diff --git "a/checkpoint-76160/trainer_state.json" "b/checkpoint-76160/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-76160/trainer_state.json"
@@ -0,0 +1,10695 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 76160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.006565126050420168,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 7.9703,
+      "step": 50
+    },
+    {
+      "epoch": 0.013130252100840336,
+      "grad_norm": 4.0,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 5.4946,
+      "step": 100
+    },
+    {
+      "epoch": 0.019695378151260504,
+      "grad_norm": 4.71875,
+      "learning_rate": 5.96e-05,
+      "loss": 4.3864,
+      "step": 150
+    },
+    {
+      "epoch": 0.026260504201680673,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.960000000000001e-05,
+      "loss": 4.1838,
+      "step": 200
+    },
+    {
+      "epoch": 0.03282563025210084,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.960000000000001e-05,
+      "loss": 4.0394,
+      "step": 250
+    },
+    {
+      "epoch": 0.03939075630252101,
+      "grad_norm": 3.9375,
+      "learning_rate": 0.00011960000000000001,
+      "loss": 3.9963,
+      "step": 300
+    },
+    {
+      "epoch": 0.04595588235294118,
+      "grad_norm": 3.1875,
+      "learning_rate": 0.0001396,
+      "loss": 3.9232,
+      "step": 350
+    },
+    {
+      "epoch": 0.052521008403361345,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001596,
+      "loss": 3.8485,
+      "step": 400
+    },
+    {
+      "epoch": 0.05908613445378151,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001796,
+      "loss": 3.8405,
+      "step": 450
+    },
+    {
+      "epoch": 0.06565126050420168,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001996,
+      "loss": 3.8045,
+      "step": 500
+    },
+    {
+      "epoch": 0.07221638655462184,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00019987047316944226,
+      "loss": 3.7991,
+      "step": 550
+    },
+    {
+      "epoch": 0.07878151260504201,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00019973830293417924,
+      "loss": 3.7579,
+      "step": 600
+    },
+    {
+      "epoch": 0.08534663865546219,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001996061326989162,
+      "loss": 3.7281,
+      "step": 650
+    },
+    {
+      "epoch": 0.09191176470588236,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001994739624636532,
+      "loss": 3.7043,
+      "step": 700
+    },
+    {
+      "epoch": 0.09847689075630252,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.0001993417922283902,
+      "loss": 3.7007,
+      "step": 750
+    },
+    {
+      "epoch": 0.10504201680672269,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00019920962199312717,
+      "loss": 3.6837,
+      "step": 800
+    },
+    {
+      "epoch": 0.11160714285714286,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00019907745175786414,
+      "loss": 3.6388,
+      "step": 850
+    },
+    {
+      "epoch": 0.11817226890756302,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00019894528152260112,
+      "loss": 3.6421,
+      "step": 900
+    },
+    {
+      "epoch": 0.1247373949579832,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00019881311128733812,
+      "loss": 3.6235,
+      "step": 950
+    },
+    {
+      "epoch": 0.13130252100840337,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.0001986809410520751,
+      "loss": 3.6121,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13786764705882354,
+      "grad_norm": 2.984375,
+      "learning_rate": 0.00019854877081681207,
+      "loss": 3.5556,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14443277310924368,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00019841660058154904,
+      "loss": 3.5299,
+      "step": 1100
+    },
+    {
+      "epoch": 0.15099789915966386,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00019828443034628602,
+      "loss": 3.4619,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15756302521008403,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00019815226011102302,
+      "loss": 3.4076,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1641281512605042,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00019802008987576,
+      "loss": 3.3649,
+      "step": 1250
+    },
+    {
+      "epoch": 0.17069327731092437,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00019788791964049697,
+      "loss": 3.3748,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17725840336134455,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00019775574940523395,
+      "loss": 3.3507,
+      "step": 1350
+    },
+    {
+      "epoch": 0.18382352941176472,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00019762357916997095,
+      "loss": 3.3366,
+      "step": 1400
+    },
+    {
+      "epoch": 0.19038865546218486,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00019749140893470792,
+      "loss": 3.3007,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19695378151260504,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.0001973592386994449,
+      "loss": 3.2944,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2035189075630252,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.00019722706846418187,
+      "loss": 3.2694,
+      "step": 1550
+    },
+    {
+      "epoch": 0.21008403361344538,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00019709489822891888,
+      "loss": 3.3025,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21664915966386555,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019696272799365585,
+      "loss": 3.278,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22321428571428573,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001968305577583928,
+      "loss": 3.2147,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22977941176470587,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001966983875231298,
+      "loss": 3.2072,
+      "step": 1750
+    },
+    {
+      "epoch": 0.23634453781512604,
+      "grad_norm": 2.75,
+      "learning_rate": 0.00019656621728786678,
+      "loss": 3.2304,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24290966386554622,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019643404705260375,
+      "loss": 3.2481,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2494747899159664,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00019630187681734073,
+      "loss": 3.194,
+      "step": 1900
+    },
+    {
+      "epoch": 0.25603991596638653,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001961697065820777,
+      "loss": 3.2233,
+      "step": 1950
+    },
+    {
+      "epoch": 0.26260504201680673,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001960375363468147,
+      "loss": 3.2257,
+      "step": 2000
+    },
+    {
+      "epoch": 0.2691701680672269,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00019590536611155168,
+      "loss": 3.1964,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2757352941176471,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00019577319587628866,
+      "loss": 3.1723,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2823004201680672,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00019564102564102563,
+      "loss": 3.1696,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28886554621848737,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019550885540576263,
+      "loss": 3.1652,
+      "step": 2200
+    },
+    {
+      "epoch": 0.29543067226890757,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001953766851704996,
+      "loss": 3.1808,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3019957983193277,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019524451493523658,
+      "loss": 3.1573,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3085609243697479,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00019511234469997356,
+      "loss": 3.1409,
+      "step": 2350
+    },
+    {
+      "epoch": 0.31512605042016806,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019498017446471056,
+      "loss": 3.1447,
+      "step": 2400
+    },
+    {
+      "epoch": 0.32169117647058826,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00019484800422944754,
+      "loss": 3.1546,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3282563025210084,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.0001947158339941845,
+      "loss": 3.1002,
+      "step": 2500
+    },
+    {
+      "epoch": 0.33482142857142855,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001945836637589215,
+      "loss": 3.1101,
+      "step": 2550
+    },
+    {
+      "epoch": 0.34138655462184875,
+      "grad_norm": 1.625,
+      "learning_rate": 0.0001944514935236585,
+      "loss": 3.1415,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3479516806722689,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00019431932328839546,
+      "loss": 3.1367,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3545168067226891,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00019418715305313244,
+      "loss": 3.1297,
+      "step": 2700
+    },
+    {
+      "epoch": 0.36108193277310924,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00019405498281786942,
+      "loss": 3.1024,
+      "step": 2750
+    },
+    {
+      "epoch": 0.36764705882352944,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00019392281258260642,
+      "loss": 3.0998,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3742121848739496,
+      "grad_norm": 1.578125,
+      "learning_rate": 0.0001937906423473434,
+      "loss": 3.1066,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3807773109243697,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00019365847211208037,
+      "loss": 3.1117,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3873424369747899,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00019352630187681734,
+      "loss": 3.0898,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3939075630252101,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00019339413164155435,
+      "loss": 3.1092,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4004726890756303,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00019326196140629132,
+      "loss": 3.0731,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4070378151260504,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001931297911710283,
+      "loss": 3.0875,
+      "step": 3100
+    },
+    {
+      "epoch": 0.41360294117647056,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00019299762093576527,
+      "loss": 3.0708,
+      "step": 3150
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00019286545070050225,
+      "loss": 3.0658,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4267331932773109,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019273328046523925,
+      "loss": 3.0819,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4332983193277311,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00019260111022997622,
+      "loss": 3.0502,
+      "step": 3300
+    },
+    {
+      "epoch": 0.43986344537815125,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.0001924689399947132,
+      "loss": 3.0808,
+      "step": 3350
+    },
+    {
+      "epoch": 0.44642857142857145,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00019233676975945017,
+      "loss": 3.0657,
+      "step": 3400
+    },
+    {
+      "epoch": 0.4529936974789916,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00019220459952418718,
+      "loss": 3.0879,
+      "step": 3450
