 0.4351,
+      "step": 2161
+    },
+    {
+      "epoch": 2.8635761589403974,
+      "grad_norm": 0.09635280610725123,
+      "learning_rate": 3.684043146755625e-05,
+      "loss": 0.4366,
+      "step": 2162
+    },
+    {
+      "epoch": 2.8649006622516557,
+      "grad_norm": 0.10071895440672987,
+      "learning_rate": 3.6803555854228504e-05,
+      "loss": 0.4333,
+      "step": 2163
+    },
+    {
+      "epoch": 2.866225165562914,
+      "grad_norm": 0.09483187328804973,
+      "learning_rate": 3.676668297475742e-05,
+      "loss": 0.4335,
+      "step": 2164
+    },
+    {
+      "epoch": 2.8675496688741724,
+      "grad_norm": 0.11329814512796528,
+      "learning_rate": 3.6729812860679646e-05,
+      "loss": 0.4356,
+      "step": 2165
+    },
+    {
+      "epoch": 2.8688741721854303,
+      "grad_norm": 0.10760280682395679,
+      "learning_rate": 3.66929455435295e-05,
+      "loss": 0.4347,
+      "step": 2166
+    },
+    {
+      "epoch": 2.8701986754966886,
+      "grad_norm": 0.10001217992400059,
+      "learning_rate": 3.6656081054838855e-05,
+      "loss": 0.432,
+      "step": 2167
+    },
+    {
+      "epoch": 2.871523178807947,
+      "grad_norm": 0.12206674150707972,
+      "learning_rate": 3.661921942613722e-05,
+      "loss": 0.4293,
+      "step": 2168
+    },
+    {
+      "epoch": 2.8728476821192053,
+      "grad_norm": 0.10801691684032066,
+      "learning_rate": 3.658236068895161e-05,
+      "loss": 0.4325,
+      "step": 2169
+    },
+    {
+      "epoch": 2.8741721854304636,
+      "grad_norm": 0.10312753803786512,
+      "learning_rate": 3.654550487480662e-05,
+      "loss": 0.4343,
+      "step": 2170
+    },
+    {
+      "epoch": 2.875496688741722,
+      "grad_norm": 0.12307129580281546,
+      "learning_rate": 3.650865201522425e-05,
+      "loss": 0.4311,
+      "step": 2171
+    },
+    {
+      "epoch": 2.8768211920529803,
+      "grad_norm": 0.118369291019685,
+      "learning_rate": 3.6471802141724114e-05,
+      "loss": 0.4311,
+      "step": 2172
+    },
+    {
+      "epoch": 2.878145695364238,
+      "grad_norm": 0.10319874141347897,
+      "learning_rate": 3.643495528582313e-05,
+      "loss": 0.4328,
+      "step": 2173
+    },
+    {
+      "epoch": 2.879470198675497,
+      "grad_norm": 0.11970221059769726,
+      "learning_rate": 3.639811147903574e-05,
+      "loss": 0.4287,
+      "step": 2174
+    },
+    {
+      "epoch": 2.880794701986755,
+      "grad_norm": 0.11055019710406717,
+      "learning_rate": 3.636127075287373e-05,
+      "loss": 0.4295,
+      "step": 2175
+    },
+    {
+      "epoch": 2.882119205298013,
+      "grad_norm": 0.09784934702595076,
+      "learning_rate": 3.632443313884623e-05,
+      "loss": 0.4291,
+      "step": 2176
+    },
+    {
+      "epoch": 2.8834437086092715,
+      "grad_norm": 0.12379912215272597,
+      "learning_rate": 3.628759866845976e-05,
+      "loss": 0.4291,
+      "step": 2177
+    },
+    {
+      "epoch": 2.88476821192053,
+      "grad_norm": 0.10500732026518177,
+      "learning_rate": 3.6250767373218124e-05,
+      "loss": 0.4297,
+      "step": 2178
+    },
+    {
+      "epoch": 2.886092715231788,
+      "grad_norm": 0.09736786192732509,
+      "learning_rate": 3.6213939284622426e-05,
+      "loss": 0.4364,
+      "step": 2179
+    },
+    {
+      "epoch": 2.8874172185430464,
+      "grad_norm": 0.09958747398648904,
+      "learning_rate": 3.617711443417099e-05,
+      "loss": 0.4336,
+      "step": 2180
+    },
+    {
+      "epoch": 2.888741721854305,
+      "grad_norm": 0.11358856665253196,
+      "learning_rate": 3.6140292853359424e-05,
+      "loss": 0.4299,
+      "step": 2181
+    },
+    {
+      "epoch": 2.8900662251655627,
+      "grad_norm": 0.09717003555904807,
+      "learning_rate": 3.610347457368049e-05,
+      "loss": 0.4295,
+      "step": 2182
+    },
+    {
+      "epoch": 2.8913907284768214,
+      "grad_norm": 0.10704396888411855,
+      "learning_rate": 3.6066659626624166e-05,
+      "loss": 0.4311,
+      "step": 2183
+    },
+    {
+      "epoch": 2.8927152317880793,
+      "grad_norm": 0.09363084387333757,
+      "learning_rate": 3.602984804367755e-05,
+      "loss": 0.4326,
+      "step": 2184
+    },
+    {
+      "epoch": 2.8940397350993377,
+      "grad_norm": 0.09212398490864294,
+      "learning_rate": 3.5993039856324864e-05,
+      "loss": 0.4296,
+      "step": 2185
+    },
+    {
+      "epoch": 2.895364238410596,
+      "grad_norm": 0.1111489076988243,
+      "learning_rate": 3.595623509604747e-05,
+      "loss": 0.431,
+      "step": 2186
+    },
+    {
+      "epoch": 2.8966887417218543,
+      "grad_norm": 0.09591487250428558,
+      "learning_rate": 3.5919433794323724e-05,
+      "loss": 0.433,
+      "step": 2187
+    },
+    {
+      "epoch": 2.8980132450331126,
+      "grad_norm": 0.1165879999632941,
+      "learning_rate": 3.5882635982629095e-05,
+      "loss": 0.4312,
+      "step": 2188
+    },
+    {
+      "epoch": 2.899337748344371,
+      "grad_norm": 0.11041219298481941,
+      "learning_rate": 3.5845841692436025e-05,
+      "loss": 0.4305,
+      "step": 2189
+    },
+    {
+      "epoch": 2.9006622516556293,
+      "grad_norm": 0.0988916875129392,
+      "learning_rate": 3.5809050955213975e-05,
+      "loss": 0.4324,
+      "step": 2190
+    },
+    {
+      "epoch": 2.901986754966887,
+      "grad_norm": 0.09180333411891947,
+      "learning_rate": 3.577226380242932e-05,
+      "loss": 0.4332,
+      "step": 2191
+    },
+    {
+      "epoch": 2.903311258278146,
+      "grad_norm": 0.10510447911673357,
+      "learning_rate": 3.573548026554542e-05,
+      "loss": 0.4309,
+      "step": 2192
+    },
+    {
+      "epoch": 2.904635761589404,
+      "grad_norm": 0.11598762855040701,
+      "learning_rate": 3.56987003760225e-05,
+      "loss": 0.4321,
+      "step": 2193
+    },
+    {
+      "epoch": 2.905960264900662,
+      "grad_norm": 0.1187974462887693,
+      "learning_rate": 3.566192416531773e-05,
+      "loss": 0.4298,
+      "step": 2194
+    },
+    {
+      "epoch": 2.9072847682119205,
+      "grad_norm": 0.10333319462023662,
+      "learning_rate": 3.5625151664885036e-05,
+      "loss": 0.4318,
+      "step": 2195
+    },
+    {
+      "epoch": 2.908609271523179,
+      "grad_norm": 0.10562018890871419,
+      "learning_rate": 3.5588382906175244e-05,
+      "loss": 0.435,
+      "step": 2196
+    },
+    {
+      "epoch": 2.909933774834437,
+      "grad_norm": 0.09156763651066957,
+      "learning_rate": 3.555161792063598e-05,
+      "loss": 0.4327,
+      "step": 2197
+    },
+    {
+      "epoch": 2.9112582781456955,
+      "grad_norm": 0.1009720292631543,
+      "learning_rate": 3.551485673971157e-05,
+      "loss": 0.4302,
+      "step": 2198
+    },
+    {
+      "epoch": 2.912582781456954,
+      "grad_norm": 0.09987625958707022,
+      "learning_rate": 3.547809939484321e-05,
+      "loss": 0.43,
+      "step": 2199
+    },
+    {
+      "epoch": 2.9139072847682117,
+      "grad_norm": 0.11215951320855698,
+      "learning_rate": 3.544134591746869e-05,
+      "loss": 0.4348,
+      "step": 2200
+    },
+    {
+      "epoch": 2.91523178807947,
+      "grad_norm": 0.10037281869098284,
+      "learning_rate": 3.5404596339022564e-05,
+      "loss": 0.4317,
+      "step": 2201
+    },
+    {
+      "epoch": 2.9165562913907284,
+      "grad_norm": 0.10619528986082252,
+      "learning_rate": 3.536785069093604e-05,
+      "loss": 0.4295,
+      "step": 2202
+    },
+    {
+      "epoch": 2.9178807947019867,
+      "grad_norm": 0.1000352259311882,
+      "learning_rate": 3.533110900463696e-05,
+      "loss": 0.434,
+      "step": 2203
+    },
+    {
+      "epoch": 2.919205298013245,
+      "grad_norm": 0.0920955582363876,
+      "learning_rate": 3.529437131154975e-05,
+      "loss": 0.4329,
+      "step": 2204
+    },
+    {
+      "epoch": 2.9205298013245033,
+      "grad_norm": 0.09830123954361845,
+      "learning_rate": 3.525763764309547e-05,
+      "loss": 0.4287,
+      "step": 2205
+    },
+    {
+      "epoch": 2.9218543046357617,
+      "grad_norm": 0.0929778307813038,
+      "learning_rate": 3.522090803069172e-05,
+      "loss": 0.4323,
+      "step": 2206
+    },
+    {
+      "epoch": 2.92317880794702,
+      "grad_norm": 0.09352720155448178,
+      "learning_rate": 3.518418250575259e-05,
+      "loss": 0.4296,
+      "step": 2207
+    },
+    {
+      "epoch": 2.9245033112582783,
+      "grad_norm": 0.1061117813142992,
+      "learning_rate": 3.5147461099688746e-05,
+      "loss": 0.4318,
+      "step": 2208
+    },
+    {
+      "epoch": 2.9258278145695362,
+      "grad_norm": 0.09803122829365099,
+      "learning_rate": 3.511074384390727e-05,
+      "loss": 0.4357,
+      "step": 2209
+    },
+    {
+      "epoch": 2.9271523178807946,
+      "grad_norm": 0.08672865145229718,
+      "learning_rate": 3.5074030769811744e-05,
+      "loss": 0.4266,
+      "step": 2210
+    },
+    {
+      "epoch": 2.928476821192053,
+      "grad_norm": 0.10289702448933082,
+      "learning_rate": 3.503732190880211e-05,
+      "loss": 0.431,
+      "step": 2211
+    },
+    {
+      "epoch": 2.929801324503311,
+      "grad_norm": 0.09624004893624487,
+      "learning_rate": 3.500061729227478e-05,
+      "loss": 0.4305,
+      "step": 2212
+    },
+    {
+      "epoch": 2.9311258278145695,
+      "grad_norm": 0.10426718111048348,
+      "learning_rate": 3.496391695162248e-05,
+      "loss": 0.43,
+      "step": 2213
+    },
+    {
+      "epoch": 2.932450331125828,
+      "grad_norm": 0.11297654478789401,
+      "learning_rate": 3.4927220918234315e-05,
+      "loss": 0.4298,
+      "step": 2214
+    },
+    {
+      "epoch": 2.933774834437086,
+      "grad_norm": 0.0973640454281217,
+      "learning_rate": 3.489052922349568e-05,
+      "loss": 0.4338,
+      "step": 2215
+    },
+    {
+      "epoch": 2.9350993377483445,
+      "grad_norm": 0.10641425660472364,
+      "learning_rate": 3.485384189878826e-05,
+      "loss": 0.4342,
+      "step": 2216
+    },
+    {
+      "epoch": 2.936423841059603,
+      "grad_norm": 0.10479684608034827,
+      "learning_rate": 3.481715897549004e-05,
+      "loss": 0.4294,
+      "step": 2217
+    },
+    {
+      "epoch": 2.9377483443708607,
+      "grad_norm": 0.11662625526617704,
+      "learning_rate": 3.478048048497519e-05,
+      "loss": 0.4316,
+      "step": 2218
+    },
+    {
+      "epoch": 2.939072847682119,
+      "grad_norm": 0.1189489649317797,
+      "learning_rate": 3.474380645861411e-05,
+      "loss": 0.4303,
+      "step": 2219
+    },
+    {
+      "epoch": 2.9403973509933774,
+      "grad_norm": 0.1117211857875321,
+      "learning_rate": 3.4707136927773385e-05,
+      "loss": 0.4305,
+      "step": 2220
+    },
+    {
+      "epoch": 2.9417218543046357,
+      "grad_norm": 0.10457546980450882,
+      "learning_rate": 3.467047192381578e-05,
+      "loss": 0.4317,
+      "step": 2221
+    },
+    {
+      "epoch": 2.943046357615894,
+      "grad_norm": 0.10658153301906093,
+      "learning_rate": 3.463381147810013e-05,
+      "loss": 0.4325,
+      "step": 2222
+    },
+    {
+      "epoch": 2.9443708609271524,
+      "grad_norm": 0.09516292271655269,
+      "learning_rate": 3.459715562198141e-05,
+      "loss": 0.4342,
+      "step": 2223
+    },
+    {
+      "epoch": 2.9456953642384107,
+      "grad_norm": 0.10565511459935187,
+      "learning_rate": 3.456050438681066e-05,
+      "loss": 0.4306,
+      "step": 2224
+    },
+    {
+      "epoch": 2.9470198675496686,
+      "grad_norm": 0.10733748480356767,
+      "learning_rate": 3.452385780393499e-05,
+      "loss": 0.4318,
+      "step": 2225
+    },
+    {
+      "epoch": 2.9483443708609274,
+      "grad_norm": 0.09592499111203541,
+      "learning_rate": 3.44872159046975e-05,
+      "loss": 0.4341,
+      "step": 2226
+    },
+    {
+      "epoch": 2.9496688741721853,
+      "grad_norm": 0.10889500182987881,
+      "learning_rate": 3.445057872043728e-05,
+      "loss": 0.4309,
+      "step": 2227
+    },
+    {
+      "epoch": 2.9509933774834436,
+      "grad_norm": 0.11098114796771495,
+      "learning_rate": 3.4413946282489417e-05,
+      "loss": 0.4271,
+      "step": 2228
+    },
+    {
+      "epoch": 2.952317880794702,
+      "grad_norm": 0.09995232765769611,
+      "learning_rate": 3.4377318622184924e-05,
+      "loss": 0.4297,
+      "step": 2229
+    },
+    {
+      "epoch": 2.9536423841059603,
+      "grad_norm": 0.12069235307527422,
+      "learning_rate": 3.434069577085074e-05,
+      "loss": 0.4309,
+      "step": 2230
+    },
+    {
+      "epoch": 2.9549668874172186,
+      "grad_norm": 0.09813339095017806,
+      "learning_rate": 3.430407775980965e-05,
+      "loss": 0.4348,
+      "step": 2231
+    },
+    {
+      "epoch": 2.956291390728477,
+      "grad_norm": 0.09617556681794964,
+      "learning_rate": 3.426746462038036e-05,
+      "loss": 0.4339,
+      "step": 2232
+    },
+    {
+      "epoch": 2.9576158940397352,
+      "grad_norm": 0.1227258901964563,
+      "learning_rate": 3.423085638387734e-05,
+      "loss": 0.4289,
+      "step": 2233
+    },
+    {
+      "epoch": 2.958940397350993,
+      "grad_norm": 0.10735601754505207,
+      "learning_rate": 3.4194253081610964e-05,
+      "loss": 0.4287,
+      "step": 2234
+    },
+    {
+      "epoch": 2.960264900662252,
+      "grad_norm": 0.10023553192272414,
+      "learning_rate": 3.415765474488727e-05,
+      "loss": 0.4327,
+      "step": 2235
+    },
+    {
+      "epoch": 2.96158940397351,
+      "grad_norm": 0.12504784856218845,
+      "learning_rate": 3.412106140500812e-05,
+      "loss": 0.4347,
+      "step": 2236
+    },
+    {
+      "epoch": 2.962913907284768,
+      "grad_norm": 0.10802755418762187,
+      "learning_rate": 3.408447309327111e-05,
+      "loss": 0.4314,
+      "step": 2237
+    },
+    {
+      "epoch": 2.9642384105960264,
+      "grad_norm": 0.12350881494672164,
+      "learning_rate": 3.404788984096948e-05,
+      "loss": 0.4312,
+      "step": 2238
+    },
+    {
+      "epoch": 2.9655629139072848,
+      "grad_norm": 0.10626929661754399,
+      "learning_rate": 3.40113116793922e-05,
+      "loss": 0.4293,
+      "step": 2239
+    },
+    {
+      "epoch": 2.966887417218543,
+      "grad_norm": 0.10868592427722641,
+      "learning_rate": 3.397473863982384e-05,
+      "loss": 0.4325,
+      "step": 2240
+    },
+    {
+      "epoch": 2.9682119205298014,
+      "grad_norm": 0.12785852108328846,
+      "learning_rate": 3.3938170753544645e-05,
+      "loss": 0.4325,
+      "step": 2241
+    },
+    {
+      "epoch": 2.9695364238410598,
+      "grad_norm": 0.08549650345244397,
+      "learning_rate": 3.390160805183038e-05,
+      "loss": 0.4286,
+      "step": 2242
+    },
+    {
+      "epoch": 2.9708609271523176,
+      "grad_norm": 0.12409113874154531,
+      "learning_rate": 3.386505056595245e-05,
+      "loss": 0.4308,
+      "step": 2243
+    },
+    {
+      "epoch": 2.9721854304635764,
+      "grad_norm": 0.10604961941358984,
+      "learning_rate": 3.3828498327177736e-05,
+      "loss": 0.4284,
+      "step": 2244
+    },
+    {
+      "epoch": 2.9735099337748343,
+      "grad_norm": 0.10402844944108959,
+      "learning_rate": 3.3791951366768685e-05,
+      "loss": 0.4363,
+      "step": 2245
+    },
+    {
+      "epoch": 2.9748344370860926,
+      "grad_norm": 0.10184832815404257,
+      "learning_rate": 3.375540971598318e-05,
+      "loss": 0.4332,
+      "step": 2246
+    },
+    {
+      "epoch": 2.976158940397351,
+      "grad_norm": 0.086620235981979,
+      "learning_rate": 3.371887340607458e-05,
+      "loss": 0.432,
+      "step": 2247
+    },
+    {
+      "epoch": 2.9774834437086093,
+      "grad_norm": 0.11542479767654115,
+      "learning_rate": 3.3682342468291715e-05,
+      "loss": 0.4303,
+      "step": 2248
+    },
+    {
+      "epoch": 2.9788079470198676,
+      "grad_norm": 0.10810864140615718,
+      "learning_rate": 3.3645816933878754e-05,
+      "loss": 0.4315,
+      "step": 2249
+    },
+    {
+      "epoch": 2.980132450331126,
+      "grad_norm": 0.10197862317422135,
+      "learning_rate": 3.3609296834075294e-05,
+      "loss": 0.4312,
+      "step": 2250
+    },
+    {
+      "epoch": 2.9814569536423843,
+      "grad_norm": 0.10069928556738235,
+      "learning_rate": 3.357278220011626e-05,
+      "loss": 0.4279,
+      "step": 2251
+    },
+    {
+      "epoch": 2.982781456953642,
+      "grad_norm": 0.09933929742171363,
+      "learning_rate": 3.353627306323193e-05,
+      "loss": 0.4335,
+      "step": 2252
+    },
+    {
+      "epoch": 2.984105960264901,
+      "grad_norm": 0.09867514713286794,
+      "learning_rate": 3.349976945464783e-05,
+      "loss": 0.4304,
+      "step": 2253
+    },
+    {
+      "epoch": 2.985430463576159,
+      "grad_norm": 0.08875679855732531,
+      "learning_rate": 3.346327140558483e-05,
+      "loss": 0.4313,
+      "step": 2254
+    },
+    {
+      "epoch": 2.986754966887417,
+      "grad_norm": 0.0946166664067538,
+      "learning_rate": 3.342677894725895e-05,
+      "loss": 0.4326,
+      "step": 2255
+    },
+    {
+      "epoch": 2.9880794701986755,
+      "grad_norm": 0.0940526010219045,
+      "learning_rate": 3.339029211088153e-05,
+      "loss": 0.4318,
+      "step": 2256
+    },
+    {
+      "epoch": 2.989403973509934,
+      "grad_norm": 0.08607734690936539,
+      "learning_rate": 3.3353810927659025e-05,
+      "loss": 0.4322,
+      "step": 2257
+    },
+    {
+      "epoch": 2.990728476821192,
+      "grad_norm": 0.08296041395919218,
+      "learning_rate": 3.3317335428793085e-05,
+      "loss": 0.4295,
+      "step": 2258
+    },
+    {
+      "epoch": 2.9920529801324505,
+      "grad_norm": 0.08863073130223639,
+      "learning_rate": 3.328086564548051e-05,
+      "loss": 0.4294,
+      "step": 2259
+    },
+    {
+      "epoch": 2.993377483443709,
+      "grad_norm": 0.08813739226048556,
+      "learning_rate": 3.324440160891318e-05,
+      "loss": 0.433,
+      "step": 2260
+    },
+    {
+      "epoch": 2.9947019867549667,
+      "grad_norm": 0.08635492982630057,
+      "learning_rate": 3.320794335027811e-05,
+      "loss": 0.4325,
+      "step": 2261
+    },
+    {
+      "epoch": 2.996026490066225,
+      "grad_norm": 0.09310818301021521,
+      "learning_rate": 3.31714909007573e-05,
+      "loss": 0.4321,
+      "step": 2262
+    },
+    {
+      "epoch": 2.9973509933774833,
+      "grad_norm": 0.08492282918863492,
+      "learning_rate": 3.313504429152785e-05,
+      "loss": 0.4328,
+      "step": 2263
+    },
+    {
+      "epoch": 2.9986754966887417,
+      "grad_norm": 0.1061506032921239,
+      "learning_rate": 3.309860355376182e-05,
+      "loss": 0.4313,
+      "step": 2264
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.08569664381173207,
+      "learning_rate": 3.306216871862629e-05,
+      "loss": 0.4292,
+      "step": 2265
+    },
+    {
+      "epoch": 3.0013245033112583,
+      "grad_norm": 0.15534664193974992,
+      "learning_rate": 3.302573981728326e-05,
+      "loss": 0.3999,
+      "step": 2266
+    },
+    {
+      "epoch": 3.0026490066225167,
+      "grad_norm": 0.29481866324564765,
+      "learning_rate": 3.298931688088964e-05,
+      "loss": 0.4058,
+      "step": 2267
+    },
+    {
+      "epoch": 3.003973509933775,
+      "grad_norm": 0.21645166221106582,
+      "learning_rate": 3.295289994059729e-05,
+      "loss": 0.4098,
+      "step": 2268
+    },
+    {
+      "epoch": 3.0052980132450333,
+      "grad_norm": 0.17084723109777153,
+      "learning_rate": 3.291648902755288e-05,
+      "loss": 0.4037,
+      "step": 2269
+    },
+    {
+      "epoch": 3.006622516556291,
+      "grad_norm": 0.1565872303553559,
+      "learning_rate": 3.2880084172898e-05,
+      "loss": 0.4096,
+      "step": 2270
+    },
+    {
+      "epoch": 3.0079470198675495,
+      "grad_norm": 0.18664904016110662,
+      "learning_rate": 3.284368540776898e-05,
+      "loss": 0.4072,
+      "step": 2271
+    },
+    {
+      "epoch": 3.009271523178808,
+      "grad_norm": 0.1681901843426961,
+      "learning_rate": 3.2807292763296996e-05,
+      "loss": 0.4089,
+      "step": 2272
+    },
+    {
+      "epoch": 3.010596026490066,
+      "grad_norm": 0.14809699855923805,
+      "learning_rate": 3.277090627060796e-05,
+      "loss": 0.4093,
+      "step": 2273
+    },
+    {
+      "epoch": 3.0119205298013245,
+      "grad_norm": 0.1750499783942725,
+      "learning_rate": 3.2734525960822545e-05,
+      "loss": 0.4057,
+      "step": 2274
+    },
+    {
+      "epoch": 3.013245033112583,
+      "grad_norm": 0.15848843233628615,
+      "learning_rate": 3.269815186505612e-05,
+      "loss": 0.4062,
+      "step": 2275
+    },
+    {
+      "epoch": 3.014569536423841,
+      "grad_norm": 0.1399088325038847,
+      "learning_rate": 3.266178401441874e-05,
+      "loss": 0.4052,
+      "step": 2276
+    },
+    {
+      "epoch": 3.0158940397350995,
+      "grad_norm": 0.1901892413145548,
+      "learning_rate": 3.2625422440015125e-05,
+      "loss": 0.4048,
+      "step": 2277
+    },
+    {
+      "epoch": 3.017218543046358,
+      "grad_norm": 0.16980518927423224,
+      "learning_rate": 3.258906717294462e-05,
+      "loss": 0.4027,
+      "step": 2278
+    },
+    {
+      "epoch": 3.0185430463576157,
+      "grad_norm": 0.1374261521306976,
+      "learning_rate": 3.2552718244301175e-05,
+      "loss": 0.4039,
+      "step": 2279
+    },
+    {
+      "epoch": 3.019867549668874,
+      "grad_norm": 0.15050724686614503,
+      "learning_rate": 3.251637568517334e-05,
+      "loss": 0.4026,
+      "step": 2280
+    },
+    {
+      "epoch": 3.0211920529801324,
+      "grad_norm": 0.16367860311024662,
+      "learning_rate": 3.24800395266442e-05,
+      "loss": 0.4061,
+      "step": 2281
+    },
+    {
+      "epoch": 3.0225165562913907,
+      "grad_norm": 0.12783334690161113,
+      "learning_rate": 3.244370979979134e-05,
+      "loss": 0.4044,
+      "step": 2282
+    },
+    {
+      "epoch": 3.023841059602649,
+      "grad_norm": 0.1664540690531511,
+      "learning_rate": 3.24073865356869e-05,
+      "loss": 0.4039,
+      "step": 2283
+    },
+    {
+      "epoch": 3.0251655629139074,
+      "grad_norm": 0.13529300672183475,
+      "learning_rate": 3.237106976539744e-05,
+      "loss": 0.4051,
+      "step": 2284
+    },
+    {
+      "epoch": 3.0264900662251657,
+      "grad_norm": 0.13854672237942603,
+      "learning_rate": 3.2334759519984014e-05,
+      "loss": 0.4026,
+      "step": 2285
+    },
+    {
+      "epoch": 3.027814569536424,
+      "grad_norm": 0.150374501214321,
+      "learning_rate": 3.2298455830502055e-05,
+      "loss": 0.4045,
+      "step": 2286
+    },
+    {
+      "epoch": 3.029139072847682,
+      "grad_norm": 0.11092529811419069,
+      "learning_rate": 3.2262158728001395e-05,
+      "loss": 0.4033,
+      "step": 2287
+    },
+    {
+      "epoch": 3.0304635761589402,
+      "grad_norm": 0.1275456726434029,
+      "learning_rate": 3.222586824352628e-05,
+      "loss": 0.3997,
+      "step": 2288
+    },
+    {
+      "epoch": 3.0317880794701986,
+      "grad_norm": 0.13759167210165402,
+      "learning_rate": 3.21895844081152e-05,
+      "loss": 0.4035,
+      "step": 2289
+    },
+    {
+      "epoch": 3.033112582781457,
+      "grad_norm": 0.12545691917908433,
+      "learning_rate": 3.2153307252801064e-05,
+      "loss": 0.4014,
+      "step": 2290
+    },
+    {
+      "epoch": 3.0344370860927152,
+      "grad_norm": 0.158285914987687,
+      "learning_rate": 3.211703680861099e-05,
+      "loss": 0.4004,
+      "step": 2291
+    },
+    {
+      "epoch": 3.0357615894039736,
+      "grad_norm": 0.09543147140905171,
+      "learning_rate": 3.208077310656641e-05,
+      "loss": 0.4015,
+      "step": 2292
+    },
+    {
+      "epoch": 3.037086092715232,
+      "grad_norm": 0.14260982164417915,
+      "learning_rate": 3.204451617768293e-05,
+      "loss": 0.4032,
+      "step": 2293
+    },
+    {
+      "epoch": 3.03841059602649,
+      "grad_norm": 0.10541459719873689,
+      "learning_rate": 3.2008266052970435e-05,
+      "loss": 0.4053,
+      "step": 2294
+    },
+    {
+      "epoch": 3.0397350993377485,
+      "grad_norm": 0.10487684609841477,
+      "learning_rate": 3.197202276343292e-05,
+      "loss": 0.4021,
+      "step": 2295
+    },
+    {
+      "epoch": 3.0410596026490064,
+      "grad_norm": 0.12764164056999686,
+      "learning_rate": 3.193578634006861e-05,
+      "loss": 0.4068,
+      "step": 2296
+    },
+    {
+      "epoch": 3.0423841059602648,
+      "grad_norm": 0.11327652833127354,
+      "learning_rate": 3.189955681386978e-05,
+      "loss": 0.4029,
+      "step": 2297
+    },
+    {
+      "epoch": 3.043708609271523,
+      "grad_norm": 0.11014122931568608,
+      "learning_rate": 3.1863334215822846e-05,
+      "loss": 0.4033,
+      "step": 2298
+    },
+    {
+      "epoch": 3.0450331125827814,
+      "grad_norm": 0.09312333084232355,
+      "learning_rate": 3.182711857690831e-05,
+      "loss": 0.4069,
+      "step": 2299
+    },
+    {
+      "epoch": 3.0463576158940397,
+      "grad_norm": 0.12652628251827716,
+      "learning_rate": 3.1790909928100674e-05,
+      "loss": 0.4017,
+      "step": 2300
+    },
+    {
+      "epoch": 3.047682119205298,
+      "grad_norm": 0.08916244265648833,
+      "learning_rate": 3.175470830036852e-05,
+      "loss": 0.4068,
+      "step": 2301
+    },
+    {
+      "epoch": 3.0490066225165564,
+      "grad_norm": 0.12489732547893252,
+      "learning_rate": 3.171851372467438e-05,
+      "loss": 0.401,
+      "step": 2302
+    },
+    {
+      "epoch": 3.0503311258278147,
+      "grad_norm": 0.11216239416049933,
+      "learning_rate": 3.1682326231974803e-05,
+      "loss": 0.4037,
+      "step": 2303
+    },
+    {
+      "epoch": 3.0516556291390726,
+      "grad_norm": 0.0816451177708888,
+      "learning_rate": 3.16461458532202e-05,
+      "loss": 0.402,
+      "step": 2304
+    },
+    {
+      "epoch": 3.052980132450331,
+      "grad_norm": 0.10188034346030879,
+      "learning_rate": 3.1609972619355005e-05,
+      "loss": 0.4015,
+      "step": 2305
+    },
+    {
+      "epoch": 3.0543046357615893,
+      "grad_norm": 0.08993549322660951,
+      "learning_rate": 3.157380656131744e-05,
+      "loss": 0.4071,
+      "step": 2306
+    },
+    {
+      "epoch": 3.0556291390728476,
+      "grad_norm": 0.10282804031755058,
+      "learning_rate": 3.153764771003965e-05,
+      "loss": 0.4041,
+      "step": 2307
+    },
+    {
+      "epoch": 3.056953642384106,
+      "grad_norm": 0.09471903355955938,
+      "learning_rate": 3.1501496096447595e-05,
+      "loss": 0.4016,
+      "step": 2308
+    },
+    {
+      "epoch": 3.0582781456953643,
+      "grad_norm": 0.09041747917235014,
+      "learning_rate": 3.1465351751461045e-05,
+      "loss": 0.4,
+      "step": 2309
+    },
+    {
+      "epoch": 3.0596026490066226,
+      "grad_norm": 0.09785695871519903,
+      "learning_rate": 3.1429214705993565e-05,
+      "loss": 0.4033,
+      "step": 2310
+    },
+    {
+      "epoch": 3.060927152317881,
+      "grad_norm": 0.08840144489586671,
+      "learning_rate": 3.1393084990952457e-05,
+      "loss": 0.4044,
+      "step": 2311
+    },
+    {
+      "epoch": 3.0622516556291393,
+      "grad_norm": 0.0806644968256852,
+      "learning_rate": 3.135696263723879e-05,
+      "loss": 0.402,
+      "step": 2312
+    },
+    {
+      "epoch": 3.063576158940397,
+      "grad_norm": 0.09322843205981043,
+      "learning_rate": 3.132084767574728e-05,
+      "loss": 0.4045,
+      "step": 2313
+    },
+    {
+      "epoch": 3.0649006622516555,
+      "grad_norm": 0.08806292399674895,
+      "learning_rate": 3.1284740137366374e-05,
+      "loss": 0.4057,
+      "step": 2314
+    },
+    {
+      "epoch": 3.066225165562914,
+      "grad_norm": 0.09152218364796835,
+      "learning_rate": 3.1248640052978144e-05,
+      "loss": 0.4028,
+      "step": 2315
+    },
+    {
+      "epoch": 3.067549668874172,
+      "grad_norm": 0.08093899964627807,
+      "learning_rate": 3.12125474534583e-05,
+      "loss": 0.402,
+      "step": 2316
+    },
+    {
+      "epoch": 3.0688741721854305,
+      "grad_norm": 0.08778200285684828,
+      "learning_rate": 3.1176462369676123e-05,
+      "loss": 0.4049,
+      "step": 2317
+    },
+    {
+      "epoch": 3.070198675496689,
+      "grad_norm": 0.07938390193403032,
+      "learning_rate": 3.1140384832494495e-05,
+      "loss": 0.4017,
+      "step": 2318
+    },
+    {
+      "epoch": 3.071523178807947,
+      "grad_norm": 0.08815761329214541,
+      "learning_rate": 3.1104314872769845e-05,
+      "loss": 0.4032,
+      "step": 2319
+    },
+    {
+      "epoch": 3.0728476821192054,
+      "grad_norm": 0.08972867123968312,
+      "learning_rate": 3.106825252135209e-05,
+      "loss": 0.4027,
+      "step": 2320
+    },
+    {
+      "epoch": 3.0741721854304638,
+      "grad_norm": 0.08995169593510406,
+      "learning_rate": 3.103219780908467e-05,
+      "loss": 0.4054,
+      "step": 2321
+    },
+    {
+      "epoch": 3.0754966887417217,
+      "grad_norm": 0.10946527522202462,
+      "learning_rate": 3.099615076680447e-05,
+      "loss": 0.4064,
+      "step": 2322
+    },
+    {
+      "epoch": 3.07682119205298,
+      "grad_norm": 0.08278334914095663,
+      "learning_rate": 3.0960111425341855e-05,
+      "loss": 0.4029,
+      "step": 2323
+    },
+    {
+      "epoch": 3.0781456953642383,
+      "grad_norm": 0.09844223721337633,
+      "learning_rate": 3.092407981552054e-05,
+      "loss": 0.4028,
+      "step": 2324
+    },
+    {
+      "epoch": 3.0794701986754967,
+      "grad_norm": 0.09181839544351093,
+      "learning_rate": 3.0888055968157686e-05,
+      "loss": 0.4027,
+      "step": 2325
+    },
+    {
+      "epoch": 3.080794701986755,
+      "grad_norm": 0.09191107546786065,
+      "learning_rate": 3.085203991406377e-05,
+      "loss": 0.4057,
+      "step": 2326
+    },
+    {
+      "epoch": 3.0821192052980133,
+      "grad_norm": 0.09494479993004162,
+      "learning_rate": 3.081603168404265e-05,
+      "loss": 0.4032,
+      "step": 2327
+    },
+    {
+      "epoch": 3.0834437086092716,
+      "grad_norm": 0.10164649364537229,
+      "learning_rate": 3.078003130889145e-05,
+      "loss": 0.3994,
+      "step": 2328
+    },
+    {
+      "epoch": 3.08476821192053,
+      "grad_norm": 0.08107204284448558,
+      "learning_rate": 3.074403881940058e-05,
+      "loss": 0.4061,
+      "step": 2329
+    },
+    {
+      "epoch": 3.0860927152317883,
+      "grad_norm": 0.10343316621550787,
+      "learning_rate": 3.070805424635375e-05,
+      "loss": 0.405,
+      "step": 2330
+    },
+    {
+      "epoch": 3.087417218543046,
+      "grad_norm": 0.1100562755455789,
+      "learning_rate": 3.067207762052783e-05,
+      "loss": 0.4033,
+      "step": 2331
+    },
+    {
+      "epoch": 3.0887417218543045,
+      "grad_norm": 0.10549308477550641,
+      "learning_rate": 3.063610897269297e-05,
+      "loss": 0.4038,
+      "step": 2332
+    },
+    {
+      "epoch": 3.090066225165563,
+      "grad_norm": 0.11083496422321341,
+      "learning_rate": 3.060014833361243e-05,
+      "loss": 0.4003,
+      "step": 2333
+    },
+    {
+      "epoch": 3.091390728476821,
+      "grad_norm": 0.11736638121310733,
+      "learning_rate": 3.056419573404265e-05,
+      "loss": 0.407,
+      "step": 2334
+    },
+    {
+      "epoch": 3.0927152317880795,
+      "grad_norm": 0.10319809795296618,
+      "learning_rate": 3.052825120473319e-05,
+      "loss": 0.4049,
+      "step": 2335
+    },
+    {
+      "epoch": 3.094039735099338,
+      "grad_norm": 0.09495061209065422,
+      "learning_rate": 3.0492314776426735e-05,
+      "loss": 0.4023,
+      "step": 2336
+    },
+    {
+      "epoch": 3.095364238410596,
+      "grad_norm": 0.10130048984188025,
+      "learning_rate": 3.0456386479858978e-05,
+      "loss": 0.4016,
+      "step": 2337
+    },
+    {
+      "epoch": 3.0966887417218545,
+      "grad_norm": 0.09351666917867457,
+      "learning_rate": 3.0420466345758705e-05,
+      "loss": 0.4012,
+      "step": 2338
+    },
+    {
+      "epoch": 3.0980132450331124,
+      "grad_norm": 0.10054963832724513,
+      "learning_rate": 3.0384554404847733e-05,
+      "loss": 0.406,
+      "step": 2339
+    },
+    {
+      "epoch": 3.0993377483443707,
+      "grad_norm": 0.11085886185996868,
+      "learning_rate": 3.034865068784081e-05,
+      "loss": 0.4051,
+      "step": 2340
+    },
+    {
+      "epoch": 3.100662251655629,
+      "grad_norm": 0.10807997334630987,
+      "learning_rate": 3.031275522544571e-05,
+      "loss": 0.4055,
+      "step": 2341
+    },
+    {
+      "epoch": 3.1019867549668874,
+      "grad_norm": 0.10574374576832271,
+      "learning_rate": 3.0276868048363122e-05,
+      "loss": 0.4052,
+      "step": 2342
+    },
+    {
+      "epoch": 3.1033112582781457,
+      "grad_norm": 0.09598601474218471,
+      "learning_rate": 3.0240989187286666e-05,
+      "loss": 0.4063,
+      "step": 2343
+    },
+    {
+      "epoch": 3.104635761589404,
+      "grad_norm": 0.10262876182004757,
+      "learning_rate": 3.0205118672902807e-05,
+      "loss": 0.4031,
+      "step": 2344
+    },
+    {
+      "epoch": 3.1059602649006623,
+      "grad_norm": 0.07963044764600813,
+      "learning_rate": 3.016925653589093e-05,
+      "loss": 0.4009,
+      "step": 2345
+    },
+    {
+      "epoch": 3.1072847682119207,
+      "grad_norm": 0.10203774403887544,
+      "learning_rate": 3.0133402806923197e-05,
+      "loss": 0.4022,
+      "step": 2346
+    },
+    {
+      "epoch": 3.108609271523179,
+      "grad_norm": 0.09604002375220365,
+      "learning_rate": 3.0097557516664643e-05,
+      "loss": 0.4047,
+      "step": 2347
+    },
+    {
+      "epoch": 3.109933774834437,
+      "grad_norm": 0.0897705999951039,
+      "learning_rate": 3.0061720695773017e-05,
+      "loss": 0.4051,
+      "step": 2348
+    },
+    {
+      "epoch": 3.111258278145695,
+      "grad_norm": 0.10201453943593268,
+      "learning_rate": 3.0025892374898857e-05,
+      "loss": 0.4058,
+      "step": 2349
+    },
+    {
+      "epoch": 3.1125827814569536,
+      "grad_norm": 0.09775598974741936,
+      "learning_rate": 2.999007258468546e-05,
+      "loss": 0.4068,
+      "step": 2350
+    },
+    {
+      "epoch": 3.113907284768212,
+      "grad_norm": 0.10312282159622557,
+      "learning_rate": 2.9954261355768765e-05,
+      "loss": 0.4027,
+      "step": 2351
+    },
+    {
+      "epoch": 3.11523178807947,
+      "grad_norm": 0.0985799766939854,
+      "learning_rate": 2.991845871877744e-05,
+      "loss": 0.4048,
+      "step": 2352
+    },
+    {
+      "epoch": 3.1165562913907285,
+      "grad_norm": 0.10636619368666118,
+      "learning_rate": 2.988266470433277e-05,
+      "loss": 0.4043,
+      "step": 2353
+    },
+    {
+      "epoch": 3.117880794701987,
+      "grad_norm": 0.09266580383345042,
+      "learning_rate": 2.9846879343048713e-05,
+      "loss": 0.4061,
+      "step": 2354
+    },
+    {
+      "epoch": 3.119205298013245,
+      "grad_norm": 0.09780072979807662,
+      "learning_rate": 2.9811102665531756e-05,
+      "loss": 0.4051,
+      "step": 2355
+    },
+    {
+      "epoch": 3.120529801324503,
+      "grad_norm": 0.09888925591705774,
+      "learning_rate": 2.9775334702381017e-05,
+      "loss": 0.4016,
+      "step": 2356
+    },
+    {
+      "epoch": 3.1218543046357614,
+      "grad_norm": 0.10733259307409741,
+      "learning_rate": 2.9739575484188133e-05,
+      "loss": 0.4049,
+      "step": 2357
+    },
+    {
+      "epoch": 3.1231788079470197,
+      "grad_norm": 0.11605373492251035,
+      "learning_rate": 2.9703825041537286e-05,
+      "loss": 0.4048,
+      "step": 2358
+    },
+    {
+      "epoch": 3.124503311258278,
+      "grad_norm": 0.10513788439628456,
+      "learning_rate": 2.9668083405005133e-05,
+      "loss": 0.4075,
+      "step": 2359
+    },
+    {
+      "epoch": 3.1258278145695364,
+      "grad_norm": 0.09171488725512197,
+      "learning_rate": 2.9632350605160772e-05,
+      "loss": 0.4036,
+      "step": 2360
+    },
+    {
+      "epoch": 3.1271523178807947,
+      "grad_norm": 0.10427198468917705,
+      "learning_rate": 2.959662667256581e-05,
+      "loss": 0.4037,
+      "step": 2361
+    },
+    {
+      "epoch": 3.128476821192053,
+      "grad_norm": 0.09965423388807089,
+      "learning_rate": 2.956091163777421e-05,
+      "loss": 0.4056,
+      "step": 2362
+    },
+    {
+      "epoch": 3.1298013245033114,
+      "grad_norm": 0.0937168128161606,
+      "learning_rate": 2.952520553133237e-05,
+      "loss": 0.4014,
+      "step": 2363