+    },
+    {
+      "epoch": 0.45955882352941174,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00019207242928892415,
+      "loss": 3.0342,
+      "step": 3500
+    },
+    {
+      "epoch": 0.46612394957983194,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00019194025905366113,
+      "loss": 3.0596,
+      "step": 3550
+    },
+    {
+      "epoch": 0.4726890756302521,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.0001918080888183981,
+      "loss": 3.0289,
+      "step": 3600
+    },
+    {
+      "epoch": 0.4792542016806723,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001916759185831351,
+      "loss": 3.025,
+      "step": 3650
+    },
+    {
+      "epoch": 0.48581932773109243,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00019154374834787208,
+      "loss": 3.0299,
+      "step": 3700
+    },
+    {
+      "epoch": 0.49238445378151263,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00019141157811260905,
+      "loss": 3.0436,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4989495798319328,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00019127940787734603,
+      "loss": 3.0206,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5055147058823529,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00019114723764208303,
+      "loss": 3.0454,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5120798319327731,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00019101506740682,
+      "loss": 3.0163,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5186449579831933,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00019088289717155698,
+      "loss": 3.0114,
+      "step": 3950
+    },
+    {
+      "epoch": 0.5252100840336135,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00019075072693629396,
+      "loss": 3.0274,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5317752100840336,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00019061855670103093,
+      "loss": 3.0153,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5383403361344538,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001904863864657679,
+      "loss": 3.0309,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5449054621848739,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019035421623050488,
+      "loss": 3.0383,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5514705882352942,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00019022204599524186,
+      "loss": 2.9907,
+      "step": 4200
+    },
+    {
+      "epoch": 0.5580357142857143,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00019008987575997886,
+      "loss": 3.0058,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5646008403361344,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018995770552471584,
+      "loss": 3.0301,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5711659663865546,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001898255352894528,
+      "loss": 3.0068,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5777310924369747,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00018969336505418979,
+      "loss": 2.9968,
+      "step": 4400
+    },
+    {
+      "epoch": 0.584296218487395,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.0001895611948189268,
+      "loss": 3.0087,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5908613445378151,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00018942902458366376,
+      "loss": 3.016,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5974264705882353,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00018929685434840074,
+      "loss": 2.9857,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6039915966386554,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00018916468411313771,
+      "loss": 2.9912,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6105567226890757,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00018903251387787472,
+      "loss": 2.9913,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6171218487394958,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001889003436426117,
+      "loss": 2.9865,
+      "step": 4700
+    },
+    {
+      "epoch": 0.623686974789916,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00018876817340734867,
+      "loss": 2.9865,
+      "step": 4750
+    },
+    {
+      "epoch": 0.6302521008403361,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018863600317208564,
+      "loss": 2.9858,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6368172268907563,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018850383293682264,
+      "loss": 2.97,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6433823529411765,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018837166270155962,
+      "loss": 2.9701,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6499474789915967,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001882394924662966,
+      "loss": 2.9929,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6565126050420168,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00018810732223103357,
+      "loss": 2.9927,
+      "step": 5000
+    },
+    {
+      "epoch": 0.663077731092437,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00018797515199577057,
+      "loss": 2.9872,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6696428571428571,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00018784298176050755,
+      "loss": 2.9825,
+      "step": 5100
+    },
+    {
+      "epoch": 0.6762079831932774,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018771081152524452,
+      "loss": 2.9643,
+      "step": 5150
+    },
+    {
+      "epoch": 0.6827731092436975,
+      "grad_norm": 0.98046875,
+      "learning_rate": 0.0001875786412899815,
+      "loss": 2.9671,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6893382352941176,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018744647105471847,
+      "loss": 2.9624,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6959033613445378,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00018731430081945548,
+      "loss": 2.9481,
+      "step": 5300
+    },
+    {
+      "epoch": 0.7024684873949579,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018718213058419245,
+      "loss": 2.9712,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7090336134453782,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018704996034892943,
+      "loss": 2.9757,
+      "step": 5400
+    },
+    {
+      "epoch": 0.7155987394957983,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001869177901136664,
+      "loss": 2.9722,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7221638655462185,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001867856198784034,
+      "loss": 2.9506,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7287289915966386,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00018665344964314038,
+      "loss": 2.9628,
+      "step": 5550
+    },
+    {
+      "epoch": 0.7352941176470589,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018652127940787735,
+      "loss": 2.9618,
+      "step": 5600
+    },
+    {
+      "epoch": 0.741859243697479,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00018638910917261433,
+      "loss": 2.9802,
+      "step": 5650
+    },
+    {
+      "epoch": 0.7484243697478992,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018625693893735133,
+      "loss": 2.9556,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7549894957983193,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.0001861247687020883,
+      "loss": 2.9388,
+      "step": 5750
+    },
+    {
+      "epoch": 0.7615546218487395,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00018599259846682528,
+      "loss": 2.9328,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7681197478991597,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00018586042823156226,
+      "loss": 2.937,
+      "step": 5850
+    },
+    {
+      "epoch": 0.7746848739495799,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00018572825799629926,
+      "loss": 2.9736,
+      "step": 5900
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00018559608776103623,
+      "loss": 2.9638,
+      "step": 5950
+    },
+    {
+      "epoch": 0.7878151260504201,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001854639175257732,
+      "loss": 2.9147,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7943802521008403,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00018533174729051018,
+      "loss": 2.9329,
+      "step": 6050
+    },
+    {
+      "epoch": 0.8009453781512605,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001851995770552472,
+      "loss": 2.923,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8075105042016807,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00018506740681998416,
+      "loss": 2.9503,
+      "step": 6150
+    },
+    {
+      "epoch": 0.8140756302521008,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00018493523658472114,
+      "loss": 2.9705,
+      "step": 6200
+    },
+    {
+      "epoch": 0.820640756302521,
+      "grad_norm": 0.96484375,
+      "learning_rate": 0.0001848030663494581,
+      "loss": 2.9244,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8272058823529411,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00018467089611419511,
+      "loss": 2.9116,
+      "step": 6300
+    },
+    {
+      "epoch": 0.8337710084033614,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001845387258789321,
+      "loss": 2.9746,
+      "step": 6350
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00018440655564366907,
+      "loss": 2.9773,
+      "step": 6400
+    },
+    {
+      "epoch": 0.8469012605042017,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.000184274385408406,
+      "loss": 2.9174,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8534663865546218,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00018414221517314302,
+      "loss": 2.9405,
+      "step": 6500
+    },
+    {
+      "epoch": 0.8600315126050421,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018401004493788,
+      "loss": 2.9203,
+      "step": 6550
+    },
+    {
+      "epoch": 0.8665966386554622,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018387787470261697,
+      "loss": 2.9295,
+      "step": 6600
+    },
+    {
+      "epoch": 0.8731617647058824,
+      "grad_norm": 1.265625,
+      "learning_rate": 0.00018374570446735394,
+      "loss": 2.9213,
+      "step": 6650
+    },
+    {
+      "epoch": 0.8797268907563025,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00018361353423209094,
+      "loss": 2.9434,
+      "step": 6700
+    },
+    {
+      "epoch": 0.8862920168067226,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00018348136399682792,
+      "loss": 2.9609,
+      "step": 6750
+    },
+    {
+      "epoch": 0.8928571428571429,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.0001833491937615649,
+      "loss": 2.9187,
+      "step": 6800
+    },
+    {
+      "epoch": 0.899422268907563,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00018321702352630187,
+      "loss": 2.9306,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9059873949579832,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00018308485329103887,
+      "loss": 2.9435,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9125525210084033,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018295268305577585,
+      "loss": 2.9247,
+      "step": 6950
+    },
+    {
+      "epoch": 0.9191176470588235,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00018282051282051282,
+      "loss": 2.9573,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9256827731092437,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001826883425852498,
+      "loss": 2.9421,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9322478991596639,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.0001825561723499868,
+      "loss": 2.9193,
+      "step": 7100
+    },
+    {
+      "epoch": 0.938813025210084,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00018242400211472377,
+      "loss": 2.8935,
+      "step": 7150
+    },
+    {
+      "epoch": 0.9453781512605042,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00018229183187946075,
+      "loss": 2.9295,
+      "step": 7200
+    },
+    {
+      "epoch": 0.9519432773109243,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00018215966164419772,
+      "loss": 2.9219,
+      "step": 7250
+    },
+    {
+      "epoch": 0.9585084033613446,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00018202749140893473,
+      "loss": 2.8695,
+      "step": 7300
+    },
+    {
+      "epoch": 0.9650735294117647,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001818953211736717,
+      "loss": 2.9135,
+      "step": 7350
+    },
+    {
+      "epoch": 0.9716386554621849,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00018176315093840868,
+      "loss": 2.9415,
+      "step": 7400
+    },
+    {
+      "epoch": 0.978203781512605,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.00018163098070314565,
+      "loss": 2.926,
+      "step": 7450
+    },
+    {
+      "epoch": 0.9847689075630253,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00018149881046788263,
+      "loss": 2.8909,
+      "step": 7500
+    },
+    {
+      "epoch": 0.9913340336134454,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00018136664023261963,
+      "loss": 2.9088,
+      "step": 7550
+    },
+    {
+      "epoch": 0.9978991596638656,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001812344699973566,
+      "loss": 2.9116,
+      "step": 7600
+    },
+    {
+      "epoch": 1.0044642857142858,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00018110229976209358,
+      "loss": 2.7498,
+      "step": 7650
+    },
+    {
+      "epoch": 1.0110294117647058,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00018097012952683056,
+      "loss": 2.7164,
+      "step": 7700
+    },
+    {
+      "epoch": 1.017594537815126,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018083795929156756,
+      "loss": 2.6681,
+      "step": 7750
+    },
+    {
+      "epoch": 1.0241596638655461,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00018070578905630453,
+      "loss": 2.7374,
+      "step": 7800
+    },
+    {
+      "epoch": 1.0307247899159664,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001805736188210415,
+      "loss": 2.6908,
+      "step": 7850
+    },
+    {
+      "epoch": 1.0372899159663866,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.00018044144858577848,
+      "loss": 2.7301,
+      "step": 7900
+    },
+    {
+      "epoch": 1.0438550420168067,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00018030927835051549,
+      "loss": 2.7097,
+      "step": 7950
+    },
+    {
+      "epoch": 1.050420168067227,
+      "grad_norm": 1.484375,
+      "learning_rate": 0.00018017710811525246,
+      "loss": 2.7109,
+      "step": 8000
+    },
+    {
+      "epoch": 1.056985294117647,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00018004493787998944,
+      "loss": 2.7128,
+      "step": 8050
+    },
+    {
+      "epoch": 1.0635504201680672,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001799127676447264,
+      "loss": 2.6722,
+      "step": 8100
+    },
+    {
+      "epoch": 1.0701155462184875,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00017978059740946341,
+      "loss": 2.7261,
+      "step": 8150
+    },
+    {
+      "epoch": 1.0766806722689075,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001796484271742004,
+      "loss": 2.7065,
+      "step": 8200
+    },
+    {
+      "epoch": 1.0832457983193278,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.00017951625693893736,
+      "loss": 2.7079,
+      "step": 8250
+    },
+    {
+      "epoch": 1.0898109243697478,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017938408670367434,
+      "loss": 2.7226,
+      "step": 8300
+    },
+    {
+      "epoch": 1.096376050420168,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017925191646841134,
+      "loss": 2.7011,
+      "step": 8350
+    },
+    {
+      "epoch": 1.1029411764705883,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017911974623314832,
+      "loss": 2.7024,
+      "step": 8400
+    },
+    {
+      "epoch": 1.1095063025210083,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.0001789875759978853,
+      "loss": 2.7094,
+      "step": 8450
+    },
+    {
+      "epoch": 1.1160714285714286,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00017885540576262227,
+      "loss": 2.7323,
+      "step": 8500
+    },
+    {
+      "epoch": 1.1226365546218486,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00017872323552735924,
+      "loss": 2.7223,
+      "step": 8550
+    },
+    {
+      "epoch": 1.129201680672269,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.00017859106529209624,
+      "loss": 2.7014,
+      "step": 8600
+    },
+    {
+      "epoch": 1.1357668067226891,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00017845889505683322,
+      "loss": 2.7235,
+      "step": 8650
+    },
+    {
+      "epoch": 1.1423319327731092,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001783267248215702,
+      "loss": 2.7127,
+      "step": 8700
+    },
+    {
+      "epoch": 1.1488970588235294,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017819455458630717,
+      "loss": 2.7256,
+      "step": 8750
+    },
+    {
+      "epoch": 1.1554621848739495,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017806238435104415,
+      "loss": 2.7401,
+      "step": 8800
+    },
+    {
+      "epoch": 1.1620273109243697,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017793021411578112,
+      "loss": 2.6972,
+      "step": 8850
+    },
+    {
+      "epoch": 1.16859243697479,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.0001777980438805181,
+      "loss": 2.7009,
+      "step": 8900
+    },
+    {
+      "epoch": 1.17515756302521,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001776658736452551,
+      "loss": 2.7225,
+      "step": 8950
+    },
+    {
+      "epoch": 1.1817226890756303,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00017753370340999207,
+      "loss": 2.7215,
+      "step": 9000
+    },
+    {
+      "epoch": 1.1882878151260505,
+      "grad_norm": 0.9609375,
+      "learning_rate": 0.00017740153317472905,
+      "loss": 2.7292,
+      "step": 9050
+    },
+    {
+      "epoch": 1.1948529411764706,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00017726936293946602,
+      "loss": 2.7364,
+      "step": 9100
+    },
+    {
+      "epoch": 1.2014180672268908,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00017713719270420303,
+      "loss": 2.7467,
+      "step": 9150
+    },
+    {
+      "epoch": 1.2079831932773109,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00017700502246894,
+      "loss": 2.7084,
+      "step": 9200
+    },
+    {
+      "epoch": 1.214548319327731,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017687285223367698,
+      "loss": 2.7161,
+      "step": 9250
+    },
+    {
+      "epoch": 1.2211134453781511,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.00017674068199841395,
+      "loss": 2.7257,
+      "step": 9300
+    },
+    {
+      "epoch": 1.2276785714285714,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017660851176315095,
+      "loss": 2.7518,
+      "step": 9350
+    },
+    {
+      "epoch": 1.2342436974789917,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00017647634152788793,
+      "loss": 2.733,
+      "step": 9400
+    },
+    {
+      "epoch": 1.2408088235294117,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001763441712926249,
+      "loss": 2.7051,
+      "step": 9450
+    },
+    {
+      "epoch": 1.247373949579832,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00017621200105736188,
+      "loss": 2.6999,
+      "step": 9500
+    },
+    {
+      "epoch": 1.2539390756302522,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00017607983082209886,
+      "loss": 2.7401,
+      "step": 9550
+    },
+    {
+      "epoch": 1.2605042016806722,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.00017594766058683586,
+      "loss": 2.7076,
+      "step": 9600
+    },
+    {
+      "epoch": 1.2670693277310925,
+      "grad_norm": 1.421875,
+      "learning_rate": 0.00017581549035157283,
+      "loss": 2.7448,
+      "step": 9650
+    },
+    {
+      "epoch": 1.2736344537815127,
+      "grad_norm": 1.375,
+      "learning_rate": 0.0001756833201163098,
+      "loss": 2.7407,
+      "step": 9700
+    },
+    {
+      "epoch": 1.2801995798319328,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00017555114988104678,
+      "loss": 2.7003,
+      "step": 9750
+    },
+    {
+      "epoch": 1.2867647058823528,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017541897964578379,
+      "loss": 2.6974,
+      "step": 9800
+    },
+    {
+      "epoch": 1.293329831932773,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00017528680941052076,
+      "loss": 2.7261,
+      "step": 9850
+    },
+    {
+      "epoch": 1.2998949579831933,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00017515463917525774,
+      "loss": 2.7052,
+      "step": 9900
+    },
+    {
+      "epoch": 1.3064600840336134,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001750224689399947,
+      "loss": 2.6775,
+      "step": 9950
+    },
+    {
+      "epoch": 1.3130252100840336,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.0001748902987047317,
+      "loss": 2.7389,
+      "step": 10000
+    },
+    {
+      "epoch": 1.3195903361344539,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.0001747581284694687,
+      "loss": 2.6958,
+      "step": 10050
+    },
+    {
+      "epoch": 1.326155462184874,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00017462595823420566,
+      "loss": 2.7364,
+      "step": 10100
+    },
+    {
+      "epoch": 1.3327205882352942,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00017449378799894264,
+      "loss": 2.7042,
+      "step": 10150
+    },
+    {
+      "epoch": 1.3392857142857144,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00017436161776367964,
+      "loss": 2.7221,
+      "step": 10200
+    },
+    {
+      "epoch": 1.3458508403361344,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00017422944752841662,
+      "loss": 2.6988,
+      "step": 10250
+    },
+    {
+      "epoch": 1.3524159663865547,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.0001740972772931536,
+      "loss": 2.7224,
+      "step": 10300
+    },
+    {
+      "epoch": 1.3589810924369747,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00017396510705789057,
+      "loss": 2.7076,
+      "step": 10350
+    },
+    {
+      "epoch": 1.365546218487395,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00017383293682262757,
+      "loss": 2.7175,
+      "step": 10400
+    },
+    {
+      "epoch": 1.372111344537815,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017370076658736454,
+      "loss": 2.7272,
+      "step": 10450