+    },
+    {
+      "epoch": 3.1311258278145697,
+      "grad_norm": 0.10711300102125722,
+      "learning_rate": 2.9489508383779005e-05,
+      "loss": 0.4016,
+      "step": 2364
+    },
+    {
+      "epoch": 3.1324503311258276,
+      "grad_norm": 0.1006770667349882,
+      "learning_rate": 2.9453820225645207e-05,
+      "loss": 0.4041,
+      "step": 2365
+    },
+    {
+      "epoch": 3.133774834437086,
+      "grad_norm": 0.09366555721496313,
+      "learning_rate": 2.941814108745436e-05,
+      "loss": 0.4047,
+      "step": 2366
+    },
+    {
+      "epoch": 3.1350993377483443,
+      "grad_norm": 0.10859334637837735,
+      "learning_rate": 2.9382470999722162e-05,
+      "loss": 0.4055,
+      "step": 2367
+    },
+    {
+      "epoch": 3.1364238410596026,
+      "grad_norm": 0.10410897454940748,
+      "learning_rate": 2.9346809992956517e-05,
+      "loss": 0.4069,
+      "step": 2368
+    },
+    {
+      "epoch": 3.137748344370861,
+      "grad_norm": 0.10703055982210374,
+      "learning_rate": 2.931115809765759e-05,
+      "loss": 0.4048,
+      "step": 2369
+    },
+    {
+      "epoch": 3.1390728476821192,
+      "grad_norm": 0.08725143666292122,
+      "learning_rate": 2.9275515344317784e-05,
+      "loss": 0.402,
+      "step": 2370
+    },
+    {
+      "epoch": 3.1403973509933776,
+      "grad_norm": 0.10555137698124473,
+      "learning_rate": 2.923988176342163e-05,
+      "loss": 0.4034,
+      "step": 2371
+    },
+    {
+      "epoch": 3.141721854304636,
+      "grad_norm": 0.10040549114761302,
+      "learning_rate": 2.9204257385445846e-05,
+      "loss": 0.4031,
+      "step": 2372
+    },
+    {
+      "epoch": 3.1430463576158942,
+      "grad_norm": 0.0971737958356472,
+      "learning_rate": 2.916864224085927e-05,
+      "loss": 0.4092,
+      "step": 2373
+    },
+    {
+      "epoch": 3.144370860927152,
+      "grad_norm": 0.09435194190510479,
+      "learning_rate": 2.9133036360122856e-05,
+      "loss": 0.4049,
+      "step": 2374
+    },
+    {
+      "epoch": 3.1456953642384105,
+      "grad_norm": 0.08575853309340882,
+      "learning_rate": 2.90974397736896e-05,
+      "loss": 0.4037,
+      "step": 2375
+    },
+    {
+      "epoch": 3.147019867549669,
+      "grad_norm": 0.08137637188877894,
+      "learning_rate": 2.9061852512004595e-05,
+      "loss": 0.4022,
+      "step": 2376
+    },
+    {
+      "epoch": 3.148344370860927,
+      "grad_norm": 0.07948522196510012,
+      "learning_rate": 2.9026274605504917e-05,
+      "loss": 0.4007,
+      "step": 2377
+    },
+    {
+      "epoch": 3.1496688741721854,
+      "grad_norm": 0.08995456027014385,
+      "learning_rate": 2.8990706084619682e-05,
+      "loss": 0.4039,
+      "step": 2378
+    },
+    {
+      "epoch": 3.1509933774834438,
+      "grad_norm": 0.0833957715717763,
+      "learning_rate": 2.895514697976993e-05,
+      "loss": 0.4035,
+      "step": 2379
+    },
+    {
+      "epoch": 3.152317880794702,
+      "grad_norm": 0.10174663222899376,
+      "learning_rate": 2.8919597321368686e-05,
+      "loss": 0.4046,
+      "step": 2380
+    },
+    {
+      "epoch": 3.1536423841059604,
+      "grad_norm": 0.0915973549487414,
+      "learning_rate": 2.8884057139820893e-05,
+      "loss": 0.4072,
+      "step": 2381
+    },
+    {
+      "epoch": 3.1549668874172188,
+      "grad_norm": 0.08503630932020943,
+      "learning_rate": 2.884852646552335e-05,
+      "loss": 0.4055,
+      "step": 2382
+    },
+    {
+      "epoch": 3.1562913907284766,
+      "grad_norm": 0.08428264721083464,
+      "learning_rate": 2.8813005328864777e-05,
+      "loss": 0.4027,
+      "step": 2383
+    },
+    {
+      "epoch": 3.157615894039735,
+      "grad_norm": 0.07571735834955415,
+      "learning_rate": 2.87774937602257e-05,
+      "loss": 0.4008,
+      "step": 2384
+    },
+    {
+      "epoch": 3.1589403973509933,
+      "grad_norm": 0.09344379060564686,
+      "learning_rate": 2.8741991789978498e-05,
+      "loss": 0.4057,
+      "step": 2385
+    },
+    {
+      "epoch": 3.1602649006622516,
+      "grad_norm": 0.08523227983046333,
+      "learning_rate": 2.8706499448487287e-05,
+      "loss": 0.4066,
+      "step": 2386
+    },
+    {
+      "epoch": 3.16158940397351,
+      "grad_norm": 0.08963240684950766,
+      "learning_rate": 2.8671016766108013e-05,
+      "loss": 0.4012,
+      "step": 2387
+    },
+    {
+      "epoch": 3.1629139072847683,
+      "grad_norm": 0.08361080440976226,
+      "learning_rate": 2.8635543773188295e-05,
+      "loss": 0.4018,
+      "step": 2388
+    },
+    {
+      "epoch": 3.1642384105960266,
+      "grad_norm": 0.08969609321802438,
+      "learning_rate": 2.8600080500067496e-05,
+      "loss": 0.4038,
+      "step": 2389
+    },
+    {
+      "epoch": 3.165562913907285,
+      "grad_norm": 0.08693476179414286,
+      "learning_rate": 2.856462697707669e-05,
+      "loss": 0.4028,
+      "step": 2390
+    },
+    {
+      "epoch": 3.1668874172185433,
+      "grad_norm": 0.08504446618802275,
+      "learning_rate": 2.852918323453856e-05,
+      "loss": 0.3995,
+      "step": 2391
+    },
+    {
+      "epoch": 3.168211920529801,
+      "grad_norm": 0.09006491409477808,
+      "learning_rate": 2.849374930276747e-05,
+      "loss": 0.4014,
+      "step": 2392
+    },
+    {
+      "epoch": 3.1695364238410595,
+      "grad_norm": 0.08088508911180523,
+      "learning_rate": 2.8458325212069362e-05,
+      "loss": 0.404,
+      "step": 2393
+    },
+    {
+      "epoch": 3.170860927152318,
+      "grad_norm": 0.08649708259344495,
+      "learning_rate": 2.8422910992741797e-05,
+      "loss": 0.4045,
+      "step": 2394
+    },
+    {
+      "epoch": 3.172185430463576,
+      "grad_norm": 0.08989222659464781,
+      "learning_rate": 2.8387506675073836e-05,
+      "loss": 0.407,
+      "step": 2395
+    },
+    {
+      "epoch": 3.1735099337748345,
+      "grad_norm": 0.0826517820092594,
+      "learning_rate": 2.8352112289346128e-05,
+      "loss": 0.4044,
+      "step": 2396
+    },
+    {
+      "epoch": 3.174834437086093,
+      "grad_norm": 0.09157612439110052,
+      "learning_rate": 2.831672786583079e-05,
+      "loss": 0.4054,
+      "step": 2397
+    },
+    {
+      "epoch": 3.176158940397351,
+      "grad_norm": 0.08573397233458208,
+      "learning_rate": 2.828135343479146e-05,
+      "loss": 0.4051,
+      "step": 2398
+    },
+    {
+      "epoch": 3.1774834437086095,
+      "grad_norm": 0.08824406528092782,
+      "learning_rate": 2.8245989026483173e-05,
+      "loss": 0.4018,
+      "step": 2399
+    },
+    {
+      "epoch": 3.1788079470198674,
+      "grad_norm": 0.10525554057487063,
+      "learning_rate": 2.8210634671152424e-05,
+      "loss": 0.4046,
+      "step": 2400
+    },
+    {
+      "epoch": 3.1801324503311257,
+      "grad_norm": 0.09387743058891224,
+      "learning_rate": 2.817529039903713e-05,
+      "loss": 0.4029,
+      "step": 2401
+    },
+    {
+      "epoch": 3.181456953642384,
+      "grad_norm": 0.08701416295713361,
+      "learning_rate": 2.8139956240366535e-05,
+      "loss": 0.4046,
+      "step": 2402
+    },
+    {
+      "epoch": 3.1827814569536423,
+      "grad_norm": 0.08786051862420713,
+      "learning_rate": 2.810463222536128e-05,
+      "loss": 0.4065,
+      "step": 2403
+    },
+    {
+      "epoch": 3.1841059602649007,
+      "grad_norm": 0.09188701964131137,
+      "learning_rate": 2.8069318384233283e-05,
+      "loss": 0.4059,
+      "step": 2404
+    },
+    {
+      "epoch": 3.185430463576159,
+      "grad_norm": 0.0891961469661525,
+      "learning_rate": 2.803401474718583e-05,
+      "loss": 0.4058,
+      "step": 2405
+    },
+    {
+      "epoch": 3.1867549668874173,
+      "grad_norm": 0.09613307883517483,
+      "learning_rate": 2.799872134441339e-05,
+      "loss": 0.4061,
+      "step": 2406
+    },
+    {
+      "epoch": 3.1880794701986757,
+      "grad_norm": 0.08949479012073,
+      "learning_rate": 2.796343820610176e-05,
+      "loss": 0.4032,
+      "step": 2407
+    },
+    {
+      "epoch": 3.1894039735099335,
+      "grad_norm": 0.09546348256342833,
+      "learning_rate": 2.7928165362427897e-05,
+      "loss": 0.405,
+      "step": 2408
+    },
+    {
+      "epoch": 3.190728476821192,
+      "grad_norm": 0.09326318572034699,
+      "learning_rate": 2.7892902843560015e-05,
+      "loss": 0.4066,
+      "step": 2409
+    },
+    {
+      "epoch": 3.19205298013245,
+      "grad_norm": 0.09381334837971277,
+      "learning_rate": 2.785765067965743e-05,
+      "loss": 0.4046,
+      "step": 2410
+    },
+    {
+      "epoch": 3.1933774834437085,
+      "grad_norm": 0.08892029209075933,
+      "learning_rate": 2.7822408900870638e-05,
+      "loss": 0.4043,
+      "step": 2411
+    },
+    {
+      "epoch": 3.194701986754967,
+      "grad_norm": 0.10588688864384158,
+      "learning_rate": 2.7787177537341274e-05,
+      "loss": 0.4052,
+      "step": 2412
+    },
+    {
+      "epoch": 3.196026490066225,
+      "grad_norm": 0.08736984933241855,
+      "learning_rate": 2.775195661920199e-05,
+      "loss": 0.4057,
+      "step": 2413
+    },
+    {
+      "epoch": 3.1973509933774835,
+      "grad_norm": 0.11229599695051605,
+      "learning_rate": 2.7716746176576612e-05,
+      "loss": 0.4056,
+      "step": 2414
+    },
+    {
+      "epoch": 3.198675496688742,
+      "grad_norm": 0.09509619366876432,
+      "learning_rate": 2.7681546239579913e-05,
+      "loss": 0.4078,
+      "step": 2415
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.1093630741437871,
+      "learning_rate": 2.7646356838317723e-05,
+      "loss": 0.4055,
+      "step": 2416
+    },
+    {
+      "epoch": 3.201324503311258,
+      "grad_norm": 0.07973058132220097,
+      "learning_rate": 2.7611178002886855e-05,
+      "loss": 0.4034,
+      "step": 2417
+    },
+    {
+      "epoch": 3.2026490066225164,
+      "grad_norm": 0.10081180921984992,
+      "learning_rate": 2.7576009763375097e-05,
+      "loss": 0.4045,
+      "step": 2418
+    },
+    {
+      "epoch": 3.2039735099337747,
+      "grad_norm": 0.10005296690387926,
+      "learning_rate": 2.7540852149861143e-05,
+      "loss": 0.4046,
+      "step": 2419
+    },
+    {
+      "epoch": 3.205298013245033,
+      "grad_norm": 0.0967361647460188,
+      "learning_rate": 2.7505705192414624e-05,
+      "loss": 0.4046,
+      "step": 2420
+    },
+    {
+      "epoch": 3.2066225165562914,
+      "grad_norm": 0.10188403607599202,
+      "learning_rate": 2.7470568921096058e-05,
+      "loss": 0.4016,
+      "step": 2421
+    },
+    {
+      "epoch": 3.2079470198675497,
+      "grad_norm": 0.07468071847468992,
+      "learning_rate": 2.74354433659568e-05,
+      "loss": 0.4038,
+      "step": 2422
+    },
+    {
+      "epoch": 3.209271523178808,
+      "grad_norm": 0.08363929810078277,
+      "learning_rate": 2.7400328557039074e-05,
+      "loss": 0.4058,
+      "step": 2423
+    },
+    {
+      "epoch": 3.2105960264900664,
+      "grad_norm": 0.07669040798895498,
+      "learning_rate": 2.7365224524375874e-05,
+      "loss": 0.4021,
+      "step": 2424
+    },
+    {
+      "epoch": 3.2119205298013247,
+      "grad_norm": 0.08475310675785717,
+      "learning_rate": 2.733013129799103e-05,
+      "loss": 0.403,
+      "step": 2425
+    },
+    {
+      "epoch": 3.2132450331125826,
+      "grad_norm": 0.08376745250314327,
+      "learning_rate": 2.7295048907899065e-05,
+      "loss": 0.4041,
+      "step": 2426
+    },
+    {
+      "epoch": 3.214569536423841,
+      "grad_norm": 0.08472499243220317,
+      "learning_rate": 2.7259977384105285e-05,
+      "loss": 0.407,
+      "step": 2427
+    },
+    {
+      "epoch": 3.2158940397350992,
+      "grad_norm": 0.0811603518193595,
+      "learning_rate": 2.7224916756605676e-05,
+      "loss": 0.406,
+      "step": 2428
+    },
+    {
+      "epoch": 3.2172185430463576,
+      "grad_norm": 0.08564726853957504,
+      "learning_rate": 2.7189867055386927e-05,
+      "loss": 0.4048,
+      "step": 2429
+    },
+    {
+      "epoch": 3.218543046357616,
+      "grad_norm": 0.08138356896221405,
+      "learning_rate": 2.7154828310426346e-05,
+      "loss": 0.4023,
+      "step": 2430
+    },
+    {
+      "epoch": 3.2198675496688742,
+      "grad_norm": 0.08016141673138501,
+      "learning_rate": 2.71198005516919e-05,
+      "loss": 0.4023,
+      "step": 2431
+    },
+    {
+      "epoch": 3.2211920529801326,
+      "grad_norm": 0.08174843838526905,
+      "learning_rate": 2.7084783809142164e-05,
+      "loss": 0.4039,
+      "step": 2432
+    },
+    {
+      "epoch": 3.222516556291391,
+      "grad_norm": 0.08875213809161796,
+      "learning_rate": 2.704977811272625e-05,
+      "loss": 0.403,
+      "step": 2433
+    },
+    {
+      "epoch": 3.223841059602649,
+      "grad_norm": 0.08945711647856756,
+      "learning_rate": 2.7014783492383885e-05,
+      "loss": 0.4034,
+      "step": 2434
+    },
+    {
+      "epoch": 3.225165562913907,
+      "grad_norm": 0.0920143730624745,
+      "learning_rate": 2.6979799978045257e-05,
+      "loss": 0.4031,
+      "step": 2435
+    },
+    {
+      "epoch": 3.2264900662251654,
+      "grad_norm": 0.09342826863570612,
+      "learning_rate": 2.6944827599631124e-05,
+      "loss": 0.4049,
+      "step": 2436
+    },
+    {
+      "epoch": 3.2278145695364238,
+      "grad_norm": 0.08250648047443322,
+      "learning_rate": 2.6909866387052635e-05,
+      "loss": 0.4081,
+      "step": 2437
+    },
+    {
+      "epoch": 3.229139072847682,
+      "grad_norm": 0.10948607175839914,
+      "learning_rate": 2.6874916370211487e-05,
+      "loss": 0.4046,
+      "step": 2438
+    },
+    {
+      "epoch": 3.2304635761589404,
+      "grad_norm": 0.07735146384996507,
+      "learning_rate": 2.683997757899973e-05,
+      "loss": 0.405,
+      "step": 2439
+    },
+    {
+      "epoch": 3.2317880794701987,
+      "grad_norm": 0.08770437654449731,
+      "learning_rate": 2.6805050043299808e-05,
+      "loss": 0.4,
+      "step": 2440
+    },
+    {
+      "epoch": 3.233112582781457,
+      "grad_norm": 0.07666579005598831,
+      "learning_rate": 2.677013379298461e-05,
+      "loss": 0.4036,
+      "step": 2441
+    },
+    {
+      "epoch": 3.2344370860927154,
+      "grad_norm": 0.08513088397798622,
+      "learning_rate": 2.673522885791729e-05,
+      "loss": 0.4016,
+      "step": 2442
+    },
+    {
+      "epoch": 3.2357615894039737,
+      "grad_norm": 0.08035773874804854,
+      "learning_rate": 2.670033526795139e-05,
+      "loss": 0.4018,
+      "step": 2443
+    },
+    {
+      "epoch": 3.2370860927152316,
+      "grad_norm": 0.0819496294597152,
+      "learning_rate": 2.6665453052930695e-05,
+      "loss": 0.4077,
+      "step": 2444
+    },
+    {
+      "epoch": 3.23841059602649,
+      "grad_norm": 0.08957249946573453,
+      "learning_rate": 2.663058224268932e-05,
+      "loss": 0.4066,
+      "step": 2445
+    },
+    {
+      "epoch": 3.2397350993377483,
+      "grad_norm": 0.0758141054717789,
+      "learning_rate": 2.6595722867051558e-05,
+      "loss": 0.4041,
+      "step": 2446
+    },
+    {
+      "epoch": 3.2410596026490066,
+      "grad_norm": 0.09894539802689038,
+      "learning_rate": 2.6560874955831972e-05,
+      "loss": 0.4025,
+      "step": 2447
+    },
+    {
+      "epoch": 3.242384105960265,
+      "grad_norm": 0.08124582446890334,
+      "learning_rate": 2.6526038538835304e-05,
+      "loss": 0.4055,
+      "step": 2448
+    },
+    {
+      "epoch": 3.2437086092715233,
+      "grad_norm": 0.09828145300451026,
+      "learning_rate": 2.6491213645856477e-05,
+      "loss": 0.4065,
+      "step": 2449
+    },
+    {
+      "epoch": 3.2450331125827816,
+      "grad_norm": 0.08385849962290601,
+      "learning_rate": 2.6456400306680526e-05,
+      "loss": 0.4056,
+      "step": 2450
+    },
+    {
+      "epoch": 3.24635761589404,
+      "grad_norm": 0.09233608457787497,
+      "learning_rate": 2.642159855108263e-05,
+      "loss": 0.4048,
+      "step": 2451
+    },
+    {
+      "epoch": 3.247682119205298,
+      "grad_norm": 0.08286231164443274,
+      "learning_rate": 2.638680840882807e-05,
+      "loss": 0.4038,
+      "step": 2452
+    },
+    {
+      "epoch": 3.249006622516556,
+      "grad_norm": 0.09037301764172137,
+      "learning_rate": 2.6352029909672156e-05,
+      "loss": 0.4027,
+      "step": 2453
+    },
+    {
+      "epoch": 3.2503311258278145,
+      "grad_norm": 0.07722612385851925,
+      "learning_rate": 2.6317263083360275e-05,
+      "loss": 0.4032,
+      "step": 2454
+    },
+    {
+      "epoch": 3.251655629139073,
+      "grad_norm": 0.08654548511326879,
+      "learning_rate": 2.628250795962782e-05,
+      "loss": 0.4031,
+      "step": 2455
+    },
+    {
+      "epoch": 3.252980132450331,
+      "grad_norm": 0.07670172380839418,
+      "learning_rate": 2.6247764568200192e-05,
+      "loss": 0.4039,
+      "step": 2456
+    },
+    {
+      "epoch": 3.2543046357615895,
+      "grad_norm": 0.08374992336224117,
+      "learning_rate": 2.6213032938792696e-05,
+      "loss": 0.4022,
+      "step": 2457
+    },
+    {
+      "epoch": 3.255629139072848,
+      "grad_norm": 0.0747857863050913,
+      "learning_rate": 2.6178313101110667e-05,
+      "loss": 0.403,
+      "step": 2458
+    },
+    {
+      "epoch": 3.256953642384106,
+      "grad_norm": 0.07903414416783318,
+      "learning_rate": 2.6143605084849268e-05,
+      "loss": 0.4034,
+      "step": 2459
+    },
+    {
+      "epoch": 3.258278145695364,
+      "grad_norm": 0.07555887184192538,
+      "learning_rate": 2.6108908919693638e-05,
+      "loss": 0.4042,
+      "step": 2460
+    },
+    {
+      "epoch": 3.2596026490066223,
+      "grad_norm": 0.07373353581247641,
+      "learning_rate": 2.6074224635318693e-05,
+      "loss": 0.4039,
+      "step": 2461
+    },
+    {
+      "epoch": 3.2609271523178807,
+      "grad_norm": 0.08393907099304199,
+      "learning_rate": 2.6039552261389237e-05,
+      "loss": 0.4034,
+      "step": 2462
+    },
+    {
+      "epoch": 3.262251655629139,
+      "grad_norm": 0.07553700877475415,
+      "learning_rate": 2.60048918275599e-05,
+      "loss": 0.4046,
+      "step": 2463
+    },
+    {
+      "epoch": 3.2635761589403973,
+      "grad_norm": 0.08221667707042135,
+      "learning_rate": 2.597024336347504e-05,
+      "loss": 0.4053,
+      "step": 2464
+    },
+    {
+      "epoch": 3.2649006622516556,
+      "grad_norm": 0.08915479117460684,
+      "learning_rate": 2.5935606898768865e-05,
+      "loss": 0.4037,
+      "step": 2465
+    },
+    {
+      "epoch": 3.266225165562914,
+      "grad_norm": 0.07931052644070692,
+      "learning_rate": 2.590098246306523e-05,
+      "loss": 0.4011,
+      "step": 2466
+    },
+    {
+      "epoch": 3.2675496688741723,
+      "grad_norm": 0.10090929510698474,
+      "learning_rate": 2.5866370085977766e-05,
+      "loss": 0.4059,
+      "step": 2467
+    },
+    {
+      "epoch": 3.2688741721854306,
+      "grad_norm": 0.07671503367625188,
+      "learning_rate": 2.5831769797109754e-05,
+      "loss": 0.4066,
+      "step": 2468
+    },
+    {
+      "epoch": 3.2701986754966885,
+      "grad_norm": 0.08676372158797792,
+      "learning_rate": 2.5797181626054176e-05,
+      "loss": 0.4035,
+      "step": 2469
+    },
+    {
+      "epoch": 3.271523178807947,
+      "grad_norm": 0.08683974960895507,
+      "learning_rate": 2.5762605602393593e-05,
+      "loss": 0.4071,
+      "step": 2470
+    },
+    {
+      "epoch": 3.272847682119205,
+      "grad_norm": 0.07726702748601384,
+      "learning_rate": 2.5728041755700207e-05,
+      "loss": 0.4008,
+      "step": 2471
+    },
+    {
+      "epoch": 3.2741721854304635,
+      "grad_norm": 0.08852326626255903,
+      "learning_rate": 2.569349011553584e-05,
+      "loss": 0.401,
+      "step": 2472
+    },
+    {
+      "epoch": 3.275496688741722,
+      "grad_norm": 0.08258150090827311,
+      "learning_rate": 2.565895071145179e-05,
+      "loss": 0.4047,
+      "step": 2473
+    },
+    {
+      "epoch": 3.27682119205298,
+      "grad_norm": 0.08753372760726123,
+      "learning_rate": 2.562442357298898e-05,
+      "loss": 0.407,
+      "step": 2474
+    },
+    {
+      "epoch": 3.2781456953642385,
+      "grad_norm": 0.08278017897927051,
+      "learning_rate": 2.5589908729677785e-05,
+      "loss": 0.4029,
+      "step": 2475
+    },
+    {
+      "epoch": 3.279470198675497,
+      "grad_norm": 0.08062898206719166,
+      "learning_rate": 2.5555406211038092e-05,
+      "loss": 0.4041,
+      "step": 2476
+    },
+    {
+      "epoch": 3.280794701986755,
+      "grad_norm": 0.08181915828489868,
+      "learning_rate": 2.552091604657921e-05,
+      "loss": 0.4061,
+      "step": 2477
+    },
+    {
+      "epoch": 3.282119205298013,
+      "grad_norm": 0.08062145557682944,
+      "learning_rate": 2.5486438265799936e-05,
+      "loss": 0.4016,
+      "step": 2478
+    },
+    {
+      "epoch": 3.2834437086092714,
+      "grad_norm": 0.08534098022781486,
+      "learning_rate": 2.5451972898188428e-05,
+      "loss": 0.4034,
+      "step": 2479
+    },
+    {
+      "epoch": 3.2847682119205297,
+      "grad_norm": 0.08099397659840993,
+      "learning_rate": 2.5417519973222273e-05,
+      "loss": 0.4045,
+      "step": 2480
+    },
+    {
+      "epoch": 3.286092715231788,
+      "grad_norm": 0.08026924962339368,
+      "learning_rate": 2.5383079520368356e-05,
+      "loss": 0.4067,
+      "step": 2481
+    },
+    {
+      "epoch": 3.2874172185430464,
+      "grad_norm": 0.0746070820033379,
+      "learning_rate": 2.5348651569082937e-05,
+      "loss": 0.4017,
+      "step": 2482
+    },
+    {
+      "epoch": 3.2887417218543047,
+      "grad_norm": 0.08531630457813028,
+      "learning_rate": 2.5314236148811594e-05,
+      "loss": 0.4015,
+      "step": 2483
+    },
+    {
+      "epoch": 3.290066225165563,
+      "grad_norm": 0.0769291602601457,
+      "learning_rate": 2.5279833288989134e-05,
+      "loss": 0.4048,
+      "step": 2484
+    },
+    {
+      "epoch": 3.2913907284768213,
+      "grad_norm": 0.08599580102634333,
+      "learning_rate": 2.5245443019039693e-05,
+      "loss": 0.4025,
+      "step": 2485
+    },
+    {
+      "epoch": 3.2927152317880797,
+      "grad_norm": 0.07600414493354074,
+      "learning_rate": 2.5211065368376574e-05,
+      "loss": 0.4052,
+      "step": 2486
+    },
+    {
+      "epoch": 3.2940397350993376,
+      "grad_norm": 0.09083475371194488,
+      "learning_rate": 2.517670036640234e-05,
+      "loss": 0.4041,
+      "step": 2487
+    },
+    {
+      "epoch": 3.295364238410596,
+      "grad_norm": 0.0773042881587003,
+      "learning_rate": 2.514234804250868e-05,
+      "loss": 0.404,
+      "step": 2488
+    },
+    {
+      "epoch": 3.296688741721854,
+      "grad_norm": 0.08035331861859786,
+      "learning_rate": 2.5108008426076495e-05,
+      "loss": 0.4047,
+      "step": 2489
+    },
+    {
+      "epoch": 3.2980132450331126,
+      "grad_norm": 0.07740102195552238,
+      "learning_rate": 2.5073681546475786e-05,
+      "loss": 0.4062,
+      "step": 2490
+    },
+    {
+      "epoch": 3.299337748344371,
+      "grad_norm": 0.07647532089016501,
+      "learning_rate": 2.5039367433065644e-05,
+      "loss": 0.4071,
+      "step": 2491
+    },
+    {
+      "epoch": 3.300662251655629,
+      "grad_norm": 0.07763253789689706,
+      "learning_rate": 2.5005066115194308e-05,
+      "loss": 0.4028,
+      "step": 2492
+    },
+    {
+      "epoch": 3.3019867549668875,
+      "grad_norm": 0.08342189176673018,
+      "learning_rate": 2.4970777622198994e-05,
+      "loss": 0.4064,
+      "step": 2493
+    },
+    {
+      "epoch": 3.303311258278146,
+      "grad_norm": 0.08475841069791651,
+      "learning_rate": 2.4936501983406008e-05,
+      "loss": 0.4093,
+      "step": 2494
+    },
+    {
+      "epoch": 3.304635761589404,
+      "grad_norm": 0.09263983120399989,
+      "learning_rate": 2.490223922813061e-05,
+      "loss": 0.4027,
+      "step": 2495
+    },
+    {
+      "epoch": 3.305960264900662,
+      "grad_norm": 0.08236086126491926,
+      "learning_rate": 2.4867989385677113e-05,
+      "loss": 0.4057,
+      "step": 2496
+    },
+    {
+      "epoch": 3.3072847682119204,
+      "grad_norm": 0.09519753800599198,
+      "learning_rate": 2.4833752485338703e-05,
+      "loss": 0.4075,
+      "step": 2497
+    },
+    {
+      "epoch": 3.3086092715231787,
+      "grad_norm": 0.08283373122677687,
+      "learning_rate": 2.4799528556397558e-05,
+      "loss": 0.4057,
+      "step": 2498
+    },
+    {
+      "epoch": 3.309933774834437,
+      "grad_norm": 0.09532071628263582,
+      "learning_rate": 2.4765317628124724e-05,
+      "loss": 0.4062,
+      "step": 2499
+    },
+    {
+      "epoch": 3.3112582781456954,
+      "grad_norm": 0.0753478752247644,
+      "learning_rate": 2.473111972978017e-05,
+      "loss": 0.404,
+      "step": 2500
+    },
+    {
+      "epoch": 3.3125827814569537,
+      "grad_norm": 0.10062977702545839,
+      "learning_rate": 2.4696934890612678e-05,
+      "loss": 0.405,
+      "step": 2501
+    },
+    {
+      "epoch": 3.313907284768212,
+      "grad_norm": 0.08166369594216884,
+      "learning_rate": 2.4662763139859864e-05,
+      "loss": 0.4072,
+      "step": 2502
+    },
+    {
+      "epoch": 3.3152317880794704,
+      "grad_norm": 0.09575513607696816,
+      "learning_rate": 2.4628604506748195e-05,
+      "loss": 0.4056,
+      "step": 2503
+    },
+    {
+      "epoch": 3.3165562913907287,
+      "grad_norm": 0.08586242508830341,
+      "learning_rate": 2.459445902049286e-05,
+      "loss": 0.405,
+      "step": 2504
+    },
+    {
+      "epoch": 3.3178807947019866,
+      "grad_norm": 0.0910000241589303,
+      "learning_rate": 2.4560326710297845e-05,
+      "loss": 0.4059,
+      "step": 2505
+    },
+    {
+      "epoch": 3.319205298013245,
+      "grad_norm": 0.08795760838776083,
+      "learning_rate": 2.452620760535584e-05,
+      "loss": 0.4068,
+      "step": 2506
+    },
+    {
+      "epoch": 3.3205298013245033,
+      "grad_norm": 0.09257086021746731,
+      "learning_rate": 2.4492101734848283e-05,
+      "loss": 0.4038,
+      "step": 2507
+    },
+    {
+      "epoch": 3.3218543046357616,
+      "grad_norm": 0.08686083908012256,
+      "learning_rate": 2.4458009127945224e-05,
+      "loss": 0.4048,
+      "step": 2508
+    },
+    {
+      "epoch": 3.32317880794702,
+      "grad_norm": 0.09067489882737383,
+      "learning_rate": 2.442392981380544e-05,
+      "loss": 0.4024,
+      "step": 2509
+    },
+    {
+      "epoch": 3.3245033112582782,
+      "grad_norm": 0.08367670090605872,
+      "learning_rate": 2.4389863821576284e-05,
+      "loss": 0.4021,
+      "step": 2510
+    },
+    {
+      "epoch": 3.3258278145695366,
+      "grad_norm": 0.10115983404230991,
+      "learning_rate": 2.4355811180393767e-05,
+      "loss": 0.4065,
+      "step": 2511
+    },
+    {
+      "epoch": 3.3271523178807945,
+      "grad_norm": 0.08228819700832994,
+      "learning_rate": 2.4321771919382428e-05,
+      "loss": 0.4032,
+      "step": 2512
+    },
+    {
+      "epoch": 3.328476821192053,
+      "grad_norm": 0.08870999012168061,
+      "learning_rate": 2.4287746067655378e-05,
+      "loss": 0.4038,
+      "step": 2513
+    },
+    {
+      "epoch": 3.329801324503311,
+      "grad_norm": 0.0767749215631213,
+      "learning_rate": 2.4253733654314305e-05,
+      "loss": 0.4034,
+      "step": 2514
+    },
+    {
+      "epoch": 3.3311258278145695,
+      "grad_norm": 0.10404582469862914,
+      "learning_rate": 2.4219734708449325e-05,
+      "loss": 0.4055,
+      "step": 2515
+    },
+    {
+      "epoch": 3.332450331125828,
+      "grad_norm": 0.07916311114847494,
+      "learning_rate": 2.4185749259139107e-05,
+      "loss": 0.4092,
+      "step": 2516
+    },
+    {
+      "epoch": 3.333774834437086,
+      "grad_norm": 0.09528305953280929,
+      "learning_rate": 2.4151777335450717e-05,
+      "loss": 0.3998,
+      "step": 2517
+    },
+    {
+      "epoch": 3.3350993377483444,
+      "grad_norm": 0.08886535483098569,
+      "learning_rate": 2.411781896643973e-05,
+      "loss": 0.4018,
+      "step": 2518
+    },
+    {
+      "epoch": 3.3364238410596028,
+      "grad_norm": 0.0832724197066739,
+      "learning_rate": 2.4083874181150017e-05,
+      "loss": 0.4037,
+      "step": 2519
+    },
+    {
+      "epoch": 3.337748344370861,
+      "grad_norm": 0.09943527816436971,
+      "learning_rate": 2.404994300861396e-05,
+      "loss": 0.4021,
+      "step": 2520
+    },
+    {
+      "epoch": 3.339072847682119,
+      "grad_norm": 0.07524089351237366,
+      "learning_rate": 2.4016025477852187e-05,
+      "loss": 0.4061,
+      "step": 2521
+    },
+    {
+      "epoch": 3.3403973509933773,
+      "grad_norm": 0.09935564849391779,
+      "learning_rate": 2.3982121617873712e-05,
+      "loss": 0.4022,
+      "step": 2522
+    },
+    {
+      "epoch": 3.3417218543046356,
+      "grad_norm": 0.07889323855661073,
+      "learning_rate": 2.394823145767587e-05,
+      "loss": 0.4036,
+      "step": 2523
+    },
+    {
+      "epoch": 3.343046357615894,
+      "grad_norm": 0.09360062673052713,
+      "learning_rate": 2.391435502624422e-05,
+      "loss": 0.4047,
+      "step": 2524
+    },
+    {
+      "epoch": 3.3443708609271523,
+      "grad_norm": 0.08216355795323498,
+      "learning_rate": 2.3880492352552654e-05,
+      "loss": 0.4032,
+      "step": 2525
+    },
+    {
+      "epoch": 3.3456953642384106,
+      "grad_norm": 0.09144980121162627,
+      "learning_rate": 2.3846643465563234e-05,
+      "loss": 0.4066,
+      "step": 2526
+    },
+    {
+      "epoch": 3.347019867549669,
+      "grad_norm": 0.08232126109526955,
+      "learning_rate": 2.381280839422628e-05,
+      "loss": 0.402,
+      "step": 2527
+    },
+    {
+      "epoch": 3.3483443708609273,
+      "grad_norm": 0.08153551391517846,
+      "learning_rate": 2.377898716748025e-05,
+      "loss": 0.4053,
+      "step": 2528
+    },
+    {
+      "epoch": 3.3496688741721856,
+      "grad_norm": 0.0938092796097717,
+      "learning_rate": 2.374517981425179e-05,
+      "loss": 0.4051,
+      "step": 2529
+    },
+    {
+      "epoch": 3.3509933774834435,
+      "grad_norm": 0.0781099526628598,
+      "learning_rate": 2.3711386363455668e-05,
+      "loss": 0.4027,
+      "step": 2530
+    },
+    {
+      "epoch": 3.352317880794702,
+      "grad_norm": 0.08780908842703235,
+      "learning_rate": 2.3677606843994797e-05,
+      "loss": 0.4025,
+      "step": 2531
+    },
+    {
+      "epoch": 3.35364238410596,
+      "grad_norm": 0.0842043209631639,
+      "learning_rate": 2.3643841284760108e-05,
+      "loss": 0.4067,
+      "step": 2532
+    },
+    {
+      "epoch": 3.3549668874172185,
+      "grad_norm": 0.07988052588524308,
+      "learning_rate": 2.3610089714630638e-05,
+      "loss": 0.4032,
+      "step": 2533
+    },
+    {
+      "epoch": 3.356291390728477,
+      "grad_norm": 0.09297135094874788,
+      "learning_rate": 2.3576352162473464e-05,
+      "loss": 0.4051,
+      "step": 2534
+    },
+    {
+      "epoch": 3.357615894039735,
+      "grad_norm": 0.08494979840858917,
+      "learning_rate": 2.3542628657143638e-05,
+      "loss": 0.4039,
+      "step": 2535
+    },
+    {
+      "epoch": 3.3589403973509935,
+      "grad_norm": 0.0930787732614263,
+      "learning_rate": 2.3508919227484244e-05,
+      "loss": 0.4037,
+      "step": 2536
+    },
+    {
+      "epoch": 3.360264900662252,
+      "grad_norm": 0.07933742633238718,
+      "learning_rate": 2.3475223902326274e-05,
+      "loss": 0.4043,
+      "step": 2537
+    },
+    {
+      "epoch": 3.36158940397351,
+      "grad_norm": 0.10109847643809014,
+      "learning_rate": 2.344154271048873e-05,
+      "loss": 0.4006,
+      "step": 2538
+    },
+    {
+      "epoch": 3.362913907284768,
+      "grad_norm": 0.08455638173566357,
+      "learning_rate": 2.3407875680778433e-05,
+      "loss": 0.4065,
+      "step": 2539
+    },
+    {
+      "epoch": 3.3642384105960264,
+      "grad_norm": 0.08949555617947444,
+      "learning_rate": 2.3374222841990168e-05,
+      "loss": 0.4022,
+      "step": 2540
+    },
+    {
+      "epoch": 3.3655629139072847,
+      "grad_norm": 0.08959964797017053,
+      "learning_rate": 2.3340584222906557e-05,
+      "loss": 0.4049,
+      "step": 2541
+    },
+    {
+      "epoch": 3.366887417218543,
+      "grad_norm": 0.07912429482889999,
+      "learning_rate": 2.330695985229804e-05,
+      "loss": 0.4056,
+      "step": 2542
+    },
+    {
+      "epoch": 3.3682119205298013,
+      "grad_norm": 0.09145502624178324,
+      "learning_rate": 2.3273349758922913e-05,
+      "loss": 0.4046,
+      "step": 2543
+    },
+    {
+      "epoch": 3.3695364238410597,
+      "grad_norm": 0.07353096026767844,
+      "learning_rate": 2.323975397152719e-05,
+      "loss": 0.4018,
+      "step": 2544
+    },
+    {
+      "epoch": 3.370860927152318,
+      "grad_norm": 0.09799012955658815,
+      "learning_rate": 2.3206172518844765e-05,
+      "loss": 0.4016,
+      "step": 2545
+    },
+    {
+      "epoch": 3.3721854304635763,
+      "grad_norm": 0.07687059349194636,
+      "learning_rate": 2.317260542959715e-05,
+      "loss": 0.4019,
+      "step": 2546
+    },
+    {
+      "epoch": 3.3735099337748347,
+      "grad_norm": 0.09629829423474826,
+      "learning_rate": 2.3139052732493658e-05,
+      "loss": 0.4055,
+      "step": 2547
+    },
+    {
+      "epoch": 3.3748344370860925,
+      "grad_norm": 0.08554568906392758,
+      "learning_rate": 2.310551445623123e-05,
+      "loss": 0.4035,
+      "step": 2548
+    },
+    {
+      "epoch": 3.376158940397351,
+      "grad_norm": 0.09448970576759666,
+      "learning_rate": 2.307199062949453e-05,
+      "loss": 0.4027,
+      "step": 2549
+    },
+    {
+      "epoch": 3.377483443708609,
+      "grad_norm": 0.081401661152439,
+      "learning_rate": 2.303848128095581e-05,
+      "loss": 0.4036,
+      "step": 2550
+    },
+    {
+      "epoch": 3.3788079470198675,
+      "grad_norm": 0.09463075136448817,
+      "learning_rate": 2.3004986439274975e-05,
+      "loss": 0.4068,
+      "step": 2551
+    },
+    {
+      "epoch": 3.380132450331126,
+      "grad_norm": 0.09652914252463071,
+      "learning_rate": 2.2971506133099544e-05,
+      "loss": 0.4045,
+      "step": 2552
+    },
+    {
+      "epoch": 3.381456953642384,
+      "grad_norm": 0.10000102335189794,
+      "learning_rate": 2.2938040391064523e-05,
+      "loss": 0.4041,
+      "step": 2553
+    },
+    {
+      "epoch": 3.3827814569536425,
+      "grad_norm": 0.08600213845256945,
+      "learning_rate": 2.2904589241792555e-05,
+      "loss": 0.402,
+      "step": 2554
+    },
+    {
+      "epoch": 3.384105960264901,
+      "grad_norm": 0.0971638220841416,
+      "learning_rate": 2.2871152713893724e-05,
+      "loss": 0.4035,
+      "step": 2555
+    },
+    {
+      "epoch": 3.385430463576159,
+      "grad_norm": 0.08253933812348893,
+      "learning_rate": 2.283773083596568e-05,
+      "loss": 0.4053,
+      "step": 2556
+    },
+    {
+      "epoch": 3.386754966887417,
+      "grad_norm": 0.09793568900050217,
+      "learning_rate": 2.2804323636593467e-05,
+      "loss": 0.402,
+      "step": 2557
+    },
+    {
+      "epoch": 3.3880794701986754,
+      "grad_norm": 0.08783955320862084,
+      "learning_rate": 2.2770931144349638e-05,
+      "loss": 0.405,
+      "step": 2558
+    },
+    {
+      "epoch": 3.3894039735099337,
+      "grad_norm": 0.09358030957275644,
+      "learning_rate": 2.2737553387794132e-05,
+      "loss": 0.4024,
+      "step": 2559
+    },
+    {
+      "epoch": 3.390728476821192,
+      "grad_norm": 0.0760040238487492,
+      "learning_rate": 2.2704190395474324e-05,
+      "loss": 0.4028,
+      "step": 2560
+    },
+    {
+      "epoch": 3.3920529801324504,
+      "grad_norm": 0.08624676667464895,
+      "learning_rate": 2.2670842195924902e-05,
+      "loss": 0.4068,
+      "step": 2561
+    },
+    {
+      "epoch": 3.3933774834437087,
+      "grad_norm": 0.07415115857215471,
+      "learning_rate": 2.2637508817667958e-05,
+      "loss": 0.4064,
+      "step": 2562
+    },
+    {
+      "epoch": 3.394701986754967,
+      "grad_norm": 0.09063419197567955,
+      "learning_rate": 2.2604190289212876e-05,
+      "loss": 0.4056,
+      "step": 2563
+    },
+    {
+      "epoch": 3.396026490066225,
+      "grad_norm": 0.08173400341816893,
+      "learning_rate": 2.2570886639056326e-05,
+      "loss": 0.403,
+      "step": 2564
+    },
+    {
+      "epoch": 3.3973509933774833,
+      "grad_norm": 0.0889018440689219,
+      "learning_rate": 2.2537597895682292e-05,