+    },
+    {
+      "epoch": 1.3786764705882353,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00017356859635210152,
+      "loss": 2.7176,
+      "step": 10500
+    },
+    {
+      "epoch": 1.3852415966386555,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001734364261168385,
+      "loss": 2.7443,
+      "step": 10550
+    },
+    {
+      "epoch": 1.3918067226890756,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001733042558815755,
+      "loss": 2.712,
+      "step": 10600
+    },
+    {
+      "epoch": 1.3983718487394958,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00017317208564631247,
+      "loss": 2.7126,
+      "step": 10650
+    },
+    {
+      "epoch": 1.404936974789916,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00017303991541104945,
+      "loss": 2.7164,
+      "step": 10700
+    },
+    {
+      "epoch": 1.4115021008403361,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00017290774517578642,
+      "loss": 2.7406,
+      "step": 10750
+    },
+    {
+      "epoch": 1.4180672268907564,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.0001727755749405234,
+      "loss": 2.7049,
+      "step": 10800
+    },
+    {
+      "epoch": 1.4246323529411764,
+      "grad_norm": 0.875,
+      "learning_rate": 0.0001726434047052604,
+      "loss": 2.7081,
+      "step": 10850
+    },
+    {
+      "epoch": 1.4311974789915967,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00017251123446999738,
+      "loss": 2.7084,
+      "step": 10900
+    },
+    {
+      "epoch": 1.4377626050420167,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00017237906423473435,
+      "loss": 2.7071,
+      "step": 10950
+    },
+    {
+      "epoch": 1.444327731092437,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00017224689399947133,
+      "loss": 2.7335,
+      "step": 11000
+    },
+    {
+      "epoch": 1.4508928571428572,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00017211472376420833,
+      "loss": 2.7514,
+      "step": 11050
+    },
+    {
+      "epoch": 1.4574579831932772,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.0001719825535289453,
+      "loss": 2.702,
+      "step": 11100
+    },
+    {
+      "epoch": 1.4640231092436975,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00017185038329368225,
+      "loss": 2.7518,
+      "step": 11150
+    },
+    {
+      "epoch": 1.4705882352941178,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00017171821305841925,
+      "loss": 2.7228,
+      "step": 11200
+    },
+    {
+      "epoch": 1.4771533613445378,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00017158604282315623,
+      "loss": 2.7007,
+      "step": 11250
+    },
+    {
+      "epoch": 1.483718487394958,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001714538725878932,
+      "loss": 2.7112,
+      "step": 11300
+    },
+    {
+      "epoch": 1.490283613445378,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00017132170235263018,
+      "loss": 2.7264,
+      "step": 11350
+    },
+    {
+      "epoch": 1.4968487394957983,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00017118953211736718,
+      "loss": 2.7302,
+      "step": 11400
+    },
+    {
+      "epoch": 1.5034138655462184,
+      "grad_norm": 1.03125,
+      "learning_rate": 0.00017105736188210416,
+      "loss": 2.7063,
+      "step": 11450
+    },
+    {
+      "epoch": 1.5099789915966386,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00017092519164684113,
+      "loss": 2.7232,
+      "step": 11500
+    },
+    {
+      "epoch": 1.5165441176470589,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.0001707930214115781,
+      "loss": 2.7323,
+      "step": 11550
+    },
+    {
+      "epoch": 1.523109243697479,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00017066085117631508,
+      "loss": 2.7504,
+      "step": 11600
+    },
+    {
+      "epoch": 1.5296743697478992,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00017052868094105208,
+      "loss": 2.7087,
+      "step": 11650
+    },
+    {
+      "epoch": 1.5362394957983194,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00017039651070578906,
+      "loss": 2.687,
+      "step": 11700
+    },
+    {
+      "epoch": 1.5428046218487395,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017026434047052603,
+      "loss": 2.7249,
+      "step": 11750
+    },
+    {
+      "epoch": 1.5493697478991597,
+      "grad_norm": 1.375,
+      "learning_rate": 0.000170132170235263,
+      "loss": 2.7173,
+      "step": 11800
+    },
+    {
+      "epoch": 1.55593487394958,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00017,
+      "loss": 2.7128,
+      "step": 11850
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000169867829764737,
+      "loss": 2.6948,
+      "step": 11900
+    },
+    {
+      "epoch": 1.56906512605042,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00016973565952947396,
+      "loss": 2.7077,
+      "step": 11950
+    },
+    {
+      "epoch": 1.5756302521008403,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00016960348929421094,
+      "loss": 2.7234,
+      "step": 12000
+    },
+    {
+      "epoch": 1.5821953781512605,
+      "grad_norm": 1.0859375,
+      "learning_rate": 0.00016947131905894794,
+      "loss": 2.7016,
+      "step": 12050
+    },
+    {
+      "epoch": 1.5887605042016806,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00016933914882368492,
+      "loss": 2.7031,
+      "step": 12100
+    },
+    {
+      "epoch": 1.5953256302521008,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001692069785884219,
+      "loss": 2.7401,
+      "step": 12150
+    },
+    {
+      "epoch": 1.601890756302521,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00016907480835315887,
+      "loss": 2.7053,
+      "step": 12200
+    },
+    {
+      "epoch": 1.6084558823529411,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016894263811789587,
+      "loss": 2.7007,
+      "step": 12250
+    },
+    {
+      "epoch": 1.6150210084033614,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.00016881046788263284,
+      "loss": 2.7229,
+      "step": 12300
+    },
+    {
+      "epoch": 1.6215861344537816,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.00016867829764736982,
+      "loss": 2.707,
+      "step": 12350
+    },
+    {
+      "epoch": 1.6281512605042017,
+      "grad_norm": 0.9375,
+      "learning_rate": 0.0001685461274121068,
+      "loss": 2.7095,
+      "step": 12400
+    },
+    {
+      "epoch": 1.6347163865546217,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001684139571768438,
+      "loss": 2.7258,
+      "step": 12450
+    },
+    {
+      "epoch": 1.6412815126050422,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00016828178694158077,
+      "loss": 2.6818,
+      "step": 12500
+    },
+    {
+      "epoch": 1.6478466386554622,
+      "grad_norm": 1.140625,
+      "learning_rate": 0.00016814961670631775,
+      "loss": 2.7119,
+      "step": 12550
+    },
+    {
+      "epoch": 1.6544117647058822,
+      "grad_norm": 1.2734375,
+      "learning_rate": 0.00016801744647105472,
+      "loss": 2.7186,
+      "step": 12600
+    },
+    {
+      "epoch": 1.6609768907563025,
+      "grad_norm": 1.3984375,
+      "learning_rate": 0.00016788527623579172,
+      "loss": 2.6939,
+      "step": 12650
+    },
+    {
+      "epoch": 1.6675420168067228,
+      "grad_norm": 1.1015625,
+      "learning_rate": 0.0001677531060005287,
+      "loss": 2.7297,
+      "step": 12700
+    },
+    {
+      "epoch": 1.6741071428571428,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016762093576526567,
+      "loss": 2.7283,
+      "step": 12750
+    },
+    {
+      "epoch": 1.680672268907563,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016748876553000265,
+      "loss": 2.6951,
+      "step": 12800
+    },
+    {
+      "epoch": 1.6872373949579833,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00016735659529473962,
+      "loss": 2.693,
+      "step": 12850
+    },
+    {
+      "epoch": 1.6938025210084033,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016722442505947663,
+      "loss": 2.7194,
+      "step": 12900
+    },
+    {
+      "epoch": 1.7003676470588234,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001670922548242136,
+      "loss": 2.6951,
+      "step": 12950
+    },
+    {
+      "epoch": 1.7069327731092439,
+      "grad_norm": 1.125,
+      "learning_rate": 0.00016696008458895058,
+      "loss": 2.7053,
+      "step": 13000
+    },
+    {
+      "epoch": 1.7134978991596639,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00016682791435368755,
+      "loss": 2.7034,
+      "step": 13050
+    },
+    {
+      "epoch": 1.720063025210084,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016669574411842455,
+      "loss": 2.7118,
+      "step": 13100
+    },
+    {
+      "epoch": 1.7266281512605042,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.00016656357388316153,
+      "loss": 2.7267,
+      "step": 13150
+    },
+    {
+      "epoch": 1.7331932773109244,
+      "grad_norm": 1.0,
+      "learning_rate": 0.0001664314036478985,
+      "loss": 2.6885,
+      "step": 13200
+    },
+    {
+      "epoch": 1.7397584033613445,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016629923341263548,
+      "loss": 2.6921,
+      "step": 13250
+    },
+    {
+      "epoch": 1.7463235294117647,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00016616706317737248,
+      "loss": 2.7048,
+      "step": 13300
+    },
+    {
+      "epoch": 1.752888655462185,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016603489294210946,
+      "loss": 2.688,
+      "step": 13350
+    },
+    {
+      "epoch": 1.759453781512605,
+      "grad_norm": 1.0546875,
+      "learning_rate": 0.00016590272270684643,
+      "loss": 2.7187,
+      "step": 13400
+    },
+    {
+      "epoch": 1.7660189075630253,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001657705524715834,
+      "loss": 2.7089,
+      "step": 13450
+    },
+    {
+      "epoch": 1.7725840336134455,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001656383822363204,
+      "loss": 2.7104,
+      "step": 13500
+    },
+    {
+      "epoch": 1.7791491596638656,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016550621200105736,
+      "loss": 2.694,
+      "step": 13550
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 1.84375,
+      "learning_rate": 0.00016537404176579433,
+      "loss": 2.7143,
+      "step": 13600
+    },
+    {
+      "epoch": 1.7922794117647058,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.0001652418715305313,
+      "loss": 2.7057,
+      "step": 13650
+    },
+    {
+      "epoch": 1.798844537815126,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001651097012952683,
+      "loss": 2.7065,
+      "step": 13700
+    },
+    {
+      "epoch": 1.8054096638655461,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001649775310600053,
+      "loss": 2.7258,
+      "step": 13750
+    },
+    {
+      "epoch": 1.8119747899159664,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00016484536082474226,
+      "loss": 2.7026,
+      "step": 13800
+    },
+    {
+      "epoch": 1.8185399159663866,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016471319058947924,
+      "loss": 2.7012,
+      "step": 13850
+    },
+    {
+      "epoch": 1.8251050420168067,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016458102035421624,
+      "loss": 2.677,
+      "step": 13900
+    },
+    {
+      "epoch": 1.831670168067227,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00016444885011895321,