+      "loss": 0.4002,
+      "step": 2565
+    },
+    {
+      "epoch": 3.3986754966887416,
+      "grad_norm": 0.09117861151728518,
+      "learning_rate": 2.250432408756199e-05,
+      "loss": 0.4013,
+      "step": 2566
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.08003009516831328,
+      "learning_rate": 2.247106524315388e-05,
+      "loss": 0.4047,
+      "step": 2567
+    },
+    {
+      "epoch": 3.4013245033112582,
+      "grad_norm": 0.09306333345547338,
+      "learning_rate": 2.243782139090357e-05,
+      "loss": 0.4046,
+      "step": 2568
+    },
+    {
+      "epoch": 3.4026490066225166,
+      "grad_norm": 0.07986939155809411,
+      "learning_rate": 2.2404592559243912e-05,
+      "loss": 0.402,
+      "step": 2569
+    },
+    {
+      "epoch": 3.403973509933775,
+      "grad_norm": 0.0819003068673905,
+      "learning_rate": 2.237137877659486e-05,
+      "loss": 0.4015,
+      "step": 2570
+    },
+    {
+      "epoch": 3.4052980132450332,
+      "grad_norm": 0.07715300562548492,
+      "learning_rate": 2.2338180071363543e-05,
+      "loss": 0.4078,
+      "step": 2571
+    },
+    {
+      "epoch": 3.4066225165562916,
+      "grad_norm": 0.08084739982476111,
+      "learning_rate": 2.230499647194413e-05,
+      "loss": 0.4,
+      "step": 2572
+    },
+    {
+      "epoch": 3.4079470198675494,
+      "grad_norm": 0.07587255732682704,
+      "learning_rate": 2.227182800671794e-05,
+      "loss": 0.4008,
+      "step": 2573
+    },
+    {
+      "epoch": 3.4092715231788078,
+      "grad_norm": 0.08303584030061825,
+      "learning_rate": 2.2238674704053328e-05,
+      "loss": 0.4053,
+      "step": 2574
+    },
+    {
+      "epoch": 3.410596026490066,
+      "grad_norm": 0.08143813556361919,
+      "learning_rate": 2.2205536592305635e-05,
+      "loss": 0.4043,
+      "step": 2575
+    },
+    {
+      "epoch": 3.4119205298013244,
+      "grad_norm": 0.08560806712744357,
+      "learning_rate": 2.2172413699817286e-05,
+      "loss": 0.4045,
+      "step": 2576
+    },
+    {
+      "epoch": 3.4132450331125828,
+      "grad_norm": 0.07631099051736634,
+      "learning_rate": 2.2139306054917604e-05,
+      "loss": 0.4023,
+      "step": 2577
+    },
+    {
+      "epoch": 3.414569536423841,
+      "grad_norm": 0.08804472983523881,
+      "learning_rate": 2.2106213685922945e-05,
+      "loss": 0.4028,
+      "step": 2578
+    },
+    {
+      "epoch": 3.4158940397350994,
+      "grad_norm": 0.08188701465533611,
+      "learning_rate": 2.2073136621136554e-05,
+      "loss": 0.4008,
+      "step": 2579
+    },
+    {
+      "epoch": 3.4172185430463577,
+      "grad_norm": 0.07489035207043672,
+      "learning_rate": 2.2040074888848634e-05,
+      "loss": 0.4073,
+      "step": 2580
+    },
+    {
+      "epoch": 3.418543046357616,
+      "grad_norm": 0.07850664611963504,
+      "learning_rate": 2.20070285173362e-05,
+      "loss": 0.4021,
+      "step": 2581
+    },
+    {
+      "epoch": 3.419867549668874,
+      "grad_norm": 0.07345081083168074,
+      "learning_rate": 2.1973997534863215e-05,
+      "loss": 0.4018,
+      "step": 2582
+    },
+    {
+      "epoch": 3.4211920529801323,
+      "grad_norm": 0.08210534416106037,
+      "learning_rate": 2.1940981969680404e-05,
+      "loss": 0.4023,
+      "step": 2583
+    },
+    {
+      "epoch": 3.4225165562913906,
+      "grad_norm": 0.07446971556294112,
+      "learning_rate": 2.1907981850025348e-05,
+      "loss": 0.4029,
+      "step": 2584
+    },
+    {
+      "epoch": 3.423841059602649,
+      "grad_norm": 0.08387035147437327,
+      "learning_rate": 2.187499720412241e-05,
+      "loss": 0.4008,
+      "step": 2585
+    },
+    {
+      "epoch": 3.4251655629139073,
+      "grad_norm": 0.07780722378592922,
+      "learning_rate": 2.184202806018272e-05,
+      "loss": 0.4021,
+      "step": 2586
+    },
+    {
+      "epoch": 3.4264900662251656,
+      "grad_norm": 0.08925099418786196,
+      "learning_rate": 2.1809074446404157e-05,
+      "loss": 0.4031,
+      "step": 2587
+    },
+    {
+      "epoch": 3.427814569536424,
+      "grad_norm": 0.08280392172299605,
+      "learning_rate": 2.1776136390971287e-05,
+      "loss": 0.4048,
+      "step": 2588
+    },
+    {
+      "epoch": 3.4291390728476823,
+      "grad_norm": 0.07321336098357038,
+      "learning_rate": 2.174321392205542e-05,
+      "loss": 0.4009,
+      "step": 2589
+    },
+    {
+      "epoch": 3.4304635761589406,
+      "grad_norm": 0.08015287238393809,
+      "learning_rate": 2.171030706781446e-05,
+      "loss": 0.4042,
+      "step": 2590
+    },
+    {
+      "epoch": 3.4317880794701985,
+      "grad_norm": 0.07433314132340209,
+      "learning_rate": 2.1677415856393045e-05,
+      "loss": 0.4044,
+      "step": 2591
+    },
+    {
+      "epoch": 3.433112582781457,
+      "grad_norm": 0.07611368009499751,
+      "learning_rate": 2.1644540315922363e-05,
+      "loss": 0.4049,
+      "step": 2592
+    },
+    {
+      "epoch": 3.434437086092715,
+      "grad_norm": 0.08408240914178432,
+      "learning_rate": 2.1611680474520228e-05,
+      "loss": 0.4033,
+      "step": 2593
+    },
+    {
+      "epoch": 3.4357615894039735,
+      "grad_norm": 0.07983054938388037,
+      "learning_rate": 2.157883636029105e-05,
+      "loss": 0.4037,
+      "step": 2594
+    },
+    {
+      "epoch": 3.437086092715232,
+      "grad_norm": 0.07779968356204135,
+      "learning_rate": 2.1546008001325724e-05,
+      "loss": 0.3995,
+      "step": 2595
+    },
+    {
+      "epoch": 3.43841059602649,
+      "grad_norm": 0.0750810966382616,
+      "learning_rate": 2.1513195425701746e-05,
+      "loss": 0.4011,
+      "step": 2596
+    },
+    {
+      "epoch": 3.4397350993377485,
+      "grad_norm": 0.08103191083441488,
+      "learning_rate": 2.148039866148304e-05,
+      "loss": 0.4071,
+      "step": 2597
+    },
+    {
+      "epoch": 3.441059602649007,
+      "grad_norm": 0.08089654189009486,
+      "learning_rate": 2.1447617736720083e-05,
+      "loss": 0.4065,
+      "step": 2598
+    },
+    {
+      "epoch": 3.442384105960265,
+      "grad_norm": 0.10070113038552912,
+      "learning_rate": 2.1414852679449704e-05,
+      "loss": 0.4063,
+      "step": 2599
+    },
+    {
+      "epoch": 3.443708609271523,
+      "grad_norm": 0.07272917766801329,
+      "learning_rate": 2.1382103517695287e-05,
+      "loss": 0.4002,
+      "step": 2600
+    },
+    {
+      "epoch": 3.4450331125827813,
+      "grad_norm": 0.10449336467436218,
+      "learning_rate": 2.1349370279466514e-05,
+      "loss": 0.4026,
+      "step": 2601
+    },
+    {
+      "epoch": 3.4463576158940397,
+      "grad_norm": 0.07719445986519574,
+      "learning_rate": 2.1316652992759515e-05,
+      "loss": 0.4032,
+      "step": 2602
+    },
+    {
+      "epoch": 3.447682119205298,
+      "grad_norm": 0.09450207292560323,
+      "learning_rate": 2.128395168555673e-05,
+      "loss": 0.4054,
+      "step": 2603
+    },
+    {
+      "epoch": 3.4490066225165563,
+      "grad_norm": 0.07853224160140512,
+      "learning_rate": 2.125126638582695e-05,
+      "loss": 0.406,
+      "step": 2604
+    },
+    {
+      "epoch": 3.4503311258278146,
+      "grad_norm": 0.08301437902799161,
+      "learning_rate": 2.121859712152528e-05,
+      "loss": 0.4028,
+      "step": 2605
+    },
+    {
+      "epoch": 3.451655629139073,
+      "grad_norm": 0.078361793580761,
+      "learning_rate": 2.118594392059312e-05,
+      "loss": 0.406,
+      "step": 2606
+    },
+    {
+      "epoch": 3.4529801324503313,
+      "grad_norm": 0.0823315634824611,
+      "learning_rate": 2.1153306810958125e-05,
+      "loss": 0.4062,
+      "step": 2607
+    },
+    {
+      "epoch": 3.4543046357615896,
+      "grad_norm": 0.08401464135515224,
+      "learning_rate": 2.1120685820534153e-05,
+      "loss": 0.4046,
+      "step": 2608
+    },
+    {
+      "epoch": 3.4556291390728475,
+      "grad_norm": 0.0724655404732036,
+      "learning_rate": 2.108808097722134e-05,
+      "loss": 0.4004,
+      "step": 2609
+    },
+    {
+      "epoch": 3.456953642384106,
+      "grad_norm": 0.08933598842306008,
+      "learning_rate": 2.105549230890594e-05,
+      "loss": 0.4044,
+      "step": 2610
+    },
+    {
+      "epoch": 3.458278145695364,
+      "grad_norm": 0.07759723325291645,
+      "learning_rate": 2.102291984346044e-05,
+      "loss": 0.4056,
+      "step": 2611
+    },
+    {
+      "epoch": 3.4596026490066225,
+      "grad_norm": 0.08087584697691291,
+      "learning_rate": 2.0990363608743406e-05,
+      "loss": 0.4056,
+      "step": 2612
+    },
+    {
+      "epoch": 3.460927152317881,
+      "grad_norm": 0.07525248872960735,
+      "learning_rate": 2.095782363259957e-05,
+      "loss": 0.405,
+      "step": 2613
+    },
+    {
+      "epoch": 3.462251655629139,
+      "grad_norm": 0.07892425561556923,
+      "learning_rate": 2.0925299942859768e-05,
+      "loss": 0.4041,
+      "step": 2614
+    },
+    {
+      "epoch": 3.4635761589403975,
+      "grad_norm": 0.07531494993246282,
+      "learning_rate": 2.0892792567340834e-05,
+      "loss": 0.4049,
+      "step": 2615
+    },
+    {
+      "epoch": 3.4649006622516554,
+      "grad_norm": 0.07029707945620313,
+      "learning_rate": 2.0860301533845737e-05,
+      "loss": 0.4044,
+      "step": 2616
+    },
+    {
+      "epoch": 3.466225165562914,
+      "grad_norm": 0.07822757654750886,
+      "learning_rate": 2.082782687016339e-05,
+      "loss": 0.404,
+      "step": 2617
+    },
+    {
+      "epoch": 3.467549668874172,
+      "grad_norm": 0.08238361211830786,
+      "learning_rate": 2.079536860406878e-05,
+      "loss": 0.404,
+      "step": 2618
+    },
+    {
+      "epoch": 3.4688741721854304,
+      "grad_norm": 0.08496194983470456,
+      "learning_rate": 2.0762926763322797e-05,
+      "loss": 0.4025,
+      "step": 2619
+    },
+    {
+      "epoch": 3.4701986754966887,
+      "grad_norm": 0.07349829415706853,
+      "learning_rate": 2.0730501375672333e-05,
+      "loss": 0.4048,
+      "step": 2620
+    },
+    {
+      "epoch": 3.471523178807947,
+      "grad_norm": 0.09100430505360024,
+      "learning_rate": 2.0698092468850194e-05,
+      "loss": 0.4019,
+      "step": 2621
+    },
+    {
+      "epoch": 3.4728476821192054,
+      "grad_norm": 0.07377240714861191,
+      "learning_rate": 2.0665700070575108e-05,
+      "loss": 0.4019,
+      "step": 2622
+    },
+    {
+      "epoch": 3.4741721854304637,
+      "grad_norm": 0.08063765454671072,
+      "learning_rate": 2.0633324208551645e-05,
+      "loss": 0.4037,
+      "step": 2623
+    },
+    {
+      "epoch": 3.475496688741722,
+      "grad_norm": 0.0783988320918644,
+      "learning_rate": 2.0600964910470243e-05,
+      "loss": 0.405,
+      "step": 2624
+    },
+    {
+      "epoch": 3.47682119205298,
+      "grad_norm": 0.070280178296158,
+      "learning_rate": 2.0568622204007203e-05,
+      "loss": 0.406,
+      "step": 2625
+    },
+    {
+      "epoch": 3.4781456953642382,
+      "grad_norm": 0.07801093979469993,
+      "learning_rate": 2.053629611682457e-05,
+      "loss": 0.4036,
+      "step": 2626
+    },
+    {
+      "epoch": 3.4794701986754966,
+      "grad_norm": 0.07361382104801392,
+      "learning_rate": 2.0503986676570288e-05,
+      "loss": 0.4049,
+      "step": 2627
+    },
+    {
+      "epoch": 3.480794701986755,
+      "grad_norm": 0.07487111733029982,
+      "learning_rate": 2.0471693910877944e-05,
+      "loss": 0.404,
+      "step": 2628
+    },
+    {
+      "epoch": 3.482119205298013,
+      "grad_norm": 0.07460123726218792,
+      "learning_rate": 2.0439417847366942e-05,
+      "loss": 0.4,
+      "step": 2629
+    },
+    {
+      "epoch": 3.4834437086092715,
+      "grad_norm": 0.0805902468331474,
+      "learning_rate": 2.040715851364235e-05,
+      "loss": 0.4051,
+      "step": 2630
+    },
+    {
+      "epoch": 3.48476821192053,
+      "grad_norm": 0.07766305563749591,
+      "learning_rate": 2.037491593729497e-05,
+      "loss": 0.4056,
+      "step": 2631
+    },
+    {
+      "epoch": 3.486092715231788,
+      "grad_norm": 0.09611889516192337,
+      "learning_rate": 2.0342690145901232e-05,
+      "loss": 0.4016,
+      "step": 2632
+    },
+    {
+      "epoch": 3.4874172185430465,
+      "grad_norm": 0.0860008765127659,
+      "learning_rate": 2.0310481167023238e-05,
+      "loss": 0.4063,
+      "step": 2633
+    },
+    {
+      "epoch": 3.4887417218543044,
+      "grad_norm": 0.08963588871511341,
+      "learning_rate": 2.027828902820872e-05,
+      "loss": 0.4036,
+      "step": 2634
+    },
+    {
+      "epoch": 3.4900662251655628,
+      "grad_norm": 0.09162223939728582,
+      "learning_rate": 2.0246113756990958e-05,
+      "loss": 0.4019,
+      "step": 2635
+    },
+    {
+      "epoch": 3.491390728476821,
+      "grad_norm": 0.09051000495671133,
+      "learning_rate": 2.021395538088887e-05,
+      "loss": 0.4016,
+      "step": 2636
+    },
+    {
+      "epoch": 3.4927152317880794,
+      "grad_norm": 0.08614678150801282,
+      "learning_rate": 2.0181813927406863e-05,
+      "loss": 0.4012,
+      "step": 2637
+    },
+    {
+      "epoch": 3.4940397350993377,
+      "grad_norm": 0.09495508823816746,
+      "learning_rate": 2.0149689424034923e-05,
+      "loss": 0.4028,
+      "step": 2638
+    },
+    {
+      "epoch": 3.495364238410596,
+      "grad_norm": 0.08999656288026055,
+      "learning_rate": 2.011758189824849e-05,
+      "loss": 0.3996,
+      "step": 2639
+    },
+    {
+      "epoch": 3.4966887417218544,
+      "grad_norm": 0.0939595600202846,
+      "learning_rate": 2.0085491377508522e-05,
+      "loss": 0.4027,
+      "step": 2640
+    },
+    {
+      "epoch": 3.4980132450331127,
+      "grad_norm": 0.10380041814904373,
+      "learning_rate": 2.0053417889261416e-05,
+      "loss": 0.4009,
+      "step": 2641
+    },
+    {
+      "epoch": 3.499337748344371,
+      "grad_norm": 0.08824629597082938,
+      "learning_rate": 2.0021361460939017e-05,
+      "loss": 0.4036,
+      "step": 2642
+    },
+    {
+      "epoch": 3.500662251655629,
+      "grad_norm": 0.09711458068734925,
+      "learning_rate": 1.998932211995856e-05,
+      "loss": 0.4047,
+      "step": 2643
+    },
+    {
+      "epoch": 3.5019867549668873,
+      "grad_norm": 0.08379503450745822,
+      "learning_rate": 1.995729989372265e-05,
+      "loss": 0.4037,
+      "step": 2644
+    },
+    {
+      "epoch": 3.5033112582781456,
+      "grad_norm": 0.08992262846726551,
+      "learning_rate": 1.9925294809619306e-05,
+      "loss": 0.4032,
+      "step": 2645
+    },
+    {
+      "epoch": 3.504635761589404,
+      "grad_norm": 0.08796831734110297,
+      "learning_rate": 1.9893306895021824e-05,
+      "loss": 0.4019,
+      "step": 2646
+    },
+    {
+      "epoch": 3.5059602649006623,
+      "grad_norm": 0.08987505339863154,
+      "learning_rate": 1.986133617728887e-05,
+      "loss": 0.4073,
+      "step": 2647
+    },
+    {
+      "epoch": 3.5072847682119206,
+      "grad_norm": 0.08009913026790369,
+      "learning_rate": 1.982938268376437e-05,
+      "loss": 0.3992,
+      "step": 2648
+    },
+    {
+      "epoch": 3.508609271523179,
+      "grad_norm": 0.08622578088909237,
+      "learning_rate": 1.9797446441777546e-05,
+      "loss": 0.4041,
+      "step": 2649
+    },
+    {
+      "epoch": 3.5099337748344372,
+      "grad_norm": 0.08836941689230868,
+      "learning_rate": 1.9765527478642817e-05,
+      "loss": 0.4057,
+      "step": 2650
+    },
+    {
+      "epoch": 3.5112582781456956,
+      "grad_norm": 0.09571780551764646,
+      "learning_rate": 1.9733625821659877e-05,
+      "loss": 0.4024,
+      "step": 2651
+    },
+    {
+      "epoch": 3.5125827814569535,
+      "grad_norm": 0.09315749426171421,
+      "learning_rate": 1.9701741498113575e-05,
+      "loss": 0.4018,
+      "step": 2652
+    },
+    {
+      "epoch": 3.513907284768212,
+      "grad_norm": 0.08984964589252316,
+      "learning_rate": 1.966987453527396e-05,
+      "loss": 0.4048,
+      "step": 2653
+    },
+    {
+      "epoch": 3.51523178807947,
+      "grad_norm": 0.09138513705625553,
+      "learning_rate": 1.9638024960396245e-05,
+      "loss": 0.404,
+      "step": 2654
+    },
+    {
+      "epoch": 3.5165562913907285,
+      "grad_norm": 0.09253766741136983,
+      "learning_rate": 1.9606192800720723e-05,
+      "loss": 0.404,
+      "step": 2655
+    },
+    {
+      "epoch": 3.517880794701987,
+      "grad_norm": 0.0977242104790117,
+      "learning_rate": 1.9574378083472854e-05,
+      "loss": 0.4037,
+      "step": 2656
+    },
+    {
+      "epoch": 3.519205298013245,
+      "grad_norm": 0.087252554793045,
+      "learning_rate": 1.9542580835863117e-05,
+      "loss": 0.4045,
+      "step": 2657
+    },
+    {
+      "epoch": 3.5205298013245034,
+      "grad_norm": 0.0996889685412119,
+      "learning_rate": 1.951080108508711e-05,
+      "loss": 0.4044,
+      "step": 2658
+    },
+    {
+      "epoch": 3.5218543046357613,
+      "grad_norm": 0.07897104792022089,
+      "learning_rate": 1.947903885832541e-05,
+      "loss": 0.404,
+      "step": 2659
+    },
+    {
+      "epoch": 3.52317880794702,
+      "grad_norm": 0.09453949729998372,
+      "learning_rate": 1.9447294182743648e-05,
+      "loss": 0.4035,
+      "step": 2660
+    },
+    {
+      "epoch": 3.524503311258278,
+      "grad_norm": 0.08451647555215465,
+      "learning_rate": 1.941556708549243e-05,
+      "loss": 0.4026,
+      "step": 2661
+    },
+    {
+      "epoch": 3.5258278145695363,
+      "grad_norm": 0.10591880220122565,
+      "learning_rate": 1.9383857593707332e-05,
+      "loss": 0.4005,
+      "step": 2662
+    },
+    {
+      "epoch": 3.5271523178807946,
+      "grad_norm": 0.0866777248839157,
+      "learning_rate": 1.935216573450885e-05,
+      "loss": 0.3997,
+      "step": 2663
+    },
+    {
+      "epoch": 3.528476821192053,
+      "grad_norm": 0.10099780695379253,
+      "learning_rate": 1.9320491535002446e-05,
+      "loss": 0.4033,
+      "step": 2664
+    },
+    {
+      "epoch": 3.5298013245033113,
+      "grad_norm": 0.09073650966816052,
+      "learning_rate": 1.9288835022278423e-05,
+      "loss": 0.4058,
+      "step": 2665
+    },
+    {
+      "epoch": 3.5311258278145696,
+      "grad_norm": 0.0967613297992203,
+      "learning_rate": 1.925719622341198e-05,
+      "loss": 0.4038,
+      "step": 2666
+    },
+    {
+      "epoch": 3.532450331125828,
+      "grad_norm": 0.08924102468799765,
+      "learning_rate": 1.922557516546317e-05,
+      "loss": 0.4015,
+      "step": 2667
+    },
+    {
+      "epoch": 3.533774834437086,
+      "grad_norm": 0.09411153272107249,
+      "learning_rate": 1.9193971875476873e-05,
+      "loss": 0.4041,
+      "step": 2668
+    },
+    {
+      "epoch": 3.5350993377483446,
+      "grad_norm": 0.08387435599893106,
+      "learning_rate": 1.9162386380482795e-05,
+      "loss": 0.4022,
+      "step": 2669
+    },
+    {
+      "epoch": 3.5364238410596025,
+      "grad_norm": 0.08779836389823381,
+      "learning_rate": 1.9130818707495358e-05,
+      "loss": 0.4031,
+      "step": 2670
+    },
+    {
+      "epoch": 3.537748344370861,
+      "grad_norm": 0.08508943794504044,
+      "learning_rate": 1.909926888351381e-05,
+      "loss": 0.4036,
+      "step": 2671
+    },
+    {
+      "epoch": 3.539072847682119,
+      "grad_norm": 0.08398020373921557,
+      "learning_rate": 1.9067736935522075e-05,
+      "loss": 0.4061,
+      "step": 2672
+    },
+    {
+      "epoch": 3.5403973509933775,
+      "grad_norm": 0.08063381052731423,
+      "learning_rate": 1.903622289048885e-05,
+      "loss": 0.404,
+      "step": 2673
+    },
+    {
+      "epoch": 3.541721854304636,
+      "grad_norm": 0.08198597553763784,
+      "learning_rate": 1.9004726775367454e-05,
+      "loss": 0.4037,
+      "step": 2674
+    },
+    {
+      "epoch": 3.543046357615894,
+      "grad_norm": 0.07779503514081716,
+      "learning_rate": 1.8973248617095918e-05,
+      "loss": 0.3974,
+      "step": 2675
+    },
+    {
+      "epoch": 3.5443708609271525,
+      "grad_norm": 0.07607279973910284,
+      "learning_rate": 1.894178844259691e-05,
+      "loss": 0.4011,
+      "step": 2676
+    },
+    {
+      "epoch": 3.5456953642384104,
+      "grad_norm": 0.08251407661063749,
+      "learning_rate": 1.8910346278777686e-05,
+      "loss": 0.4022,
+      "step": 2677
+    },
+    {
+      "epoch": 3.547019867549669,
+      "grad_norm": 0.08321576467223794,
+      "learning_rate": 1.8878922152530142e-05,
+      "loss": 0.4007,
+      "step": 2678
+    },
+    {
+      "epoch": 3.548344370860927,
+      "grad_norm": 0.07953789552769344,
+      "learning_rate": 1.8847516090730696e-05,
+      "loss": 0.4021,
+      "step": 2679
+    },
+    {
+      "epoch": 3.5496688741721854,
+      "grad_norm": 0.07133868691780867,
+      "learning_rate": 1.8816128120240356e-05,
+      "loss": 0.4041,
+      "step": 2680
+    },
+    {
+      "epoch": 3.5509933774834437,
+      "grad_norm": 0.07301019245322564,
+      "learning_rate": 1.878475826790465e-05,
+      "loss": 0.4036,
+      "step": 2681
+    },
+    {
+      "epoch": 3.552317880794702,
+      "grad_norm": 0.07524190363748468,
+      "learning_rate": 1.8753406560553613e-05,
+      "loss": 0.4022,
+      "step": 2682
+    },
+    {
+      "epoch": 3.5536423841059603,
+      "grad_norm": 0.07026533680139667,
+      "learning_rate": 1.8722073025001727e-05,
+      "loss": 0.4042,
+      "step": 2683
+    },
+    {
+      "epoch": 3.5549668874172187,
+      "grad_norm": 0.07538507158992452,
+      "learning_rate": 1.8690757688047983e-05,
+      "loss": 0.4063,
+      "step": 2684
+    },
+    {
+      "epoch": 3.556291390728477,
+      "grad_norm": 0.06883051346212138,
+      "learning_rate": 1.8659460576475764e-05,
+      "loss": 0.4048,
+      "step": 2685
+    },
+    {
+      "epoch": 3.557615894039735,
+      "grad_norm": 0.07700161934789786,
+      "learning_rate": 1.8628181717052872e-05,
+      "loss": 0.4042,
+      "step": 2686
+    },
+    {
+      "epoch": 3.558940397350993,
+      "grad_norm": 0.0681048312890505,
+      "learning_rate": 1.859692113653153e-05,
+      "loss": 0.4014,
+      "step": 2687
+    },
+    {
+      "epoch": 3.5602649006622515,
+      "grad_norm": 0.08836472740082771,
+      "learning_rate": 1.8565678861648295e-05,
+      "loss": 0.4017,
+      "step": 2688
+    },
+    {
+      "epoch": 3.56158940397351,
+      "grad_norm": 0.07383230716771019,
+      "learning_rate": 1.8534454919124095e-05,
+      "loss": 0.4047,
+      "step": 2689
+    },
+    {
+      "epoch": 3.562913907284768,
+      "grad_norm": 0.0880204932747777,
+      "learning_rate": 1.8503249335664134e-05,
+      "loss": 0.4037,
+      "step": 2690
+    },
+    {
+      "epoch": 3.5642384105960265,
+      "grad_norm": 0.07009828475104407,
+      "learning_rate": 1.8472062137957972e-05,
+      "loss": 0.4011,
+      "step": 2691
+    },
+    {
+      "epoch": 3.565562913907285,
+      "grad_norm": 0.09360091477075817,
+      "learning_rate": 1.844089335267939e-05,
+      "loss": 0.4043,
+      "step": 2692
+    },
+    {
+      "epoch": 3.566887417218543,
+      "grad_norm": 0.0725015730348533,
+      "learning_rate": 1.840974300648646e-05,
+      "loss": 0.3996,
+      "step": 2693
+    },
+    {
+      "epoch": 3.5682119205298015,
+      "grad_norm": 0.0943716960402556,
+      "learning_rate": 1.8378611126021457e-05,
+      "loss": 0.4049,
+      "step": 2694
+    },
+    {
+      "epoch": 3.5695364238410594,
+      "grad_norm": 0.06926116977288063,
+      "learning_rate": 1.834749773791088e-05,
+      "loss": 0.4046,
+      "step": 2695
+    },
+    {
+      "epoch": 3.5708609271523177,
+      "grad_norm": 0.08566219522045808,
+      "learning_rate": 1.8316402868765422e-05,
+      "loss": 0.4033,
+      "step": 2696
+    },
+    {
+      "epoch": 3.572185430463576,
+      "grad_norm": 0.07485449156382122,
+      "learning_rate": 1.8285326545179892e-05,
+      "loss": 0.4024,
+      "step": 2697
+    },
+    {
+      "epoch": 3.5735099337748344,
+      "grad_norm": 0.08017384834028503,
+      "learning_rate": 1.8254268793733296e-05,
+      "loss": 0.4012,
+      "step": 2698
+    },
+    {
+      "epoch": 3.5748344370860927,
+      "grad_norm": 0.08149937810732648,
+      "learning_rate": 1.8223229640988694e-05,
+      "loss": 0.4013,
+      "step": 2699
+    },
+    {
+      "epoch": 3.576158940397351,
+      "grad_norm": 0.08336638386834645,
+      "learning_rate": 1.8192209113493307e-05,
+      "loss": 0.4049,
+      "step": 2700
+    },
+    {
+      "epoch": 3.5774834437086094,
+      "grad_norm": 0.07832624863539846,
+      "learning_rate": 1.8161207237778358e-05,
+      "loss": 0.4026,
+      "step": 2701
+    },
+    {
+      "epoch": 3.5788079470198677,
+      "grad_norm": 0.08135285175112106,
+      "learning_rate": 1.8130224040359162e-05,
+      "loss": 0.4014,
+      "step": 2702
+    },
+    {
+      "epoch": 3.580132450331126,
+      "grad_norm": 0.07706987205620607,
+      "learning_rate": 1.809925954773505e-05,
+      "loss": 0.399,
+      "step": 2703
+    },
+    {
+      "epoch": 3.581456953642384,
+      "grad_norm": 0.0789720407720559,
+      "learning_rate": 1.8068313786389363e-05,
+      "loss": 0.4006,
+      "step": 2704
+    },
+    {
+      "epoch": 3.5827814569536423,
+      "grad_norm": 0.07335026901441832,
+      "learning_rate": 1.8037386782789397e-05,
+      "loss": 0.4045,
+      "step": 2705
+    },
+    {
+      "epoch": 3.5841059602649006,
+      "grad_norm": 0.07757236456071089,
+      "learning_rate": 1.800647856338641e-05,
+      "loss": 0.4041,
+      "step": 2706
+    },
+    {
+      "epoch": 3.585430463576159,
+      "grad_norm": 0.07595803302691703,
+      "learning_rate": 1.7975589154615606e-05,
+      "loss": 0.4044,
+      "step": 2707
+    },
+    {
+      "epoch": 3.5867549668874172,
+      "grad_norm": 0.07580454099914491,
+      "learning_rate": 1.7944718582896104e-05,
+      "loss": 0.4063,
+      "step": 2708
+    },
+    {
+      "epoch": 3.5880794701986756,
+      "grad_norm": 0.07268115419051686,
+      "learning_rate": 1.7913866874630915e-05,
+      "loss": 0.4005,
+      "step": 2709
+    },
+    {
+      "epoch": 3.589403973509934,
+      "grad_norm": 0.07124330965477502,
+      "learning_rate": 1.788303405620688e-05,
+      "loss": 0.4035,
+      "step": 2710
+    },
+    {
+      "epoch": 3.590728476821192,
+      "grad_norm": 0.07337201494756067,
+      "learning_rate": 1.7852220153994743e-05,
+      "loss": 0.403,
+      "step": 2711
+    },
+    {
+      "epoch": 3.5920529801324506,
+      "grad_norm": 0.07334584910784764,
+      "learning_rate": 1.7821425194349006e-05,
+      "loss": 0.4049,
+      "step": 2712
+    },
+    {
+      "epoch": 3.5933774834437084,
+      "grad_norm": 0.08182649797987217,
+      "learning_rate": 1.7790649203608028e-05,
+      "loss": 0.3992,
+      "step": 2713
+    },
+    {
+      "epoch": 3.5947019867549668,
+      "grad_norm": 0.06970528201350043,
+      "learning_rate": 1.775989220809388e-05,
+      "loss": 0.401,
+      "step": 2714
+    },
+    {
+      "epoch": 3.596026490066225,
+      "grad_norm": 0.07985243454019562,
+      "learning_rate": 1.7729154234112457e-05,
+      "loss": 0.4018,
+      "step": 2715
+    },
+    {
+      "epoch": 3.5973509933774834,
+      "grad_norm": 0.06488416144150094,
+      "learning_rate": 1.769843530795335e-05,
+      "loss": 0.4033,
+      "step": 2716
+    },
+    {
+      "epoch": 3.5986754966887418,
+      "grad_norm": 0.08229960238674111,
+      "learning_rate": 1.7667735455889835e-05,
+      "loss": 0.4032,
+      "step": 2717
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.06848296865646226,
+      "learning_rate": 1.763705470417893e-05,
+      "loss": 0.4041,
+      "step": 2718
+    },
+    {
+      "epoch": 3.6013245033112584,
+      "grad_norm": 0.08283535747104054,
+      "learning_rate": 1.7606393079061265e-05,
+      "loss": 0.4004,
+      "step": 2719
+    },
+    {
+      "epoch": 3.6026490066225163,
+      "grad_norm": 0.06651856197465372,
+      "learning_rate": 1.757575060676115e-05,
+      "loss": 0.3999,
+      "step": 2720
+    },
+    {
+      "epoch": 3.603973509933775,
+      "grad_norm": 0.07370764730428896,
+      "learning_rate": 1.7545127313486477e-05,
+      "loss": 0.4018,
+      "step": 2721
+    },
+    {
+      "epoch": 3.605298013245033,
+      "grad_norm": 0.07026121480171515,
+      "learning_rate": 1.7514523225428762e-05,
+      "loss": 0.4026,
+      "step": 2722
+    },
+    {
+      "epoch": 3.6066225165562913,
+      "grad_norm": 0.07228040348821091,
+      "learning_rate": 1.748393836876309e-05,
+      "loss": 0.4033,
+      "step": 2723
+    },
+    {
+      "epoch": 3.6079470198675496,
+      "grad_norm": 0.07748128456453411,
+      "learning_rate": 1.7453372769648113e-05,
+      "loss": 0.4008,
+      "step": 2724
+    },
+    {
+      "epoch": 3.609271523178808,
+      "grad_norm": 0.0703958225407095,
+      "learning_rate": 1.742282645422598e-05,
+      "loss": 0.4018,
+      "step": 2725
+    },
+    {
+      "epoch": 3.6105960264900663,
+      "grad_norm": 0.07850769680003031,
+      "learning_rate": 1.7392299448622347e-05,
+      "loss": 0.4025,
+      "step": 2726
+    },
+    {
+      "epoch": 3.6119205298013246,
+      "grad_norm": 0.06854576068213458,
+      "learning_rate": 1.73617917789464e-05,
+      "loss": 0.4055,
+      "step": 2727
+    },
+    {
+      "epoch": 3.613245033112583,
+      "grad_norm": 0.0706325161266164,
+      "learning_rate": 1.733130347129072e-05,
+      "loss": 0.399,
+      "step": 2728
+    },
+    {
+      "epoch": 3.614569536423841,
+      "grad_norm": 0.07696769185387745,
+      "learning_rate": 1.730083455173139e-05,
+      "loss": 0.4018,
+      "step": 2729
+    },
+    {
+      "epoch": 3.6158940397350996,
+      "grad_norm": 0.07479941719273167,
+      "learning_rate": 1.727038504632788e-05,
+      "loss": 0.4029,
+      "step": 2730
+    },
+    {
+      "epoch": 3.6172185430463575,
+      "grad_norm": 0.07106488019642584,
+      "learning_rate": 1.7239954981123076e-05,
+      "loss": 0.4049,
+      "step": 2731
+    },
+    {
+      "epoch": 3.618543046357616,
+      "grad_norm": 0.0836414067268976,
+      "learning_rate": 1.7209544382143196e-05,
+      "loss": 0.4032,
+      "step": 2732
+    },
+    {
+      "epoch": 3.619867549668874,
+      "grad_norm": 0.07281084218370452,
+      "learning_rate": 1.7179153275397856e-05,
+      "loss": 0.4025,
+      "step": 2733
+    },
+    {
+      "epoch": 3.6211920529801325,
+      "grad_norm": 0.07716061056675297,
+      "learning_rate": 1.7148781686879968e-05,
+      "loss": 0.4043,
+      "step": 2734
+    },
+    {
+      "epoch": 3.622516556291391,
+      "grad_norm": 0.07151140893227753,
+      "learning_rate": 1.7118429642565758e-05,
+      "loss": 0.4007,
+      "step": 2735
+    },
+    {
+      "epoch": 3.623841059602649,
+      "grad_norm": 0.07414586305580802,
+      "learning_rate": 1.708809716841477e-05,
+      "loss": 0.4015,
+      "step": 2736
+    },
+    {
+      "epoch": 3.6251655629139075,
+      "grad_norm": 0.07016451748538376,
+      "learning_rate": 1.7057784290369743e-05,
+      "loss": 0.4032,
+      "step": 2737
+    },
+    {
+      "epoch": 3.6264900662251653,
+      "grad_norm": 0.07558952873064954,
+      "learning_rate": 1.702749103435672e-05,
+      "loss": 0.404,
+      "step": 2738
+    },
+    {
+      "epoch": 3.627814569536424,
+      "grad_norm": 0.06593774648589777,
+      "learning_rate": 1.6997217426284916e-05,
+      "loss": 0.4051,
+      "step": 2739
+    },
+    {
+      "epoch": 3.629139072847682,
+      "grad_norm": 0.08016451077735068,
+      "learning_rate": 1.6966963492046783e-05,
+      "loss": 0.4032,
+      "step": 2740
+    },
+    {
+      "epoch": 3.6304635761589403,
+      "grad_norm": 0.07060948051286069,
+      "learning_rate": 1.69367292575179e-05,
+      "loss": 0.4024,
+      "step": 2741
+    },
+    {
+      "epoch": 3.6317880794701987,
+      "grad_norm": 0.07774477926220501,
+      "learning_rate": 1.690651474855703e-05,
+      "loss": 0.4061,
+      "step": 2742
+    },
+    {
+      "epoch": 3.633112582781457,
+      "grad_norm": 0.07783736375851423,
+      "learning_rate": 1.687631999100606e-05,
+      "loss": 0.4034,
+      "step": 2743
+    },
+    {
+      "epoch": 3.6344370860927153,
+      "grad_norm": 0.07296862828773823,
+      "learning_rate": 1.684614501068999e-05,
+      "loss": 0.4017,
+      "step": 2744
+    },
+    {
+      "epoch": 3.6357615894039736,
+      "grad_norm": 0.06591107379753554,
+      "learning_rate": 1.681598983341689e-05,
+      "loss": 0.4019,
+      "step": 2745
+    },
+    {
+      "epoch": 3.637086092715232,
+      "grad_norm": 0.08607665219207229,
+      "learning_rate": 1.678585448497787e-05,
+      "loss": 0.4031,
+      "step": 2746
+    },
+    {
+      "epoch": 3.63841059602649,
+      "grad_norm": 0.06691730760116338,
+      "learning_rate": 1.6755738991147153e-05,
+      "loss": 0.4056,
+      "step": 2747
+    },
+    {
+      "epoch": 3.639735099337748,
+      "grad_norm": 0.08935393347053226,
+      "learning_rate": 1.6725643377681893e-05,
+      "loss": 0.4018,
+      "step": 2748
+    },
+    {
+      "epoch": 3.6410596026490065,
+      "grad_norm": 0.07568046807996637,
+      "learning_rate": 1.669556767032231e-05,
+      "loss": 0.4014,
+      "step": 2749
+    },
+    {
+      "epoch": 3.642384105960265,
+      "grad_norm": 0.10332894268802692,
+      "learning_rate": 1.666551189479157e-05,
+      "loss": 0.4019,
+      "step": 2750
+    },
+    {
+      "epoch": 3.643708609271523,
+      "grad_norm": 0.07217789688232593,
+      "learning_rate": 1.663547607679581e-05,
+      "loss": 0.4033,
+      "step": 2751
+    },
+    {
+      "epoch": 3.6450331125827815,
+      "grad_norm": 0.08657831479781186,
+      "learning_rate": 1.6605460242024048e-05,
+      "loss": 0.4036,
+      "step": 2752
+    },
+    {
+      "epoch": 3.64635761589404,
+      "grad_norm": 0.07609773210052878,
+      "learning_rate": 1.657546441614828e-05,
+      "loss": 0.4002,
+      "step": 2753
+    },
+    {
+      "epoch": 3.647682119205298,
+      "grad_norm": 0.09093471459670913,
+      "learning_rate": 1.654548862482332e-05,
+      "loss": 0.4021,
+      "step": 2754
+    },
+    {
+      "epoch": 3.6490066225165565,
+      "grad_norm": 0.07462455463484126,
+      "learning_rate": 1.6515532893686918e-05,
+      "loss": 0.4011,
+      "step": 2755
+    },
+    {
+      "epoch": 3.6503311258278144,
+      "grad_norm": 0.0828284892399292,
+      "learning_rate": 1.6485597248359586e-05,
+      "loss": 0.4,
+      "step": 2756
+    },
+    {
+      "epoch": 3.6516556291390727,
+      "grad_norm": 0.0743878122957689,
+      "learning_rate": 1.6455681714444732e-05,
+      "loss": 0.4025,
+      "step": 2757
+    },
+    {
+      "epoch": 3.652980132450331,
+      "grad_norm": 0.07668467194157602,
+      "learning_rate": 1.6425786317528535e-05,
+      "loss": 0.4034,
+      "step": 2758
+    },
+    {
+      "epoch": 3.6543046357615894,
+      "grad_norm": 0.079311432569385,
+      "learning_rate": 1.6395911083179926e-05,
+      "loss": 0.4018,
+      "step": 2759
+    },
+    {
+      "epoch": 3.6556291390728477,
+      "grad_norm": 0.07407826761946497,
+      "learning_rate": 1.636605603695064e-05,
+      "loss": 0.4001,
+      "step": 2760
+    },
+    {
+      "epoch": 3.656953642384106,
+      "grad_norm": 0.07445256421047851,
+      "learning_rate": 1.633622120437508e-05,
+      "loss": 0.3958,
+      "step": 2761
+    },
+    {
+      "epoch": 3.6582781456953644,
+      "grad_norm": 0.0763006704023334,
+      "learning_rate": 1.630640661097043e-05,
+      "loss": 0.4044,
+      "step": 2762
+    },
+    {
+      "epoch": 3.6596026490066222,
+      "grad_norm": 0.08047147204784517,
+      "learning_rate": 1.6276612282236523e-05,
+      "loss": 0.4045,
+      "step": 2763
+    },
+    {
+      "epoch": 3.660927152317881,
+      "grad_norm": 0.06540292112272512,
+      "learning_rate": 1.624683824365589e-05,
+      "loss": 0.4033,
+      "step": 2764
+    },
+    {
+      "epoch": 3.662251655629139,
+      "grad_norm": 0.08620961249945797,
+      "learning_rate": 1.621708452069366e-05,
+      "loss": 0.4052,
+      "step": 2765
+    },
+    {
+      "epoch": 3.6635761589403972,
+      "grad_norm": 0.07132411017299689,
+      "learning_rate": 1.6187351138797637e-05,
+      "loss": 0.4023,
+      "step": 2766
+    },
+    {
+      "epoch": 3.6649006622516556,
+      "grad_norm": 0.06973657830998323,
+      "learning_rate": 1.6157638123398192e-05,
+      "loss": 0.4017,
+      "step": 2767
+    },
+    {
+      "epoch": 3.666225165562914,
+      "grad_norm": 0.07762098306774316,
+      "learning_rate": 1.6127945499908286e-05,
+      "loss": 0.4014,
+      "step": 2768
+    },
+    {
+      "epoch": 3.667549668874172,
+      "grad_norm": 0.07358935781601827,
+      "learning_rate": 1.609827329372345e-05,
+      "loss": 0.4018,
+      "step": 2769
+    },
+    {
+      "epoch": 3.6688741721854305,
+      "grad_norm": 0.06865442939954434,
+      "learning_rate": 1.6068621530221752e-05,
+      "loss": 0.405,
+      "step": 2770
+    },
+    {
+      "epoch": 3.670198675496689,
+      "grad_norm": 0.07179642477356453,
+      "learning_rate": 1.6038990234763776e-05,
+      "loss": 0.4028,
+      "step": 2771
+    },
+    {
+      "epoch": 3.6715231788079468,
+      "grad_norm": 0.07686177188064675,
+      "learning_rate": 1.600937943269257e-05,
+      "loss": 0.4041,
+      "step": 2772
+    },
+    {
+      "epoch": 3.6728476821192055,
+      "grad_norm": 0.06623236057363172,
+      "learning_rate": 1.5979789149333713e-05,
+      "loss": 0.4041,
+      "step": 2773
+    },
+    {
+      "epoch": 3.6741721854304634,
+      "grad_norm": 0.06837633817160997,
+      "learning_rate": 1.595021940999516e-05,
+      "loss": 0.4033,
+      "step": 2774
+    },
+    {
+      "epoch": 3.6754966887417218,
+      "grad_norm": 0.07654948959291114,
+      "learning_rate": 1.5920670239967374e-05,
+      "loss": 0.3982,
+      "step": 2775
+    },
+    {
+      "epoch": 3.67682119205298,
+      "grad_norm": 0.07054686274010741,
+      "learning_rate": 1.589114166452315e-05,
+      "loss": 0.4058,
+      "step": 2776
+    },
+    {
+      "epoch": 3.6781456953642384,
+      "grad_norm": 0.07110592523732093,
+      "learning_rate": 1.5861633708917724e-05,
+      "loss": 0.4024,
+      "step": 2777
+    },
+    {
+      "epoch": 3.6794701986754967,
+      "grad_norm": 0.07549245522826879,
+      "learning_rate": 1.5832146398388695e-05,
+      "loss": 0.4052,
+      "step": 2778
+    },
+    {
+      "epoch": 3.680794701986755,
+      "grad_norm": 0.07874510149912005,
+      "learning_rate": 1.5802679758155958e-05,
+      "loss": 0.4023,
+      "step": 2779
+    },
+    {
+      "epoch": 3.6821192052980134,
+      "grad_norm": 0.06821501288061792,
+      "learning_rate": 1.5773233813421782e-05,
+      "loss": 0.4026,
+      "step": 2780
+    },
+    {
+      "epoch": 3.6834437086092713,
+      "grad_norm": 0.0794990233672452,
+      "learning_rate": 1.5743808589370687e-05,
+      "loss": 0.4005,
+      "step": 2781
+    },
+    {
+      "epoch": 3.68476821192053,
+      "grad_norm": 0.08221482620339925,
+      "learning_rate": 1.571440411116953e-05,
+      "loss": 0.403,
+      "step": 2782
+    },
+    {
+      "epoch": 3.686092715231788,
+      "grad_norm": 0.06896464001867653,
+      "learning_rate": 1.5685020403967354e-05,
+      "loss": 0.4004,
+      "step": 2783
+    },
+    {
+      "epoch": 3.6874172185430463,
+      "grad_norm": 0.07184632003759685,
+      "learning_rate": 1.565565749289549e-05,
+      "loss": 0.4026,
+      "step": 2784
+    },
+    {
+      "epoch": 3.6887417218543046,
+      "grad_norm": 0.07075113320732666,
+      "learning_rate": 1.5626315403067473e-05,
+      "loss": 0.403,
+      "step": 2785
+    },
+    {
+      "epoch": 3.690066225165563,
+      "grad_norm": 0.06644153262214307,
+      "learning_rate": 1.5596994159579025e-05,
+      "loss": 0.4018,
+      "step": 2786
+    },
+    {
+      "epoch": 3.6913907284768213,
+      "grad_norm": 0.07214710750338038,
+      "learning_rate": 1.5567693787508036e-05,
+      "loss": 0.4028,
+      "step": 2787
+    },
+    {
+      "epoch": 3.6927152317880796,
+      "grad_norm": 0.07270094867834175,
+      "learning_rate": 1.553841431191452e-05,
+      "loss": 0.4045,
+      "step": 2788
+    },
+    {
+      "epoch": 3.694039735099338,
+      "grad_norm": 0.07293122136430172,
+      "learning_rate": 1.5509155757840667e-05,
+      "loss": 0.4045,
+      "step": 2789
+    },
+    {
+      "epoch": 3.695364238410596,
+      "grad_norm": 0.06993685955522767,
+      "learning_rate": 1.5479918150310743e-05,
+      "loss": 0.4016,
+      "step": 2790
+    },
+    {
+      "epoch": 3.6966887417218546,
+      "grad_norm": 0.072723615634079,
+      "learning_rate": 1.5450701514331132e-05,
+      "loss": 0.401,
+      "step": 2791
+    },
+    {
+      "epoch": 3.6980132450331125,
+      "grad_norm": 0.073546654867496,
+      "learning_rate": 1.5421505874890216e-05,
+      "loss": 0.4003,
+      "step": 2792
+    },
+    {
+      "epoch": 3.699337748344371,
+      "grad_norm": 0.0807147097475966,
+      "learning_rate": 1.5392331256958493e-05,
+      "loss": 0.4032,
+      "step": 2793
+    },
+    {
+      "epoch": 3.700662251655629,
+      "grad_norm": 0.06779163002249068,
+      "learning_rate": 1.5363177685488415e-05,
+      "loss": 0.3985,
+      "step": 2794
+    },
+    {
+      "epoch": 3.7019867549668874,
+      "grad_norm": 0.07180568153742786,
+      "learning_rate": 1.5334045185414498e-05,
+      "loss": 0.3979,
+      "step": 2795
+    },
+    {
+      "epoch": 3.703311258278146,
+      "grad_norm": 0.0643375732267641,
+      "learning_rate": 1.5304933781653178e-05,
+      "loss": 0.4033,
+      "step": 2796
+    },
+    {
+      "epoch": 3.704635761589404,
+      "grad_norm": 0.07078196083238612,
+      "learning_rate": 1.5275843499102886e-05,
+      "loss": 0.4031,
+      "step": 2797
+    },
+    {
+      "epoch": 3.7059602649006624,
+      "grad_norm": 0.06440837519656563,
+      "learning_rate": 1.5246774362643985e-05,
+      "loss": 0.4081,
+      "step": 2798
+    },
+    {
+      "epoch": 3.7072847682119203,
+      "grad_norm": 0.07502340363062797,
+      "learning_rate": 1.5217726397138725e-05,
+      "loss": 0.4011,
+      "step": 2799
+    },
+    {
+      "epoch": 3.7086092715231787,
+      "grad_norm": 0.06471849483872212,
+      "learning_rate": 1.5188699627431288e-05,
+      "loss": 0.4009,
+      "step": 2800
+    },
+    {
+      "epoch": 3.709933774834437,
+      "grad_norm": 0.068858833788632,
+      "learning_rate": 1.5159694078347685e-05,
+      "loss": 0.3997,
+      "step": 2801
+    },
+    {
+      "epoch": 3.7112582781456953,
+      "grad_norm": 0.0666313163388003,
+      "learning_rate": 1.5130709774695817e-05,
+      "loss": 0.4023,
+      "step": 2802
+    },
+    {
+      "epoch": 3.7125827814569536,
+      "grad_norm": 0.07185486205039612,
+      "learning_rate": 1.510174674126538e-05,
+      "loss": 0.4002,
+      "step": 2803
+    },
+    {
+      "epoch": 3.713907284768212,
+      "grad_norm": 0.06964861293617845,
+      "learning_rate": 1.5072805002827902e-05,
+      "loss": 0.4041,
+      "step": 2804
+    },
+    {
+      "epoch": 3.7152317880794703,
+      "grad_norm": 0.06694884591648373,
+      "learning_rate": 1.5043884584136695e-05,
+      "loss": 0.3959,
+      "step": 2805
+    },
+    {
+      "epoch": 3.7165562913907286,
+      "grad_norm": 0.06789205963259771,
+      "learning_rate": 1.5014985509926843e-05,
+      "loss": 0.4015,
+      "step": 2806
+    },
+    {
+      "epoch": 3.717880794701987,
+      "grad_norm": 0.061976463693040953,
+      "learning_rate": 1.4986107804915148e-05,
+      "loss": 0.406,
+      "step": 2807
+    },
+    {
+      "epoch": 3.719205298013245,
+      "grad_norm": 0.0812872379004388,
+      "learning_rate": 1.495725149380014e-05,
+      "loss": 0.4029,
+      "step": 2808
+    },
+    {
+      "epoch": 3.720529801324503,
+      "grad_norm": 0.06607593220399426,
+      "learning_rate": 1.4928416601262089e-05,
+      "loss": 0.4008,
+      "step": 2809
+    },
+    {
+      "epoch": 3.7218543046357615,
+      "grad_norm": 0.07193701154740731,
+      "learning_rate": 1.4899603151962886e-05,
+      "loss": 0.4033,
+      "step": 2810
+    },
+    {
+      "epoch": 3.72317880794702,
+      "grad_norm": 0.05982563096229403,
+      "learning_rate": 1.487081117054614e-05,
+      "loss": 0.4038,
+      "step": 2811
+    },
+    {
+      "epoch": 3.724503311258278,
+      "grad_norm": 0.06608875655225581,
+      "learning_rate": 1.484204068163707e-05,
+      "loss": 0.4014,
+      "step": 2812
+    },
+    {
+      "epoch": 3.7258278145695365,
+      "grad_norm": 0.06534542567622434,
+      "learning_rate": 1.4813291709842524e-05,
+      "loss": 0.4046,
+      "step": 2813
+    },
+    {
+      "epoch": 3.727152317880795,
+      "grad_norm": 0.07469406805893081,
+      "learning_rate": 1.4784564279750919e-05,
+      "loss": 0.3991,
+      "step": 2814
+    },
+    {
+      "epoch": 3.7284768211920527,
+      "grad_norm": 0.061688705341919114,
+      "learning_rate": 1.4755858415932295e-05,
+      "loss": 0.4,
+      "step": 2815
+    },
+    {
+      "epoch": 3.7298013245033115,
+      "grad_norm": 0.0721078072479029,
+      "learning_rate": 1.4727174142938191e-05,
+      "loss": 0.4027,
+      "step": 2816
+    },
+    {
+      "epoch": 3.7311258278145694,
+      "grad_norm": 0.06533759872608708,
+      "learning_rate": 1.4698511485301725e-05,
+      "loss": 0.4035,
+      "step": 2817
+    },
+    {
+      "epoch": 3.7324503311258277,
+      "grad_norm": 0.06969503565047425,
+      "learning_rate": 1.4669870467537526e-05,
+      "loss": 0.3981,
+      "step": 2818
+    },
+    {
+      "epoch": 3.733774834437086,
+      "grad_norm": 0.06669516456809087,
+      "learning_rate": 1.464125111414167e-05,
+      "loss": 0.4078,
+      "step": 2819
+    },
+    {
+      "epoch": 3.7350993377483444,
+      "grad_norm": 0.07351350977742928,
+      "learning_rate": 1.4612653449591764e-05,
+      "loss": 0.4044,
+      "step": 2820
+    },
+    {
+      "epoch": 3.7364238410596027,
+      "grad_norm": 0.07065438760324726,
+      "learning_rate": 1.4584077498346814e-05,
+      "loss": 0.4054,
+      "step": 2821
+    },
+    {
+      "epoch": 3.737748344370861,
+      "grad_norm": 0.0634130919048118,
+      "learning_rate": 1.4555523284847293e-05,
+      "loss": 0.4008,
+      "step": 2822
+    },
+    {
+      "epoch": 3.7390728476821193,
+      "grad_norm": 0.07038715165800637,
+      "learning_rate": 1.452699083351505e-05,
+      "loss": 0.4041,
+      "step": 2823
+    },
+    {
+      "epoch": 3.7403973509933772,
+      "grad_norm": 0.06581420032144297,
+      "learning_rate": 1.449848016875334e-05,
+      "loss": 0.4009,
+      "step": 2824
+    },
+    {
+      "epoch": 3.741721854304636,
+      "grad_norm": 0.0648992757840232,
+      "learning_rate": 1.4469991314946788e-05,
+      "loss": 0.4028,
+      "step": 2825
+    },
+    {
+      "epoch": 3.743046357615894,
+      "grad_norm": 0.06881568747936907,
+      "learning_rate": 1.4441524296461368e-05,
+      "loss": 0.4016,
+      "step": 2826
+    },
+    {
+      "epoch": 3.744370860927152,
+      "grad_norm": 0.06262526248630869,
+      "learning_rate": 1.4413079137644358e-05,
+      "loss": 0.4016,
+      "step": 2827
+    },
+    {
+      "epoch": 3.7456953642384105,
+      "grad_norm": 0.07247915966395997,
+      "learning_rate": 1.4384655862824337e-05,
+      "loss": 0.3998,
+      "step": 2828
+    },
+    {
+      "epoch": 3.747019867549669,
+      "grad_norm": 0.06599771551520396,
+      "learning_rate": 1.4356254496311208e-05,
+      "loss": 0.4057,
+      "step": 2829
+    },
+    {
+      "epoch": 3.748344370860927,
+      "grad_norm": 0.0710820095746011,
+      "learning_rate": 1.432787506239608e-05,
+      "loss": 0.4023,
+      "step": 2830
+    },
+    {
+      "epoch": 3.7496688741721855,
+      "grad_norm": 0.06413366108623028,
+      "learning_rate": 1.4299517585351343e-05,
+      "loss": 0.4043,
+      "step": 2831
+    },
+    {
+      "epoch": 3.750993377483444,
+      "grad_norm": 0.06684087170283486,
+      "learning_rate": 1.4271182089430605e-05,
+      "loss": 0.4018,
+      "step": 2832
+    },
+    {
+      "epoch": 3.7523178807947017,
+      "grad_norm": 0.06539395561722647,
+      "learning_rate": 1.4242868598868672e-05,
+      "loss": 0.4024,
+      "step": 2833
+    },
+    {
+      "epoch": 3.7536423841059605,
+      "grad_norm": 0.07526035213150198,
+      "learning_rate": 1.4214577137881493e-05,
+      "loss": 0.403,
+      "step": 2834
+    },
+    {
+      "epoch": 3.7549668874172184,
+      "grad_norm": 0.06262257079312994,
+      "learning_rate": 1.4186307730666244e-05,
+      "loss": 0.4017,
+      "step": 2835
+    },
+    {
+      "epoch": 3.7562913907284767,
+      "grad_norm": 0.06558538534162613,
+      "learning_rate": 1.4158060401401167e-05,
+      "loss": 0.3995,
+      "step": 2836
+    },
+    {
+      "epoch": 3.757615894039735,
+      "grad_norm": 0.06528123769707747,
+      "learning_rate": 1.4129835174245682e-05,
+      "loss": 0.4023,
+      "step": 2837
+    },
+    {
+      "epoch": 3.7589403973509934,
+      "grad_norm": 0.06904399939683989,
+      "learning_rate": 1.4101632073340254e-05,
+      "loss": 0.4036,
+      "step": 2838
+    },
+    {
+      "epoch": 3.7602649006622517,
+      "grad_norm": 0.056364821460412474,
+      "learning_rate": 1.4073451122806456e-05,
+      "loss": 0.4076,
+      "step": 2839
+    },
+    {
+      "epoch": 3.76158940397351,
+      "grad_norm": 0.07016764674336255,
+      "learning_rate": 1.4045292346746929e-05,
+      "loss": 0.4039,
+      "step": 2840
+    },
+    {
+      "epoch": 3.7629139072847684,
+      "grad_norm": 0.06738547784881524,
+      "learning_rate": 1.4017155769245289e-05,
+      "loss": 0.4034,
+      "step": 2841
+    },
+    {
+      "epoch": 3.7642384105960263,
+      "grad_norm": 0.06648120358279239,
+      "learning_rate": 1.3989041414366246e-05,
+      "loss": 0.3996,
+      "step": 2842
+    },
+    {
+      "epoch": 3.765562913907285,
+      "grad_norm": 0.06894065810927363,
+      "learning_rate": 1.3960949306155432e-05,
+      "loss": 0.4035,
+      "step": 2843
+    },
+    {
+      "epoch": 3.766887417218543,
+      "grad_norm": 0.06598234674446392,
+      "learning_rate": 1.3932879468639487e-05,
+      "loss": 0.3993,
+      "step": 2844
+    },
+    {
+      "epoch": 3.7682119205298013,
+      "grad_norm": 0.07441278582184649,
+      "learning_rate": 1.3904831925826003e-05,
+      "loss": 0.4062,
+      "step": 2845
+    },
+    {
+      "epoch": 3.7695364238410596,
+      "grad_norm": 0.06582317666608718,
+      "learning_rate": 1.3876806701703519e-05,
+      "loss": 0.4032,
+      "step": 2846
+    },
+    {
+      "epoch": 3.770860927152318,
+      "grad_norm": 0.06660863163990607,
+      "learning_rate": 1.3848803820241443e-05,
+      "loss": 0.3997,
+      "step": 2847
+    },
+    {
+      "epoch": 3.7721854304635762,
+      "grad_norm": 0.0734658925558501,
+      "learning_rate": 1.3820823305390088e-05,
+      "loss": 0.399,
+      "step": 2848
+    },
+    {
+      "epoch": 3.7735099337748346,
+      "grad_norm": 0.06763713432897966,
+      "learning_rate": 1.3792865181080668e-05,
+      "loss": 0.4019,
+      "step": 2849
+    },
+    {
+      "epoch": 3.774834437086093,
+      "grad_norm": 0.06442647731821995,
+      "learning_rate": 1.3764929471225203e-05,
+      "loss": 0.401,
+      "step": 2850
+    },
+    {
+      "epoch": 3.776158940397351,
+      "grad_norm": 0.07436068658248038,
+      "learning_rate": 1.3737016199716574e-05,
+      "loss": 0.3976,
+      "step": 2851
+    },
+    {
+      "epoch": 3.777483443708609,
+      "grad_norm": 0.0631584281884298,
+      "learning_rate": 1.3709125390428462e-05,
+      "loss": 0.4033,
+      "step": 2852
+    },
+    {
+      "epoch": 3.7788079470198674,
+      "grad_norm": 0.07053623746831919,
+      "learning_rate": 1.3681257067215347e-05,
+      "loss": 0.3992,
+      "step": 2853
+    },
+    {
+      "epoch": 3.7801324503311258,
+      "grad_norm": 0.06480153686810405,
+      "learning_rate": 1.3653411253912436e-05,
+      "loss": 0.4023,
+      "step": 2854
+    },
+    {
+      "epoch": 3.781456953642384,
+      "grad_norm": 0.06733203697127803,
+      "learning_rate": 1.3625587974335748e-05,
+      "loss": 0.3979,
+      "step": 2855
+    },
+    {
+      "epoch": 3.7827814569536424,
+      "grad_norm": 0.06092869697991686,
+      "learning_rate": 1.3597787252281958e-05,
+      "loss": 0.4032,
+      "step": 2856
+    },
+    {
+      "epoch": 3.7841059602649008,
+      "grad_norm": 0.06195198671209535,
+      "learning_rate": 1.3570009111528517e-05,
+      "loss": 0.4061,
+      "step": 2857
+    },
+    {
+      "epoch": 3.785430463576159,
+      "grad_norm": 0.06488078289048507,
+      "learning_rate": 1.3542253575833498e-05,
+      "loss": 0.4034,
+      "step": 2858
+    },
+    {
+      "epoch": 3.7867549668874174,
+      "grad_norm": 0.0708713869044741,
+      "learning_rate": 1.3514520668935692e-05,
+      "loss": 0.4018,
+      "step": 2859
+    },
+    {
+      "epoch": 3.7880794701986753,
+      "grad_norm": 0.060739157109561584,
+      "learning_rate": 1.3486810414554525e-05,
+      "loss": 0.4047,
+      "step": 2860
+    },
+    {
+      "epoch": 3.7894039735099336,
+      "grad_norm": 0.07054389540384395,
+      "learning_rate": 1.3459122836390015e-05,
+      "loss": 0.4017,
+      "step": 2861
+    },
+    {
+      "epoch": 3.790728476821192,
+      "grad_norm": 0.06268726919103913,
+      "learning_rate": 1.3431457958122835e-05,
+      "loss": 0.4017,
+      "step": 2862
+    },
+    {
+      "epoch": 3.7920529801324503,
+      "grad_norm": 0.060297673953707207,
+      "learning_rate": 1.3403815803414192e-05,
+      "loss": 0.403,
+      "step": 2863
+    },
+    {
+      "epoch": 3.7933774834437086,
+      "grad_norm": 0.06019005567731638,
+      "learning_rate": 1.3376196395905918e-05,
+      "loss": 0.4019,
+      "step": 2864
+    },
+    {
+      "epoch": 3.794701986754967,
+      "grad_norm": 0.06519379840774463,
+      "learning_rate": 1.3348599759220307e-05,
+      "loss": 0.4039,
+      "step": 2865
+    },
+    {
+      "epoch": 3.7960264900662253,
+      "grad_norm": 0.05915791383246683,
+      "learning_rate": 1.3321025916960286e-05,
+      "loss": 0.3987,
+      "step": 2866
+    },
+    {
+      "epoch": 3.7973509933774836,
+      "grad_norm": 0.06516088120495411,
+      "learning_rate": 1.3293474892709188e-05,
+      "loss": 0.3995,
+      "step": 2867
+    },
+    {
+      "epoch": 3.798675496688742,
+      "grad_norm": 0.059784318277652466,
+      "learning_rate": 1.3265946710030893e-05,
+      "loss": 0.4028,
+      "step": 2868
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.06702964991952103,
+      "learning_rate": 1.3238441392469708e-05,
+      "loss": 0.4031,
+      "step": 2869
+    },
+    {
+      "epoch": 3.801324503311258,
+      "grad_norm": 0.06445990209842625,
+      "learning_rate": 1.3210958963550384e-05,
+      "loss": 0.403,
+      "step": 2870
+    },
+    {
+      "epoch": 3.8026490066225165,
+      "grad_norm": 0.06633688027232391,
+      "learning_rate": 1.3183499446778125e-05,
+      "loss": 0.4037,
+      "step": 2871
+    },
+    {
+      "epoch": 3.803973509933775,
+      "grad_norm": 0.0625918823156593,
+      "learning_rate": 1.3156062865638526e-05,
+      "loss": 0.403,
+      "step": 2872
+    },
+    {
+      "epoch": 3.805298013245033,
+      "grad_norm": 0.06525654946338523,
+      "learning_rate": 1.3128649243597567e-05,
+      "loss": 0.3993,
+      "step": 2873
+    },
+    {
+      "epoch": 3.8066225165562915,
+      "grad_norm": 0.06434562989451502,
+      "learning_rate": 1.3101258604101568e-05,
+      "loss": 0.4031,
+      "step": 2874
+    },
+    {
+      "epoch": 3.80794701986755,
+      "grad_norm": 0.06575214871385654,
+      "learning_rate": 1.3073890970577234e-05,
+      "loss": 0.4026,
+      "step": 2875
+    },
+    {
+      "epoch": 3.8092715231788077,
+      "grad_norm": 0.07380032021357245,
+      "learning_rate": 1.3046546366431549e-05,
+      "loss": 0.4035,
+      "step": 2876
+    },
+    {
+      "epoch": 3.8105960264900665,
+      "grad_norm": 0.06263214799852966,
+      "learning_rate": 1.301922481505184e-05,
+      "loss": 0.4043,
+      "step": 2877
+    },
+    {
+      "epoch": 3.8119205298013243,
+      "grad_norm": 0.06771087959986266,
+      "learning_rate": 1.2991926339805683e-05,
+      "loss": 0.4041,
+      "step": 2878
+    },
+    {
+      "epoch": 3.8132450331125827,
+      "grad_norm": 0.06652103492747177,
+      "learning_rate": 1.2964650964040945e-05,
+      "loss": 0.4017,
+      "step": 2879
+    },
+    {
+      "epoch": 3.814569536423841,
+      "grad_norm": 0.06119414319828468,
+      "learning_rate": 1.2937398711085738e-05,
+      "loss": 0.4003,
+      "step": 2880
+    },
+    {
+      "epoch": 3.8158940397350993,
+      "grad_norm": 0.06525472492079559,
+      "learning_rate": 1.2910169604248357e-05,
+      "loss": 0.4003,
+      "step": 2881
+    },
+    {
+      "epoch": 3.8172185430463577,
+      "grad_norm": 0.06980531189910043,
+      "learning_rate": 1.2882963666817356e-05,
+      "loss": 0.403,
+      "step": 2882
+    },
+    {
+      "epoch": 3.818543046357616,
+      "grad_norm": 0.06253117911526623,
+      "learning_rate": 1.2855780922061417e-05,
+      "loss": 0.4026,
+      "step": 2883
+    },
+    {
+      "epoch": 3.8198675496688743,
+      "grad_norm": 0.060701887317399596,
+      "learning_rate": 1.2828621393229445e-05,
+      "loss": 0.3989,
+      "step": 2884
+    },
+    {
+      "epoch": 3.821192052980132,
+      "grad_norm": 0.06213558551447263,
+      "learning_rate": 1.2801485103550433e-05,
+      "loss": 0.4013,
+      "step": 2885
+    },
+    {
+      "epoch": 3.822516556291391,
+      "grad_norm": 0.07216928103504071,
+      "learning_rate": 1.2774372076233532e-05,
+      "loss": 0.4021,
+      "step": 2886
+    },
+    {
+      "epoch": 3.823841059602649,
+      "grad_norm": 0.058932988084277466,
+      "learning_rate": 1.274728233446799e-05,
+      "loss": 0.4006,
+      "step": 2887
+    },
+    {
+      "epoch": 3.825165562913907,
+      "grad_norm": 0.06528832416745486,
+      "learning_rate": 1.2720215901423147e-05,
+      "loss": 0.4035,
+      "step": 2888
+    },
+    {
+      "epoch": 3.8264900662251655,
+      "grad_norm": 0.06668834936406041,
+      "learning_rate": 1.2693172800248394e-05,
+      "loss": 0.3964,
+      "step": 2889
+    },
+    {
+      "epoch": 3.827814569536424,
+      "grad_norm": 0.07239039880088244,
+      "learning_rate": 1.2666153054073149e-05,
+      "loss": 0.4006,
+      "step": 2890
+    },
+    {
+      "epoch": 3.829139072847682,
+      "grad_norm": 0.058772308726574955,
+      "learning_rate": 1.2639156686006904e-05,
+      "loss": 0.4013,
+      "step": 2891
+    },
+    {
+      "epoch": 3.8304635761589405,
+      "grad_norm": 0.06491816841926848,
+      "learning_rate": 1.2612183719139091e-05,
+      "loss": 0.4012,
+      "step": 2892
+    },
+    {
+      "epoch": 3.831788079470199,
+      "grad_norm": 0.06587801214109976,
+      "learning_rate": 1.2585234176539221e-05,
+      "loss": 0.397,
+      "step": 2893
+    },
+    {
+      "epoch": 3.8331125827814567,
+      "grad_norm": 0.06312981737963859,
+      "learning_rate": 1.255830808125667e-05,
+      "loss": 0.4024,
+      "step": 2894
+    },
+    {
+      "epoch": 3.8344370860927155,
+      "grad_norm": 0.07780481319480968,
+      "learning_rate": 1.2531405456320837e-05,
+      "loss": 0.4028,
+      "step": 2895
+    },
+    {
+      "epoch": 3.8357615894039734,
+      "grad_norm": 0.06496230089588123,
+      "learning_rate": 1.2504526324740982e-05,
+      "loss": 0.4011,
+      "step": 2896
+    },
+    {
+      "epoch": 3.8370860927152317,
+      "grad_norm": 0.07155067009632768,
+      "learning_rate": 1.2477670709506335e-05,
+      "loss": 0.4013,
+      "step": 2897
+    },
+    {
+      "epoch": 3.83841059602649,
+      "grad_norm": 0.058830116304708205,
+      "learning_rate": 1.245083863358596e-05,
+      "loss": 0.4013,
+      "step": 2898
+    },
+    {
+      "epoch": 3.8397350993377484,
+      "grad_norm": 0.06964879614837148,
+      "learning_rate": 1.2424030119928819e-05,
+      "loss": 0.4005,
+      "step": 2899
+    },
+    {
+      "epoch": 3.8410596026490067,
+      "grad_norm": 0.06348006799526085,
+      "learning_rate": 1.2397245191463724e-05,
+      "loss": 0.4,
+      "step": 2900
+    },
+    {
+      "epoch": 3.842384105960265,
+      "grad_norm": 0.06275456952972068,
+      "learning_rate": 1.2370483871099284e-05,
+      "loss": 0.4031,
+      "step": 2901
+    },
+    {
+      "epoch": 3.8437086092715234,
+      "grad_norm": 0.06818468342949366,
+      "learning_rate": 1.2343746181723964e-05,
+      "loss": 0.4001,
+      "step": 2902
+    },
+    {
+      "epoch": 3.8450331125827812,
+      "grad_norm": 0.07436351933747067,
+      "learning_rate": 1.231703214620596e-05,
+      "loss": 0.4049,
+      "step": 2903
+    },
+    {
+      "epoch": 3.8463576158940396,
+      "grad_norm": 0.06878802607309677,
+      "learning_rate": 1.2290341787393297e-05,
+      "loss": 0.401,
+      "step": 2904
+    },
+    {
+      "epoch": 3.847682119205298,
+      "grad_norm": 0.06364786266364714,
+      "learning_rate": 1.2263675128113697e-05,
+      "loss": 0.4036,
+      "step": 2905
+    },
+    {
+      "epoch": 3.8490066225165562,
+      "grad_norm": 0.07627436180337482,
+      "learning_rate": 1.2237032191174642e-05,
+      "loss": 0.4012,
+      "step": 2906
+    },
+    {
+      "epoch": 3.8503311258278146,
+      "grad_norm": 0.06253121301036035,
+      "learning_rate": 1.2210412999363333e-05,
+      "loss": 0.4,
+      "step": 2907
+    },
+    {
+      "epoch": 3.851655629139073,
+      "grad_norm": 0.06715042022516957,
+      "learning_rate": 1.2183817575446647e-05,
+      "loss": 0.4029,
+      "step": 2908
+    },
+    {
+      "epoch": 3.852980132450331,
+      "grad_norm": 0.06322280517868324,
+      "learning_rate": 1.2157245942171128e-05,
+      "loss": 0.4004,
+      "step": 2909
+    },
+    {
+      "epoch": 3.8543046357615895,
+      "grad_norm": 0.06608744296781426,
+      "learning_rate": 1.213069812226296e-05,
+      "loss": 0.401,
+      "step": 2910
+    },
+    {
+      "epoch": 3.855629139072848,
+      "grad_norm": 0.06599654266497208,
+      "learning_rate": 1.2104174138428015e-05,
+      "loss": 0.4042,
+      "step": 2911
+    },
+    {
+      "epoch": 3.8569536423841058,
+      "grad_norm": 0.06175836477289363,
+      "learning_rate": 1.2077674013351706e-05,
+      "loss": 0.4029,
+      "step": 2912
+    },
+    {
+      "epoch": 3.858278145695364,
+      "grad_norm": 0.06618570288721021,
+      "learning_rate": 1.2051197769699092e-05,
+      "loss": 0.403,
+      "step": 2913
+    },
+    {
+      "epoch": 3.8596026490066224,
+      "grad_norm": 0.0627099537428428,
+      "learning_rate": 1.202474543011479e-05,
+      "loss": 0.4026,
+      "step": 2914
+    },
+    {
+      "epoch": 3.8609271523178808,
+      "grad_norm": 0.06656252835961865,
+      "learning_rate": 1.1998317017222992e-05,
+      "loss": 0.4044,
+      "step": 2915
+    },
+    {
+      "epoch": 3.862251655629139,
+      "grad_norm": 0.06115536109659195,
+      "learning_rate": 1.1971912553627369e-05,
+      "loss": 0.4019,
+      "step": 2916
+    },
+    {
+      "epoch": 3.8635761589403974,
+      "grad_norm": 0.0710633460827064,
+      "learning_rate": 1.194553206191118e-05,
+      "loss": 0.4019,
+      "step": 2917
+    },
+    {
+      "epoch": 3.8649006622516557,
+      "grad_norm": 0.06311112242575166,
+      "learning_rate": 1.1919175564637122e-05,
+      "loss": 0.4023,
+      "step": 2918
+    },
+    {
+      "epoch": 3.866225165562914,
+      "grad_norm": 0.06482487646318136,
+      "learning_rate": 1.1892843084347407e-05,
+      "loss": 0.4042,
+      "step": 2919
+    },
+    {
+      "epoch": 3.8675496688741724,
+      "grad_norm": 0.06992011678453404,
+      "learning_rate": 1.1866534643563706e-05,
+      "loss": 0.4006,
+      "step": 2920
+    },
+    {
+      "epoch": 3.8688741721854303,
+      "grad_norm": 0.06597471605412433,
+      "learning_rate": 1.1840250264787096e-05,
+      "loss": 0.4006,
+      "step": 2921
+    },
+    {
+      "epoch": 3.8701986754966886,
+      "grad_norm": 0.05905664602544527,
+      "learning_rate": 1.1813989970498119e-05,
+      "loss": 0.4035,
+      "step": 2922
+    },
+    {
+      "epoch": 3.871523178807947,
+      "grad_norm": 0.06597405318810448,
+      "learning_rate": 1.1787753783156669e-05,
+      "loss": 0.4037,
+      "step": 2923
+    },
+    {
+      "epoch": 3.8728476821192053,
+      "grad_norm": 0.06190616704643098,
+      "learning_rate": 1.1761541725202075e-05,
+      "loss": 0.4035,
+      "step": 2924
+    },
+    {
+      "epoch": 3.8741721854304636,
+      "grad_norm": 0.0609760777026225,
+      "learning_rate": 1.173535381905298e-05,
+      "loss": 0.3994,
+      "step": 2925
+    },
+    {
+      "epoch": 3.875496688741722,
+      "grad_norm": 0.06163050666579174,
+      "learning_rate": 1.1709190087107398e-05,
+      "loss": 0.4061,
+      "step": 2926
+    },
+    {
+      "epoch": 3.8768211920529803,
+      "grad_norm": 0.06846530064285344,
+      "learning_rate": 1.1683050551742662e-05,
+      "loss": 0.404,
+      "step": 2927
+    },
+    {
+      "epoch": 3.878145695364238,
+      "grad_norm": 0.05715470489543213,
+      "learning_rate": 1.1656935235315427e-05,
+      "loss": 0.4036,
+      "step": 2928
+    },
+    {
+      "epoch": 3.879470198675497,
+      "grad_norm": 0.06509091228308139,
+      "learning_rate": 1.1630844160161603e-05,
+      "loss": 0.4029,
+      "step": 2929
+    },
+    {
+      "epoch": 3.880794701986755,
+      "grad_norm": 0.06359031357703979,
+      "learning_rate": 1.1604777348596365e-05,
+      "loss": 0.3989,
+      "step": 2930
+    },
+    {
+      "epoch": 3.882119205298013,
+      "grad_norm": 0.07262675046802262,
+      "learning_rate": 1.1578734822914175e-05,
+      "loss": 0.398,
+      "step": 2931
+    },
+    {
+      "epoch": 3.8834437086092715,
+      "grad_norm": 0.05994216617642183,
+      "learning_rate": 1.155271660538868e-05,
+      "loss": 0.3997,
+      "step": 2932
+    },
+    {
+      "epoch": 3.88476821192053,
+      "grad_norm": 0.0666557115481596,
+      "learning_rate": 1.1526722718272766e-05,
+      "loss": 0.4019,
+      "step": 2933
+    },
+    {
+      "epoch": 3.886092715231788,
+      "grad_norm": 0.07852832766515284,
+      "learning_rate": 1.1500753183798499e-05,
+      "loss": 0.405,
+      "step": 2934
+    },
+    {
+      "epoch": 3.8874172185430464,
+      "grad_norm": 0.059240929744946984,
+      "learning_rate": 1.1474808024177132e-05,
+      "loss": 0.4064,
+      "step": 2935
+    },
+    {
+      "epoch": 3.888741721854305,
+      "grad_norm": 0.06831900734674434,
+      "learning_rate": 1.1448887261599034e-05,
+      "loss": 0.4017,
+      "step": 2936
+    },
+    {
+      "epoch": 3.8900662251655627,
+      "grad_norm": 0.06753495113327727,
+      "learning_rate": 1.1422990918233752e-05,
+      "loss": 0.4027,
+      "step": 2937
+    },
+    {
+      "epoch": 3.8913907284768214,
+      "grad_norm": 0.06577174866630586,
+      "learning_rate": 1.1397119016229903e-05,
+      "loss": 0.4004,
+      "step": 2938
+    },
+    {
+      "epoch": 3.8927152317880793,
+      "grad_norm": 0.06689933583299842,
+      "learning_rate": 1.1371271577715249e-05,
+      "loss": 0.4023,
+      "step": 2939
+    },
+    {
+      "epoch": 3.8940397350993377,
+      "grad_norm": 0.06776043717803303,
+      "learning_rate": 1.1345448624796572e-05,
+      "loss": 0.4037,
+      "step": 2940
+    },
+    {
+      "epoch": 3.895364238410596,
+      "grad_norm": 0.061402305925796416,
+      "learning_rate": 1.1319650179559761e-05,
+      "loss": 0.4044,
+      "step": 2941
+    },
+    {
+      "epoch": 3.8966887417218543,
+      "grad_norm": 0.06818618672286345,
+      "learning_rate": 1.1293876264069734e-05,
+      "loss": 0.3977,
+      "step": 2942
+    },
+    {
+      "epoch": 3.8980132450331126,
+      "grad_norm": 0.06124390710992125,
+      "learning_rate": 1.1268126900370401e-05,
+      "loss": 0.4008,
+      "step": 2943
+    },
+    {
+      "epoch": 3.899337748344371,
+      "grad_norm": 0.06161523820611098,
+      "learning_rate": 1.1242402110484716e-05,
+      "loss": 0.399,
+      "step": 2944
+    },
+    {
+      "epoch": 3.9006622516556293,
+      "grad_norm": 0.0640593527096696,
+      "learning_rate": 1.1216701916414564e-05,
+      "loss": 0.399,
+      "step": 2945
+    },
+    {
+      "epoch": 3.901986754966887,
+      "grad_norm": 0.06380241985729908,
+      "learning_rate": 1.1191026340140852e-05,
+      "loss": 0.4009,
+      "step": 2946
+    },
+    {
+      "epoch": 3.903311258278146,
+      "grad_norm": 0.05913068674121376,
+      "learning_rate": 1.1165375403623364e-05,
+      "loss": 0.401,
+      "step": 2947
+    },
+    {
+      "epoch": 3.904635761589404,
+      "grad_norm": 0.05389093787501384,
+      "learning_rate": 1.1139749128800896e-05,
+      "loss": 0.4045,
+      "step": 2948
+    },
+    {
+      "epoch": 3.905960264900662,
+      "grad_norm": 0.06382994673707466,
+      "learning_rate": 1.111414753759108e-05,
+      "loss": 0.4097,
+      "step": 2949
+    },
+    {
+      "epoch": 3.9072847682119205,
+      "grad_norm": 0.06321668720505338,
+      "learning_rate": 1.108857065189045e-05,
+      "loss": 0.3987,
+      "step": 2950
+    },
+    {
+      "epoch": 3.908609271523179,
+      "grad_norm": 0.05931442094388174,
+      "learning_rate": 1.1063018493574442e-05,
+      "loss": 0.4019,
+      "step": 2951
+    },
+    {
+      "epoch": 3.909933774834437,
+      "grad_norm": 0.06422845754127897,
+      "learning_rate": 1.1037491084497299e-05,
+      "loss": 0.4003,
+      "step": 2952
+    },
+    {
+      "epoch": 3.9112582781456955,
+      "grad_norm": 0.060414036944261465,
+      "learning_rate": 1.1011988446492125e-05,
+      "loss": 0.4019,
+      "step": 2953
+    },
+    {
+      "epoch": 3.912582781456954,
+      "grad_norm": 0.060968627089956026,
+      "learning_rate": 1.0986510601370837e-05,
+      "loss": 0.3961,
+      "step": 2954
+    },
+    {
+      "epoch": 3.9139072847682117,
+      "grad_norm": 0.06609594704744401,
+      "learning_rate": 1.0961057570924157e-05,
+      "loss": 0.3994,
+      "step": 2955
+    },
+    {
+      "epoch": 3.91523178807947,
+      "grad_norm": 0.062460764171944246,
+      "learning_rate": 1.0935629376921537e-05,
+      "loss": 0.3997,
+      "step": 2956
+    },
+    {
+      "epoch": 3.9165562913907284,
+      "grad_norm": 0.06316948318268853,
+      "learning_rate": 1.0910226041111244e-05,
+      "loss": 0.4019,
+      "step": 2957
+    },
+    {
+      "epoch": 3.9178807947019867,
+      "grad_norm": 0.05719162373289322,
+      "learning_rate": 1.0884847585220237e-05,
+      "loss": 0.4003,
+      "step": 2958
+    },
+    {
+      "epoch": 3.919205298013245,
+      "grad_norm": 0.06337218500672874,
+      "learning_rate": 1.0859494030954237e-05,
+      "loss": 0.3971,
+      "step": 2959
+    },
+    {
+      "epoch": 3.9205298013245033,
+      "grad_norm": 0.05495251216124657,
+      "learning_rate": 1.083416539999763e-05,
+      "loss": 0.4013,
+      "step": 2960
+    },
+    {
+      "epoch": 3.9218543046357617,
+      "grad_norm": 0.06156376939448068,
+      "learning_rate": 1.0808861714013505e-05,
+      "loss": 0.4004,
+      "step": 2961
+    },
+    {
+      "epoch": 3.92317880794702,
+      "grad_norm": 0.06758317859073212,
+      "learning_rate": 1.078358299464363e-05,
+      "loss": 0.4018,
+      "step": 2962
+    },
+    {
+      "epoch": 3.9245033112582783,
+      "grad_norm": 0.062171916589283505,
+      "learning_rate": 1.075832926350838e-05,
+      "loss": 0.4009,
+      "step": 2963
+    },
+    {
+      "epoch": 3.9258278145695362,
+      "grad_norm": 0.06451604631079402,
+      "learning_rate": 1.0733100542206807e-05,
+      "loss": 0.3986,
+      "step": 2964
+    },
+    {
+      "epoch": 3.9271523178807946,
+      "grad_norm": 0.07602063692530693,
+      "learning_rate": 1.0707896852316519e-05,
+      "loss": 0.4018,
+      "step": 2965
+    },
+    {
+      "epoch": 3.928476821192053,
+      "grad_norm": 0.05973084042710855,
+      "learning_rate": 1.0682718215393773e-05,
+      "loss": 0.4013,
+      "step": 2966
+    },
+    {
+      "epoch": 3.929801324503311,
+      "grad_norm": 0.06810494271190727,
+      "learning_rate": 1.0657564652973336e-05,
+      "loss": 0.4063,
+      "step": 2967
+    },
+    {
+      "epoch": 3.9311258278145695,
+      "grad_norm": 0.07344790317234755,
+      "learning_rate": 1.0632436186568582e-05,
+      "loss": 0.4021,
+      "step": 2968
+    },
+    {
+      "epoch": 3.932450331125828,