+      "loss": 2.6752,
+      "step": 13950
+    },
+    {
+      "epoch": 1.8382352941176472,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.0001643166798836902,
+      "loss": 2.71,
+      "step": 14000
+    },
+    {
+      "epoch": 1.8448004201680672,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00016418450964842716,
+      "loss": 2.7039,
+      "step": 14050
+    },
+    {
+      "epoch": 1.8513655462184873,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00016405233941316417,
+      "loss": 2.6974,
+      "step": 14100
+    },
+    {
+      "epoch": 1.8579306722689075,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00016392016917790114,
+      "loss": 2.7124,
+      "step": 14150
+    },
+    {
+      "epoch": 1.8644957983193278,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00016378799894263812,
+      "loss": 2.7038,
+      "step": 14200
+    },
+    {
+      "epoch": 1.8710609243697478,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.0001636558287073751,
+      "loss": 2.6604,
+      "step": 14250
+    },
+    {
+      "epoch": 1.877626050420168,
+      "grad_norm": 1.2109375,
+      "learning_rate": 0.0001635236584721121,
+      "loss": 2.6858,
+      "step": 14300
+    },
+    {
+      "epoch": 1.8841911764705883,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00016339148823684907,
+      "loss": 2.7067,
+      "step": 14350
+    },
+    {
+      "epoch": 1.8907563025210083,
+      "grad_norm": 1.25,
+      "learning_rate": 0.00016325931800158605,
+      "loss": 2.7035,
+      "step": 14400
+    },
+    {
+      "epoch": 1.8973214285714286,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00016312714776632302,
+      "loss": 2.6906,
+      "step": 14450
+    },
+    {
+      "epoch": 1.9038865546218489,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00016299497753106002,
+      "loss": 2.6797,
+      "step": 14500
+    },
+    {
+      "epoch": 1.910451680672269,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000162862807295797,
+      "loss": 2.6928,
+      "step": 14550
+    },
+    {
+      "epoch": 1.917016806722689,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00016273063706053397,
+      "loss": 2.7051,
+      "step": 14600
+    },
+    {
+      "epoch": 1.9235819327731094,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.00016259846682527095,
+      "loss": 2.6779,
+      "step": 14650
+    },
+    {
+      "epoch": 1.9301470588235294,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.00016246629659000795,
+      "loss": 2.6788,
+      "step": 14700
+    },
+    {
+      "epoch": 1.9367121848739495,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00016233412635474493,
+      "loss": 2.6799,
+      "step": 14750
+    },
+    {
+      "epoch": 1.9432773109243697,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.0001622019561194819,
+      "loss": 2.6975,
+      "step": 14800
+    },
+    {
+      "epoch": 1.94984243697479,
+      "grad_norm": 1.546875,
+      "learning_rate": 0.00016206978588421888,
+      "loss": 2.7005,
+      "step": 14850
+    },
+    {
+      "epoch": 1.95640756302521,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00016193761564895585,
+      "loss": 2.7031,
+      "step": 14900
+    },
+    {
+      "epoch": 1.9629726890756303,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00016180544541369285,
+      "loss": 2.6751,
+      "step": 14950
+    },
+    {
+      "epoch": 1.9695378151260505,
+      "grad_norm": 0.984375,
+      "learning_rate": 0.00016167327517842983,
+      "loss": 2.7391,
+      "step": 15000
+    },
+    {
+      "epoch": 1.9761029411764706,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.0001615411049431668,
+      "loss": 2.7114,
+      "step": 15050
+    },
+    {
+      "epoch": 1.9826680672268906,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00016140893470790378,
+      "loss": 2.6763,
+      "step": 15100
+    },
+    {
+      "epoch": 1.989233193277311,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00016127676447264078,
+      "loss": 2.7001,
+      "step": 15150
+    },
+    {
+      "epoch": 1.995798319327731,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00016114459423737776,
+      "loss": 2.697,
+      "step": 15200
+    },
+    {
+      "epoch": 2.002363445378151,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00016101242400211473,
+      "loss": 2.5799,
+      "step": 15250
+    },
+    {
+      "epoch": 2.0089285714285716,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001608802537668517,
+      "loss": 2.4102,
+      "step": 15300
+    },
+    {
+      "epoch": 2.0154936974789917,
+      "grad_norm": 1.1171875,
+      "learning_rate": 0.0001607480835315887,
+      "loss": 2.4198,
+      "step": 15350
+    },
+    {
+      "epoch": 2.0220588235294117,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00016061591329632568,
+      "loss": 2.398,
+      "step": 15400
+    },
+    {
+      "epoch": 2.0286239495798317,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00016048374306106266,
+      "loss": 2.3839,
+      "step": 15450
+    },
+    {
+      "epoch": 2.035189075630252,
+      "grad_norm": 1.171875,
+      "learning_rate": 0.00016035157282579964,
+      "loss": 2.3944,
+      "step": 15500
+    },
+    {
+      "epoch": 2.0417542016806722,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00016021940259053664,
+      "loss": 2.3921,
+      "step": 15550
+    },
+    {
+      "epoch": 2.0483193277310923,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001600872323552736,
+      "loss": 2.3781,
+      "step": 15600
+    },
+    {
+      "epoch": 2.0548844537815127,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.0001599550621200106,
+      "loss": 2.3969,
+      "step": 15650
+    },
+    {
+      "epoch": 2.0614495798319328,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00015982289188474756,
+      "loss": 2.4146,
+      "step": 15700
+    },
+    {
+      "epoch": 2.068014705882353,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015969072164948457,
+      "loss": 2.4412,
+      "step": 15750
+    },
+    {
+      "epoch": 2.0745798319327733,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00015955855141422154,
+      "loss": 2.417,
+      "step": 15800
+    },
+    {
+      "epoch": 2.0811449579831933,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00015942638117895852,
+      "loss": 2.4068,
+      "step": 15850
+    },
+    {
+      "epoch": 2.0877100840336134,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00015929421094369546,
+      "loss": 2.4335,
+      "step": 15900
+    },
+    {
+      "epoch": 2.0942752100840334,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015916204070843247,
+      "loss": 2.4253,
+      "step": 15950
+    },
+    {
+      "epoch": 2.100840336134454,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00015902987047316944,
+      "loss": 2.4018,
+      "step": 16000
+    },
+    {
+      "epoch": 2.107405462184874,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.00015889770023790642,
+      "loss": 2.4285,
+      "step": 16050
+    },
+    {
+      "epoch": 2.113970588235294,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.0001587655300026434,
+      "loss": 2.4368,
+      "step": 16100
+    },
+    {
+      "epoch": 2.1205357142857144,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.0001586333597673804,
+      "loss": 2.4384,
+      "step": 16150
+    },
+    {
+      "epoch": 2.1271008403361344,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00015850118953211737,
+      "loss": 2.414,
+      "step": 16200
+    },
+    {
+      "epoch": 2.1336659663865545,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00015836901929685434,
+      "loss": 2.4166,
+      "step": 16250
+    },
+    {
+      "epoch": 2.140231092436975,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.00015823684906159132,
+      "loss": 2.4319,
+      "step": 16300
+    },
+    {
+      "epoch": 2.146796218487395,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015810467882632832,
+      "loss": 2.4421,
+      "step": 16350
+    },
+    {
+      "epoch": 2.153361344537815,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.0001579725085910653,
+      "loss": 2.4383,
+      "step": 16400
+    },
+    {
+      "epoch": 2.1599264705882355,
+      "grad_norm": 1.1796875,
+      "learning_rate": 0.00015784033835580227,
+      "loss": 2.4301,
+      "step": 16450
+    },
+    {
+      "epoch": 2.1664915966386555,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015770816812053925,
+      "loss": 2.4249,
+      "step": 16500
+    },
+    {
+      "epoch": 2.1730567226890756,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015757599788527625,
+      "loss": 2.4465,
+      "step": 16550
+    },
+    {
+      "epoch": 2.1796218487394956,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00015744382765001323,
+      "loss": 2.4457,
+      "step": 16600
+    },
+    {
+      "epoch": 2.186186974789916,
+      "grad_norm": 1.5,
+      "learning_rate": 0.0001573116574147502,
+      "loss": 2.435,
+      "step": 16650
+    },
+    {
+      "epoch": 2.192752100840336,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00015717948717948718,
+      "loss": 2.4267,
+      "step": 16700
+    },
+    {
+      "epoch": 2.199317226890756,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.00015704731694422418,
+      "loss": 2.4513,
+      "step": 16750
+    },
+    {
+      "epoch": 2.2058823529411766,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015691514670896115,
+      "loss": 2.4385,
+      "step": 16800
+    },
+    {
+      "epoch": 2.2124474789915967,
+      "grad_norm": 1.4140625,
+      "learning_rate": 0.00015678297647369813,
+      "loss": 2.4747,
+      "step": 16850
+    },
+    {
+      "epoch": 2.2190126050420167,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001566508062384351,
+      "loss": 2.4497,
+      "step": 16900
+    },
+    {
+      "epoch": 2.225577731092437,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.0001565186360031721,
+      "loss": 2.4393,
+      "step": 16950
+    },
+    {
+      "epoch": 2.232142857142857,
+      "grad_norm": 1.203125,
+      "learning_rate": 0.00015638646576790908,
+      "loss": 2.4628,
+      "step": 17000
+    },
+    {
+      "epoch": 2.2387079831932772,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015625429553264606,
+      "loss": 2.426,
+      "step": 17050
+    },
+    {
+      "epoch": 2.2452731092436973,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00015612212529738303,
+      "loss": 2.4226,
+      "step": 17100
+    },
+    {
+      "epoch": 2.2518382352941178,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00015598995506212,
+      "loss": 2.461,
+      "step": 17150
+    },
+    {
+      "epoch": 2.258403361344538,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.000155857784826857,
+      "loss": 2.4646,
+      "step": 17200
+    },
+    {
+      "epoch": 2.264968487394958,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015572561459159398,
+      "loss": 2.443,
+      "step": 17250
+    },
+    {
+      "epoch": 2.2715336134453783,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00015559344435633096,
+      "loss": 2.4328,
+      "step": 17300
+    },
+    {
+      "epoch": 2.2780987394957983,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015546127412106793,
+      "loss": 2.4707,
+      "step": 17350
+    },
+    {
+      "epoch": 2.2846638655462184,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00015532910388580494,