+      "grad_norm": 0.06539182723036895,
+      "learning_rate": 1.0607332837671395e-05,
+      "loss": 0.4061,
+      "step": 2969
+    },
+    {
+      "epoch": 3.933774834437086,
+      "grad_norm": 0.0619830575506605,
+      "learning_rate": 1.0582254627752197e-05,
+      "loss": 0.4018,
+      "step": 2970
+    },
+    {
+      "epoch": 3.9350993377483445,
+      "grad_norm": 0.06631144157680846,
+      "learning_rate": 1.0557201578259883e-05,
+      "loss": 0.4041,
+      "step": 2971
+    },
+    {
+      "epoch": 3.936423841059603,
+      "grad_norm": 0.06280539877552122,
+      "learning_rate": 1.0532173710621825e-05,
+      "loss": 0.4008,
+      "step": 2972
+    },
+    {
+      "epoch": 3.9377483443708607,
+      "grad_norm": 0.06453113472756093,
+      "learning_rate": 1.0507171046243898e-05,
+      "loss": 0.4004,
+      "step": 2973
+    },
+    {
+      "epoch": 3.939072847682119,
+      "grad_norm": 0.06535213162269121,
+      "learning_rate": 1.0482193606510363e-05,
+      "loss": 0.4002,
+      "step": 2974
+    },
+    {
+      "epoch": 3.9403973509933774,
+      "grad_norm": 0.0596571391299545,
+      "learning_rate": 1.0457241412783988e-05,
+      "loss": 0.3991,
+      "step": 2975
+    },
+    {
+      "epoch": 3.9417218543046357,
+      "grad_norm": 0.06141198544586345,
+      "learning_rate": 1.0432314486405856e-05,
+      "loss": 0.4033,
+      "step": 2976
+    },
+    {
+      "epoch": 3.943046357615894,
+      "grad_norm": 0.060576719975864675,
+      "learning_rate": 1.0407412848695513e-05,
+      "loss": 0.3993,
+      "step": 2977
+    },
+    {
+      "epoch": 3.9443708609271524,
+      "grad_norm": 0.062228957895003593,
+      "learning_rate": 1.0382536520950825e-05,
+      "loss": 0.3988,
+      "step": 2978
+    },
+    {
+      "epoch": 3.9456953642384107,
+      "grad_norm": 0.056635168594248095,
+      "learning_rate": 1.0357685524448055e-05,
+      "loss": 0.4008,
+      "step": 2979
+    },
+    {
+      "epoch": 3.9470198675496686,
+      "grad_norm": 0.06141241640519426,
+      "learning_rate": 1.0332859880441753e-05,
+      "loss": 0.4008,
+      "step": 2980
+    },
+    {
+      "epoch": 3.9483443708609274,
+      "grad_norm": 0.06254128306159139,
+      "learning_rate": 1.030805961016482e-05,
+      "loss": 0.3993,
+      "step": 2981
+    },
+    {
+      "epoch": 3.9496688741721853,
+      "grad_norm": 0.06570201272106455,
+      "learning_rate": 1.0283284734828461e-05,
+      "loss": 0.3983,
+      "step": 2982
+    },
+    {
+      "epoch": 3.9509933774834436,
+      "grad_norm": 0.05961346928443802,
+      "learning_rate": 1.0258535275622123e-05,
+      "loss": 0.3988,
+      "step": 2983
+    },
+    {
+      "epoch": 3.952317880794702,
+      "grad_norm": 0.06156581140222551,
+      "learning_rate": 1.0233811253713561e-05,
+      "loss": 0.4008,
+      "step": 2984
+    },
+    {
+      "epoch": 3.9536423841059603,
+      "grad_norm": 0.07116410876796253,
+      "learning_rate": 1.0209112690248726e-05,
+      "loss": 0.4033,
+      "step": 2985
+    },
+    {
+      "epoch": 3.9549668874172186,
+      "grad_norm": 0.05930190480445743,
+      "learning_rate": 1.0184439606351847e-05,
+      "loss": 0.3996,
+      "step": 2986
+    },
+    {
+      "epoch": 3.956291390728477,
+      "grad_norm": 0.06202121066767995,
+      "learning_rate": 1.0159792023125314e-05,
+      "loss": 0.4008,
+      "step": 2987
+    },
+    {
+      "epoch": 3.9576158940397352,
+      "grad_norm": 0.06261530679039023,
+      "learning_rate": 1.0135169961649728e-05,
+      "loss": 0.4032,
+      "step": 2988
+    },
+    {
+      "epoch": 3.958940397350993,
+      "grad_norm": 0.059769916134108834,
+      "learning_rate": 1.0110573442983868e-05,
+      "loss": 0.4003,
+      "step": 2989
+    },
+    {
+      "epoch": 3.960264900662252,
+      "grad_norm": 0.06346785535900609,
+      "learning_rate": 1.0086002488164666e-05,
+      "loss": 0.4016,
+      "step": 2990
+    },
+    {
+      "epoch": 3.96158940397351,
+      "grad_norm": 0.06321146188689296,
+      "learning_rate": 1.0061457118207176e-05,
+      "loss": 0.402,
+      "step": 2991
+    },
+    {
+      "epoch": 3.962913907284768,
+      "grad_norm": 0.06267095244138975,
+      "learning_rate": 1.0036937354104563e-05,
+      "loss": 0.3991,
+      "step": 2992
+    },
+    {
+      "epoch": 3.9642384105960264,
+      "grad_norm": 0.06437251429753697,
+      "learning_rate": 1.0012443216828132e-05,
+      "loss": 0.4041,
+      "step": 2993
+    },
+    {
+      "epoch": 3.9655629139072848,
+      "grad_norm": 0.06458567410261234,
+      "learning_rate": 9.987974727327212e-06,
+      "loss": 0.398,
+      "step": 2994
+    },
+    {
+      "epoch": 3.966887417218543,
+      "grad_norm": 0.058006495307883085,
+      "learning_rate": 9.963531906529246e-06,
+      "loss": 0.4001,
+      "step": 2995
+    },
+    {
+      "epoch": 3.9682119205298014,
+      "grad_norm": 0.06331318159586626,
+      "learning_rate": 9.939114775339704e-06,
+      "loss": 0.4023,
+      "step": 2996
+    },
+    {
+      "epoch": 3.9695364238410598,
+      "grad_norm": 0.0610199637423256,
+      "learning_rate": 9.914723354642097e-06,
+      "loss": 0.3975,
+      "step": 2997
+    },
+    {
+      "epoch": 3.9708609271523176,
+      "grad_norm": 0.05936349880383175,
+      "learning_rate": 9.890357665297903e-06,
+      "loss": 0.4048,
+      "step": 2998
+    },
+    {
+      "epoch": 3.9721854304635764,
+      "grad_norm": 0.057308284153708255,
+      "learning_rate": 9.866017728146655e-06,
+      "loss": 0.4004,
+      "step": 2999
+    },
+    {
+      "epoch": 3.9735099337748343,
+      "grad_norm": 0.05896363593646538,
+      "learning_rate": 9.841703564005818e-06,
+      "loss": 0.3987,
+      "step": 3000
+    },
+    {
+      "epoch": 3.9748344370860926,
+      "grad_norm": 0.05614728344514775,
+      "learning_rate": 9.817415193670791e-06,
+      "loss": 0.406,
+      "step": 3001
+    },
+    {
+      "epoch": 3.976158940397351,
+      "grad_norm": 0.05616145865396552,
+      "learning_rate": 9.793152637915004e-06,
+      "loss": 0.4016,
+      "step": 3002
+    },
+    {
+      "epoch": 3.9774834437086093,
+      "grad_norm": 0.05778063671424077,
+      "learning_rate": 9.76891591748971e-06,
+      "loss": 0.4012,
+      "step": 3003
+    },
+    {
+      "epoch": 3.9788079470198676,
+      "grad_norm": 0.05528322820293477,
+      "learning_rate": 9.74470505312413e-06,
+      "loss": 0.3984,
+      "step": 3004
+    },
+    {
+      "epoch": 3.980132450331126,
+      "grad_norm": 0.05879923440184878,
+      "learning_rate": 9.720520065525321e-06,
+      "loss": 0.4002,
+      "step": 3005
+    },
+    {
+      "epoch": 3.9814569536423843,
+      "grad_norm": 0.055471067974861316,
+      "learning_rate": 9.696360975378262e-06,
+      "loss": 0.398,
+      "step": 3006
+    },
+    {
+      "epoch": 3.982781456953642,
+      "grad_norm": 0.05532189869308287,
+      "learning_rate": 9.672227803345726e-06,
+      "loss": 0.4006,
+      "step": 3007
+    },
+    {
+      "epoch": 3.984105960264901,
+      "grad_norm": 0.0568921199219698,
+      "learning_rate": 9.64812057006836e-06,
+      "loss": 0.4035,
+      "step": 3008
+    },
+    {
+      "epoch": 3.985430463576159,
+      "grad_norm": 0.062223317418673554,
+      "learning_rate": 9.624039296164617e-06,
+      "loss": 0.4004,
+      "step": 3009
+    },
+    {
+      "epoch": 3.986754966887417,
+      "grad_norm": 0.05746419052122922,
+      "learning_rate": 9.599984002230748e-06,
+      "loss": 0.4016,
+      "step": 3010
+    },
+    {
+      "epoch": 3.9880794701986755,
+      "grad_norm": 0.06085328045807441,
+      "learning_rate": 9.57595470884077e-06,
+      "loss": 0.4,
+      "step": 3011
+    },
+    {
+      "epoch": 3.989403973509934,
+      "grad_norm": 0.057727919398356024,
+      "learning_rate": 9.551951436546466e-06,
+      "loss": 0.3987,
+      "step": 3012
+    },
+    {
+      "epoch": 3.990728476821192,
+      "grad_norm": 0.06324586793181358,
+      "learning_rate": 9.527974205877383e-06,
+      "loss": 0.3975,
+      "step": 3013
+    },
+    {
+      "epoch": 3.9920529801324505,
+      "grad_norm": 0.06028147205670741,
+      "learning_rate": 9.504023037340762e-06,
+      "loss": 0.4024,
+      "step": 3014
+    },
+    {
+      "epoch": 3.993377483443709,
+      "grad_norm": 0.05872052365889469,
+      "learning_rate": 9.480097951421587e-06,
+      "loss": 0.4011,
+      "step": 3015
+    },
+    {
+      "epoch": 3.9947019867549667,
+      "grad_norm": 0.06095713056087473,
+      "learning_rate": 9.456198968582515e-06,
+      "loss": 0.4039,
+      "step": 3016
+    },
+    {
+      "epoch": 3.996026490066225,
+      "grad_norm": 0.06120314155444866,
+      "learning_rate": 9.432326109263897e-06,
+      "loss": 0.4016,
+      "step": 3017
+    },
+    {
+      "epoch": 3.9973509933774833,
+      "grad_norm": 0.06139496668900677,
+      "learning_rate": 9.408479393883701e-06,
+      "loss": 0.4005,
+      "step": 3018
+    },
+    {
+      "epoch": 3.9986754966887417,
+      "grad_norm": 0.06059674978678238,
+      "learning_rate": 9.384658842837587e-06,
+      "loss": 0.3988,
+      "step": 3019
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.05612553341882397,
+      "learning_rate": 9.360864476498781e-06,
+      "loss": 0.3964,
+      "step": 3020
+    },
+    {
+      "epoch": 4.001324503311258,
+      "grad_norm": 0.15821692692040762,
+      "learning_rate": 9.337096315218175e-06,
+      "loss": 0.376,
+      "step": 3021
+    },
+    {
+      "epoch": 4.002649006622517,
+      "grad_norm": 0.14221915164464313,
+      "learning_rate": 9.313354379324187e-06,
+      "loss": 0.3748,
+      "step": 3022
+    },
+    {
+      "epoch": 4.0039735099337745,
+      "grad_norm": 0.0851408403207247,
+      "learning_rate": 9.289638689122844e-06,
+      "loss": 0.3746,
+      "step": 3023
+    },
+    {
+      "epoch": 4.005298013245033,
+      "grad_norm": 0.13531830286106578,
+      "learning_rate": 9.265949264897728e-06,
+      "loss": 0.3727,
+      "step": 3024
+    },
+    {
+      "epoch": 4.006622516556291,
+      "grad_norm": 0.1230233941555162,
+      "learning_rate": 9.242286126909925e-06,
+      "loss": 0.3747,
+      "step": 3025
+    },
+    {
+      "epoch": 4.00794701986755,
+      "grad_norm": 0.11843645448240152,
+      "learning_rate": 9.218649295398077e-06,
+      "loss": 0.3773,
+      "step": 3026
+    },
+    {
+      "epoch": 4.009271523178808,
+      "grad_norm": 0.10594736590521961,
+      "learning_rate": 9.19503879057829e-06,
+      "loss": 0.3729,
+      "step": 3027
+    },
+    {
+      "epoch": 4.010596026490067,
+      "grad_norm": 0.12294770918824384,
+      "learning_rate": 9.171454632644176e-06,
+      "loss": 0.3746,
+      "step": 3028
+    },
+    {
+      "epoch": 4.0119205298013245,
+      "grad_norm": 0.1092678693407112,
+      "learning_rate": 9.147896841766806e-06,
+      "loss": 0.3725,
+      "step": 3029
+    },
+    {
+      "epoch": 4.013245033112582,
+      "grad_norm": 0.11010977039862803,
+      "learning_rate": 9.124365438094722e-06,
+      "loss": 0.372,
+      "step": 3030
+    },
+    {
+      "epoch": 4.014569536423841,
+      "grad_norm": 0.1025293326146463,
+      "learning_rate": 9.100860441753859e-06,
+      "loss": 0.373,
+      "step": 3031
+    },
+    {
+      "epoch": 4.015894039735099,
+      "grad_norm": 0.09325638915264459,
+      "learning_rate": 9.077381872847573e-06,
+      "loss": 0.3739,
+      "step": 3032
+    },
+    {
+      "epoch": 4.017218543046358,
+      "grad_norm": 0.10163465491976643,
+      "learning_rate": 9.053929751456652e-06,
+      "loss": 0.3727,
+      "step": 3033
+    },
+    {
+      "epoch": 4.018543046357616,
+      "grad_norm": 0.08771397369104895,
+      "learning_rate": 9.03050409763922e-06,
+      "loss": 0.3734,
+      "step": 3034
+    },
+    {
+      "epoch": 4.0198675496688745,
+      "grad_norm": 0.08273167416484159,
+      "learning_rate": 9.007104931430795e-06,
+      "loss": 0.3723,
+      "step": 3035
+    },
+    {
+      "epoch": 4.021192052980132,
+      "grad_norm": 0.08914847297182826,
+      "learning_rate": 8.983732272844228e-06,
+      "loss": 0.3753,
+      "step": 3036
+    },
+    {
+      "epoch": 4.022516556291391,
+      "grad_norm": 0.07595964944494685,
+      "learning_rate": 8.96038614186971e-06,
+      "loss": 0.3746,
+      "step": 3037
+    },
+    {
+      "epoch": 4.023841059602649,
+      "grad_norm": 0.08132827007370205,
+      "learning_rate": 8.937066558474714e-06,
+      "loss": 0.3742,
+      "step": 3038
+    },
+    {
+      "epoch": 4.025165562913907,
+      "grad_norm": 0.07217024099236759,
+      "learning_rate": 8.913773542604053e-06,
+      "loss": 0.3715,
+      "step": 3039
+    },
+    {
+      "epoch": 4.026490066225166,
+      "grad_norm": 0.07692781939636459,
+      "learning_rate": 8.890507114179763e-06,
+      "loss": 0.3692,
+      "step": 3040
+    },
+    {
+      "epoch": 4.027814569536424,
+      "grad_norm": 0.08687114163932284,
+      "learning_rate": 8.867267293101195e-06,
+      "loss": 0.3735,
+      "step": 3041
+    },
+    {
+      "epoch": 4.029139072847682,
+      "grad_norm": 0.07275724978733855,
+      "learning_rate": 8.844054099244892e-06,
+      "loss": 0.3756,
+      "step": 3042
+    },
+    {
+      "epoch": 4.03046357615894,
+      "grad_norm": 0.07880730322066151,
+      "learning_rate": 8.82086755246466e-06,
+      "loss": 0.372,
+      "step": 3043
+    },
+    {
+      "epoch": 4.031788079470199,
+      "grad_norm": 0.07996481669066002,
+      "learning_rate": 8.79770767259152e-06,
+      "loss": 0.3717,
+      "step": 3044
+    },
+    {
+      "epoch": 4.033112582781457,
+      "grad_norm": 0.07837170739922389,
+      "learning_rate": 8.774574479433634e-06,
+      "loss": 0.3706,
+      "step": 3045
+    },
+    {
+      "epoch": 4.034437086092716,
+      "grad_norm": 0.07422869161711917,
+      "learning_rate": 8.75146799277641e-06,
+      "loss": 0.3714,
+      "step": 3046
+    },
+    {
+      "epoch": 4.035761589403974,
+      "grad_norm": 0.07111086134857852,
+      "learning_rate": 8.728388232382344e-06,
+      "loss": 0.3758,
+      "step": 3047
+    },
+    {
+      "epoch": 4.0370860927152314,
+      "grad_norm": 0.06998576112585891,
+      "learning_rate": 8.705335217991129e-06,
+      "loss": 0.3745,
+      "step": 3048
+    },
+    {
+      "epoch": 4.03841059602649,
+      "grad_norm": 0.06855373522005107,
+      "learning_rate": 8.682308969319551e-06,
+      "loss": 0.3708,
+      "step": 3049
+    },
+    {
+      "epoch": 4.039735099337748,
+      "grad_norm": 0.07497103707253672,
+      "learning_rate": 8.659309506061508e-06,
+      "loss": 0.3768,
+      "step": 3050
+    },
+    {
+      "epoch": 4.041059602649007,
+      "grad_norm": 0.0718160439805932,
+      "learning_rate": 8.636336847888018e-06,
+      "loss": 0.3789,
+      "step": 3051
+    },
+    {
+      "epoch": 4.042384105960265,
+      "grad_norm": 0.06742589536332153,
+      "learning_rate": 8.613391014447128e-06,
+      "loss": 0.3724,
+      "step": 3052
+    },
+    {
+      "epoch": 4.0437086092715235,
+      "grad_norm": 0.07187480934996475,
+      "learning_rate": 8.590472025363978e-06,
+      "loss": 0.376,
+      "step": 3053
+    },
+    {
+      "epoch": 4.045033112582781,
+      "grad_norm": 0.06573224371407344,
+      "learning_rate": 8.567579900240721e-06,
+      "loss": 0.3733,
+      "step": 3054
+    },
+    {
+      "epoch": 4.046357615894039,
+      "grad_norm": 0.0672405582184344,
+      "learning_rate": 8.544714658656551e-06,
+      "loss": 0.3711,
+      "step": 3055
+    },
+    {
+      "epoch": 4.047682119205298,
+      "grad_norm": 0.0683123591986567,
+      "learning_rate": 8.521876320167673e-06,
+      "loss": 0.3734,
+      "step": 3056
+    },
+    {
+      "epoch": 4.049006622516556,
+      "grad_norm": 0.07215581685429465,
+      "learning_rate": 8.49906490430728e-06,
+      "loss": 0.3721,
+      "step": 3057
+    },
+    {
+      "epoch": 4.050331125827815,
+      "grad_norm": 0.0716019555387714,
+      "learning_rate": 8.476280430585508e-06,
+      "loss": 0.3741,
+      "step": 3058
+    },
+    {
+      "epoch": 4.051655629139073,
+      "grad_norm": 0.05788127143918414,
+      "learning_rate": 8.453522918489509e-06,
+      "loss": 0.3715,
+      "step": 3059
+    },
+    {
+      "epoch": 4.052980132450331,
+      "grad_norm": 0.07079848600292896,
+      "learning_rate": 8.430792387483313e-06,
+      "loss": 0.3675,
+      "step": 3060
+    },
+    {
+      "epoch": 4.054304635761589,
+      "grad_norm": 0.06968721500625365,
+      "learning_rate": 8.408088857007923e-06,
+      "loss": 0.3717,
+      "step": 3061
+    },
+    {
+      "epoch": 4.055629139072848,
+      "grad_norm": 0.0645502707426834,
+      "learning_rate": 8.385412346481208e-06,
+      "loss": 0.3739,
+      "step": 3062
+    },
+    {
+      "epoch": 4.056953642384106,
+      "grad_norm": 0.07321866280167781,
+      "learning_rate": 8.362762875297959e-06,
+      "loss": 0.3729,
+      "step": 3063
+    },
+    {
+      "epoch": 4.058278145695364,
+      "grad_norm": 0.0614070855778223,
+      "learning_rate": 8.34014046282984e-06,
+      "loss": 0.37,
+      "step": 3064
+    },
+    {
+      "epoch": 4.059602649006623,
+      "grad_norm": 0.06535728526803684,
+      "learning_rate": 8.317545128425339e-06,
+      "loss": 0.3721,
+      "step": 3065
+    },
+    {
+      "epoch": 4.0609271523178805,
+      "grad_norm": 0.0663519342382353,
+      "learning_rate": 8.294976891409825e-06,
+      "loss": 0.3714,
+      "step": 3066
+    },
+    {
+      "epoch": 4.062251655629139,
+      "grad_norm": 0.06233022699867169,
+      "learning_rate": 8.27243577108546e-06,
+      "loss": 0.3717,
+      "step": 3067
+    },
+    {
+      "epoch": 4.063576158940397,
+      "grad_norm": 0.06287336369081675,
+      "learning_rate": 8.249921786731243e-06,
+      "loss": 0.366,
+      "step": 3068
+    },
+    {
+      "epoch": 4.064900662251656,
+      "grad_norm": 0.06642922146799224,
+      "learning_rate": 8.227434957602929e-06,
+      "loss": 0.3768,
+      "step": 3069
+    },
+    {
+      "epoch": 4.066225165562914,
+      "grad_norm": 0.06135737201131537,
+      "learning_rate": 8.204975302933072e-06,
+      "loss": 0.3724,
+      "step": 3070
+    },
+    {
+      "epoch": 4.067549668874173,
+      "grad_norm": 0.061571769595304554,
+      "learning_rate": 8.182542841930985e-06,
+      "loss": 0.3748,
+      "step": 3071
+    },
+    {
+      "epoch": 4.0688741721854305,
+      "grad_norm": 0.058768426858174246,
+      "learning_rate": 8.160137593782727e-06,
+      "loss": 0.3753,
+      "step": 3072
+    },
+    {
+      "epoch": 4.070198675496688,
+      "grad_norm": 0.05802858212210787,
+      "learning_rate": 8.137759577651057e-06,
+      "loss": 0.3736,
+      "step": 3073
+    },
+    {
+      "epoch": 4.071523178807947,
+      "grad_norm": 0.05818652331326773,
+      "learning_rate": 8.115408812675452e-06,
+      "loss": 0.3738,
+      "step": 3074
+    },
+    {
+      "epoch": 4.072847682119205,
+      "grad_norm": 0.059543957955595084,
+      "learning_rate": 8.093085317972105e-06,
+      "loss": 0.3719,
+      "step": 3075
+    },
+    {
+      "epoch": 4.074172185430464,
+      "grad_norm": 0.0634430486501205,
+      "learning_rate": 8.070789112633846e-06,
+      "loss": 0.3722,
+      "step": 3076
+    },
+    {
+      "epoch": 4.075496688741722,
+      "grad_norm": 0.06273617945383256,
+      "learning_rate": 8.048520215730207e-06,
+      "loss": 0.3707,
+      "step": 3077
+    },
+    {
+      "epoch": 4.07682119205298,
+      "grad_norm": 0.055462634569067575,
+      "learning_rate": 8.026278646307331e-06,
+      "loss": 0.3729,
+      "step": 3078
+    },
+    {
+      "epoch": 4.078145695364238,
+      "grad_norm": 0.06235126559920844,
+      "learning_rate": 8.004064423388018e-06,
+      "loss": 0.3733,
+      "step": 3079
+    },
+    {
+      "epoch": 4.079470198675497,
+      "grad_norm": 0.06187259375435491,
+      "learning_rate": 7.981877565971641e-06,
+      "loss": 0.3704,
+      "step": 3080
+    },
+    {
+      "epoch": 4.080794701986755,
+      "grad_norm": 0.061411693176465054,
+      "learning_rate": 7.95971809303421e-06,
+      "loss": 0.3758,
+      "step": 3081
+    },
+    {
+      "epoch": 4.082119205298013,
+      "grad_norm": 0.056881163904113956,
+      "learning_rate": 7.937586023528268e-06,
+      "loss": 0.3711,
+      "step": 3082
+    },
+    {
+      "epoch": 4.083443708609272,
+      "grad_norm": 0.06992594576705861,
+      "learning_rate": 7.915481376382956e-06,
+      "loss": 0.3725,
+      "step": 3083
+    },
+    {
+      "epoch": 4.0847682119205295,
+      "grad_norm": 0.05851534685192537,
+      "learning_rate": 7.893404170503962e-06,
+      "loss": 0.3687,
+      "step": 3084
+    },
+    {
+      "epoch": 4.086092715231788,
+      "grad_norm": 0.060905279542296596,
+      "learning_rate": 7.87135442477347e-06,
+      "loss": 0.3708,
+      "step": 3085
+    },
+    {
+      "epoch": 4.087417218543046,
+      "grad_norm": 0.06752605255303604,
+      "learning_rate": 7.849332158050221e-06,
+      "loss": 0.3732,
+      "step": 3086
+    },
+    {
+      "epoch": 4.088741721854305,
+      "grad_norm": 0.06187669614731747,
+      "learning_rate": 7.82733738916941e-06,
+      "loss": 0.3728,
+      "step": 3087
+    },
+    {
+      "epoch": 4.090066225165563,
+      "grad_norm": 0.06126660168661204,
+      "learning_rate": 7.805370136942759e-06,
+      "loss": 0.3749,
+      "step": 3088
+    },
+    {
+      "epoch": 4.091390728476822,
+      "grad_norm": 0.06482907705800582,
+      "learning_rate": 7.783430420158411e-06,
+      "loss": 0.376,
+      "step": 3089
+    },
+    {
+      "epoch": 4.0927152317880795,
+      "grad_norm": 0.060115231914684875,
+      "learning_rate": 7.761518257580994e-06,
+      "loss": 0.3762,
+      "step": 3090
+    },
+    {
+      "epoch": 4.094039735099337,
+      "grad_norm": 0.06293520992394078,
+      "learning_rate": 7.739633667951549e-06,
+      "loss": 0.3775,
+      "step": 3091
+    },
+    {
+      "epoch": 4.095364238410596,
+      "grad_norm": 0.06730084279102812,
+      "learning_rate": 7.717776669987556e-06,
+      "loss": 0.3713,
+      "step": 3092
+    },
+    {
+      "epoch": 4.096688741721854,
+      "grad_norm": 0.0648307957194961,
+      "learning_rate": 7.695947282382867e-06,
+      "loss": 0.3707,
+      "step": 3093
+    },
+    {
+      "epoch": 4.098013245033113,
+      "grad_norm": 0.06063431063818551,
+      "learning_rate": 7.674145523807727e-06,
+      "loss": 0.3721,
+      "step": 3094
+    },
+    {
+      "epoch": 4.099337748344371,
+      "grad_norm": 0.06616116680794415,
+      "learning_rate": 7.652371412908777e-06,
+      "loss": 0.3773,
+      "step": 3095
+    },
+    {
+      "epoch": 4.1006622516556295,
+      "grad_norm": 0.06482872178692521,
+      "learning_rate": 7.63062496830897e-06,
+      "loss": 0.3724,
+      "step": 3096
+    },
+    {
+      "epoch": 4.101986754966887,
+      "grad_norm": 0.06424814947114406,
+      "learning_rate": 7.6089062086076316e-06,
+      "loss": 0.3726,
+      "step": 3097
+    },
+    {
+      "epoch": 4.103311258278145,
+      "grad_norm": 0.06861228597096203,
+      "learning_rate": 7.587215152380394e-06,
+      "loss": 0.3726,
+      "step": 3098
+    },
+    {
+      "epoch": 4.104635761589404,
+      "grad_norm": 0.06018118425971416,
+      "learning_rate": 7.5655518181791995e-06,
+      "loss": 0.3734,
+      "step": 3099
+    },
+    {
+      "epoch": 4.105960264900662,
+      "grad_norm": 0.06454724002920333,
+      "learning_rate": 7.543916224532268e-06,
+      "loss": 0.3689,
+      "step": 3100
+    },
+    {
+      "epoch": 4.107284768211921,
+      "grad_norm": 0.06933729965878117,
+      "learning_rate": 7.522308389944122e-06,
+      "loss": 0.3719,
+      "step": 3101
+    },
+    {
+      "epoch": 4.108609271523179,
+      "grad_norm": 0.0625338520670997,
+      "learning_rate": 7.500728332895506e-06,
+      "loss": 0.3761,
+      "step": 3102
+    },
+    {
+      "epoch": 4.109933774834437,
+      "grad_norm": 0.06996766881328337,
+      "learning_rate": 7.4791760718434166e-06,
+      "loss": 0.3705,
+      "step": 3103
+    },
+    {
+      "epoch": 4.111258278145695,
+      "grad_norm": 0.0656126603206766,
+      "learning_rate": 7.457651625221101e-06,
+      "loss": 0.3716,
+      "step": 3104
+    },
+    {
+      "epoch": 4.112582781456954,
+      "grad_norm": 0.058670976600393485,
+      "learning_rate": 7.436155011437995e-06,
+      "loss": 0.3724,
+      "step": 3105
+    },
+    {
+      "epoch": 4.113907284768212,
+      "grad_norm": 0.07201950437357969,
+      "learning_rate": 7.414686248879746e-06,
+      "loss": 0.3703,
+      "step": 3106
+    },
+    {
+      "epoch": 4.11523178807947,
+      "grad_norm": 0.06131799414760589,
+      "learning_rate": 7.393245355908147e-06,
+      "loss": 0.3724,
+      "step": 3107
+    },
+    {
+      "epoch": 4.1165562913907285,
+      "grad_norm": 0.0649947555361676,
+      "learning_rate": 7.371832350861203e-06,
+      "loss": 0.3679,
+      "step": 3108
+    },
+    {
+      "epoch": 4.117880794701986,
+      "grad_norm": 0.06405033117417876,
+      "learning_rate": 7.350447252053019e-06,
+      "loss": 0.3783,
+      "step": 3109
+    },
+    {
+      "epoch": 4.119205298013245,
+      "grad_norm": 0.05707509842574818,
+      "learning_rate": 7.329090077773871e-06,
+      "loss": 0.3695,
+      "step": 3110
+    },
+    {
+      "epoch": 4.120529801324503,
+      "grad_norm": 0.06186759630374367,
+      "learning_rate": 7.307760846290124e-06,
+      "loss": 0.3738,
+      "step": 3111
+    },
+    {
+      "epoch": 4.121854304635762,
+      "grad_norm": 0.061541664487942224,
+      "learning_rate": 7.28645957584428e-06,
+      "loss": 0.3733,
+      "step": 3112
+    },
+    {
+      "epoch": 4.12317880794702,
+      "grad_norm": 0.05493436315308323,
+      "learning_rate": 7.265186284654881e-06,
+      "loss": 0.3704,
+      "step": 3113
+    },
+    {
+      "epoch": 4.1245033112582785,
+      "grad_norm": 0.0600630663373046,
+      "learning_rate": 7.2439409909165515e-06,
+      "loss": 0.3731,
+      "step": 3114
+    },
+    {
+      "epoch": 4.125827814569536,
+      "grad_norm": 0.059542912975278514,
+      "learning_rate": 7.222723712800008e-06,
+      "loss": 0.372,
+      "step": 3115
+    },
+    {
+      "epoch": 4.127152317880794,
+      "grad_norm": 0.05504456859993287,
+      "learning_rate": 7.201534468451945e-06,
+      "loss": 0.3709,
+      "step": 3116
+    },
+    {
+      "epoch": 4.128476821192053,
+      "grad_norm": 0.06410000492703405,
+      "learning_rate": 7.180373275995132e-06,
+      "loss": 0.3735,
+      "step": 3117
+    },
+    {
+      "epoch": 4.129801324503311,
+      "grad_norm": 0.059311823149836546,
+      "learning_rate": 7.159240153528322e-06,
+      "loss": 0.3732,
+      "step": 3118
+    },
+    {
+      "epoch": 4.13112582781457,
+      "grad_norm": 0.06192183162162323,
+      "learning_rate": 7.13813511912627e-06,
+      "loss": 0.3727,
+      "step": 3119
+    },
+    {
+      "epoch": 4.132450331125828,
+      "grad_norm": 0.05908065139852093,
+      "learning_rate": 7.1170581908396944e-06,
+      "loss": 0.3735,
+      "step": 3120
+    },
+    {
+      "epoch": 4.133774834437086,
+      "grad_norm": 0.062123902951400696,
+      "learning_rate": 7.096009386695293e-06,
+      "loss": 0.3722,
+      "step": 3121
+    },
+    {
+      "epoch": 4.135099337748344,
+      "grad_norm": 0.06089161804138,
+      "learning_rate": 7.074988724695679e-06,
+      "loss": 0.3728,
+      "step": 3122
+    },
+    {
+      "epoch": 4.136423841059603,
+      "grad_norm": 0.06491379618375508,
+      "learning_rate": 7.053996222819446e-06,
+      "loss": 0.3746,
+      "step": 3123
+    },
+    {
+      "epoch": 4.137748344370861,
+      "grad_norm": 0.060319830067484696,
+      "learning_rate": 7.033031899021044e-06,
+      "loss": 0.367,
+      "step": 3124
+    },
+    {
+      "epoch": 4.139072847682119,
+      "grad_norm": 0.06152984353838788,
+      "learning_rate": 7.012095771230867e-06,
+      "loss": 0.3708,
+      "step": 3125
+    },
+    {
+      "epoch": 4.140397350993378,
+      "grad_norm": 0.06986703114172534,
+      "learning_rate": 6.991187857355188e-06,
+      "loss": 0.3749,
+      "step": 3126
+    },
+    {
+      "epoch": 4.1417218543046355,
+      "grad_norm": 0.06520531259401417,
+      "learning_rate": 6.970308175276121e-06,
+      "loss": 0.3696,
+      "step": 3127
+    },
+    {
+      "epoch": 4.143046357615894,
+      "grad_norm": 0.06253450589822676,
+      "learning_rate": 6.949456742851675e-06,
+      "loss": 0.3741,
+      "step": 3128
+    },
+    {
+      "epoch": 4.144370860927152,
+      "grad_norm": 0.06111710983248197,
+      "learning_rate": 6.928633577915649e-06,
+      "loss": 0.3754,
+      "step": 3129
+    },
+    {
+      "epoch": 4.145695364238411,
+      "grad_norm": 0.062242718700375156,
+      "learning_rate": 6.907838698277718e-06,
+      "loss": 0.3735,
+      "step": 3130
+    },
+    {
+      "epoch": 4.147019867549669,
+      "grad_norm": 0.06278748127987815,
+      "learning_rate": 6.8870721217233175e-06,
+      "loss": 0.3734,
+      "step": 3131
+    },
+    {
+      "epoch": 4.1483443708609276,
+      "grad_norm": 0.06513322445413756,
+      "learning_rate": 6.8663338660137055e-06,
+      "loss": 0.3727,
+      "step": 3132
+    },
+    {
+      "epoch": 4.149668874172185,
+      "grad_norm": 0.055224283105361385,
+      "learning_rate": 6.845623948885922e-06,
+      "loss": 0.3714,
+      "step": 3133
+    },
+    {
+      "epoch": 4.150993377483443,
+      "grad_norm": 0.06345765104449927,
+      "learning_rate": 6.824942388052735e-06,
+      "loss": 0.3716,
+      "step": 3134
+    },
+    {
+      "epoch": 4.152317880794702,
+      "grad_norm": 0.06802787409597888,
+      "learning_rate": 6.804289201202707e-06,
+      "loss": 0.3715,
+      "step": 3135
+    },
+    {
+      "epoch": 4.15364238410596,
+      "grad_norm": 0.05440458517724995,
+      "learning_rate": 6.783664406000085e-06,
+      "loss": 0.373,
+      "step": 3136
+    },
+    {
+      "epoch": 4.154966887417219,
+      "grad_norm": 0.06730651862795374,
+      "learning_rate": 6.763068020084866e-06,
+      "loss": 0.3695,
+      "step": 3137
+    },
+    {
+      "epoch": 4.156291390728477,
+      "grad_norm": 0.06525108214211425,
+      "learning_rate": 6.742500061072737e-06,
+      "loss": 0.375,
+      "step": 3138
+    },
+    {
+      "epoch": 4.157615894039735,
+      "grad_norm": 0.05479711577107501,
+      "learning_rate": 6.721960546555086e-06,
+      "loss": 0.3693,
+      "step": 3139
+    },
+    {
+      "epoch": 4.158940397350993,
+      "grad_norm": 0.06492660684394942,
+      "learning_rate": 6.701449494098944e-06,
+      "loss": 0.3669,
+      "step": 3140
+    },
+    {
+      "epoch": 4.160264900662252,
+      "grad_norm": 0.06103757117644582,
+      "learning_rate": 6.680966921247036e-06,
+      "loss": 0.3739,
+      "step": 3141
+    },
+    {
+      "epoch": 4.16158940397351,
+      "grad_norm": 0.05906983012794455,
+      "learning_rate": 6.660512845517684e-06,
+      "loss": 0.3719,
+      "step": 3142
+    },
+    {
+      "epoch": 4.162913907284768,
+      "grad_norm": 0.056421270542410536,
+      "learning_rate": 6.640087284404888e-06,
+      "loss": 0.3733,
+      "step": 3143
+    },
+    {
+      "epoch": 4.164238410596027,
+      "grad_norm": 0.05699401958829645,
+      "learning_rate": 6.619690255378213e-06,
+      "loss": 0.3705,
+      "step": 3144
+    },
+    {
+      "epoch": 4.1655629139072845,
+      "grad_norm": 0.05971974120744329,
+      "learning_rate": 6.599321775882845e-06,
+      "loss": 0.3715,
+      "step": 3145
+    },
+    {
+      "epoch": 4.166887417218543,
+      "grad_norm": 0.05478976953004014,
+      "learning_rate": 6.57898186333957e-06,
+      "loss": 0.3739,
+      "step": 3146
+    },
+    {
+      "epoch": 4.168211920529801,
+      "grad_norm": 0.06065986658549768,
+      "learning_rate": 6.558670535144695e-06,
+      "loss": 0.3723,
+      "step": 3147
+    },
+    {
+      "epoch": 4.16953642384106,
+      "grad_norm": 0.06084261086635381,
+      "learning_rate": 6.538387808670128e-06,
+      "loss": 0.3753,
+      "step": 3148
+    },
+    {
+      "epoch": 4.170860927152318,
+      "grad_norm": 0.06040999044893852,
+      "learning_rate": 6.518133701263268e-06,
+      "loss": 0.3743,
+      "step": 3149
+    },
+    {
+      "epoch": 4.172185430463577,
+      "grad_norm": 0.06216967916049152,
+      "learning_rate": 6.497908230247078e-06,
+      "loss": 0.3751,
+      "step": 3150
+    },
+    {
+      "epoch": 4.1735099337748345,
+      "grad_norm": 0.05700289330687671,
+      "learning_rate": 6.477711412919996e-06,
+      "loss": 0.3724,
+      "step": 3151
+    },
+    {
+      "epoch": 4.174834437086092,
+      "grad_norm": 0.05774817416243034,
+      "learning_rate": 6.457543266555979e-06,
+      "loss": 0.3736,
+      "step": 3152
+    },
+    {
+      "epoch": 4.176158940397351,
+      "grad_norm": 0.06362559788173278,
+      "learning_rate": 6.437403808404456e-06,
+      "loss": 0.3708,
+      "step": 3153
+    },
+    {
+      "epoch": 4.177483443708609,
+      "grad_norm": 0.05907541331984116,
+      "learning_rate": 6.417293055690295e-06,
+      "loss": 0.3757,
+      "step": 3154
+    },
+    {
+      "epoch": 4.178807947019868,
+      "grad_norm": 0.058323059680660636,
+      "learning_rate": 6.397211025613858e-06,
+      "loss": 0.3702,
+      "step": 3155
+    },
+    {
+      "epoch": 4.180132450331126,
+      "grad_norm": 0.061738413497363316,
+      "learning_rate": 6.377157735350894e-06,
+      "loss": 0.3698,
+      "step": 3156
+    },
+    {
+      "epoch": 4.1814569536423845,
+      "grad_norm": 0.054975766545683875,
+      "learning_rate": 6.35713320205261e-06,
+      "loss": 0.3725,
+      "step": 3157
+    },
+    {
+      "epoch": 4.182781456953642,
+      "grad_norm": 0.06285743009657514,
+      "learning_rate": 6.337137442845582e-06,
+      "loss": 0.3723,
+      "step": 3158
+    },
+    {
+      "epoch": 4.184105960264901,
+      "grad_norm": 0.057028726809819505,
+      "learning_rate": 6.317170474831811e-06,
+      "loss": 0.3718,
+      "step": 3159
+    },
+    {
+      "epoch": 4.185430463576159,
+      "grad_norm": 0.05827579695129068,
+      "learning_rate": 6.297232315088648e-06,
+      "loss": 0.3731,
+      "step": 3160
+    },
+    {
+      "epoch": 4.186754966887417,
+      "grad_norm": 0.06040102372377136,
+      "learning_rate": 6.277322980668827e-06,
+      "loss": 0.3727,
+      "step": 3161
+    },
+    {
+      "epoch": 4.188079470198676,
+      "grad_norm": 0.059875601074714166,
+      "learning_rate": 6.2574424886003985e-06,
+      "loss": 0.3689,
+      "step": 3162
+    },
+    {
+      "epoch": 4.1894039735099335,
+      "grad_norm": 0.0577220304850499,
+      "learning_rate": 6.237590855886777e-06,
+      "loss": 0.3749,
+      "step": 3163
+    },
+    {
+      "epoch": 4.190728476821192,
+      "grad_norm": 0.06477029244821768,
+      "learning_rate": 6.217768099506667e-06,
+      "loss": 0.3702,
+      "step": 3164
+    },
+    {
+      "epoch": 4.19205298013245,
+      "grad_norm": 0.056269699801823454,
+      "learning_rate": 6.197974236414088e-06,
+      "loss": 0.3699,
+      "step": 3165
+    },
+    {
+      "epoch": 4.193377483443709,
+      "grad_norm": 0.05824041607938708,
+      "learning_rate": 6.178209283538361e-06,
+      "loss": 0.3696,
+      "step": 3166
+    },
+    {
+      "epoch": 4.194701986754967,
+      "grad_norm": 0.0635616424756616,
+      "learning_rate": 6.158473257784052e-06,
+      "loss": 0.3716,
+      "step": 3167
+    },
+    {
+      "epoch": 4.196026490066225,
+      "grad_norm": 0.0639220473714096,
+      "learning_rate": 6.13876617603101e-06,
+      "loss": 0.3739,
+      "step": 3168
+    },
+    {
+      "epoch": 4.1973509933774835,
+      "grad_norm": 0.059486738190905435,
+      "learning_rate": 6.119088055134308e-06,
+      "loss": 0.3722,
+      "step": 3169
+    },
+    {
+      "epoch": 4.198675496688741,
+      "grad_norm": 0.06798978186892249,
+      "learning_rate": 6.099438911924278e-06,
+      "loss": 0.3727,
+      "step": 3170
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.0574939507168004,
+      "learning_rate": 6.079818763206433e-06,
+      "loss": 0.374,