+      "loss": 2.4422,
+      "step": 17400
+    },
+    {
+      "epoch": 2.291228991596639,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.0001551969336505419,
+      "loss": 2.4504,
+      "step": 17450
+    },
+    {
+      "epoch": 2.297794117647059,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001550647634152789,
+      "loss": 2.4589,
+      "step": 17500
+    },
+    {
+      "epoch": 2.304359243697479,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00015493259318001586,
+      "loss": 2.4549,
+      "step": 17550
+    },
+    {
+      "epoch": 2.310924369747899,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00015480042294475286,
+      "loss": 2.4549,
+      "step": 17600
+    },
+    {
+      "epoch": 2.3174894957983194,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015466825270948984,
+      "loss": 2.4304,
+      "step": 17650
+    },
+    {
+      "epoch": 2.3240546218487395,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015453608247422681,
+      "loss": 2.4605,
+      "step": 17700
+    },
+    {
+      "epoch": 2.3306197478991595,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.0001544039122389638,
+      "loss": 2.4587,
+      "step": 17750
+    },
+    {
+      "epoch": 2.33718487394958,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.0001542717420037008,
+      "loss": 2.4478,
+      "step": 17800
+    },
+    {
+      "epoch": 2.34375,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00015413957176843777,
+      "loss": 2.4473,
+      "step": 17850
+    },
+    {
+      "epoch": 2.35031512605042,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015400740153317474,
+      "loss": 2.4497,
+      "step": 17900
+    },
+    {
+      "epoch": 2.3568802521008405,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00015387523129791172,
+      "loss": 2.463,
+      "step": 17950
+    },
+    {
+      "epoch": 2.3634453781512605,
+      "grad_norm": 1.1328125,
+      "learning_rate": 0.00015374306106264872,
+      "loss": 2.4603,
+      "step": 18000
+    },
+    {
+      "epoch": 2.3700105042016806,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.0001536108908273857,
+      "loss": 2.431,
+      "step": 18050
+    },
+    {
+      "epoch": 2.376575630252101,
+      "grad_norm": 1.671875,
+      "learning_rate": 0.00015347872059212267,
+      "loss": 2.4369,
+      "step": 18100
+    },
+    {
+      "epoch": 2.383140756302521,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00015334655035685965,
+      "loss": 2.4516,
+      "step": 18150
+    },
+    {
+      "epoch": 2.389705882352941,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00015321438012159662,
+      "loss": 2.4478,
+      "step": 18200
+    },
+    {
+      "epoch": 2.396271008403361,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001530822098863336,
+      "loss": 2.4471,
+      "step": 18250
+    },
+    {
+      "epoch": 2.4028361344537816,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00015295003965107057,
+      "loss": 2.4637,
+      "step": 18300
+    },
+    {
+      "epoch": 2.4094012605042017,
+      "grad_norm": 1.0390625,
+      "learning_rate": 0.00015281786941580755,
+      "loss": 2.4466,
+      "step": 18350
+    },
+    {
+      "epoch": 2.4159663865546217,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00015268569918054455,
+      "loss": 2.4603,
+      "step": 18400
+    },
+    {
+      "epoch": 2.422531512605042,
+      "grad_norm": 1.5234375,
+      "learning_rate": 0.00015255352894528152,
+      "loss": 2.4641,
+      "step": 18450
+    },
+    {
+      "epoch": 2.429096638655462,
+      "grad_norm": 0.97265625,
+      "learning_rate": 0.0001524213587100185,
+      "loss": 2.4667,
+      "step": 18500
+    },
+    {
+      "epoch": 2.4356617647058822,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00015228918847475547,
+      "loss": 2.4812,
+      "step": 18550
+    },
+    {
+      "epoch": 2.4422268907563023,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015215701823949248,
+      "loss": 2.458,
+      "step": 18600
+    },
+    {
+      "epoch": 2.4487920168067228,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00015202484800422945,
+      "loss": 2.4618,
+      "step": 18650
+    },
+    {
+      "epoch": 2.455357142857143,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00015189267776896643,
+      "loss": 2.4799,
+      "step": 18700
+    },
+    {
+      "epoch": 2.4619222689075633,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001517605075337034,
+      "loss": 2.4915,
+      "step": 18750
+    },
+    {
+      "epoch": 2.4684873949579833,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001516283372984404,
+      "loss": 2.4903,
+      "step": 18800
+    },
+    {
+      "epoch": 2.4750525210084033,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015149616706317738,
+      "loss": 2.4655,
+      "step": 18850
+    },
+    {
+      "epoch": 2.4816176470588234,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00015136399682791436,
+      "loss": 2.4695,
+      "step": 18900
+    },
+    {
+      "epoch": 2.488182773109244,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00015123182659265133,
+      "loss": 2.4624,
+      "step": 18950
+    },
+    {
+      "epoch": 2.494747899159664,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015109965635738833,
+      "loss": 2.505,
+      "step": 19000
+    },
+    {
+      "epoch": 2.501313025210084,
+      "grad_norm": 1.125,
+      "learning_rate": 0.0001509674861221253,
+      "loss": 2.4549,
+      "step": 19050
+    },
+    {
+      "epoch": 2.5078781512605044,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00015083531588686228,
+      "loss": 2.4784,
+      "step": 19100
+    },
+    {
+      "epoch": 2.5144432773109244,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.00015070314565159926,
+      "loss": 2.4385,
+      "step": 19150
+    },
+    {
+      "epoch": 2.5210084033613445,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00015057097541633623,
+      "loss": 2.4701,
+      "step": 19200
+    },
+    {
+      "epoch": 2.5275735294117645,
+      "grad_norm": 1.390625,
+      "learning_rate": 0.00015043880518107324,
+      "loss": 2.4518,
+      "step": 19250
+    },
+    {
+      "epoch": 2.534138655462185,
+      "grad_norm": 1.6484375,
+      "learning_rate": 0.0001503066349458102,
+      "loss": 2.4421,
+      "step": 19300
+    },
+    {
+      "epoch": 2.540703781512605,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00015017446471054719,
+      "loss": 2.4659,
+      "step": 19350
+    },
+    {
+      "epoch": 2.5472689075630255,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00015004229447528416,
+      "loss": 2.444,
+      "step": 19400
+    },
+    {
+      "epoch": 2.5538340336134455,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00014991012424002116,
+      "loss": 2.4648,
+      "step": 19450
+    },
+    {
+      "epoch": 2.5603991596638656,
+      "grad_norm": 1.3359375,
+      "learning_rate": 0.00014977795400475814,
+      "loss": 2.4668,
+      "step": 19500
+    },
+    {
+      "epoch": 2.5669642857142856,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00014964578376949511,
+      "loss": 2.467,
+      "step": 19550
+    },
+    {
+      "epoch": 2.5735294117647056,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001495136135342321,
+      "loss": 2.5089,
+      "step": 19600
+    },
+    {
+      "epoch": 2.580094537815126,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.0001493814432989691,
+      "loss": 2.4665,
+      "step": 19650
+    },
+    {
+      "epoch": 2.586659663865546,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014924927306370607,
+      "loss": 2.4865,
+      "step": 19700
+    },
+    {
+      "epoch": 2.5932247899159666,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00014911710282844304,
+      "loss": 2.4689,
+      "step": 19750
+    },
+    {
+      "epoch": 2.5997899159663866,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014898493259318002,
+      "loss": 2.4494,
+      "step": 19800
+    },
+    {
+      "epoch": 2.6063550420168067,
+      "grad_norm": 1.75,
+      "learning_rate": 0.00014885276235791702,
+      "loss": 2.4745,
+      "step": 19850
+    },
+    {
+      "epoch": 2.6129201680672267,
+      "grad_norm": 1.6171875,
+      "learning_rate": 0.000148720592122654,
+      "loss": 2.4324,
+      "step": 19900
+    },
+    {
+      "epoch": 2.619485294117647,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00014858842188739097,
+      "loss": 2.4776,
+      "step": 19950
+    },
+    {
+      "epoch": 2.6260504201680672,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00014845625165212794,
+      "loss": 2.4833,
+      "step": 20000
+    },
+    {
+      "epoch": 2.6326155462184873,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.00014832408141686495,
+      "loss": 2.5234,
+      "step": 20050
+    },
+    {
+      "epoch": 2.6391806722689077,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00014819191118160192,
+      "loss": 2.4781,
+      "step": 20100
+    },
+    {
+      "epoch": 2.6457457983193278,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.0001480597409463389,
+      "loss": 2.456,
+      "step": 20150
+    },
+    {
+      "epoch": 2.652310924369748,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00014792757071107587,
+      "loss": 2.5011,
+      "step": 20200
+    },
+    {
+      "epoch": 2.658876050420168,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00014779540047581288,
+      "loss": 2.4785,
+      "step": 20250
+    },
+    {
+      "epoch": 2.6654411764705883,
+      "grad_norm": 1.4296875,
+      "learning_rate": 0.00014766323024054985,
+      "loss": 2.516,
+      "step": 20300
+    },
+    {
+      "epoch": 2.6720063025210083,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00014753106000528683,
+      "loss": 2.5028,
+      "step": 20350
+    },
+    {
+      "epoch": 2.678571428571429,
+      "grad_norm": 1.25,
+      "learning_rate": 0.0001473988897700238,
+      "loss": 2.4378,
+      "step": 20400
+    },
+    {
+      "epoch": 2.685136554621849,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00014726671953476078,
+      "loss": 2.4658,
+      "step": 20450
+    },
+    {
+      "epoch": 2.691701680672269,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00014713454929949778,
+      "loss": 2.4844,
+      "step": 20500
+    },
+    {
+      "epoch": 2.698266806722689,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00014700237906423475,
+      "loss": 2.4709,
+      "step": 20550
+    },
+    {
+      "epoch": 2.7048319327731094,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.0001468702088289717,
+      "loss": 2.4797,
+      "step": 20600
+    },
+    {
+      "epoch": 2.7113970588235294,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.0001467380385937087,
+      "loss": 2.5037,
+      "step": 20650
+    },
+    {
+      "epoch": 2.7179621848739495,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00014660586835844568,
+      "loss": 2.4743,
+      "step": 20700
+    },
+    {
+      "epoch": 2.72452731092437,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00014647369812318265,
+      "loss": 2.5179,
+      "step": 20750
+    },
+    {
+      "epoch": 2.73109243697479,
+      "grad_norm": 1.3828125,
+      "learning_rate": 0.00014634152788791963,
+      "loss": 2.4847,
+      "step": 20800
+    },
+    {
+      "epoch": 2.73765756302521,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00014620935765265663,
+      "loss": 2.5035,