+      "step": 3171
+    },
+    {
+      "epoch": 4.201324503311258,
+      "grad_norm": 0.05846783721538231,
+      "learning_rate": 6.0602276257615146e-06,
+      "loss": 0.3715,
+      "step": 3172
+    },
+    {
+      "epoch": 4.202649006622517,
+      "grad_norm": 0.06278764476666787,
+      "learning_rate": 6.040665516345439e-06,
+      "loss": 0.3689,
+      "step": 3173
+    },
+    {
+      "epoch": 4.203973509933775,
+      "grad_norm": 0.05753587882295615,
+      "learning_rate": 6.021132451689315e-06,
+      "loss": 0.3742,
+      "step": 3174
+    },
+    {
+      "epoch": 4.2052980132450335,
+      "grad_norm": 0.06114792151240317,
+      "learning_rate": 6.001628448499377e-06,
+      "loss": 0.3719,
+      "step": 3175
+    },
+    {
+      "epoch": 4.206622516556291,
+      "grad_norm": 0.06298828218862924,
+      "learning_rate": 5.982153523457017e-06,
+      "loss": 0.3713,
+      "step": 3176
+    },
+    {
+      "epoch": 4.207947019867549,
+      "grad_norm": 0.06289137378060614,
+      "learning_rate": 5.962707693218779e-06,
+      "loss": 0.3764,
+      "step": 3177
+    },
+    {
+      "epoch": 4.209271523178808,
+      "grad_norm": 0.05598516655514256,
+      "learning_rate": 5.943290974416278e-06,
+      "loss": 0.3735,
+      "step": 3178
+    },
+    {
+      "epoch": 4.210596026490066,
+      "grad_norm": 0.06212220042976007,
+      "learning_rate": 5.92390338365628e-06,
+      "loss": 0.371,
+      "step": 3179
+    },
+    {
+      "epoch": 4.211920529801325,
+      "grad_norm": 0.06573838002106222,
+      "learning_rate": 5.9045449375206e-06,
+      "loss": 0.3758,
+      "step": 3180
+    },
+    {
+      "epoch": 4.213245033112583,
+      "grad_norm": 0.06523044219356447,
+      "learning_rate": 5.885215652566163e-06,
+      "loss": 0.3699,
+      "step": 3181
+    },
+    {
+      "epoch": 4.214569536423841,
+      "grad_norm": 0.05845724684821544,
+      "learning_rate": 5.865915545324905e-06,
+      "loss": 0.3725,
+      "step": 3182
+    },
+    {
+      "epoch": 4.215894039735099,
+      "grad_norm": 0.0618103652029999,
+      "learning_rate": 5.846644632303857e-06,
+      "loss": 0.3751,
+      "step": 3183
+    },
+    {
+      "epoch": 4.217218543046358,
+      "grad_norm": 0.06413661148073764,
+      "learning_rate": 5.8274029299850445e-06,
+      "loss": 0.368,
+      "step": 3184
+    },
+    {
+      "epoch": 4.218543046357616,
+      "grad_norm": 0.05883577117763686,
+      "learning_rate": 5.808190454825515e-06,
+      "loss": 0.3711,
+      "step": 3185
+    },
+    {
+      "epoch": 4.219867549668874,
+      "grad_norm": 0.059551749689320026,
+      "learning_rate": 5.78900722325733e-06,
+      "loss": 0.3752,
+      "step": 3186
+    },
+    {
+      "epoch": 4.221192052980133,
+      "grad_norm": 0.06296455534606159,
+      "learning_rate": 5.769853251687543e-06,
+      "loss": 0.3664,
+      "step": 3187
+    },
+    {
+      "epoch": 4.22251655629139,
+      "grad_norm": 0.0670940439789761,
+      "learning_rate": 5.750728556498177e-06,
+      "loss": 0.3734,
+      "step": 3188
+    },
+    {
+      "epoch": 4.223841059602649,
+      "grad_norm": 0.05634249488848627,
+      "learning_rate": 5.731633154046199e-06,
+      "loss": 0.3703,
+      "step": 3189
+    },
+    {
+      "epoch": 4.225165562913907,
+      "grad_norm": 0.05479907923898322,
+      "learning_rate": 5.712567060663552e-06,
+      "loss": 0.3729,
+      "step": 3190
+    },
+    {
+      "epoch": 4.226490066225166,
+      "grad_norm": 0.059773611483370175,
+      "learning_rate": 5.6935302926570816e-06,
+      "loss": 0.3692,
+      "step": 3191
+    },
+    {
+      "epoch": 4.227814569536424,
+      "grad_norm": 0.05719807745752411,
+      "learning_rate": 5.6745228663085675e-06,
+      "loss": 0.3727,
+      "step": 3192
+    },
+    {
+      "epoch": 4.2291390728476825,
+      "grad_norm": 0.05678878373299473,
+      "learning_rate": 5.655544797874699e-06,
+      "loss": 0.3734,
+      "step": 3193
+    },
+    {
+      "epoch": 4.23046357615894,
+      "grad_norm": 0.054032808273799186,
+      "learning_rate": 5.6365961035870534e-06,
+      "loss": 0.3717,
+      "step": 3194
+    },
+    {
+      "epoch": 4.231788079470198,
+      "grad_norm": 0.0532411255168714,
+      "learning_rate": 5.617676799652079e-06,
+      "loss": 0.3743,
+      "step": 3195
+    },
+    {
+      "epoch": 4.233112582781457,
+      "grad_norm": 0.055317710116551316,
+      "learning_rate": 5.5987869022510764e-06,
+      "loss": 0.3692,
+      "step": 3196
+    },
+    {
+      "epoch": 4.234437086092715,
+      "grad_norm": 0.05944057927201677,
+      "learning_rate": 5.579926427540226e-06,
+      "loss": 0.3684,
+      "step": 3197
+    },
+    {
+      "epoch": 4.235761589403974,
+      "grad_norm": 0.05834049820660612,
+      "learning_rate": 5.56109539165051e-06,
+      "loss": 0.3727,
+      "step": 3198
+    },
+    {
+      "epoch": 4.237086092715232,
+      "grad_norm": 0.060962387551973196,
+      "learning_rate": 5.542293810687756e-06,
+      "loss": 0.3747,
+      "step": 3199
+    },
+    {
+      "epoch": 4.23841059602649,
+      "grad_norm": 0.062119358202115275,
+      "learning_rate": 5.52352170073259e-06,
+      "loss": 0.373,
+      "step": 3200
+    },
+    {
+      "epoch": 4.239735099337748,
+      "grad_norm": 0.05521160198820213,
+      "learning_rate": 5.504779077840448e-06,
+      "loss": 0.3725,
+      "step": 3201
+    },
+    {
+      "epoch": 4.241059602649006,
+      "grad_norm": 0.05855187934188443,
+      "learning_rate": 5.486065958041504e-06,
+      "loss": 0.3713,
+      "step": 3202
+    },
+    {
+      "epoch": 4.242384105960265,
+      "grad_norm": 0.06312646806084335,
+      "learning_rate": 5.4673823573407535e-06,
+      "loss": 0.37,
+      "step": 3203
+    },
+    {
+      "epoch": 4.243708609271523,
+      "grad_norm": 0.05644641792750668,
+      "learning_rate": 5.4487282917179064e-06,
+      "loss": 0.3721,
+      "step": 3204
+    },
+    {
+      "epoch": 4.245033112582782,
+      "grad_norm": 0.05615745630703661,
+      "learning_rate": 5.430103777127414e-06,
+      "loss": 0.3734,
+      "step": 3205
+    },
+    {
+      "epoch": 4.2463576158940395,
+      "grad_norm": 0.06277354436628364,
+      "learning_rate": 5.4115088294984665e-06,
+      "loss": 0.3695,
+      "step": 3206
+    },
+    {
+      "epoch": 4.247682119205298,
+      "grad_norm": 0.058724543294833814,
+      "learning_rate": 5.392943464734965e-06,
+      "loss": 0.3705,
+      "step": 3207
+    },
+    {
+      "epoch": 4.249006622516556,
+      "grad_norm": 0.057277312129951165,
+      "learning_rate": 5.374407698715516e-06,
+      "loss": 0.3716,
+      "step": 3208
+    },
+    {
+      "epoch": 4.250331125827815,
+      "grad_norm": 0.06298976631930867,
+      "learning_rate": 5.355901547293383e-06,
+      "loss": 0.372,
+      "step": 3209
+    },
+    {
+      "epoch": 4.251655629139073,
+      "grad_norm": 0.05849921899377683,
+      "learning_rate": 5.3374250262965325e-06,
+      "loss": 0.3721,
+      "step": 3210
+    },
+    {
+      "epoch": 4.252980132450331,
+      "grad_norm": 0.055455133413586065,
+      "learning_rate": 5.318978151527563e-06,
+      "loss": 0.372,
+      "step": 3211
+    },
+    {
+      "epoch": 4.2543046357615895,
+      "grad_norm": 0.06168684746267043,
+      "learning_rate": 5.300560938763739e-06,
+      "loss": 0.3738,
+      "step": 3212
+    },
+    {
+      "epoch": 4.255629139072847,
+      "grad_norm": 0.06639995708092984,
+      "learning_rate": 5.2821734037569275e-06,
+      "loss": 0.3705,
+      "step": 3213
+    },
+    {
+      "epoch": 4.256953642384106,
+      "grad_norm": 0.05491043760222792,
+      "learning_rate": 5.263815562233654e-06,
+      "loss": 0.3697,
+      "step": 3214
+    },
+    {
+      "epoch": 4.258278145695364,
+      "grad_norm": 0.06557867578520253,
+      "learning_rate": 5.245487429895017e-06,
+      "loss": 0.3742,
+      "step": 3215
+    },
+    {
+      "epoch": 4.259602649006623,
+      "grad_norm": 0.06540461738735372,
+      "learning_rate": 5.227189022416701e-06,
+      "loss": 0.3755,
+      "step": 3216
+    },
+    {
+      "epoch": 4.260927152317881,
+      "grad_norm": 0.056225633849636095,
+      "learning_rate": 5.208920355448994e-06,
+      "loss": 0.3711,
+      "step": 3217
+    },
+    {
+      "epoch": 4.262251655629139,
+      "grad_norm": 0.06029665884154871,
+      "learning_rate": 5.190681444616714e-06,
+      "loss": 0.3695,
+      "step": 3218
+    },
+    {
+      "epoch": 4.263576158940397,
+      "grad_norm": 0.0629888103756,
+      "learning_rate": 5.172472305519262e-06,
+      "loss": 0.3697,
+      "step": 3219
+    },
+    {
+      "epoch": 4.264900662251655,
+      "grad_norm": 0.06056429995279433,
+      "learning_rate": 5.154292953730555e-06,
+      "loss": 0.3727,
+      "step": 3220
+    },
+    {
+      "epoch": 4.266225165562914,
+      "grad_norm": 0.05637337689703233,
+      "learning_rate": 5.136143404799052e-06,
+      "loss": 0.3682,
+      "step": 3221
+    },
+    {
+      "epoch": 4.267549668874172,
+      "grad_norm": 0.06355880973916127,
+      "learning_rate": 5.118023674247692e-06,
+      "loss": 0.3703,
+      "step": 3222
+    },
+    {
+      "epoch": 4.268874172185431,
+      "grad_norm": 0.058458367354189064,
+      "learning_rate": 5.099933777573949e-06,
+      "loss": 0.3742,
+      "step": 3223
+    },
+    {
+      "epoch": 4.2701986754966885,
+      "grad_norm": 0.05802664453211581,
+      "learning_rate": 5.08187373024974e-06,
+      "loss": 0.3716,
+      "step": 3224
+    },
+    {
+      "epoch": 4.271523178807947,
+      "grad_norm": 0.05901740597090468,
+      "learning_rate": 5.063843547721492e-06,
+      "loss": 0.3681,
+      "step": 3225
+    },
+    {
+      "epoch": 4.272847682119205,
+      "grad_norm": 0.05542066215480775,
+      "learning_rate": 5.045843245410051e-06,
+      "loss": 0.3742,
+      "step": 3226
+    },
+    {
+      "epoch": 4.274172185430464,
+      "grad_norm": 0.05610601625212525,
+      "learning_rate": 5.027872838710735e-06,
+      "loss": 0.3727,
+      "step": 3227
+    },
+    {
+      "epoch": 4.275496688741722,
+      "grad_norm": 0.05923527708808371,
+      "learning_rate": 5.009932342993296e-06,
+      "loss": 0.3747,
+      "step": 3228
+    },
+    {
+      "epoch": 4.27682119205298,
+      "grad_norm": 0.05536010001358956,
+      "learning_rate": 4.9920217736018736e-06,
+      "loss": 0.3705,
+      "step": 3229
+    },
+    {
+      "epoch": 4.2781456953642385,
+      "grad_norm": 0.05449068248806535,
+      "learning_rate": 4.9741411458550425e-06,
+      "loss": 0.3736,
+      "step": 3230
+    },
+    {
+      "epoch": 4.279470198675496,
+      "grad_norm": 0.060327157219830625,
+      "learning_rate": 4.956290475045746e-06,
+      "loss": 0.3733,
+      "step": 3231
+    },
+    {
+      "epoch": 4.280794701986755,
+      "grad_norm": 0.05841315545179947,
+      "learning_rate": 4.938469776441333e-06,
+      "loss": 0.3722,
+      "step": 3232
+    },
+    {
+      "epoch": 4.282119205298013,
+      "grad_norm": 0.05492656698512546,
+      "learning_rate": 4.920679065283485e-06,
+      "loss": 0.3733,
+      "step": 3233
+    },
+    {
+      "epoch": 4.283443708609272,
+      "grad_norm": 0.055240003360881286,
+      "learning_rate": 4.902918356788253e-06,
+      "loss": 0.3711,
+      "step": 3234
+    },
+    {
+      "epoch": 4.28476821192053,
+      "grad_norm": 0.05262126718433263,
+      "learning_rate": 4.885187666146043e-06,
+      "loss": 0.3692,
+      "step": 3235
+    },
+    {
+      "epoch": 4.2860927152317885,
+      "grad_norm": 0.051548669503388087,
+      "learning_rate": 4.867487008521549e-06,
+      "loss": 0.376,
+      "step": 3236
+    },
+    {
+      "epoch": 4.287417218543046,
+      "grad_norm": 0.05585987041157282,
+      "learning_rate": 4.849816399053819e-06,
+      "loss": 0.3773,
+      "step": 3237
+    },
+    {
+      "epoch": 4.288741721854304,
+      "grad_norm": 0.05581031739693372,
+      "learning_rate": 4.832175852856167e-06,
+      "loss": 0.3753,
+      "step": 3238
+    },
+    {
+      "epoch": 4.290066225165563,
+      "grad_norm": 0.056178049819487756,
+      "learning_rate": 4.8145653850162255e-06,
+      "loss": 0.3767,
+      "step": 3239
+    },
+    {
+      "epoch": 4.291390728476821,
+      "grad_norm": 0.054545784626811075,
+      "learning_rate": 4.7969850105958585e-06,
+      "loss": 0.371,
+      "step": 3240
+    },
+    {
+      "epoch": 4.29271523178808,
+      "grad_norm": 0.05905961530580799,
+      "learning_rate": 4.779434744631259e-06,
+      "loss": 0.3698,
+      "step": 3241
+    },
+    {
+      "epoch": 4.294039735099338,
+      "grad_norm": 0.057726235283124996,
+      "learning_rate": 4.761914602132796e-06,
+      "loss": 0.3701,
+      "step": 3242
+    },
+    {
+      "epoch": 4.295364238410596,
+      "grad_norm": 0.055841332219419296,
+      "learning_rate": 4.744424598085129e-06,
+      "loss": 0.3714,
+      "step": 3243
+    },
+    {
+      "epoch": 4.296688741721854,
+      "grad_norm": 0.053829397746671157,
+      "learning_rate": 4.726964747447098e-06,
+      "loss": 0.3729,
+      "step": 3244
+    },
+    {
+      "epoch": 4.298013245033113,
+      "grad_norm": 0.05830368091699477,
+      "learning_rate": 4.709535065151798e-06,
+      "loss": 0.3716,
+      "step": 3245
+    },
+    {
+      "epoch": 4.299337748344371,
+      "grad_norm": 0.05624043827788624,
+      "learning_rate": 4.692135566106477e-06,
+      "loss": 0.3714,
+      "step": 3246
+    },
+    {
+      "epoch": 4.300662251655629,
+      "grad_norm": 0.05499569442042148,
+      "learning_rate": 4.674766265192596e-06,
+      "loss": 0.3666,
+      "step": 3247
+    },
+    {
+      "epoch": 4.3019867549668875,
+      "grad_norm": 0.05347884712774157,
+      "learning_rate": 4.657427177265797e-06,
+      "loss": 0.3716,
+      "step": 3248
+    },
+    {
+      "epoch": 4.303311258278145,
+      "grad_norm": 0.05623747810335483,
+      "learning_rate": 4.640118317155837e-06,
+      "loss": 0.3705,
+      "step": 3249
+    },
+    {
+      "epoch": 4.304635761589404,
+      "grad_norm": 0.05539813162826488,
+      "learning_rate": 4.622839699666677e-06,
+      "loss": 0.3741,
+      "step": 3250
+    },
+    {
+      "epoch": 4.305960264900662,
+      "grad_norm": 0.05542648493409444,
+      "learning_rate": 4.605591339576356e-06,
+      "loss": 0.3719,
+      "step": 3251
+    },
+    {
+      "epoch": 4.307284768211921,
+      "grad_norm": 0.05506588465492518,
+      "learning_rate": 4.588373251637084e-06,
+      "loss": 0.3707,
+      "step": 3252
+    },
+    {
+      "epoch": 4.308609271523179,
+      "grad_norm": 0.05846798328041569,
+      "learning_rate": 4.571185450575142e-06,
+      "loss": 0.3739,
+      "step": 3253
+    },
+    {
+      "epoch": 4.3099337748344375,
+      "grad_norm": 0.05429663460973701,
+      "learning_rate": 4.554027951090932e-06,
+      "loss": 0.3743,
+      "step": 3254
+    },
+    {
+      "epoch": 4.311258278145695,
+      "grad_norm": 0.05430602705388554,
+      "learning_rate": 4.536900767858932e-06,
+      "loss": 0.3725,
+      "step": 3255
+    },
+    {
+      "epoch": 4.312582781456953,
+      "grad_norm": 0.053407887047379414,
+      "learning_rate": 4.519803915527679e-06,
+      "loss": 0.3705,
+      "step": 3256
+    },
+    {
+      "epoch": 4.313907284768212,
+      "grad_norm": 0.055060635276879516,
+      "learning_rate": 4.50273740871979e-06,
+      "loss": 0.368,
+      "step": 3257
+    },
+    {
+      "epoch": 4.31523178807947,
+      "grad_norm": 0.054756377142523086,
+      "learning_rate": 4.4857012620319075e-06,
+      "loss": 0.373,
+      "step": 3258
+    },
+    {
+      "epoch": 4.316556291390729,
+      "grad_norm": 0.05784575186416977,
+      "learning_rate": 4.468695490034733e-06,
+      "loss": 0.369,
+      "step": 3259
+    },
+    {
+      "epoch": 4.317880794701987,
+      "grad_norm": 0.052821086509685006,
+      "learning_rate": 4.45172010727295e-06,
+      "loss": 0.3719,
+      "step": 3260
+    },
+    {
+      "epoch": 4.319205298013245,
+      "grad_norm": 0.054171759214417546,
+      "learning_rate": 4.43477512826529e-06,
+      "loss": 0.3778,
+      "step": 3261
+    },
+    {
+      "epoch": 4.320529801324503,
+      "grad_norm": 0.05412059564547874,
+      "learning_rate": 4.417860567504462e-06,
+      "loss": 0.369,
+      "step": 3262
+    },
+    {
+      "epoch": 4.321854304635762,
+      "grad_norm": 0.05426458429512897,
+      "learning_rate": 4.400976439457165e-06,
+      "loss": 0.3744,
+      "step": 3263
+    },
+    {
+      "epoch": 4.32317880794702,
+      "grad_norm": 0.05426930599937499,
+      "learning_rate": 4.384122758564058e-06,
+      "loss": 0.3713,
+      "step": 3264
+    },
+    {
+      "epoch": 4.324503311258278,
+      "grad_norm": 0.05298336899121976,
+      "learning_rate": 4.367299539239778e-06,
+      "loss": 0.3715,
+      "step": 3265
+    },
+    {
+      "epoch": 4.325827814569537,
+      "grad_norm": 0.05484142043110527,
+      "learning_rate": 4.350506795872895e-06,
+      "loss": 0.3749,
+      "step": 3266
+    },
+    {
+      "epoch": 4.3271523178807945,
+      "grad_norm": 0.0533013627098048,
+      "learning_rate": 4.333744542825895e-06,
+      "loss": 0.3729,
+      "step": 3267
+    },
+    {
+      "epoch": 4.328476821192053,
+      "grad_norm": 0.06292519130372307,
+      "learning_rate": 4.317012794435243e-06,
+      "loss": 0.3705,
+      "step": 3268
+    },
+    {
+      "epoch": 4.329801324503311,
+      "grad_norm": 0.05917673954139492,
+      "learning_rate": 4.300311565011251e-06,
+      "loss": 0.3751,
+      "step": 3269
+    },
+    {
+      "epoch": 4.33112582781457,
+      "grad_norm": 0.050798658161655894,
+      "learning_rate": 4.283640868838177e-06,
+      "loss": 0.3719,
+      "step": 3270
+    },
+    {
+      "epoch": 4.332450331125828,
+      "grad_norm": 0.05785124927907248,
+      "learning_rate": 4.267000720174128e-06,
+      "loss": 0.3736,
+      "step": 3271
+    },
+    {
+      "epoch": 4.3337748344370866,
+      "grad_norm": 0.056398270553440336,
+      "learning_rate": 4.250391133251115e-06,
+      "loss": 0.3719,
+      "step": 3272
+    },
+    {
+      "epoch": 4.335099337748344,
+      "grad_norm": 0.05963086072458534,
+      "learning_rate": 4.233812122274979e-06,
+      "loss": 0.3705,
+      "step": 3273
+    },
+    {
+      "epoch": 4.336423841059602,
+      "grad_norm": 0.052510865516351896,
+      "learning_rate": 4.217263701425434e-06,
+      "loss": 0.3712,
+      "step": 3274
+    },
+    {
+      "epoch": 4.337748344370861,
+      "grad_norm": 0.05267339862668582,
+      "learning_rate": 4.200745884856021e-06,
+      "loss": 0.3707,
+      "step": 3275
+    },
+    {
+      "epoch": 4.339072847682119,
+      "grad_norm": 0.05400225391215884,
+      "learning_rate": 4.184258686694125e-06,
+      "loss": 0.3727,
+      "step": 3276
+    },
+    {
+      "epoch": 4.340397350993378,
+      "grad_norm": 0.057177484212471055,
+      "learning_rate": 4.167802121040909e-06,
+      "loss": 0.371,
+      "step": 3277
+    },
+    {
+      "epoch": 4.341721854304636,
+      "grad_norm": 0.05258235892107664,
+      "learning_rate": 4.151376201971351e-06,
+      "loss": 0.3765,
+      "step": 3278
+    },
+    {
+      "epoch": 4.343046357615894,
+      "grad_norm": 0.05310068458271093,
+      "learning_rate": 4.134980943534235e-06,
+      "loss": 0.3735,
+      "step": 3279
+    },
+    {
+      "epoch": 4.344370860927152,
+      "grad_norm": 0.054223231553374406,
+      "learning_rate": 4.1186163597520945e-06,
+      "loss": 0.371,
+      "step": 3280
+    },
+    {
+      "epoch": 4.34569536423841,
+      "grad_norm": 0.058334681180233466,
+      "learning_rate": 4.1022824646212415e-06,
+      "loss": 0.3709,
+      "step": 3281
+    },
+    {
+      "epoch": 4.347019867549669,
+      "grad_norm": 0.05486893414544714,
+      "learning_rate": 4.085979272111749e-06,
+      "loss": 0.3709,
+      "step": 3282
+    },
+    {
+      "epoch": 4.348344370860927,
+      "grad_norm": 0.060756354682925756,
+      "learning_rate": 4.069706796167423e-06,
+      "loss": 0.3754,
+      "step": 3283
+    },
+    {
+      "epoch": 4.349668874172186,
+      "grad_norm": 0.06080318752738516,
+      "learning_rate": 4.0534650507057846e-06,
+      "loss": 0.3693,
+      "step": 3284
+    },
+    {
+      "epoch": 4.3509933774834435,
+      "grad_norm": 0.05918215596837122,
+      "learning_rate": 4.037254049618096e-06,
+      "loss": 0.3704,
+      "step": 3285
+    },
+    {
+      "epoch": 4.352317880794702,
+      "grad_norm": 0.056305174387794955,
+      "learning_rate": 4.0210738067693045e-06,
+      "loss": 0.3722,
+      "step": 3286
+    },
+    {
+      "epoch": 4.35364238410596,
+      "grad_norm": 0.05878210524968779,
+      "learning_rate": 4.004924335998052e-06,
+      "loss": 0.3723,
+      "step": 3287
+    },
+    {
+      "epoch": 4.354966887417219,
+      "grad_norm": 0.06031598895561373,
+      "learning_rate": 3.988805651116674e-06,
+      "loss": 0.3771,
+      "step": 3288
+    },
+    {
+      "epoch": 4.356291390728477,
+      "grad_norm": 0.06079687694473692,
+      "learning_rate": 3.972717765911171e-06,
+      "loss": 0.3752,
+      "step": 3289
+    },
+    {
+      "epoch": 4.357615894039735,
+      "grad_norm": 0.05865430966504273,
+      "learning_rate": 3.956660694141201e-06,
+      "loss": 0.372,
+      "step": 3290
+    },
+    {
+      "epoch": 4.3589403973509935,
+      "grad_norm": 0.057959231621634855,
+      "learning_rate": 3.940634449540057e-06,
+      "loss": 0.3753,
+      "step": 3291
+    },
+    {
+      "epoch": 4.360264900662251,
+      "grad_norm": 0.056583953824116684,
+      "learning_rate": 3.924639045814686e-06,
+      "loss": 0.3727,
+      "step": 3292
+    },
+    {
+      "epoch": 4.36158940397351,
+      "grad_norm": 0.06098362212766336,
+      "learning_rate": 3.908674496645635e-06,
+      "loss": 0.3692,
+      "step": 3293
+    },
+    {
+      "epoch": 4.362913907284768,
+      "grad_norm": 0.06475229432656446,
+      "learning_rate": 3.892740815687081e-06,
+      "loss": 0.3668,
+      "step": 3294
+    },
+    {
+      "epoch": 4.364238410596027,
+      "grad_norm": 0.060241809898571086,
+      "learning_rate": 3.8768380165667884e-06,
+      "loss": 0.3707,
+      "step": 3295
+    },
+    {
+      "epoch": 4.365562913907285,
+      "grad_norm": 0.05640658621946441,
+      "learning_rate": 3.86096611288612e-06,
+      "loss": 0.3723,
+      "step": 3296
+    },
+    {
+      "epoch": 4.3668874172185435,
+      "grad_norm": 0.06543909805472091,
+      "learning_rate": 3.845125118220003e-06,
+      "loss": 0.3738,
+      "step": 3297
+    },
+    {
+      "epoch": 4.368211920529801,
+      "grad_norm": 0.061473430122810375,
+      "learning_rate": 3.8293150461169215e-06,
+      "loss": 0.3705,
+      "step": 3298
+    },
+    {
+      "epoch": 4.369536423841059,
+      "grad_norm": 0.056379981306383595,
+      "learning_rate": 3.8135359100989424e-06,
+      "loss": 0.3724,
+      "step": 3299
+    },
+    {
+      "epoch": 4.370860927152318,
+      "grad_norm": 0.05443835433695726,
+      "learning_rate": 3.797787723661639e-06,
+      "loss": 0.3738,
+      "step": 3300
+    },
+    {
+      "epoch": 4.372185430463576,
+      "grad_norm": 0.05573816619182568,
+      "learning_rate": 3.7820705002741353e-06,
+      "loss": 0.3737,
+      "step": 3301
+    },
+    {
+      "epoch": 4.373509933774835,
+      "grad_norm": 0.05857003747472218,
+      "learning_rate": 3.7663842533790697e-06,
+      "loss": 0.3722,
+      "step": 3302
+    },
+    {
+      "epoch": 4.3748344370860925,
+      "grad_norm": 0.055844776249709044,
+      "learning_rate": 3.7507289963925897e-06,
+      "loss": 0.3708,
+      "step": 3303
+    },
+    {
+      "epoch": 4.376158940397351,
+      "grad_norm": 0.056476283414995686,
+      "learning_rate": 3.7351047427043186e-06,
+      "loss": 0.3708,
+      "step": 3304
+    },
+    {
+      "epoch": 4.377483443708609,
+      "grad_norm": 0.054587964768334024,
+      "learning_rate": 3.7195115056773933e-06,
+      "loss": 0.3744,
+      "step": 3305
+    },
+    {
+      "epoch": 4.378807947019867,
+      "grad_norm": 0.05990812489393697,
+      "learning_rate": 3.703949298648399e-06,
+      "loss": 0.3697,
+      "step": 3306
+    },
+    {
+      "epoch": 4.380132450331126,
+      "grad_norm": 0.05820559450819373,
+      "learning_rate": 3.688418134927383e-06,
+      "loss": 0.3725,
+      "step": 3307
+    },
+    {
+      "epoch": 4.381456953642384,
+      "grad_norm": 0.05974933480929865,
+      "learning_rate": 3.672918027797856e-06,
+      "loss": 0.3707,
+      "step": 3308
+    },
+    {
+      "epoch": 4.3827814569536425,
+      "grad_norm": 0.054125817555478725,
+      "learning_rate": 3.6574489905167565e-06,
+      "loss": 0.3661,
+      "step": 3309
+    },
+    {
+      "epoch": 4.3841059602649,
+      "grad_norm": 0.06023973037298084,
+      "learning_rate": 3.6420110363144612e-06,
+      "loss": 0.373,
+      "step": 3310
+    },
+    {
+      "epoch": 4.385430463576159,
+      "grad_norm": 0.06035523827745607,
+      "learning_rate": 3.626604178394737e-06,
+      "loss": 0.3698,
+      "step": 3311
+    },
+    {
+      "epoch": 4.386754966887417,
+      "grad_norm": 0.05258775778820736,
+      "learning_rate": 3.6112284299347855e-06,
+      "loss": 0.3727,
+      "step": 3312
+    },
+    {
+      "epoch": 4.388079470198676,
+      "grad_norm": 0.1081631737972739,
+      "learning_rate": 3.595883804085172e-06,
+      "loss": 0.3674,
+      "step": 3313
+    },
+    {
+      "epoch": 4.389403973509934,
+      "grad_norm": 0.05567773005247213,
+      "learning_rate": 3.5805703139698687e-06,
+      "loss": 0.3675,
+      "step": 3314
+    },
+    {
+      "epoch": 4.390728476821192,
+      "grad_norm": 0.06071449617395098,
+      "learning_rate": 3.5652879726862e-06,
+      "loss": 0.3695,
+      "step": 3315
+    },
+    {
+      "epoch": 4.39205298013245,
+      "grad_norm": 0.054506272485323505,
+      "learning_rate": 3.5500367933048563e-06,
+      "loss": 0.3699,
+      "step": 3316
+    },
+    {
+      "epoch": 4.393377483443708,
+      "grad_norm": 0.05885052277206538,
+      "learning_rate": 3.5348167888698837e-06,
+      "loss": 0.3719,
+      "step": 3317
+    },
+    {
+      "epoch": 4.394701986754967,
+      "grad_norm": 0.05643164597813604,
+      "learning_rate": 3.5196279723986425e-06,
+      "loss": 0.3682,
+      "step": 3318
+    },
+    {
+      "epoch": 4.396026490066225,
+      "grad_norm": 0.055307138662794395,
+      "learning_rate": 3.5044703568818482e-06,
+      "loss": 0.3703,
+      "step": 3319
+    },
+    {
+      "epoch": 4.397350993377484,
+      "grad_norm": 0.05958713609665852,
+      "learning_rate": 3.4893439552834995e-06,
+      "loss": 0.3668,
+      "step": 3320
+    },
+    {
+      "epoch": 4.398675496688742,
+      "grad_norm": 0.053225552374267196,
+      "learning_rate": 3.474248780540932e-06,
+      "loss": 0.371,
+      "step": 3321
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.05138231996516966,
+      "learning_rate": 3.4591848455647294e-06,
+      "loss": 0.3735,
+      "step": 3322
+    },
+    {
+      "epoch": 4.401324503311258,
+      "grad_norm": 0.052151328899043,
+      "learning_rate": 3.4441521632388165e-06,
+      "loss": 0.3705,
+      "step": 3323
+    },
+    {
+      "epoch": 4.402649006622516,
+      "grad_norm": 0.05649826585964485,
+      "learning_rate": 3.4291507464203265e-06,
+      "loss": 0.3673,
+      "step": 3324
+    },
+    {
+      "epoch": 4.403973509933775,
+      "grad_norm": 0.05311831507501756,
+      "learning_rate": 3.4141806079397033e-06,
+      "loss": 0.3722,
+      "step": 3325
+    },
+    {
+      "epoch": 4.405298013245033,
+      "grad_norm": 0.05975279378786029,
+      "learning_rate": 3.399241760600598e-06,
+      "loss": 0.3733,
+      "step": 3326
+    },
+    {
+      "epoch": 4.406622516556292,
+      "grad_norm": 0.05314603748681124,
+      "learning_rate": 3.3843342171799274e-06,
+      "loss": 0.3727,
+      "step": 3327
+    },
+    {
+      "epoch": 4.407947019867549,
+      "grad_norm": 0.055408807481035695,
+      "learning_rate": 3.3694579904278134e-06,
+      "loss": 0.3747,
+      "step": 3328
+    },
+    {
+      "epoch": 4.409271523178808,
+      "grad_norm": 0.052729833830710796,
+      "learning_rate": 3.3546130930676114e-06,
+      "loss": 0.3683,
+      "step": 3329
+    },
+    {
+      "epoch": 4.410596026490066,
+      "grad_norm": 0.05469023487727927,
+      "learning_rate": 3.3397995377958803e-06,
+      "loss": 0.3711,
+      "step": 3330
+    },
+    {
+      "epoch": 4.411920529801325,
+      "grad_norm": 0.05029105291105915,
+      "learning_rate": 3.3250173372823525e-06,
+      "loss": 0.3701,
+      "step": 3331
+    },
+    {
+      "epoch": 4.413245033112583,
+      "grad_norm": 0.053797192199714926,
+      "learning_rate": 3.3102665041699677e-06,
+      "loss": 0.3698,
+      "step": 3332
+    },
+    {
+      "epoch": 4.414569536423841,
+      "grad_norm": 0.05409608674297109,
+      "learning_rate": 3.295547051074821e-06,
+      "loss": 0.3741,
+      "step": 3333
+    },
+    {
+      "epoch": 4.415894039735099,
+      "grad_norm": 0.0531005784247034,
+      "learning_rate": 3.2808589905861886e-06,
+      "loss": 0.3722,
+      "step": 3334
+    },
+    {
+      "epoch": 4.417218543046357,
+      "grad_norm": 0.0550120664950517,
+      "learning_rate": 3.2662023352664616e-06,
+      "loss": 0.3719,
+      "step": 3335
+    },
+    {
+      "epoch": 4.418543046357616,
+      "grad_norm": 0.051956375777313164,
+      "learning_rate": 3.251577097651204e-06,
+      "loss": 0.3688,
+      "step": 3336
+    },
+    {
+      "epoch": 4.419867549668874,
+      "grad_norm": 0.05317577016464127,
+      "learning_rate": 3.236983290249107e-06,
+      "loss": 0.3738,
+      "step": 3337
+    },
+    {
+      "epoch": 4.421192052980133,
+      "grad_norm": 0.051536787982733305,
+      "learning_rate": 3.2224209255419557e-06,
+      "loss": 0.3697,
+      "step": 3338
+    },
+    {
+      "epoch": 4.422516556291391,
+      "grad_norm": 0.054601373967059504,
+      "learning_rate": 3.2078900159846717e-06,
+      "loss": 0.3704,
+      "step": 3339
+    },
+    {
+      "epoch": 4.423841059602649,
+      "grad_norm": 0.052995398769948854,
+      "learning_rate": 3.193390574005246e-06,
+      "loss": 0.3721,
+      "step": 3340
+    },
+    {
+      "epoch": 4.425165562913907,
+      "grad_norm": 0.053219030669820706,
+      "learning_rate": 3.1789226120047824e-06,
+      "loss": 0.3712,
+      "step": 3341
+    },
+    {
+      "epoch": 4.426490066225165,
+      "grad_norm": 0.05297663381234791,
+      "learning_rate": 3.1644861423574346e-06,
+      "loss": 0.3693,
+      "step": 3342
+    },
+    {
+      "epoch": 4.427814569536424,
+      "grad_norm": 0.05535122779246543,
+      "learning_rate": 3.1500811774104467e-06,
+      "loss": 0.3695,
+      "step": 3343
+    },
+    {
+      "epoch": 4.429139072847682,
+      "grad_norm": 0.050760853130584205,
+      "learning_rate": 3.135707729484101e-06,
+      "loss": 0.3716,
+      "step": 3344
+    },
+    {
+      "epoch": 4.430463576158941,
+      "grad_norm": 0.051235818016288084,
+      "learning_rate": 3.1213658108717372e-06,
+      "loss": 0.3711,
+      "step": 3345
+    },
+    {
+      "epoch": 4.4317880794701985,
+      "grad_norm": 0.051584836380401275,
+      "learning_rate": 3.107055433839703e-06,
+      "loss": 0.3718,
+      "step": 3346
+    },
+    {
+      "epoch": 4.433112582781457,
+      "grad_norm": 0.05222197031791826,
+      "learning_rate": 3.0927766106274036e-06,
+      "loss": 0.3708,
+      "step": 3347
+    },
+    {
+      "epoch": 4.434437086092715,
+      "grad_norm": 0.056205163675413464,
+      "learning_rate": 3.078529353447235e-06,
+      "loss": 0.3692,
+      "step": 3348
+    },
+    {
+      "epoch": 4.435761589403974,
+      "grad_norm": 0.05530333459355767,
+      "learning_rate": 3.064313674484578e-06,
+      "loss": 0.3723,
+      "step": 3349
+    },
+    {
+      "epoch": 4.437086092715232,
+      "grad_norm": 0.05241580868335749,
+      "learning_rate": 3.050129585897863e-06,
+      "loss": 0.3729,
+      "step": 3350
+    },
+    {
+      "epoch": 4.43841059602649,
+      "grad_norm": 0.053308596416170494,
+      "learning_rate": 3.0359770998184433e-06,
+      "loss": 0.3699,
+      "step": 3351
+    },
+    {
+      "epoch": 4.4397350993377485,
+      "grad_norm": 0.05351718784658625,
+      "learning_rate": 3.0218562283506723e-06,
+      "loss": 0.3729,
+      "step": 3352
+    },
+    {
+      "epoch": 4.441059602649006,
+      "grad_norm": 0.05336197412523471,
+      "learning_rate": 3.0077669835718535e-06,
+      "loss": 0.3706,
+      "step": 3353
+    },
+    {
+      "epoch": 4.442384105960265,
+      "grad_norm": 0.055018194536030574,
+      "learning_rate": 2.99370937753225e-06,
+      "loss": 0.3702,
+      "step": 3354
+    },
+    {
+      "epoch": 4.443708609271523,
+      "grad_norm": 0.06083516287211753,
+      "learning_rate": 2.979683422255053e-06,
+      "loss": 0.371,
+      "step": 3355
+    },
+    {
+      "epoch": 4.445033112582782,
+      "grad_norm": 0.05331786500889179,
+      "learning_rate": 2.9656891297363953e-06,
+      "loss": 0.3684,
+      "step": 3356
+    },
+    {
+      "epoch": 4.44635761589404,
+      "grad_norm": 0.05618508662088391,
+      "learning_rate": 2.9517265119453254e-06,
+      "loss": 0.3722,
+      "step": 3357
+    },
+    {
+      "epoch": 4.447682119205298,
+      "grad_norm": 0.05493157751162418,
+      "learning_rate": 2.937795580823788e-06,
+      "loss": 0.3667,
+      "step": 3358
+    },
+    {
+      "epoch": 4.449006622516556,
+      "grad_norm": 0.06764550185222543,
+      "learning_rate": 2.9238963482866567e-06,
+      "loss": 0.3767,
+      "step": 3359
+    },
+    {
+      "epoch": 4.450331125827814,
+      "grad_norm": 0.04938965762459705,
+      "learning_rate": 2.9100288262216535e-06,
+      "loss": 0.3701,
+      "step": 3360
+    },
+    {
+      "epoch": 4.451655629139073,
+      "grad_norm": 0.05233893465724733,
+      "learning_rate": 2.8961930264894243e-06,
+      "loss": 0.3704,
+      "step": 3361
+    },
+    {
+      "epoch": 4.452980132450331,
+      "grad_norm": 0.05697719455524134,
+      "learning_rate": 2.882388960923437e-06,
+      "loss": 0.3723,
+      "step": 3362
+    },
+    {
+      "epoch": 4.45430463576159,
+      "grad_norm": 0.0518338470488526,
+      "learning_rate": 2.8686166413300553e-06,
+      "loss": 0.3709,
+      "step": 3363
+    },
+    {
+      "epoch": 4.4556291390728475,
+      "grad_norm": 0.05313634294369378,
+      "learning_rate": 2.8548760794884713e-06,
+      "loss": 0.3701,
+      "step": 3364
+    },
+    {
+      "epoch": 4.456953642384106,
+      "grad_norm": 0.05181751678338091,
+      "learning_rate": 2.8411672871507325e-06,
+      "loss": 0.373,
+      "step": 3365
+    },
+    {
+      "epoch": 4.458278145695364,
+      "grad_norm": 0.053622010506040055,
+      "learning_rate": 2.8274902760416866e-06,
+      "loss": 0.3719,
+      "step": 3366
+    },
+    {
+      "epoch": 4.459602649006623,
+      "grad_norm": 0.05390141829871152,
+      "learning_rate": 2.8138450578590302e-06,
+      "loss": 0.369,
+      "step": 3367
+    },
+    {
+      "epoch": 4.460927152317881,
+      "grad_norm": 0.05565143220263221,
+      "learning_rate": 2.80023164427325e-06,
+      "loss": 0.3695,
+      "step": 3368
+    },
+    {
+      "epoch": 4.462251655629139,
+      "grad_norm": 0.05399458599992344,
+      "learning_rate": 2.786650046927619e-06,
+      "loss": 0.3689,
+      "step": 3369
+    },
+    {
+      "epoch": 4.4635761589403975,
+      "grad_norm": 0.05229507908096956,
+      "learning_rate": 2.773100277438232e-06,
+      "loss": 0.369,
+      "step": 3370
+    },
+    {
+      "epoch": 4.464900662251655,
+      "grad_norm": 0.05656871548439233,
+      "learning_rate": 2.759582347393934e-06,
+      "loss": 0.3737,
+      "step": 3371
+    },
+    {
+      "epoch": 4.466225165562914,
+      "grad_norm": 0.051641791719630126,
+      "learning_rate": 2.746096268356362e-06,
+      "loss": 0.3729,
+      "step": 3372
+    },
+    {
+      "epoch": 4.467549668874172,
+      "grad_norm": 0.05248484170981964,
+      "learning_rate": 2.7326420518598795e-06,
+      "loss": 0.373,
+      "step": 3373
+    },
+    {
+      "epoch": 4.468874172185431,
+      "grad_norm": 0.053407662031936254,
+      "learning_rate": 2.7192197094116334e-06,
+      "loss": 0.3724,
+      "step": 3374
+    },
+    {
+      "epoch": 4.470198675496689,
+      "grad_norm": 0.05177219680025056,
+      "learning_rate": 2.705829252491481e-06,
+      "loss": 0.3681,
+      "step": 3375
+    },
+    {
+      "epoch": 4.4715231788079475,
+      "grad_norm": 0.05492841726274564,
+      "learning_rate": 2.6924706925520207e-06,
+      "loss": 0.3731,
+      "step": 3376
+    },
+    {
+      "epoch": 4.472847682119205,
+      "grad_norm": 0.051469166104163806,
+      "learning_rate": 2.6791440410185753e-06,
+      "loss": 0.3715,
+      "step": 3377
+    },
+    {
+      "epoch": 4.474172185430463,
+      "grad_norm": 0.05407033903112173,
+      "learning_rate": 2.6658493092891745e-06,
+      "loss": 0.3711,
+      "step": 3378
+    },
+    {
+      "epoch": 4.475496688741722,
+      "grad_norm": 0.05232074922244117,
+      "learning_rate": 2.65258650873454e-06,
+      "loss": 0.3738,
+      "step": 3379
+    },
+    {
+      "epoch": 4.47682119205298,
+      "grad_norm": 0.0549981945359926,
+      "learning_rate": 2.6393556506980834e-06,
+      "loss": 0.3742,
+      "step": 3380
+    },
+    {
+      "epoch": 4.478145695364239,
+      "grad_norm": 0.05393946315170122,
+      "learning_rate": 2.626156746495907e-06,
+      "loss": 0.3711,
+      "step": 3381
+    },
+    {
+      "epoch": 4.479470198675497,
+      "grad_norm": 0.051360740242812085,
+      "learning_rate": 2.612989807416768e-06,
+      "loss": 0.3731,
+      "step": 3382
+    },
+    {
+      "epoch": 4.480794701986755,
+      "grad_norm": 0.05317098267527241,
+      "learning_rate": 2.5998548447221028e-06,
+      "loss": 0.3664,
+      "step": 3383
+    },
+    {
+      "epoch": 4.482119205298013,
+      "grad_norm": 0.05173866753654059,
+      "learning_rate": 2.5867518696459915e-06,
+      "loss": 0.3723,
+      "step": 3384
+    },
+    {
+      "epoch": 4.483443708609272,
+      "grad_norm": 0.052825756443002185,
+      "learning_rate": 2.573680893395154e-06,
+      "loss": 0.3745,
+      "step": 3385
+    },
+    {
+      "epoch": 4.48476821192053,
+      "grad_norm": 0.05061042753792952,
+      "learning_rate": 2.5606419271489324e-06,
+      "loss": 0.3701,
+      "step": 3386
+    },
+    {
+      "epoch": 4.486092715231788,
+      "grad_norm": 0.05368816641121618,
+      "learning_rate": 2.5476349820593216e-06,
+      "loss": 0.3718,
+      "step": 3387
+    },
+    {
+      "epoch": 4.4874172185430465,
+      "grad_norm": 0.05040825109729622,
+      "learning_rate": 2.5346600692508894e-06,
+      "loss": 0.3715,
+      "step": 3388
+    },
+    {
+      "epoch": 4.488741721854304,
+      "grad_norm": 0.054267528485775325,
+      "learning_rate": 2.5217171998208344e-06,
+      "loss": 0.3706,
+      "step": 3389
+    },
+    {
+      "epoch": 4.490066225165563,
+      "grad_norm": 0.052808843848268484,
+      "learning_rate": 2.5088063848389377e-06,
+      "loss": 0.3686,
+      "step": 3390
+    },
+    {
+      "epoch": 4.491390728476821,
+      "grad_norm": 0.0568120763685595,
+      "learning_rate": 2.4959276353475746e-06,
+      "loss": 0.3708,
+      "step": 3391
+    },
+    {
+      "epoch": 4.49271523178808,
+      "grad_norm": 0.055791904435398734,
+      "learning_rate": 2.4830809623616947e-06,
+      "loss": 0.3728,
+      "step": 3392
+    },
+    {
+      "epoch": 4.494039735099338,
+      "grad_norm": 0.05247451276320937,
+      "learning_rate": 2.4702663768687973e-06,
+      "loss": 0.3728,
+      "step": 3393
+    },
+    {
+      "epoch": 4.495364238410596,
+      "grad_norm": 0.05204138352830065,
+      "learning_rate": 2.4574838898289557e-06,
+      "loss": 0.3682,
+      "step": 3394
+    },
+    {
+      "epoch": 4.496688741721854,
+      "grad_norm": 0.053947184316665425,
+      "learning_rate": 2.4447335121747794e-06,
+      "loss": 0.372,
+      "step": 3395
+    },
+    {
+      "epoch": 4.498013245033112,
+      "grad_norm": 0.053493949151805845,
+      "learning_rate": 2.432015254811422e-06,
+      "loss": 0.3668,
+      "step": 3396
+    },
+    {
+      "epoch": 4.499337748344371,
+      "grad_norm": 0.05321170964936278,
+      "learning_rate": 2.4193291286165545e-06,
+      "loss": 0.3717,
+      "step": 3397
+    },
+    {
+      "epoch": 4.500662251655629,
+      "grad_norm": 0.052715833374386,
+      "learning_rate": 2.4066751444403824e-06,
+      "loss": 0.3737,
+      "step": 3398
+    },
+    {
+      "epoch": 4.501986754966888,
+      "grad_norm": 0.052370569842419425,
+      "learning_rate": 2.394053313105613e-06,
+      "loss": 0.3728,
+      "step": 3399
+    },
+    {
+      "epoch": 4.503311258278146,
+      "grad_norm": 0.049837316006815145,
+      "learning_rate": 2.3814636454074467e-06,
+      "loss": 0.3682,
+      "step": 3400
+    },
+    {
+      "epoch": 4.5046357615894035,
+      "grad_norm": 0.04891162580946775,
+      "learning_rate": 2.368906152113595e-06,
+      "loss": 0.3688,
+      "step": 3401
+    },
+    {
+      "epoch": 4.505960264900662,
+      "grad_norm": 0.05041903793650895,
+      "learning_rate": 2.3563808439642255e-06,
+      "loss": 0.3707,
+      "step": 3402
+    },
+    {
+      "epoch": 4.50728476821192,
+      "grad_norm": 0.05099276218975096,
+      "learning_rate": 2.343887731671992e-06,
+      "loss": 0.3745,
+      "step": 3403
+    },
+    {
+      "epoch": 4.508609271523179,
+      "grad_norm": 0.056776943712872174,
+      "learning_rate": 2.3314268259220187e-06,
+      "loss": 0.3751,
+      "step": 3404
+    },
+    {
+      "epoch": 4.509933774834437,
+      "grad_norm": 0.05202737885423659,
+      "learning_rate": 2.318998137371877e-06,
+      "loss": 0.3736,
+      "step": 3405
+    },
+    {
+      "epoch": 4.511258278145696,
+      "grad_norm": 0.04849146662399287,
+      "learning_rate": 2.306601676651572e-06,
+      "loss": 0.3711,
+      "step": 3406
+    },
+    {
+      "epoch": 4.5125827814569535,
+      "grad_norm": 0.054980406646380944,
+      "learning_rate": 2.2942374543635725e-06,
+      "loss": 0.3694,
+      "step": 3407
+    },
+    {
+      "epoch": 4.513907284768212,
+      "grad_norm": 0.05157219882113787,
+      "learning_rate": 2.2819054810827445e-06,
+      "loss": 0.3722,
+      "step": 3408
+    },
+    {
+      "epoch": 4.51523178807947,
+      "grad_norm": 0.05020024460015974,
+      "learning_rate": 2.269605767356384e-06,
+      "loss": 0.3717,
+      "step": 3409
+    },
+    {
+      "epoch": 4.516556291390728,
+      "grad_norm": 0.05131778640141878,
+      "learning_rate": 2.2573383237042056e-06,
+      "loss": 0.37,
+      "step": 3410
+    },
+    {
+      "epoch": 4.517880794701987,
+      "grad_norm": 0.052115626505021856,
+      "learning_rate": 2.2451031606183137e-06,
+      "loss": 0.3719,
+      "step": 3411
+    },
+    {
+      "epoch": 4.519205298013245,
+      "grad_norm": 0.050001477154474276,
+      "learning_rate": 2.2329002885632088e-06,
+      "loss": 0.3693,
+      "step": 3412
+    },
+    {
+      "epoch": 4.520529801324503,
+      "grad_norm": 0.049819418611397784,
+      "learning_rate": 2.2207297179757602e-06,
+      "loss": 0.372,
+      "step": 3413
+    },
+    {
+      "epoch": 4.521854304635761,
+      "grad_norm": 0.05217401442161094,
+      "learning_rate": 2.208591459265237e-06,
+      "loss": 0.3726,
+      "step": 3414
+    },
+    {
+      "epoch": 4.52317880794702,
+      "grad_norm": 0.050767625492970286,
+      "learning_rate": 2.196485522813241e-06,
+      "loss": 0.3744,
+      "step": 3415
+    },
+    {
+      "epoch": 4.524503311258278,
+      "grad_norm": 0.05160102138388155,
+      "learning_rate": 2.1844119189737567e-06,
+      "loss": 0.3777,
+      "step": 3416
+    },
+    {
+      "epoch": 4.525827814569537,
+      "grad_norm": 0.05098830393489293,
+      "learning_rate": 2.1723706580730886e-06,
+      "loss": 0.3727,
+      "step": 3417
+    },
+    {
+      "epoch": 4.527152317880795,
+      "grad_norm": 0.05831197762771162,
+      "learning_rate": 2.160361750409905e-06,
+      "loss": 0.3703,
+      "step": 3418
+    },
+    {
+      "epoch": 4.5284768211920525,
+      "grad_norm": 0.0540859874868686,
+      "learning_rate": 2.1483852062551945e-06,
+      "loss": 0.3695,
+      "step": 3419
+    },
+    {
+      "epoch": 4.529801324503311,
+      "grad_norm": 0.049334765959467425,
+      "learning_rate": 2.1364410358522527e-06,
+      "loss": 0.3692,
+      "step": 3420
+    },
+    {
+      "epoch": 4.531125827814569,
+      "grad_norm": 0.05035317099895468,
+      "learning_rate": 2.124529249416707e-06,
+      "loss": 0.3709,
+      "step": 3421
+    },
+    {
+      "epoch": 4.532450331125828,
+      "grad_norm": 0.05379807662777689,
+      "learning_rate": 2.112649857136466e-06,
+      "loss": 0.3718,
+      "step": 3422
+    },
+    {
+      "epoch": 4.533774834437086,
+      "grad_norm": 0.05550503604854219,
+      "learning_rate": 2.100802869171754e-06,
+      "loss": 0.3768,
+      "step": 3423
+    },
+    {
+      "epoch": 4.535099337748345,
+      "grad_norm": 0.04992841853714559,
+      "learning_rate": 2.0889882956550612e-06,
+      "loss": 0.3694,
+      "step": 3424
+    },
+    {
+      "epoch": 4.5364238410596025,
+      "grad_norm": 0.04895232495690382,
+      "learning_rate": 2.0772061466911664e-06,
+      "loss": 0.3712,
+      "step": 3425
+    },
+    {
+      "epoch": 4.537748344370861,
+      "grad_norm": 0.051020302548840095,
+      "learning_rate": 2.0654564323571158e-06,
+      "loss": 0.3689,
+      "step": 3426
+    },
+    {
+      "epoch": 4.539072847682119,
+      "grad_norm": 0.05297744640946457,
+      "learning_rate": 2.0537391627022173e-06,
+      "loss": 0.3726,
+      "step": 3427
+    },
+    {
+      "epoch": 4.540397350993377,
+      "grad_norm": 0.0517666573608898,
+      "learning_rate": 2.042054347748006e-06,
+      "loss": 0.3754,
+      "step": 3428
+    },
+    {
+      "epoch": 4.541721854304636,
+      "grad_norm": 0.04988199748150816,
+      "learning_rate": 2.030401997488296e-06,
+      "loss": 0.3732,
+      "step": 3429
+    },
+    {
+      "epoch": 4.543046357615894,
+      "grad_norm": 0.04988102010689479,
+      "learning_rate": 2.0187821218891067e-06,
+      "loss": 0.3748,
+      "step": 3430
+    },
+    {
+      "epoch": 4.5443708609271525,
+      "grad_norm": 0.04922332250329625,
+      "learning_rate": 2.007194730888693e-06,
+      "loss": 0.37,
+      "step": 3431
+    },
+    {
+      "epoch": 4.54569536423841,
+      "grad_norm": 0.05528435682266555,
+      "learning_rate": 1.9956398343975315e-06,
+      "loss": 0.3728,
+      "step": 3432
+    },
+    {
+      "epoch": 4.547019867549669,
+      "grad_norm": 0.051686874594079016,
+      "learning_rate": 1.98411744229829e-06,
+      "loss": 0.3696,
+      "step": 3433
+    },
+    {
+      "epoch": 4.548344370860927,
+      "grad_norm": 0.05161749173472687,
+      "learning_rate": 1.9726275644458638e-06,
+      "loss": 0.3726,
+      "step": 3434
+    },
+    {
+      "epoch": 4.549668874172186,
+      "grad_norm": 0.05189600225030318,
+      "learning_rate": 1.9611702106673024e-06,
+      "loss": 0.3716,
+      "step": 3435
+    },
+    {
+      "epoch": 4.550993377483444,
+      "grad_norm": 0.05134059316928688,
+      "learning_rate": 1.9497453907618746e-06,
+      "loss": 0.3755,
+      "step": 3436
+    },
+    {
+      "epoch": 4.552317880794702,
+      "grad_norm": 0.055536520630817404,
+      "learning_rate": 1.938353114500999e-06,
+      "loss": 0.3711,
+      "step": 3437
+    },
+    {
+      "epoch": 4.55364238410596,
+      "grad_norm": 0.051704156219609006,
+      "learning_rate": 1.9269933916282644e-06,
+      "loss": 0.3716,
+      "step": 3438
+    },
+    {
+      "epoch": 4.554966887417218,
+      "grad_norm": 0.04857095781293348,
+      "learning_rate": 1.9156662318594366e-06,
+      "loss": 0.3724,
+      "step": 3439
+    },
+    {
+      "epoch": 4.556291390728477,
+      "grad_norm": 0.051143126164429835,
+      "learning_rate": 1.9043716448824012e-06,
+      "loss": 0.3695,
+      "step": 3440
+    },
+    {
+      "epoch": 4.557615894039735,
+      "grad_norm": 0.051380049330315404,
+      "learning_rate": 1.8931096403572092e-06,
+      "loss": 0.3711,
+      "step": 3441
+    },
+    {
+      "epoch": 4.558940397350994,
+      "grad_norm": 0.050890056618983416,
+      "learning_rate": 1.881880227916022e-06,
+      "loss": 0.3722,
+      "step": 3442
+    },
+    {
+      "epoch": 4.5602649006622515,
+      "grad_norm": 0.05216875342696687,
+      "learning_rate": 1.8706834171631528e-06,
+      "loss": 0.3672,
+      "step": 3443
+    },
+    {
+      "epoch": 4.56158940397351,
+      "grad_norm": 0.051070841304210114,
+      "learning_rate": 1.8595192176750032e-06,
+      "loss": 0.3711,
+      "step": 3444
+    },
+    {
+      "epoch": 4.562913907284768,
+      "grad_norm": 0.04991508184594716,
+      "learning_rate": 1.8483876390001043e-06,
+      "loss": 0.3709,
+      "step": 3445
+    },
+    {
+      "epoch": 4.564238410596026,
+      "grad_norm": 0.049604254460112886,
+      "learning_rate": 1.837288690659076e-06,
+      "loss": 0.3762,
+      "step": 3446
+    },
+    {
+      "epoch": 4.565562913907285,
+      "grad_norm": 0.05174042863249963,
+      "learning_rate": 1.8262223821446445e-06,
+      "loss": 0.3758,
+      "step": 3447
+    },
+    {
+      "epoch": 4.566887417218543,
+      "grad_norm": 0.05235496870566877,
+      "learning_rate": 1.8151887229215902e-06,
+      "loss": 0.3683,
+      "step": 3448
+    },
+    {
+      "epoch": 4.5682119205298015,
+      "grad_norm": 0.05205629446361908,
+      "learning_rate": 1.804187722426809e-06,
+      "loss": 0.374,
+      "step": 3449
+    },
+    {
+      "epoch": 4.569536423841059,
+      "grad_norm": 0.04948060325245394,
+      "learning_rate": 1.7932193900692274e-06,
+      "loss": 0.37,
+      "step": 3450
+    },
+    {
+      "epoch": 4.570860927152318,
+      "grad_norm": 0.049218202327598666,
+      "learning_rate": 1.782283735229844e-06,
+      "loss": 0.3735,
+      "step": 3451
+    },
+    {
+      "epoch": 4.572185430463576,
+      "grad_norm": 0.047372622034994365,
+      "learning_rate": 1.7713807672617233e-06,
+      "loss": 0.3699,
+      "step": 3452
+    },
+    {
+      "epoch": 4.573509933774835,
+      "grad_norm": 0.04990840302763397,
+      "learning_rate": 1.7605104954899577e-06,
+      "loss": 0.371,
+      "step": 3453
+    },
+    {
+      "epoch": 4.574834437086093,
+      "grad_norm": 0.05091606059544923,
+      "learning_rate": 1.7496729292116832e-06,
+      "loss": 0.3688,
+      "step": 3454
+    },
+    {
+      "epoch": 4.576158940397351,
+      "grad_norm": 0.04834880443866868,
+      "learning_rate": 1.7388680776960543e-06,
+      "loss": 0.3731,
+      "step": 3455
+    },
+    {
+      "epoch": 4.577483443708609,
+      "grad_norm": 0.05031022481881592,
+      "learning_rate": 1.728095950184261e-06,
+      "loss": 0.3742,
+      "step": 3456
+    },
+    {
+      "epoch": 4.578807947019867,
+      "grad_norm": 0.04915310890512325,
+      "learning_rate": 1.71735655588948e-06,
+      "loss": 0.3708,
+      "step": 3457
+    },
+    {
+      "epoch": 4.580132450331126,
+      "grad_norm": 0.048660631717810786,
+      "learning_rate": 1.7066499039969154e-06,
+      "loss": 0.373,
+      "step": 3458
+    },
+    {
+      "epoch": 4.581456953642384,
+      "grad_norm": 0.04886223109078645,
+      "learning_rate": 1.6959760036637662e-06,
+      "loss": 0.3712,
+      "step": 3459
+    },
+    {
+      "epoch": 4.582781456953643,
+      "grad_norm": 0.05014415709557918,
+      "learning_rate": 1.685334864019197e-06,
+      "loss": 0.3731,
+      "step": 3460
+    },
+    {
+      "epoch": 4.584105960264901,
+      "grad_norm": 0.05000393956226562,
+      "learning_rate": 1.6747264941643803e-06,
+      "loss": 0.3702,
+      "step": 3461
+    },
+    {
+      "epoch": 4.585430463576159,
+      "grad_norm": 0.04917834207832912,
+      "learning_rate": 1.6641509031724368e-06,
+      "loss": 0.3727,
+      "step": 3462
+    },
+    {
+      "epoch": 4.586754966887417,
+      "grad_norm": 0.04958989712710885,
+      "learning_rate": 1.6536081000884818e-06,
+      "loss": 0.3675,
+      "step": 3463
+    },
+    {
+      "epoch": 4.588079470198675,
+      "grad_norm": 0.0501848146818985,
+      "learning_rate": 1.6430980939295515e-06,
+      "loss": 0.3687,
+      "step": 3464
+    },
+    {
+      "epoch": 4.589403973509934,
+      "grad_norm": 0.04839017947407933,
+      "learning_rate": 1.6326208936846598e-06,
+      "loss": 0.3745,
+      "step": 3465
+    },
+    {
+      "epoch": 4.590728476821192,
+      "grad_norm": 0.05243982138602747,
+      "learning_rate": 1.62217650831475e-06,
+      "loss": 0.3644,
+      "step": 3466
+    },
+    {
+      "epoch": 4.592052980132451,
+      "grad_norm": 0.04963566322982005,
+      "learning_rate": 1.6117649467527119e-06,
+      "loss": 0.3714,
+      "step": 3467
+    },
+    {
+      "epoch": 4.593377483443708,
+      "grad_norm": 0.050363474383515745,
+      "learning_rate": 1.6013862179033379e-06,
+      "loss": 0.3696,
+      "step": 3468
+    },
+    {
+      "epoch": 4.594701986754967,
+      "grad_norm": 0.04800037065925949,
+      "learning_rate": 1.5910403306433675e-06,
+      "loss": 0.3676,
+      "step": 3469
+    },
+    {
+      "epoch": 4.596026490066225,
+      "grad_norm": 0.048689967723190966,
+      "learning_rate": 1.580727293821429e-06,
+      "loss": 0.3723,
+      "step": 3470
+    },
+    {
+      "epoch": 4.597350993377484,
+      "grad_norm": 0.04942010599070712,
+      "learning_rate": 1.5704471162580582e-06,
+      "loss": 0.3679,
+      "step": 3471
+    },
+    {
+      "epoch": 4.598675496688742,
+      "grad_norm": 0.050355369138197756,
+      "learning_rate": 1.5601998067456925e-06,
+      "loss": 0.369,
+      "step": 3472
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.05071760566747442,
+      "learning_rate": 1.549985374048668e-06,
+      "loss": 0.3694,
+      "step": 3473
+    },
+    {
+      "epoch": 4.601324503311258,
+      "grad_norm": 0.050225596821260994,
+      "learning_rate": 1.5398038269031834e-06,
+      "loss": 0.373,
+      "step": 3474
+    },
+    {
+      "epoch": 4.602649006622516,
+      "grad_norm": 0.047787735514135427,
+      "learning_rate": 1.5296551740173171e-06,
+      "loss": 0.373,
+      "step": 3475
+    },
+    {
+      "epoch": 4.603973509933775,
+      "grad_norm": 0.05039812550094974,
+      "learning_rate": 1.5195394240710194e-06,
+      "loss": 0.3698,
+      "step": 3476
+    },
+    {
+      "epoch": 4.605298013245033,
+      "grad_norm": 0.049071262152577,
+      "learning_rate": 1.5094565857160891e-06,
+      "loss": 0.37,
+      "step": 3477
+    },
+    {
+      "epoch": 4.606622516556292,
+      "grad_norm": 0.04803109511992792,
+      "learning_rate": 1.499406667576193e-06,
+      "loss": 0.3705,
+      "step": 3478
+    },
+    {
+      "epoch": 4.60794701986755,
+      "grad_norm": 0.04791653562313853,
+      "learning_rate": 1.489389678246811e-06,
+      "loss": 0.3727,
+      "step": 3479
+    },
+    {
+      "epoch": 4.609271523178808,
+      "grad_norm": 0.04811184753111811,
+      "learning_rate": 1.4794056262953026e-06,
+      "loss": 0.3737,
+      "step": 3480
+    },
+    {
+      "epoch": 4.610596026490066,
+      "grad_norm": 0.052100083737956444,
+      "learning_rate": 1.4694545202608246e-06,
+      "loss": 0.3723,
+      "step": 3481
+    },
+    {
+      "epoch": 4.611920529801324,
+      "grad_norm": 0.049695103769025355,
+      "learning_rate": 1.4595363686543596e-06,
+      "loss": 0.372,
+      "step": 3482
+    },
+    {
+      "epoch": 4.613245033112583,
+      "grad_norm": 0.050120943983876814,
+      "learning_rate": 1.4496511799587175e-06,
+      "loss": 0.3671,
+      "step": 3483
+    },
+    {
+      "epoch": 4.614569536423841,
+      "grad_norm": 0.051848222141524666,
+      "learning_rate": 1.4397989626285048e-06,
+      "loss": 0.3714,
+      "step": 3484
+    },
+    {
+      "epoch": 4.6158940397351,
+      "grad_norm": 0.049995970369596035,
+      "learning_rate": 1.429979725090127e-06,
+      "loss": 0.368,
+      "step": 3485
+    },
+    {
+      "epoch": 4.6172185430463575,
+      "grad_norm": 0.04925766634469774,
+      "learning_rate": 1.4201934757417956e-06,
+      "loss": 0.3688,
+      "step": 3486
+    },
+    {
+      "epoch": 4.618543046357616,
+      "grad_norm": 0.04886643782304452,
+      "learning_rate": 1.4104402229535042e-06,
+      "loss": 0.3748,
+      "step": 3487
+    },
+    {
+      "epoch": 4.619867549668874,
+      "grad_norm": 0.0508888316715253,
+      "learning_rate": 1.4007199750670064e-06,
+      "loss": 0.3736,
+      "step": 3488
+    },
+    {
+      "epoch": 4.621192052980133,
+      "grad_norm": 0.049141969980363374,
+      "learning_rate": 1.3910327403958522e-06,
+      "loss": 0.3743,
+      "step": 3489
+    },
+    {
+      "epoch": 4.622516556291391,
+      "grad_norm": 0.050052105671774004,
+      "learning_rate": 1.3813785272253478e-06,
+      "loss": 0.37,
+      "step": 3490
+    },
+    {
+      "epoch": 4.623841059602649,
+      "grad_norm": 0.05175989058274671,
+      "learning_rate": 1.3717573438125408e-06,
+      "loss": 0.3715,
+      "step": 3491
+    },
+    {
+      "epoch": 4.6251655629139075,
+      "grad_norm": 0.050386655015118395,
+      "learning_rate": 1.3621691983862495e-06,
+      "loss": 0.3684,
+      "step": 3492
+    },
+    {
+      "epoch": 4.626490066225165,
+      "grad_norm": 0.04994898422365207,
+      "learning_rate": 1.3526140991470294e-06,
+      "loss": 0.371,
+      "step": 3493
+    },
+    {
+      "epoch": 4.627814569536424,
+      "grad_norm": 0.05075298528312515,
+      "learning_rate": 1.3430920542671788e-06,
+      "loss": 0.3668,
+      "step": 3494
+    },
+    {
+      "epoch": 4.629139072847682,
+      "grad_norm": 0.048785433513188106,
+      "learning_rate": 1.3336030718907077e-06,
+      "loss": 0.3726,
+      "step": 3495
+    },
+    {
+      "epoch": 4.630463576158941,
+      "grad_norm": 0.051949232409483515,
+      "learning_rate": 1.3241471601333644e-06,
+      "loss": 0.375,
+      "step": 3496
+    },
+    {
+      "epoch": 4.631788079470199,
+      "grad_norm": 0.050265074677427454,
+      "learning_rate": 1.3147243270825994e-06,
+      "loss": 0.3712,
+      "step": 3497
+    },
+    {
+      "epoch": 4.633112582781457,
+      "grad_norm": 0.04799202047242098,
+      "learning_rate": 1.3053345807975926e-06,
+      "loss": 0.3739,
+      "step": 3498
+    },
+    {
+      "epoch": 4.634437086092715,
+      "grad_norm": 0.04982888010820935,
+      "learning_rate": 1.2959779293091957e-06,
+      "loss": 0.3697,
+      "step": 3499
+    },
+    {
+      "epoch": 4.635761589403973,
+      "grad_norm": 0.05016674431714092,
+      "learning_rate": 1.2866543806199805e-06,
+      "loss": 0.37,
+      "step": 3500
+    },
+    {
+      "epoch": 4.637086092715232,
+      "grad_norm": 0.05002597030282589,
+      "learning_rate": 1.2773639427041951e-06,
+      "loss": 0.3664,
+      "step": 3501
+    },
+    {
+      "epoch": 4.63841059602649,
+      "grad_norm": 0.047203640296451785,
+      "learning_rate": 1.268106623507772e-06,
+      "loss": 0.3731,
+      "step": 3502
+    },
+    {
+      "epoch": 4.639735099337749,
+      "grad_norm": 0.04918094486918962,
+      "learning_rate": 1.2588824309483204e-06,
+      "loss": 0.3727,
+      "step": 3503
+    },
+    {
+      "epoch": 4.6410596026490065,
+      "grad_norm": 0.0512125018834427,
+      "learning_rate": 1.2496913729151028e-06,
+      "loss": 0.3723,
+      "step": 3504
+    },
+    {
+      "epoch": 4.642384105960264,
+      "grad_norm": 0.049941326306949946,
+      "learning_rate": 1.2405334572690619e-06,
+      "loss": 0.3731,
+      "step": 3505
+    },
+    {
+      "epoch": 4.643708609271523,
+      "grad_norm": 0.051047440074706636,
+      "learning_rate": 1.2314086918427815e-06,
+      "loss": 0.3702,
+      "step": 3506
+    },
+    {
+      "epoch": 4.645033112582782,
+      "grad_norm": 0.049798122766014564,
+      "learning_rate": 1.222317084440494e-06,
+      "loss": 0.3697,
+      "step": 3507
+    },
+    {
+      "epoch": 4.64635761589404,
+      "grad_norm": 0.053106523161334705,
+      "learning_rate": 1.2132586428380777e-06,
+      "loss": 0.3677,
+      "step": 3508
+    },
+    {
+      "epoch": 4.647682119205298,
+      "grad_norm": 0.04911688237993245,
+      "learning_rate": 1.2042333747830504e-06,
+      "loss": 0.3691,
+      "step": 3509
+    },
+    {
+      "epoch": 4.6490066225165565,
+      "grad_norm": 0.049381968731474185,
+      "learning_rate": 1.1952412879945397e-06,
+      "loss": 0.3704,
+      "step": 3510
+    },
+    {
+      "epoch": 4.650331125827814,
+      "grad_norm": 0.04939667277298799,
+      "learning_rate": 1.186282390163296e-06,
+      "loss": 0.371,
+      "step": 3511
+    },
+    {
+      "epoch": 4.651655629139073,
+      "grad_norm": 0.04942472452228384,
+      "learning_rate": 1.1773566889517008e-06,
+      "loss": 0.3687,
+      "step": 3512
+    },
+    {
+      "epoch": 4.652980132450331,
+      "grad_norm": 0.048377595420302226,
+      "learning_rate": 1.1684641919937322e-06,
+      "loss": 0.373,
+      "step": 3513
+    },
+    {
+      "epoch": 4.654304635761589,
+      "grad_norm": 0.05546456146262929,
+      "learning_rate": 1.1596049068949732e-06,
+      "loss": 0.3715,
+      "step": 3514
+    },
+    {
+      "epoch": 4.655629139072848,
+      "grad_norm": 0.04956735351867367,
+      "learning_rate": 1.1507788412325894e-06,
+      "loss": 0.3704,
+      "step": 3515
+    },
+    {
+      "epoch": 4.656953642384106,
+      "grad_norm": 0.04968657595121092,
+      "learning_rate": 1.1419860025553509e-06,
+      "loss": 0.3675,
+      "step": 3516
+    },
+    {
+      "epoch": 4.658278145695364,
+      "grad_norm": 0.0498454380646188,
+      "learning_rate": 1.1332263983835978e-06,
+      "loss": 0.3701,
+      "step": 3517
+    },
+    {
+      "epoch": 4.659602649006622,
+      "grad_norm": 0.04899941744568456,
+      "learning_rate": 1.1245000362092573e-06,
+      "loss": 0.3656,
+      "step": 3518
+    },
+    {
+      "epoch": 4.660927152317881,
+      "grad_norm": 0.050932666413500846,
+      "learning_rate": 1.115806923495808e-06,
+      "loss": 0.3714,
+      "step": 3519
+    },
+    {
+      "epoch": 4.662251655629139,
+      "grad_norm": 0.05187373352951874,
+      "learning_rate": 1.1071470676783024e-06,
+      "loss": 0.3726,
+      "step": 3520
+    },
+    {
+      "epoch": 4.663576158940398,
+      "grad_norm": 0.049766689626379604,
+      "learning_rate": 1.0985204761633538e-06,
+      "loss": 0.3738,
+      "step": 3521
+    },
+    {
+      "epoch": 4.664900662251656,
+      "grad_norm": 0.04853281443411676,
+      "learning_rate": 1.0899271563291091e-06,
+      "loss": 0.3678,
+      "step": 3522
+    },
+    {
+      "epoch": 4.6662251655629134,
+      "grad_norm": 0.04845023692430414,
+      "learning_rate": 1.0813671155252758e-06,
+      "loss": 0.3709,
+      "step": 3523
+    },
+    {
+      "epoch": 4.667549668874172,
+      "grad_norm": 0.04852697160297893,
+      "learning_rate": 1.0728403610730863e-06,
+      "loss": 0.3686,
+      "step": 3524
+    },
+    {
+      "epoch": 4.66887417218543,
+      "grad_norm": 0.04770678221744561,
+      "learning_rate": 1.0643469002653161e-06,
+      "loss": 0.3719,
+      "step": 3525
+    },
+    {
+      "epoch": 4.670198675496689,
+      "grad_norm": 0.048741352501966595,
+      "learning_rate": 1.0558867403662432e-06,
+      "loss": 0.372,
+      "step": 3526
+    },
+    {
+      "epoch": 4.671523178807947,
+      "grad_norm": 0.048885052978858895,
+      "learning_rate": 1.0474598886116883e-06,
+      "loss": 0.3699,
+      "step": 3527
+    },
+    {
+      "epoch": 4.6728476821192055,
+      "grad_norm": 0.050525376950444116,
+      "learning_rate": 1.0390663522089705e-06,
+      "loss": 0.3706,
+      "step": 3528
+    },
+    {
+      "epoch": 4.674172185430463,
+      "grad_norm": 0.048472959463595984,
+      "learning_rate": 1.0307061383369255e-06,
+      "loss": 0.3714,
+      "step": 3529
+    },
+    {
+      "epoch": 4.675496688741722,
+      "grad_norm": 0.04930480253169775,
+      "learning_rate": 1.0223792541458733e-06,
+      "loss": 0.372,
+      "step": 3530
+    },
+    {
+      "epoch": 4.67682119205298,
+      "grad_norm": 0.05231336495413311,
+      "learning_rate": 1.0140857067576415e-06,
+      "loss": 0.3677,
+      "step": 3531
+    },
+    {
+      "epoch": 4.678145695364238,
+      "grad_norm": 0.04928066364068643,
+      "learning_rate": 1.0058255032655383e-06,
+      "loss": 0.3712,
+      "step": 3532
+    },
+    {
+      "epoch": 4.679470198675497,
+      "grad_norm": 0.04900635258710766,
+      "learning_rate": 9.975986507343527e-07,
+      "loss": 0.373,
+      "step": 3533
+    },
+    {
+      "epoch": 4.680794701986755,
+      "grad_norm": 0.04732502154673254,
+      "learning_rate": 9.89405156200358e-07,
+      "loss": 0.3669,
+      "step": 3534
+    },
+    {
+      "epoch": 4.682119205298013,
+      "grad_norm": 0.04747764818180715,
+      "learning_rate": 9.81245026671287e-07,
+      "loss": 0.3705,
+      "step": 3535
+    },
+    {
+      "epoch": 4.683443708609271,
+      "grad_norm": 0.047894221546503656,
+      "learning_rate": 9.73118269126343e-07,
+      "loss": 0.3661,
+      "step": 3536
+    },
+    {
+      "epoch": 4.68476821192053,
+      "grad_norm": 0.049483137213733056,
+      "learning_rate": 9.650248905161797e-07,
+      "loss": 0.3728,
+      "step": 3537
+    },
+    {
+      "epoch": 4.686092715231788,
+      "grad_norm": 0.05020169484069155,
+      "learning_rate": 9.569648977629176e-07,
+      "loss": 0.3698,
+      "step": 3538
+    },
+    {
+      "epoch": 4.687417218543047,
+      "grad_norm": 0.048607497777383005,
+      "learning_rate": 9.489382977600959e-07,
+      "loss": 0.3718,
+      "step": 3539
+    },
+    {
+      "epoch": 4.688741721854305,
+      "grad_norm": 0.04912387353521574,
+      "learning_rate": 9.409450973727207e-07,
+      "loss": 0.3662,
+      "step": 3540
+    },
+    {
+      "epoch": 4.6900662251655625,
+      "grad_norm": 0.04818106718037408,
+      "learning_rate": 9.329853034372172e-07,
+      "loss": 0.3695,
+      "step": 3541
+    },
+    {
+      "epoch": 4.691390728476821,
+      "grad_norm": 0.04809359604833861,
+      "learning_rate": 9.250589227614459e-07,
+      "loss": 0.3684,
+      "step": 3542
+    },
+    {
+      "epoch": 4.692715231788079,
+      "grad_norm": 0.04889993026131669,
+      "learning_rate": 9.171659621246865e-07,
+      "loss": 0.3697,
+      "step": 3543
+    },
+    {
+      "epoch": 4.694039735099338,
+      "grad_norm": 0.0484395358046598,
+      "learning_rate": 9.09306428277632e-07,
+      "loss": 0.3722,
+      "step": 3544
+    },
+    {
+      "epoch": 4.695364238410596,
+      "grad_norm": 0.048839062924069115,
+      "learning_rate": 9.014803279423901e-07,
+      "loss": 0.3729,
+      "step": 3545
+    },
+    {
+      "epoch": 4.696688741721855,
+      "grad_norm": 0.046927555658942,
+      "learning_rate": 8.936876678124728e-07,
+      "loss": 0.3675,
+      "step": 3546
+    },
+    {
+      "epoch": 4.6980132450331125,
+      "grad_norm": 0.04780022897326259,
+      "learning_rate": 8.859284545527891e-07,
+      "loss": 0.3712,
+      "step": 3547
+    },
+    {
+      "epoch": 4.699337748344371,
+      "grad_norm": 0.047317362304307074,
+      "learning_rate": 8.782026947996481e-07,
+      "loss": 0.3703,
+      "step": 3548
+    },
+    {
+      "epoch": 4.700662251655629,
+      "grad_norm": 0.04878837823924596,
+      "learning_rate": 8.705103951607418e-07,
+      "loss": 0.3735,
+      "step": 3549
+    },
+    {
+      "epoch": 4.701986754966887,
+      "grad_norm": 0.04762951226738576,
+      "learning_rate": 8.628515622151368e-07,
+      "loss": 0.3729,
+      "step": 3550
+    },
+    {
+      "epoch": 4.703311258278146,
+      "grad_norm": 0.04807651493133192,
+      "learning_rate": 8.552262025132996e-07,
+      "loss": 0.374,
+      "step": 3551
+    },
+    {
+      "epoch": 4.704635761589404,
+      "grad_norm": 0.04983235480107215,
+      "learning_rate": 8.476343225770401e-07,
+      "loss": 0.3734,
+      "step": 3552
+    },
+    {
+      "epoch": 4.705960264900662,
+      "grad_norm": 0.05046176202333655,
+      "learning_rate": 8.400759288995464e-07,
+      "loss": 0.3718,
+      "step": 3553
+    },
+    {
+      "epoch": 4.70728476821192,
+      "grad_norm": 0.04960956736734803,
+      "learning_rate": 8.325510279453674e-07,
+      "loss": 0.3724,
+      "step": 3554
+    },
+    {
+      "epoch": 4.708609271523179,
+      "grad_norm": 0.04937264644890309,
+      "learning_rate": 8.250596261504085e-07,
+      "loss": 0.3716,
+      "step": 3555
+    },
+    {
+      "epoch": 4.709933774834437,
+      "grad_norm": 0.04873964444847916,
+      "learning_rate": 8.176017299219175e-07,
+      "loss": 0.3696,
+      "step": 3556
+    },
+    {
+      "epoch": 4.711258278145696,
+      "grad_norm": 0.047993764835541984,
+      "learning_rate": 8.101773456384898e-07,
+      "loss": 0.3715,
+      "step": 3557
+    },
+    {
+      "epoch": 4.712582781456954,
+      "grad_norm": 0.04860123910483785,
+      "learning_rate": 8.027864796500595e-07,
+      "loss": 0.3722,
+      "step": 3558
+    },
+    {
+      "epoch": 4.7139072847682115,
+      "grad_norm": 0.04932047543683835,
+      "learning_rate": 7.954291382778812e-07,
+      "loss": 0.3706,
+      "step": 3559
+    },
+    {
+      "epoch": 4.71523178807947,
+      "grad_norm": 0.047720651106326464,
+      "learning_rate": 7.881053278145567e-07,
+      "loss": 0.3704,
+      "step": 3560
+    },
+    {
+      "epoch": 4.716556291390728,
+      "grad_norm": 0.04922503816729467,
+      "learning_rate": 7.808150545239912e-07,
+      "loss": 0.3695,
+      "step": 3561
+    },
+    {
+      "epoch": 4.717880794701987,
+      "grad_norm": 0.04831595416134412,
+      "learning_rate": 7.735583246414191e-07,
+      "loss": 0.3711,
+      "step": 3562
+    },
+    {
+      "epoch": 4.719205298013245,
+      "grad_norm": 0.048285874125365225,
+      "learning_rate": 7.663351443733868e-07,
+      "loss": 0.3709,
+      "step": 3563
+    },
+    {
+      "epoch": 4.720529801324504,
+      "grad_norm": 0.04937375362642654,
+      "learning_rate": 7.591455198977215e-07,
+      "loss": 0.3676,
+      "step": 3564
+    },
+    {
+      "epoch": 4.7218543046357615,
+      "grad_norm": 0.04789227773290257,
+      "learning_rate": 7.519894573635888e-07,
+      "loss": 0.3695,
+      "step": 3565
+    },
+    {
+      "epoch": 4.72317880794702,
+      "grad_norm": 0.04760748668535866,
+      "learning_rate": 7.448669628914218e-07,
+      "loss": 0.3711,
+      "step": 3566
+    },
+    {
+      "epoch": 4.724503311258278,
+      "grad_norm": 0.04736149458300514,
+      "learning_rate": 7.377780425729475e-07,
+      "loss": 0.3705,
+      "step": 3567
+    },
+    {
+      "epoch": 4.725827814569536,
+      "grad_norm": 0.047384493222244965,
+      "learning_rate": 7.307227024711872e-07,
+      "loss": 0.3703,
+      "step": 3568
+    },
+    {
+      "epoch": 4.727152317880795,
+      "grad_norm": 0.0485214449441692,
+      "learning_rate": 7.237009486204427e-07,
+      "loss": 0.369,
+      "step": 3569
+    },
+    {
+      "epoch": 4.728476821192053,
+      "grad_norm": 0.04897495473197847,
+      "learning_rate": 7.167127870262703e-07,
+      "loss": 0.3736,
+      "step": 3570
+    },
+    {
+      "epoch": 4.7298013245033115,
+      "grad_norm": 0.04811346540950375,
+      "learning_rate": 7.097582236655199e-07,
+      "loss": 0.3647,
+      "step": 3571
+    },
+    {
+      "epoch": 4.731125827814569,
+      "grad_norm": 0.047293713053865295,
+      "learning_rate": 7.028372644862913e-07,
+      "loss": 0.3714,
+      "step": 3572
+    },
+    {
+      "epoch": 4.732450331125828,
+      "grad_norm": 0.04798591622109246,
+      "learning_rate": 6.959499154079386e-07,
+      "loss": 0.3705,
+      "step": 3573
+    },
+    {
+      "epoch": 4.733774834437086,
+      "grad_norm": 0.04597169068654359,
+      "learning_rate": 6.890961823210873e-07,
+      "loss": 0.3723,
+      "step": 3574
+    },
+    {
+      "epoch": 4.735099337748345,
+      "grad_norm": 0.04751882365726145,
+      "learning_rate": 6.822760710876042e-07,
+      "loss": 0.3708,
+      "step": 3575
+    },
+    {
+      "epoch": 4.736423841059603,