+      "step": 20850
+    },
+    {
+      "epoch": 2.74422268907563,
+      "grad_norm": 1.34375,
+      "learning_rate": 0.0001460771874173936,
+      "loss": 2.4859,
+      "step": 20900
+    },
+    {
+      "epoch": 2.7507878151260505,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00014594501718213058,
+      "loss": 2.466,
+      "step": 20950
+    },
+    {
+      "epoch": 2.7573529411764706,
+      "grad_norm": 1.1640625,
+      "learning_rate": 0.00014581284694686756,
+      "loss": 2.4741,
+      "step": 21000
+    },
+    {
+      "epoch": 2.7639180672268906,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.00014568067671160456,
+      "loss": 2.4844,
+      "step": 21050
+    },
+    {
+      "epoch": 2.770483193277311,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00014554850647634153,
+      "loss": 2.4834,
+      "step": 21100
+    },
+    {
+      "epoch": 2.777048319327731,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.0001454163362410785,
+      "loss": 2.4813,
+      "step": 21150
+    },
+    {
+      "epoch": 2.783613445378151,
+      "grad_norm": 1.453125,
+      "learning_rate": 0.00014528416600581549,
+      "loss": 2.4598,
+      "step": 21200
+    },
+    {
+      "epoch": 2.790178571428571,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00014515199577055246,
+      "loss": 2.4787,
+      "step": 21250
+    },
+    {
+      "epoch": 2.7967436974789917,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014501982553528946,
+      "loss": 2.5089,
+      "step": 21300
+    },
+    {
+      "epoch": 2.8033088235294117,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00014488765530002644,
+      "loss": 2.4685,
+      "step": 21350
+    },
+    {
+      "epoch": 2.809873949579832,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001447554850647634,
+      "loss": 2.495,
+      "step": 21400
+    },
+    {
+      "epoch": 2.816439075630252,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.0001446233148295004,
+      "loss": 2.5034,
+      "step": 21450
+    },
+    {
+      "epoch": 2.8230042016806722,
+      "grad_norm": 1.5546875,
+      "learning_rate": 0.0001444911445942374,
+      "loss": 2.4668,
+      "step": 21500
+    },
+    {
+      "epoch": 2.8295693277310923,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.00014435897435897437,
+      "loss": 2.4759,
+      "step": 21550
+    },
+    {
+      "epoch": 2.8361344537815127,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014422680412371134,
+      "loss": 2.5021,
+      "step": 21600
+    },
+    {
+      "epoch": 2.8426995798319328,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00014409463388844832,
+      "loss": 2.4722,
+      "step": 21650
+    },
+    {
+      "epoch": 2.849264705882353,
+      "grad_norm": 1.3125,
+      "learning_rate": 0.00014396246365318532,
+      "loss": 2.4932,
+      "step": 21700
+    },
+    {
+      "epoch": 2.8558298319327733,
+      "grad_norm": 1.2265625,
+      "learning_rate": 0.0001438302934179223,
+      "loss": 2.4748,
+      "step": 21750
+    },
+    {
+      "epoch": 2.8623949579831933,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00014369812318265927,
+      "loss": 2.4599,
+      "step": 21800
+    },
+    {
+      "epoch": 2.8689600840336134,
+      "grad_norm": 1.296875,
+      "learning_rate": 0.00014356595294739624,
+      "loss": 2.4848,
+      "step": 21850
+    },
+    {
+      "epoch": 2.8755252100840334,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.00014343378271213325,
+      "loss": 2.4876,
+      "step": 21900
+    },
+    {
+      "epoch": 2.882090336134454,
+      "grad_norm": 1.28125,
+      "learning_rate": 0.00014330161247687022,
+      "loss": 2.4786,
+      "step": 21950
+    },
+    {
+      "epoch": 2.888655462184874,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0001431694422416072,
+      "loss": 2.4788,
+      "step": 22000
+    },
+    {
+      "epoch": 2.8952205882352944,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00014303727200634417,
+      "loss": 2.4968,
+      "step": 22050
+    },
+    {
+      "epoch": 2.9017857142857144,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.00014290510177108117,
+      "loss": 2.4782,
+      "step": 22100
+    },
+    {
+      "epoch": 2.9083508403361344,
+      "grad_norm": 1.4453125,
+      "learning_rate": 0.00014277293153581815,
+      "loss": 2.4784,
+      "step": 22150
+    },
+    {
+      "epoch": 2.9149159663865545,
+      "grad_norm": 1.109375,
+      "learning_rate": 0.00014264076130055512,
+      "loss": 2.5067,
+      "step": 22200
+    },
+    {
+      "epoch": 2.9214810924369745,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.0001425085910652921,
+      "loss": 2.4691,
+      "step": 22250
+    },
+    {
+      "epoch": 2.928046218487395,
+      "grad_norm": 1.15625,
+      "learning_rate": 0.0001423764208300291,
+      "loss": 2.4585,
+      "step": 22300
+    },
+    {
+      "epoch": 2.934611344537815,
+      "grad_norm": 1.2890625,
+      "learning_rate": 0.00014224425059476608,
+      "loss": 2.4524,
+      "step": 22350
+    },
+    {
+      "epoch": 2.9411764705882355,
+      "grad_norm": 1.328125,
+      "learning_rate": 0.00014211208035950305,
+      "loss": 2.4853,
+      "step": 22400
+    },
+    {
+      "epoch": 2.9477415966386555,
+      "grad_norm": 1.3046875,
+      "learning_rate": 0.00014197991012424003,
+      "loss": 2.4661,
+      "step": 22450
+    },
+    {
+      "epoch": 2.9543067226890756,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.000141847739888977,
+      "loss": 2.4908,
+      "step": 22500
+    },
+    {
+      "epoch": 2.9608718487394956,
+      "grad_norm": 1.46875,
+      "learning_rate": 0.000141715569653714,
+      "loss": 2.471,
+      "step": 22550
+    },
+    {
+      "epoch": 2.967436974789916,
+      "grad_norm": 1.3515625,
+      "learning_rate": 0.00014158339941845098,
+      "loss": 2.4783,
+      "step": 22600
+    },
+    {
+      "epoch": 2.974002100840336,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00014145122918318796,
+      "loss": 2.4755,
+      "step": 22650
+    },
+    {
+      "epoch": 2.980567226890756,
+      "grad_norm": 1.1953125,
+      "learning_rate": 0.00014131905894792493,
+      "loss": 2.4862,
+      "step": 22700
+    },
+    {
+      "epoch": 2.9871323529411766,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00014118688871266193,
+      "loss": 2.4763,
+      "step": 22750
+    },
+    {
+      "epoch": 2.9936974789915967,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.0001410547184773989,
+      "loss": 2.4538,
+      "step": 22800
+    },
+    {
+      "epoch": 3.0002626050420167,
+      "grad_norm": 1.359375,
+      "learning_rate": 0.00014092254824213588,
+      "loss": 2.4822,
+      "step": 22850
+    },
+    {
+      "epoch": 3.006827731092437,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00014079037800687286,
+      "loss": 2.1167,
+      "step": 22900
+    },
+    {
+      "epoch": 3.013392857142857,
+      "grad_norm": 1.984375,
+      "learning_rate": 0.00014065820777160986,
+      "loss": 2.1005,
+      "step": 22950
+    },
+    {
+      "epoch": 3.0199579831932772,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.0001405260375363468,
+      "loss": 2.0618,
+      "step": 23000
+    },
+    {
+      "epoch": 3.0265231092436973,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00014039386730108378,
+      "loss": 2.0603,
+      "step": 23050
+    },
+    {
+      "epoch": 3.0330882352941178,
+      "grad_norm": 1.3203125,
+      "learning_rate": 0.0001402616970658208,
+      "loss": 2.0654,
+      "step": 23100
+    },
+    {
+      "epoch": 3.039653361344538,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.00014012952683055776,
+      "loss": 2.0578,
+      "step": 23150
+    },
+    {
+      "epoch": 3.046218487394958,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013999735659529474,
+      "loss": 2.1235,
+      "step": 23200
+    },
+    {
+      "epoch": 3.0527836134453783,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.0001398651863600317,
+      "loss": 2.0834,
+      "step": 23250
+    },
+    {
+      "epoch": 3.0593487394957983,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.0001397330161247687,
+      "loss": 2.1192,
+      "step": 23300
+    },
+    {
+      "epoch": 3.0659138655462184,
+      "grad_norm": 1.40625,
+      "learning_rate": 0.0001396008458895057,
+      "loss": 2.0848,
+      "step": 23350
+    },
+    {
+      "epoch": 3.072478991596639,
+      "grad_norm": 1.7421875,
+      "learning_rate": 0.00013946867565424266,
+      "loss": 2.0731,
+      "step": 23400
+    },
+    {
+      "epoch": 3.079044117647059,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00013933650541897964,
+      "loss": 2.1142,
+      "step": 23450
+    },
+    {
+      "epoch": 3.085609243697479,
+      "grad_norm": 1.1875,
+      "learning_rate": 0.00013920433518371662,
+      "loss": 2.1067,
+      "step": 23500
+    },
+    {
+      "epoch": 3.092174369747899,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00013907216494845362,
+      "loss": 2.0934,
+      "step": 23550
+    },
+    {
+      "epoch": 3.0987394957983194,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001389399947131906,
+      "loss": 2.1451,
+      "step": 23600
+    },
+    {
+      "epoch": 3.1053046218487395,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013880782447792757,
+      "loss": 2.1069,
+      "step": 23650
+    },
+    {
+      "epoch": 3.1118697478991595,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00013867565424266454,
+      "loss": 2.0934,
+      "step": 23700
+    },
+    {
+      "epoch": 3.11843487394958,
+      "grad_norm": 1.2578125,
+      "learning_rate": 0.00013854348400740155,
+      "loss": 2.1632,
+      "step": 23750
+    },
+    {
+      "epoch": 3.125,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.00013841131377213852,
+      "loss": 2.119,
+      "step": 23800
+    },
+    {
+      "epoch": 3.13156512605042,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.0001382791435368755,
+      "loss": 2.0729,
+      "step": 23850
+    },
+    {
+      "epoch": 3.1381302521008405,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00013814697330161247,
+      "loss": 2.1073,
+      "step": 23900
+    },
+    {
+      "epoch": 3.1446953781512605,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013801480306634947,
+      "loss": 2.1055,
+      "step": 23950
+    },
+    {
+      "epoch": 3.1512605042016806,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013788263283108645,
+      "loss": 2.1141,
+      "step": 24000
+    },
+    {
+      "epoch": 3.157825630252101,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00013775046259582342,
+      "loss": 2.1012,
+      "step": 24050
+    },
+    {
+      "epoch": 3.164390756302521,
+      "grad_norm": 1.078125,
+      "learning_rate": 0.0001376182923605604,
+      "loss": 2.1016,
+      "step": 24100
+    },
+    {
+      "epoch": 3.170955882352941,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.0001374861221252974,
+      "loss": 2.1559,
+      "step": 24150
+    },
+    {
+      "epoch": 3.177521008403361,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00013735395189003438,
+      "loss": 2.1268,
+      "step": 24200
+    },
+    {
+      "epoch": 3.1840861344537816,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00013722178165477135,
+      "loss": 2.159,
+      "step": 24250
+    },
+    {
+      "epoch": 3.1906512605042017,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00013708961141950833,