+      "grad_norm": 0.0469040361978882,
+      "learning_rate": 6.75489587540592e-07,
+      "loss": 0.3699,
+      "step": 3576
+    },
+    {
+      "epoch": 4.737748344370861,
+      "grad_norm": 0.04941063246451634,
+      "learning_rate": 6.687367374843989e-07,
+      "loss": 0.3709,
+      "step": 3577
+    },
+    {
+      "epoch": 4.739072847682119,
+      "grad_norm": 0.05130980274338694,
+      "learning_rate": 6.620175266946138e-07,
+      "loss": 0.3729,
+      "step": 3578
+    },
+    {
+      "epoch": 4.740397350993377,
+      "grad_norm": 0.0474658844965183,
+      "learning_rate": 6.553319609180353e-07,
+      "loss": 0.3697,
+      "step": 3579
+    },
+    {
+      "epoch": 4.741721854304636,
+      "grad_norm": 0.048375532728327904,
+      "learning_rate": 6.486800458727072e-07,
+      "loss": 0.3684,
+      "step": 3580
+    },
+    {
+      "epoch": 4.743046357615894,
+      "grad_norm": 0.04974114327652747,
+      "learning_rate": 6.420617872478785e-07,
+      "loss": 0.3699,
+      "step": 3581
+    },
+    {
+      "epoch": 4.744370860927153,
+      "grad_norm": 0.04864105245703009,
+      "learning_rate": 6.354771907040213e-07,
+      "loss": 0.3727,
+      "step": 3582
+    },
+    {
+      "epoch": 4.7456953642384105,
+      "grad_norm": 0.0499811693470824,
+      "learning_rate": 6.289262618728131e-07,
+      "loss": 0.369,
+      "step": 3583
+    },
+    {
+      "epoch": 4.747019867549669,
+      "grad_norm": 0.04912895478464192,
+      "learning_rate": 6.22409006357132e-07,
+      "loss": 0.3712,
+      "step": 3584
+    },
+    {
+      "epoch": 4.748344370860927,
+      "grad_norm": 0.04825961790934474,
+      "learning_rate": 6.159254297310657e-07,
+      "loss": 0.3658,
+      "step": 3585
+    },
+    {
+      "epoch": 4.749668874172185,
+      "grad_norm": 0.05099022679084235,
+      "learning_rate": 6.094755375398808e-07,
+      "loss": 0.3649,
+      "step": 3586
+    },
+    {
+      "epoch": 4.750993377483444,
+      "grad_norm": 0.050082938142726484,
+      "learning_rate": 6.030593353000625e-07,
+      "loss": 0.3683,
+      "step": 3587
+    },
+    {
+      "epoch": 4.752317880794702,
+      "grad_norm": 0.048115325667882854,
+      "learning_rate": 5.966768284992474e-07,
+      "loss": 0.3728,
+      "step": 3588
+    },
+    {
+      "epoch": 4.7536423841059605,
+      "grad_norm": 0.0458778455839087,
+      "learning_rate": 5.903280225962782e-07,
+      "loss": 0.3697,
+      "step": 3589
+    },
+    {
+      "epoch": 4.754966887417218,
+      "grad_norm": 0.047463931481874666,
+      "learning_rate": 5.840129230211666e-07,
+      "loss": 0.37,
+      "step": 3590
+    },
+    {
+      "epoch": 4.756291390728477,
+      "grad_norm": 0.04937359494833054,
+      "learning_rate": 5.777315351750901e-07,
+      "loss": 0.3655,
+      "step": 3591
+    },
+    {
+      "epoch": 4.757615894039735,
+      "grad_norm": 0.050882310187741364,
+      "learning_rate": 5.714838644304044e-07,
+      "loss": 0.3703,
+      "step": 3592
+    },
+    {
+      "epoch": 4.758940397350994,
+      "grad_norm": 0.04855671824284819,
+      "learning_rate": 5.652699161306085e-07,
+      "loss": 0.3692,
+      "step": 3593
+    },
+    {
+      "epoch": 4.760264900662252,
+      "grad_norm": 0.046586614590526966,
+      "learning_rate": 5.590896955903846e-07,
+      "loss": 0.3732,
+      "step": 3594
+    },
+    {
+      "epoch": 4.76158940397351,
+      "grad_norm": 0.047093724546377995,
+      "learning_rate": 5.529432080955443e-07,
+      "loss": 0.3675,
+      "step": 3595
+    },
+    {
+      "epoch": 4.762913907284768,
+      "grad_norm": 0.04777405169244871,
+      "learning_rate": 5.468304589030738e-07,
+      "loss": 0.3689,
+      "step": 3596
+    },
+    {
+      "epoch": 4.764238410596026,
+      "grad_norm": 0.04944973805944234,
+      "learning_rate": 5.407514532410752e-07,
+      "loss": 0.3664,
+      "step": 3597
+    },
+    {
+      "epoch": 4.765562913907285,
+      "grad_norm": 0.04857472418444432,
+      "learning_rate": 5.347061963088162e-07,
+      "loss": 0.3686,
+      "step": 3598
+    },
+    {
+      "epoch": 4.766887417218543,
+      "grad_norm": 0.046680044972905545,
+      "learning_rate": 5.286946932766723e-07,
+      "loss": 0.3677,
+      "step": 3599
+    },
+    {
+      "epoch": 4.768211920529802,
+      "grad_norm": 0.04886873319642435,
+      "learning_rate": 5.227169492861839e-07,
+      "loss": 0.3661,
+      "step": 3600
+    },
+    {
+      "epoch": 4.76953642384106,
+      "grad_norm": 0.04873242582857316,
+      "learning_rate": 5.167729694499812e-07,
+      "loss": 0.3699,
+      "step": 3601
+    },
+    {
+      "epoch": 4.770860927152318,
+      "grad_norm": 0.04991579793811218,
+      "learning_rate": 5.108627588518467e-07,
+      "loss": 0.3764,
+      "step": 3602
+    },
+    {
+      "epoch": 4.772185430463576,
+      "grad_norm": 0.04712666148063696,
+      "learning_rate": 5.049863225466656e-07,
+      "loss": 0.3701,
+      "step": 3603
+    },
+    {
+      "epoch": 4.773509933774834,
+      "grad_norm": 0.04846706342557157,
+      "learning_rate": 4.991436655604353e-07,
+      "loss": 0.3735,
+      "step": 3604
+    },
+    {
+      "epoch": 4.774834437086093,
+      "grad_norm": 0.04800978076674294,
+      "learning_rate": 4.933347928902787e-07,
+      "loss": 0.3692,
+      "step": 3605
+    },
+    {
+      "epoch": 4.776158940397351,
+      "grad_norm": 0.04950499950525531,
+      "learning_rate": 4.875597095043994e-07,
+      "loss": 0.374,
+      "step": 3606
+    },
+    {
+      "epoch": 4.77748344370861,
+      "grad_norm": 0.04920992069307053,
+      "learning_rate": 4.818184203421217e-07,
+      "loss": 0.3682,
+      "step": 3607
+    },
+    {
+      "epoch": 4.778807947019867,
+      "grad_norm": 0.04812788200525432,
+      "learning_rate": 4.761109303138556e-07,
+      "loss": 0.369,
+      "step": 3608
+    },
+    {
+      "epoch": 4.780132450331126,
+      "grad_norm": 0.04711094993579624,
+      "learning_rate": 4.7043724430110516e-07,
+      "loss": 0.3693,
+      "step": 3609
+    },
+    {
+      "epoch": 4.781456953642384,
+      "grad_norm": 0.04637655432627991,
+      "learning_rate": 4.647973671564687e-07,
+      "loss": 0.3704,
+      "step": 3610
+    },
+    {
+      "epoch": 4.782781456953643,
+      "grad_norm": 0.04953608198206592,
+      "learning_rate": 4.5919130370362554e-07,
+      "loss": 0.3688,
+      "step": 3611
+    },
+    {
+      "epoch": 4.784105960264901,
+      "grad_norm": 0.04900494050039164,
+      "learning_rate": 4.536190587373268e-07,
+      "loss": 0.369,
+      "step": 3612
+    },
+    {
+      "epoch": 4.785430463576159,
+      "grad_norm": 0.04869888366795608,
+      "learning_rate": 4.4808063702340475e-07,
+      "loss": 0.3743,
+      "step": 3613
+    },
+    {
+      "epoch": 4.786754966887417,
+      "grad_norm": 0.04720311083594899,
+      "learning_rate": 4.4257604329876803e-07,
+      "loss": 0.3727,
+      "step": 3614
+    },
+    {
+      "epoch": 4.788079470198675,
+      "grad_norm": 0.04769878881885072,
+      "learning_rate": 4.371052822713839e-07,
+      "loss": 0.3711,
+      "step": 3615
+    },
+    {
+      "epoch": 4.789403973509934,
+      "grad_norm": 0.046528965875488694,
+      "learning_rate": 4.3166835862029634e-07,
+      "loss": 0.3732,
+      "step": 3616
+    },
+    {
+      "epoch": 4.790728476821192,
+      "grad_norm": 0.04825093296891921,
+      "learning_rate": 4.2626527699558996e-07,
+      "loss": 0.3707,
+      "step": 3617
+    },
+    {
+      "epoch": 4.79205298013245,
+      "grad_norm": 0.06582826933803297,
+      "learning_rate": 4.208960420184305e-07,
+      "loss": 0.3734,
+      "step": 3618
+    },
+    {
+      "epoch": 4.793377483443709,
+      "grad_norm": 0.047036741846439085,
+      "learning_rate": 4.155606582810068e-07,
+      "loss": 0.3699,
+      "step": 3619
+    },
+    {
+      "epoch": 4.7947019867549665,
+      "grad_norm": 0.04751411056812309,
+      "learning_rate": 4.1025913034657527e-07,
+      "loss": 0.3706,
+      "step": 3620
+    },
+    {
+      "epoch": 4.796026490066225,
+      "grad_norm": 0.04919668782287666,
+      "learning_rate": 4.049914627494289e-07,
+      "loss": 0.3733,
+      "step": 3621
+    },
+    {
+      "epoch": 4.797350993377483,
+      "grad_norm": 0.048364057668510936,
+      "learning_rate": 3.997576599948971e-07,
+      "loss": 0.3728,
+      "step": 3622
+    },
+    {
+      "epoch": 4.798675496688742,
+      "grad_norm": 0.04803655621936056,
+      "learning_rate": 3.945577265593592e-07,
+      "loss": 0.3695,
+      "step": 3623
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.04873530743409393,
+      "learning_rate": 3.8939166689021313e-07,
+      "loss": 0.3685,
+      "step": 3624
+    },
+    {
+      "epoch": 4.801324503311259,
+      "grad_norm": 0.046544882579237605,
+      "learning_rate": 3.8425948540588454e-07,
+      "loss": 0.3682,
+      "step": 3625
+    },
+    {
+      "epoch": 4.8026490066225165,
+      "grad_norm": 0.04653373966422693,
+      "learning_rate": 3.7916118649583554e-07,
+      "loss": 0.3703,
+      "step": 3626
+    },
+    {
+      "epoch": 4.803973509933774,
+      "grad_norm": 0.045700868811443154,
+      "learning_rate": 3.7409677452053816e-07,
+      "loss": 0.3698,
+      "step": 3627
+    },
+    {
+      "epoch": 4.805298013245033,
+      "grad_norm": 0.048251356475818435,
+      "learning_rate": 3.6906625381148755e-07,
+      "loss": 0.3716,
+      "step": 3628
+    },
+    {
+      "epoch": 4.806622516556291,
+      "grad_norm": 0.04740692746609153,
+      "learning_rate": 3.6406962867118867e-07,
+      "loss": 0.3663,
+      "step": 3629
+    },
+    {
+      "epoch": 4.80794701986755,
+      "grad_norm": 0.04819921241438522,
+      "learning_rate": 3.591069033731609e-07,
+      "loss": 0.3711,
+      "step": 3630
+    },
+    {
+      "epoch": 4.809271523178808,
+      "grad_norm": 0.048826451538358094,
+      "learning_rate": 3.541780821619245e-07,
+      "loss": 0.3696,
+      "step": 3631
+    },
+    {
+      "epoch": 4.8105960264900665,
+      "grad_norm": 0.04690044134800836,
+      "learning_rate": 3.492831692530052e-07,
+      "loss": 0.373,
+      "step": 3632
+    },
+    {
+      "epoch": 4.811920529801324,
+      "grad_norm": 0.046722872875169234,
+      "learning_rate": 3.444221688329297e-07,
+      "loss": 0.3672,
+      "step": 3633
+    },
+    {
+      "epoch": 4.813245033112583,
+      "grad_norm": 0.04833901145330257,
+      "learning_rate": 3.3959508505921666e-07,
+      "loss": 0.3675,
+      "step": 3634
+    },
+    {
+      "epoch": 4.814569536423841,
+      "grad_norm": 0.04608509377592555,
+      "learning_rate": 3.348019220603682e-07,
+      "loss": 0.3709,
+      "step": 3635
+    },
+    {
+      "epoch": 4.815894039735099,
+      "grad_norm": 0.04857720151327524,
+      "learning_rate": 3.3004268393589613e-07,
+      "loss": 0.3695,
+      "step": 3636
+    },
+    {
+      "epoch": 4.817218543046358,
+      "grad_norm": 0.04875853020777623,
+      "learning_rate": 3.2531737475627766e-07,
+      "loss": 0.3684,
+      "step": 3637
+    },
+    {
+      "epoch": 4.8185430463576155,
+      "grad_norm": 0.046764137949407314,
+      "learning_rate": 3.2062599856298227e-07,
+      "loss": 0.3725,
+      "step": 3638
+    },
+    {
+      "epoch": 4.819867549668874,
+      "grad_norm": 0.048262278585526355,
+      "learning_rate": 3.159685593684447e-07,
+      "loss": 0.3702,
+      "step": 3639
+    },
+    {
+      "epoch": 4.821192052980132,
+      "grad_norm": 0.05064286568567621,
+      "learning_rate": 3.1134506115609644e-07,
+      "loss": 0.375,
+      "step": 3640
+    },
+    {
+      "epoch": 4.822516556291391,
+      "grad_norm": 0.04808766788873912,
+      "learning_rate": 3.0675550788031195e-07,
+      "loss": 0.3716,
+      "step": 3641
+    },
+    {
+      "epoch": 4.823841059602649,
+      "grad_norm": 0.04787131849111675,
+      "learning_rate": 3.0219990346644914e-07,
+      "loss": 0.3704,
+      "step": 3642
+    },
+    {
+      "epoch": 4.825165562913908,
+      "grad_norm": 0.04694946343603785,
+      "learning_rate": 2.976782518108401e-07,
+      "loss": 0.3716,
+      "step": 3643
+    },
+    {
+      "epoch": 4.8264900662251655,
+      "grad_norm": 0.046796931576667575,
+      "learning_rate": 2.9319055678075135e-07,
+      "loss": 0.3686,
+      "step": 3644
+    },
+    {
+      "epoch": 4.827814569536423,
+      "grad_norm": 0.04884603774473881,
+      "learning_rate": 2.8873682221442823e-07,
+      "loss": 0.3702,
+      "step": 3645
+    },
+    {
+      "epoch": 4.829139072847682,
+      "grad_norm": 0.04692344137164246,
+      "learning_rate": 2.843170519210636e-07,
+      "loss": 0.3703,
+      "step": 3646
+    },
+    {
+      "epoch": 4.83046357615894,
+      "grad_norm": 0.049418540869306665,
+      "learning_rate": 2.7993124968080265e-07,
+      "loss": 0.3724,
+      "step": 3647
+    },
+    {
+      "epoch": 4.831788079470199,
+      "grad_norm": 0.046114149629414014,
+      "learning_rate": 2.755794192447292e-07,
+      "loss": 0.3698,
+      "step": 3648
+    },
+    {
+      "epoch": 4.833112582781457,
+      "grad_norm": 0.04584358332797208,
+      "learning_rate": 2.7126156433488815e-07,
+      "loss": 0.3712,
+      "step": 3649
+    },
+    {
+      "epoch": 4.8344370860927155,
+      "grad_norm": 0.046558768516828386,
+      "learning_rate": 2.669776886442499e-07,
+      "loss": 0.3704,
+      "step": 3650
+    },
+    {
+      "epoch": 4.835761589403973,
+      "grad_norm": 0.04578210638678965,
+      "learning_rate": 2.627277958367369e-07,
+      "loss": 0.3715,
+      "step": 3651
+    },
+    {
+      "epoch": 4.837086092715232,
+      "grad_norm": 0.04768292118643917,
+      "learning_rate": 2.585118895471972e-07,
+      "loss": 0.3692,
+      "step": 3652
+    },
+    {
+      "epoch": 4.83841059602649,
+      "grad_norm": 0.04680337600365054,
+      "learning_rate": 2.543299733814131e-07,
+      "loss": 0.3695,
+      "step": 3653
+    },
+    {
+      "epoch": 4.839735099337748,
+      "grad_norm": 0.047192432300817995,
+      "learning_rate": 2.5018205091609684e-07,
+      "loss": 0.3725,
+      "step": 3654
+    },
+    {
+      "epoch": 4.841059602649007,
+      "grad_norm": 0.04717099785740153,
+      "learning_rate": 2.4606812569888615e-07,
+      "loss": 0.3696,
+      "step": 3655
+    },
+    {
+      "epoch": 4.842384105960265,
+      "grad_norm": 0.04855336047742059,
+      "learning_rate": 2.419882012483399e-07,
+      "loss": 0.3714,
+      "step": 3656
+    },
+    {
+      "epoch": 4.843708609271523,
+      "grad_norm": 0.04684355861860895,
+      "learning_rate": 2.379422810539378e-07,
+      "loss": 0.3707,
+      "step": 3657
+    },
+    {
+      "epoch": 4.845033112582781,
+      "grad_norm": 0.04664054126909951,
+      "learning_rate": 2.3393036857608076e-07,
+      "loss": 0.3735,
+      "step": 3658
+    },
+    {
+      "epoch": 4.84635761589404,
+      "grad_norm": 0.04736763114622123,
+      "learning_rate": 2.2995246724607733e-07,
+      "loss": 0.3703,
+      "step": 3659
+    },
+    {
+      "epoch": 4.847682119205298,
+      "grad_norm": 0.04638631816294135,
+      "learning_rate": 2.260085804661527e-07,
+      "loss": 0.3748,
+      "step": 3660
+    },
+    {
+      "epoch": 4.849006622516557,
+      "grad_norm": 0.046818227014014537,
+      "learning_rate": 2.2209871160943085e-07,
+      "loss": 0.3673,
+      "step": 3661
+    },
+    {
+      "epoch": 4.850331125827815,
+      "grad_norm": 0.047850046002581345,
+      "learning_rate": 2.1822286401994796e-07,
+      "loss": 0.3749,
+      "step": 3662
+    },
+    {
+      "epoch": 4.8516556291390724,
+      "grad_norm": 0.04792428933454056,
+      "learning_rate": 2.143810410126479e-07,
+      "loss": 0.3691,
+      "step": 3663
+    },
+    {
+      "epoch": 4.852980132450331,
+      "grad_norm": 0.046066844307477144,
+      "learning_rate": 2.1057324587335559e-07,
+      "loss": 0.3695,
+      "step": 3664
+    },
+    {
+      "epoch": 4.854304635761589,
+      "grad_norm": 0.04880915494750946,
+      "learning_rate": 2.0679948185881705e-07,
+      "loss": 0.3719,
+      "step": 3665
+    },
+    {
+      "epoch": 4.855629139072848,
+      "grad_norm": 0.046383753235713754,
+      "learning_rate": 2.0305975219665042e-07,
+      "loss": 0.3683,
+      "step": 3666
+    },
+    {
+      "epoch": 4.856953642384106,
+      "grad_norm": 0.04792566181916634,
+      "learning_rate": 1.9935406008537717e-07,
+      "loss": 0.3735,
+      "step": 3667
+    },
+    {
+      "epoch": 4.8582781456953645,
+      "grad_norm": 0.04699694014713094,
+      "learning_rate": 1.9568240869440424e-07,
+      "loss": 0.3743,
+      "step": 3668
+    },
+    {
+      "epoch": 4.859602649006622,
+      "grad_norm": 0.049196063459245064,
+      "learning_rate": 1.9204480116402413e-07,
+      "loss": 0.3716,
+      "step": 3669
+    },
+    {
+      "epoch": 4.860927152317881,
+      "grad_norm": 0.04909709505457214,
+      "learning_rate": 1.8844124060540593e-07,
+      "loss": 0.3694,
+      "step": 3670
+    },
+    {
+      "epoch": 4.862251655629139,
+      "grad_norm": 0.047686979967442374,
+      "learning_rate": 1.8487173010061755e-07,
+      "loss": 0.3695,
+      "step": 3671
+    },
+    {
+      "epoch": 4.863576158940397,
+      "grad_norm": 0.04647343697930622,
+      "learning_rate": 1.8133627270258136e-07,
+      "loss": 0.37,
+      "step": 3672
+    },
+    {
+      "epoch": 4.864900662251656,
+      "grad_norm": 0.06329082135044949,
+      "learning_rate": 1.778348714351097e-07,
+      "loss": 0.3713,
+      "step": 3673
+    },
+    {
+      "epoch": 4.866225165562914,
+      "grad_norm": 0.04617358636388155,
+      "learning_rate": 1.7436752929288258e-07,
+      "loss": 0.3702,
+      "step": 3674
+    },
+    {
+      "epoch": 4.867549668874172,
+      "grad_norm": 0.04647372999899742,
+      "learning_rate": 1.7093424924144787e-07,
+      "loss": 0.3688,
+      "step": 3675
+    },
+    {
+      "epoch": 4.86887417218543,
+      "grad_norm": 0.04877961834105387,
+      "learning_rate": 1.675350342172255e-07,
+      "loss": 0.3687,
+      "step": 3676
+    },
+    {
+      "epoch": 4.870198675496689,
+      "grad_norm": 0.04668808676999443,
+      "learning_rate": 1.6416988712750327e-07,
+      "loss": 0.3703,
+      "step": 3677
+    },
+    {
+      "epoch": 4.871523178807947,
+      "grad_norm": 0.04751412802930304,
+      "learning_rate": 1.6083881085041886e-07,
+      "loss": 0.3668,
+      "step": 3678
+    },
+    {
+      "epoch": 4.872847682119206,
+      "grad_norm": 0.04877891259655651,
+      "learning_rate": 1.5754180823497778e-07,
+      "loss": 0.3698,
+      "step": 3679
+    },
+    {
+      "epoch": 4.874172185430464,
+      "grad_norm": 0.046503605467622694,
+      "learning_rate": 1.5427888210104436e-07,
+      "loss": 0.3687,
+      "step": 3680
+    },
+    {
+      "epoch": 4.8754966887417215,
+      "grad_norm": 0.04851355586911947,
+      "learning_rate": 1.5105003523933738e-07,
+      "loss": 0.3681,
+      "step": 3681
+    },
+    {
+      "epoch": 4.87682119205298,
+      "grad_norm": 0.048461091640610646,
+      "learning_rate": 1.4785527041142555e-07,
+      "loss": 0.3726,
+      "step": 3682
+    },
+    {
+      "epoch": 4.878145695364238,
+      "grad_norm": 0.04893503701782518,
+      "learning_rate": 1.4469459034972766e-07,
+      "loss": 0.3728,
+      "step": 3683
+    },
+    {
+      "epoch": 4.879470198675497,
+      "grad_norm": 0.04780306792952429,
+      "learning_rate": 1.4156799775751684e-07,
+      "loss": 0.3746,
+      "step": 3684
+    },
+    {
+      "epoch": 4.880794701986755,
+      "grad_norm": 0.04726873498015125,
+      "learning_rate": 1.3847549530889849e-07,
+      "loss": 0.3707,
+      "step": 3685
+    },
+    {
+      "epoch": 4.882119205298014,
+      "grad_norm": 0.04573488900669311,
+      "learning_rate": 1.3541708564884126e-07,
+      "loss": 0.373,
+      "step": 3686
+    },
+    {
+      "epoch": 4.8834437086092715,
+      "grad_norm": 0.04719413910931922,
+      "learning_rate": 1.3239277139313277e-07,
+      "loss": 0.367,
+      "step": 3687
+    },
+    {
+      "epoch": 4.88476821192053,
+      "grad_norm": 0.04812825338324902,
+      "learning_rate": 1.2940255512841947e-07,
+      "loss": 0.3683,
+      "step": 3688
+    },
+    {
+      "epoch": 4.886092715231788,
+      "grad_norm": 0.047501746294703766,
+      "learning_rate": 1.2644643941216673e-07,
+      "loss": 0.3713,
+      "step": 3689
+    },
+    {
+      "epoch": 4.887417218543046,
+      "grad_norm": 0.04676100003938261,
+      "learning_rate": 1.2352442677268984e-07,
+      "loss": 0.3706,
+      "step": 3690
+    },
+    {
+      "epoch": 4.888741721854305,
+      "grad_norm": 0.04754846508183037,
+      "learning_rate": 1.2063651970912304e-07,
+      "loss": 0.3716,
+      "step": 3691
+    },
+    {
+      "epoch": 4.890066225165563,
+      "grad_norm": 0.048999085586982814,
+      "learning_rate": 1.1778272069144614e-07,
+      "loss": 0.3704,
+      "step": 3692
+    },
+    {
+      "epoch": 4.891390728476821,
+      "grad_norm": 0.047197024511427606,
+      "learning_rate": 1.149630321604489e-07,
+      "loss": 0.3727,
+      "step": 3693
+    },
+    {
+      "epoch": 4.892715231788079,
+      "grad_norm": 0.048739887127799156,
+      "learning_rate": 1.1217745652776668e-07,
+      "loss": 0.3677,
+      "step": 3694
+    },
+    {
+      "epoch": 4.894039735099338,
+      "grad_norm": 0.048160368620255675,
+      "learning_rate": 1.0942599617583594e-07,
+      "loss": 0.3738,
+      "step": 3695
+    },
+    {
+      "epoch": 4.895364238410596,
+      "grad_norm": 0.046962334850135685,
+      "learning_rate": 1.0670865345793425e-07,
+      "loss": 0.3722,
+      "step": 3696
+    },
+    {
+      "epoch": 4.896688741721855,
+      "grad_norm": 0.04662992435567471,
+      "learning_rate": 1.0402543069814919e-07,
+      "loss": 0.3695,
+      "step": 3697
+    },
+    {
+      "epoch": 4.898013245033113,
+      "grad_norm": 0.048445232073126156,
+      "learning_rate": 1.0137633019139615e-07,
+      "loss": 0.3743,
+      "step": 3698
+    },
+    {
+      "epoch": 4.8993377483443705,
+      "grad_norm": 0.04722856601068438,
+      "learning_rate": 9.876135420339161e-08,
+      "loss": 0.3758,
+      "step": 3699
+    },
+    {
+      "epoch": 4.900662251655629,
+      "grad_norm": 0.04776468130262501,
+      "learning_rate": 9.618050497067544e-08,
+      "loss": 0.3705,
+      "step": 3700
+    },
+    {
+      "epoch": 4.901986754966887,
+      "grad_norm": 0.046248246400698925,
+      "learning_rate": 9.363378470060192e-08,
+      "loss": 0.3713,
+      "step": 3701
+    },
+    {
+      "epoch": 4.903311258278146,
+      "grad_norm": 0.047046502398667644,
+      "learning_rate": 9.112119557132648e-08,
+      "loss": 0.3731,
+      "step": 3702
+    },
+    {
+      "epoch": 4.904635761589404,
+      "grad_norm": 0.046254677228170095,
+      "learning_rate": 8.864273973181902e-08,
+      "loss": 0.3698,
+      "step": 3703
+    },
+    {
+      "epoch": 4.905960264900663,
+      "grad_norm": 0.04722018901980159,
+      "learning_rate": 8.619841930185946e-08,
+      "loss": 0.3743,
+      "step": 3704
+    },
+    {
+      "epoch": 4.9072847682119205,
+      "grad_norm": 0.0472606939699218,
+      "learning_rate": 8.378823637201993e-08,
+      "loss": 0.3736,
+      "step": 3705
+    },
+    {
+      "epoch": 4.908609271523179,
+      "grad_norm": 0.04655513358264326,
+      "learning_rate": 8.141219300369152e-08,
+      "loss": 0.3754,
+      "step": 3706
+    },
+    {
+      "epoch": 4.909933774834437,
+      "grad_norm": 0.047059255175671354,
+      "learning_rate": 7.907029122904863e-08,
+      "loss": 0.3719,
+      "step": 3707
+    },
+    {
+      "epoch": 4.911258278145695,
+      "grad_norm": 0.047765702019858795,
+      "learning_rate": 7.676253305108461e-08,
+      "loss": 0.371,
+      "step": 3708
+    },
+    {
+      "epoch": 4.912582781456954,
+      "grad_norm": 0.048253473895017704,
+      "learning_rate": 7.44889204435717e-08,
+      "loss": 0.3719,
+      "step": 3709
+    },
+    {
+      "epoch": 4.913907284768212,
+      "grad_norm": 0.047204830226535634,
+      "learning_rate": 7.224945535108774e-08,
+      "loss": 0.3706,
+      "step": 3710
+    },
+    {
+      "epoch": 4.9152317880794705,
+      "grad_norm": 0.0471398908591822,
+      "learning_rate": 7.004413968901169e-08,
+      "loss": 0.3707,
+      "step": 3711
+    },
+    {
+      "epoch": 4.916556291390728,
+      "grad_norm": 0.04867578885729391,
+      "learning_rate": 6.787297534349701e-08,
+      "loss": 0.3718,
+      "step": 3712
+    },
+    {
+      "epoch": 4.917880794701987,
+      "grad_norm": 0.05045239100368609,
+      "learning_rate": 6.573596417150275e-08,
+      "loss": 0.3691,
+      "step": 3713
+    },
+    {
+      "epoch": 4.919205298013245,
+      "grad_norm": 0.04641087130555902,
+      "learning_rate": 6.363310800077572e-08,
+      "loss": 0.3658,
+      "step": 3714
+    },
+    {
+      "epoch": 4.920529801324504,
+      "grad_norm": 0.04534015619071944,
+      "learning_rate": 6.156440862984613e-08,
+      "loss": 0.3686,
+      "step": 3715
+    },
+    {
+      "epoch": 4.921854304635762,
+      "grad_norm": 0.04621955163619716,
+      "learning_rate": 5.952986782803205e-08,
+      "loss": 0.3694,
+      "step": 3716
+    },
+    {
+      "epoch": 4.92317880794702,
+      "grad_norm": 0.04747756046607046,
+      "learning_rate": 5.752948733543484e-08,
+      "loss": 0.3705,
+      "step": 3717
+    },
+    {
+      "epoch": 4.924503311258278,
+      "grad_norm": 0.04638090392029211,
+      "learning_rate": 5.556326886294372e-08,
+      "loss": 0.3716,
+      "step": 3718
+    },
+    {
+      "epoch": 4.925827814569536,
+      "grad_norm": 0.04532709876844367,
+      "learning_rate": 5.363121409222238e-08,
+      "loss": 0.3743,
+      "step": 3719
+    },
+    {
+      "epoch": 4.927152317880795,
+      "grad_norm": 0.046382008797592245,
+      "learning_rate": 5.173332467573122e-08,
+      "loss": 0.3707,
+      "step": 3720
+    },
+    {
+      "epoch": 4.928476821192053,
+      "grad_norm": 0.04765622830830869,
+      "learning_rate": 4.986960223668291e-08,
+      "loss": 0.3713,
+      "step": 3721
+    },
+    {
+      "epoch": 4.929801324503311,
+      "grad_norm": 0.04765563104282191,
+      "learning_rate": 4.8040048369095706e-08,
+      "loss": 0.374,
+      "step": 3722
+    },
+    {
+      "epoch": 4.9311258278145695,
+      "grad_norm": 0.047042569570849825,
+      "learning_rate": 4.624466463774901e-08,
+      "loss": 0.3701,
+      "step": 3723
+    },
+    {
+      "epoch": 4.932450331125828,
+      "grad_norm": 0.04805216311411699,
+      "learning_rate": 4.448345257819231e-08,
+      "loss": 0.3718,
+      "step": 3724
+    },
+    {
+      "epoch": 4.933774834437086,
+      "grad_norm": 0.047881250518282036,
+      "learning_rate": 4.275641369676287e-08,
+      "loss": 0.3705,
+      "step": 3725
+    },
+    {
+      "epoch": 4.935099337748344,
+      "grad_norm": 0.04626329267094269,
+      "learning_rate": 4.106354947056357e-08,
+      "loss": 0.3709,
+      "step": 3726
+    },
+    {
+      "epoch": 4.936423841059603,
+      "grad_norm": 0.047493561407992324,
+      "learning_rate": 3.940486134746735e-08,
+      "loss": 0.3704,
+      "step": 3727
+    },
+    {
+      "epoch": 4.937748344370861,
+      "grad_norm": 0.04657634331509248,
+      "learning_rate": 3.778035074611719e-08,
+      "loss": 0.3645,
+      "step": 3728
+    },
+    {
+      "epoch": 4.9390728476821195,
+      "grad_norm": 0.04879897810017914,
+      "learning_rate": 3.6190019055921676e-08,
+      "loss": 0.369,
+      "step": 3729
+    },
+    {
+      "epoch": 4.940397350993377,
+      "grad_norm": 0.04809085263492152,
+      "learning_rate": 3.463386763706833e-08,
+      "loss": 0.373,
+      "step": 3730
+    },
+    {
+      "epoch": 4.941721854304635,
+      "grad_norm": 0.045819498742298216,
+      "learning_rate": 3.311189782049695e-08,
+      "loss": 0.3708,
+      "step": 3731
+    },
+    {
+      "epoch": 4.943046357615894,
+      "grad_norm": 0.04601576703370423,
+      "learning_rate": 3.162411090792627e-08,
+      "loss": 0.374,
+      "step": 3732
+    },
+    {
+      "epoch": 4.944370860927152,
+      "grad_norm": 0.04802058004030118,
+      "learning_rate": 3.017050817182288e-08,
+      "loss": 0.3692,
+      "step": 3733
+    },
+    {
+      "epoch": 4.945695364238411,
+      "grad_norm": 0.048108120736139195,
+      "learning_rate": 2.875109085542782e-08,
+      "loss": 0.3687,
+      "step": 3734
+    },
+    {
+      "epoch": 4.947019867549669,
+      "grad_norm": 0.04649102216518732,
+      "learning_rate": 2.7365860172747784e-08,
+      "loss": 0.3745,
+      "step": 3735
+    },
+    {
+      "epoch": 4.948344370860927,
+      "grad_norm": 0.04629617261410206,
+      "learning_rate": 2.6014817308532836e-08,
+      "loss": 0.3699,
+      "step": 3736
+    },
+    {
+      "epoch": 4.949668874172185,
+      "grad_norm": 0.0470715428986015,
+      "learning_rate": 2.4697963418311988e-08,
+      "loss": 0.3738,
+      "step": 3737
+    },
+    {
+      "epoch": 4.950993377483444,
+      "grad_norm": 0.046338919206578434,
+      "learning_rate": 2.3415299628357645e-08,
+      "loss": 0.3687,
+      "step": 3738
+    },
+    {
+      "epoch": 4.952317880794702,
+      "grad_norm": 0.0462237870583418,
+      "learning_rate": 2.216682703571671e-08,
+      "loss": 0.3691,
+      "step": 3739
+    },
+    {
+      "epoch": 4.95364238410596,
+      "grad_norm": 0.04676070455409389,
+      "learning_rate": 2.0952546708175038e-08,
+      "loss": 0.3723,
+      "step": 3740
+    },
+    {
+      "epoch": 4.954966887417219,
+      "grad_norm": 0.04659436830061062,
+      "learning_rate": 1.9772459684288537e-08,
+      "loss": 0.3709,
+      "step": 3741
+    },
+    {
+      "epoch": 4.9562913907284765,
+      "grad_norm": 0.04664386270322658,
+      "learning_rate": 1.862656697335652e-08,
+      "loss": 0.3733,
+      "step": 3742
+    },
+    {
+      "epoch": 4.957615894039735,
+      "grad_norm": 0.04622077432849068,
+      "learning_rate": 1.7514869555448343e-08,
+      "loss": 0.3739,
+      "step": 3743
+    },
+    {
+      "epoch": 4.958940397350993,
+      "grad_norm": 0.046167352208646,
+      "learning_rate": 1.643736838136789e-08,
+      "loss": 0.3699,
+      "step": 3744
+    },
+    {
+      "epoch": 4.960264900662252,
+      "grad_norm": 0.046641788101578506,
+      "learning_rate": 1.5394064372684646e-08,
+      "loss": 0.3721,
+      "step": 3745
+    },
+    {
+      "epoch": 4.96158940397351,
+      "grad_norm": 0.04862333930991028,
+      "learning_rate": 1.4384958421715944e-08,
+      "loss": 0.3714,
+      "step": 3746
+    },
+    {
+      "epoch": 4.9629139072847686,
+      "grad_norm": 0.047236812508370286,
+      "learning_rate": 1.3410051391535838e-08,
+      "loss": 0.3736,
+      "step": 3747
+    },
+    {
+      "epoch": 4.964238410596026,
+      "grad_norm": 0.04733671272109708,
+      "learning_rate": 1.2469344115952909e-08,
+      "loss": 0.3698,
+      "step": 3748
+    },
+    {
+      "epoch": 4.965562913907284,
+      "grad_norm": 0.0455616328591315,
+      "learning_rate": 1.1562837399545779e-08,
+      "loss": 0.3665,
+      "step": 3749
+    },
+    {
+      "epoch": 4.966887417218543,
+      "grad_norm": 0.047537379699613726,
+      "learning_rate": 1.0690532017627597e-08,
+      "loss": 0.3688,
+      "step": 3750
+    },
+    {
+      "epoch": 4.968211920529801,
+      "grad_norm": 0.045830111254534636,
+      "learning_rate": 9.852428716263796e-09,
+      "loss": 0.3718,
+      "step": 3751
+    },
+    {
+      "epoch": 4.96953642384106,
+      "grad_norm": 0.046335293232699,
+      "learning_rate": 9.048528212267648e-09,
+      "loss": 0.3719,
+      "step": 3752
+    },
+    {
+      "epoch": 4.970860927152318,
+      "grad_norm": 0.047021122232693745,
+      "learning_rate": 8.278831193200276e-09,
+      "loss": 0.3688,
+      "step": 3753
+    },
+    {
+      "epoch": 4.972185430463576,
+      "grad_norm": 0.046409807435437614,
+      "learning_rate": 7.5433383173662e-09,
+      "loss": 0.3677,
+      "step": 3754
+    },
+    {
+      "epoch": 4.973509933774834,
+      "grad_norm": 0.04698514974700297,
+      "learning_rate": 6.842050213826668e-09,
+      "loss": 0.3719,
+      "step": 3755
+    },
+    {
+      "epoch": 4.974834437086093,
+      "grad_norm": 0.04654283997096874,
+      "learning_rate": 6.174967482368566e-09,
+      "loss": 0.3708,
+      "step": 3756
+    },
+    {
+      "epoch": 4.976158940397351,
+      "grad_norm": 0.047080159282755656,
+      "learning_rate": 5.542090693544389e-09,
+      "loss": 0.3674,
+      "step": 3757
+    },
+    {
+      "epoch": 4.977483443708609,
+      "grad_norm": 0.046935022583489173,
+      "learning_rate": 4.9434203886367106e-09,
+      "loss": 0.3732,
+      "step": 3758
+    },
+    {
+      "epoch": 4.978807947019868,
+      "grad_norm": 0.046399573079073386,
+      "learning_rate": 4.378957079675949e-09,
+      "loss": 0.3693,
+      "step": 3759
+    },
+    {
+      "epoch": 4.9801324503311255,
+      "grad_norm": 0.04785486937301826,
+      "learning_rate": 3.848701249435927e-09,
+      "loss": 0.3708,
+      "step": 3760
+    },
+    {
+      "epoch": 4.981456953642384,
+      "grad_norm": 0.046058463643434404,
+      "learning_rate": 3.3526533514383064e-09,
+      "loss": 0.3709,
+      "step": 3761
+    },
+    {
+      "epoch": 4.982781456953642,
+      "grad_norm": 0.04653463530369661,
+      "learning_rate": 2.8908138099437154e-09,
+      "loss": 0.3744,
+      "step": 3762
+    },
+    {
+      "epoch": 4.984105960264901,
+      "grad_norm": 0.047243481723143285,
+      "learning_rate": 2.4631830199473015e-09,
+      "loss": 0.3763,
+      "step": 3763
+    },
+    {
+      "epoch": 4.985430463576159,
+      "grad_norm": 0.047377690111723755,
+      "learning_rate": 2.069761347200938e-09,
+      "loss": 0.3748,
+      "step": 3764
+    },
+    {
+      "epoch": 4.986754966887418,
+      "grad_norm": 0.04601724418987812,
+      "learning_rate": 1.7105491281865783e-09,
+      "loss": 0.3728,
+      "step": 3765
+    },
+    {
+      "epoch": 4.9880794701986755,
+      "grad_norm": 0.04909167014669262,
+      "learning_rate": 1.3855466701295784e-09,
+      "loss": 0.3724,
+      "step": 3766
+    },
+    {
+      "epoch": 4.989403973509933,
+      "grad_norm": 0.046360217073978284,
+      "learning_rate": 1.0947542510031383e-09,
+      "loss": 0.3707,
+      "step": 3767
+    },
+    {
+      "epoch": 4.990728476821192,
+      "grad_norm": 0.049199741075449555,
+      "learning_rate": 8.381721195149795e-10,
+      "loss": 0.3732,
+      "step": 3768
+    },
+    {
+      "epoch": 4.99205298013245,
+      "grad_norm": 0.046372263486766284,
+      "learning_rate": 6.158004951162255e-10,
+      "loss": 0.3752,
+      "step": 3769
+    },
+    {
+      "epoch": 4.993377483443709,
+      "grad_norm": 0.04902918919614194,
+      "learning_rate": 4.276395679925216e-10,
+      "loss": 0.3683,
+      "step": 3770
+    },
+    {
+      "epoch": 4.994701986754967,
+      "grad_norm": 0.04625760405818134,
+      "learning_rate": 2.7368949907735645e-10,
+      "loss": 0.3699,
+      "step": 3771
+    },
+    {
+      "epoch": 4.9960264900662255,
+      "grad_norm": 0.04559588647259882,
+      "learning_rate": 1.5395042003873985e-10,
+      "loss": 0.369,
+      "step": 3772
+    },
+    {
+      "epoch": 4.997350993377483,
+      "grad_norm": 0.04717160748705134,
+      "learning_rate": 6.842243328808452e-11,
+      "loss": 0.3741,
+      "step": 3773
+    },
+    {
+      "epoch": 4.998675496688742,
+      "grad_norm": 0.047753159792208316,
+      "learning_rate": 1.7105611980205994e-11,
+      "loss": 0.3708,
+      "step": 3774
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.046848406976358725,
+      "learning_rate": 0.0,
+      "loss": 0.369,
+      "step": 3775
+    },
+    {
+      "epoch": 5.0,
+      "step": 3775,
+      "total_flos": 2.126562862301184e+17,
+      "train_loss": 0.0742928122447816,
+      "train_runtime": 34892.4018,
+      "train_samples_per_second": 55.384,
+      "train_steps_per_second": 0.108
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3775,
+  "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": 2.126562862301184e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}