+      "loss": 2.1487,
+      "step": 24300
+    },
+    {
+      "epoch": 3.1972163865546217,
+      "grad_norm": 1.3671875,
+      "learning_rate": 0.00013695744118424533,
+      "loss": 2.1238,
+      "step": 24350
+    },
+    {
+      "epoch": 3.203781512605042,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.0001368252709489823,
+      "loss": 2.135,
+      "step": 24400
+    },
+    {
+      "epoch": 3.210346638655462,
+      "grad_norm": 1.5078125,
+      "learning_rate": 0.00013669310071371928,
+      "loss": 2.1486,
+      "step": 24450
+    },
+    {
+      "epoch": 3.2169117647058822,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00013656093047845625,
+      "loss": 2.138,
+      "step": 24500
+    },
+    {
+      "epoch": 3.2234768907563027,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00013642876024319323,
+      "loss": 2.1057,
+      "step": 24550
+    },
+    {
+      "epoch": 3.2300420168067228,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00013629659000793023,
+      "loss": 2.116,
+      "step": 24600
+    },
+    {
+      "epoch": 3.236607142857143,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.0001361644197726672,
+      "loss": 2.1226,
+      "step": 24650
+    },
+    {
+      "epoch": 3.243172268907563,
+      "grad_norm": 1.9375,
+      "learning_rate": 0.00013603224953740418,
+      "loss": 2.1724,
+      "step": 24700
+    },
+    {
+      "epoch": 3.2497373949579833,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00013590007930214116,
+      "loss": 2.1682,
+      "step": 24750
+    },
+    {
+      "epoch": 3.2563025210084033,
+      "grad_norm": 1.734375,
+      "learning_rate": 0.00013576790906687816,
+      "loss": 2.1606,
+      "step": 24800
+    },
+    {
+      "epoch": 3.2628676470588234,
+      "grad_norm": 1.78125,
+      "learning_rate": 0.00013563573883161514,
+      "loss": 2.1585,
+      "step": 24850
+    },
+    {
+      "epoch": 3.269432773109244,
+      "grad_norm": 1.4375,
+      "learning_rate": 0.0001355035685963521,
+      "loss": 2.1539,
+      "step": 24900
+    },
+    {
+      "epoch": 3.275997899159664,
+      "grad_norm": 1.5,
+      "learning_rate": 0.00013537139836108909,
+      "loss": 2.1504,
+      "step": 24950
+    },
+    {
+      "epoch": 3.282563025210084,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.0001352392281258261,
+      "loss": 2.1417,
+      "step": 25000
+    },
+    {
+      "epoch": 3.2891281512605044,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00013510705789056306,
+      "loss": 2.1579,
+      "step": 25050
+    },
+    {
+      "epoch": 3.2956932773109244,
+      "grad_norm": 1.703125,
+      "learning_rate": 0.00013497488765530004,
+      "loss": 2.1358,
+      "step": 25100
+    },
+    {
+      "epoch": 3.3022584033613445,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013484271742003701,
+      "loss": 2.1612,
+      "step": 25150
+    },
+    {
+      "epoch": 3.3088235294117645,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00013471054718477402,
+      "loss": 2.1567,
+      "step": 25200
+    },
+    {
+      "epoch": 3.315388655462185,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.000134578376949511,
+      "loss": 2.1374,
+      "step": 25250
+    },
+    {
+      "epoch": 3.321953781512605,
+      "grad_norm": 2.46875,
+      "learning_rate": 0.00013444620671424797,
+      "loss": 2.1922,
+      "step": 25300
+    },
+    {
+      "epoch": 3.328518907563025,
+      "grad_norm": 1.375,
+      "learning_rate": 0.00013431403647898494,
+      "loss": 2.1584,
+      "step": 25350
+    },
+    {
+      "epoch": 3.3350840336134455,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00013418186624372192,
+      "loss": 2.1268,
+      "step": 25400
+    },
+    {
+      "epoch": 3.3416491596638656,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001340496960084589,
+      "loss": 2.1464,
+      "step": 25450
+    },
+    {
+      "epoch": 3.3482142857142856,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00013391752577319587,
+      "loss": 2.1527,
+      "step": 25500
+    },
+    {
+      "epoch": 3.354779411764706,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00013378535553793284,
+      "loss": 2.1261,
+      "step": 25550
+    },
+    {
+      "epoch": 3.361344537815126,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00013365318530266984,
+      "loss": 2.1905,
+      "step": 25600
+    },
+    {
+      "epoch": 3.367909663865546,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00013352101506740682,
+      "loss": 2.1369,
+      "step": 25650
+    },
+    {
+      "epoch": 3.3744747899159666,
+      "grad_norm": 1.7109375,
+      "learning_rate": 0.0001333888448321438,
+      "loss": 2.1478,
+      "step": 25700
+    },
+    {
+      "epoch": 3.3810399159663866,
+      "grad_norm": 1.5625,
+      "learning_rate": 0.00013325667459688077,
+      "loss": 2.1798,
+      "step": 25750
+    },
+    {
+      "epoch": 3.3876050420168067,
+      "grad_norm": 1.6640625,
+      "learning_rate": 0.00013312450436161777,
+      "loss": 2.1629,
+      "step": 25800
+    },
+    {
+      "epoch": 3.3941701680672267,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00013299233412635475,
+      "loss": 2.2038,
+      "step": 25850
+    },
+    {
+      "epoch": 3.400735294117647,
+      "grad_norm": 1.4765625,
+      "learning_rate": 0.00013286016389109172,
+      "loss": 2.1782,
+      "step": 25900
+    },
+    {
+      "epoch": 3.4073004201680672,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.0001327279936558287,
+      "loss": 2.1563,
+      "step": 25950
+    },
+    {
+      "epoch": 3.4138655462184873,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.0001325958234205657,
+      "loss": 2.1824,
+      "step": 26000
+    },
+    {
+      "epoch": 3.4204306722689077,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00013246365318530268,
+      "loss": 1.8937,
+      "step": 26050
+    },
+    {
+      "epoch": 3.4269957983193278,
+      "grad_norm": 1.5703125,
+      "learning_rate": 0.00013233148295003965,
+      "loss": 1.9172,
+      "step": 26100
+    },
+    {
+      "epoch": 3.433560924369748,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00013219931271477663,
+      "loss": 1.8899,
+      "step": 26150
+    },
+    {
+      "epoch": 3.440126050420168,
+      "grad_norm": 1.90625,
+      "learning_rate": 0.00013206714247951363,
+      "loss": 1.8592,
+      "step": 26200
+    },
+    {
+      "epoch": 3.4466911764705883,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.0001319349722442506,
+      "loss": 1.9016,
+      "step": 26250
+    },
+    {
+      "epoch": 3.4532563025210083,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00013180280200898758,
+      "loss": 1.9376,
+      "step": 26300
+    },
+    {
+      "epoch": 3.4598214285714284,
+      "grad_norm": 1.640625,
+      "learning_rate": 0.00013167063177372455,
+      "loss": 1.9256,
+      "step": 26350
+    },
+    {
+      "epoch": 3.466386554621849,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00013153846153846156,
+      "loss": 1.8968,
+      "step": 26400
+    },
+    {
+      "epoch": 3.472951680672269,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00013140629130319853,
+      "loss": 1.8637,
+      "step": 26450
+    },
+    {
+      "epoch": 3.479516806722689,
+      "grad_norm": 1.6328125,
+      "learning_rate": 0.0001312741210679355,
+      "loss": 1.8956,
+      "step": 26500
+    },
+    {
+      "epoch": 3.4860819327731094,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00013114195083267248,
+      "loss": 1.9324,
+      "step": 26550
+    },
+    {
+      "epoch": 3.4926470588235294,
+      "grad_norm": 1.625,
+      "learning_rate": 0.00013100978059740948,
+      "loss": 1.921,
+      "step": 26600
+    },
+    {
+      "epoch": 3.4992121848739495,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00013087761036214646,
+      "loss": 1.8565,
+      "step": 26650
+    },
+    {
+      "epoch": 3.50577731092437,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00013074544012688343,
+      "loss": 1.9227,
+      "step": 26700
+    },
+    {
+      "epoch": 3.51234243697479,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001306132698916204,
+      "loss": 1.8476,
+      "step": 26750
+    },
+    {
+      "epoch": 3.51890756302521,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00013048109965635738,
+      "loss": 1.9273,
+      "step": 26800
+    },
+    {
+      "epoch": 3.52547268907563,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.0001303489294210944,
+      "loss": 1.9214,
+      "step": 26850
+    },
+    {
+      "epoch": 3.5320378151260505,
+      "grad_norm": 2.765625,
+      "learning_rate": 0.00013021675918583136,
+      "loss": 1.9062,
+      "step": 26900
+    },
+    {
+      "epoch": 3.5386029411764706,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00013008458895056834,
+      "loss": 1.8992,
+      "step": 26950
+    },
+    {
+      "epoch": 3.5451680672268906,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.0001299524187153053,
+      "loss": 1.9827,
+      "step": 27000
+    },
+    {
+      "epoch": 3.551733193277311,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00012982024848004231,
+      "loss": 1.8935,
+      "step": 27050
+    },
+    {
+      "epoch": 3.558298319327731,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.0001296880782447793,
+      "loss": 1.9372,
+      "step": 27100
+    },
+    {
+      "epoch": 3.564863445378151,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00012955590800951627,
+      "loss": 1.918,
+      "step": 27150
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00012942373777425324,
+      "loss": 1.9758,
+      "step": 27200
+    },
+    {
+      "epoch": 3.5779936974789917,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00012929156753899024,
+      "loss": 1.9002,
+      "step": 27250
+    },
+    {
+      "epoch": 3.5845588235294117,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00012915939730372722,
+      "loss": 1.9045,
+      "step": 27300
+    },
+    {
+      "epoch": 3.591123949579832,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001290272270684642,
+      "loss": 1.9385,
+      "step": 27350
+    },
+    {
+      "epoch": 3.597689075630252,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00012889505683320117,
+      "loss": 1.9586,
+      "step": 27400
+    },
+    {
+      "epoch": 3.6042542016806722,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00012876288659793817,
+      "loss": 1.9139,
+      "step": 27450
+    },
+    {
+      "epoch": 3.6108193277310923,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00012863071636267515,
+      "loss": 1.949,
+      "step": 27500
+    },
+    {
+      "epoch": 3.6173844537815127,
+      "grad_norm": 1.8515625,
+      "learning_rate": 0.00012849854612741212,
+      "loss": 1.9769,
+      "step": 27550
+    },
+    {
+      "epoch": 3.6239495798319328,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001283663758921491,
+      "loss": 1.9476,
+      "step": 27600
+    },
+    {
+      "epoch": 3.630514705882353,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.0001282342056568861,
+      "loss": 1.9785,
+      "step": 27650
+    },
+    {
+      "epoch": 3.6370798319327733,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00012810203542162305,
+      "loss": 1.9009,
+      "step": 27700
+    },
+    {
+      "epoch": 3.6436449579831933,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012796986518636002,
+      "loss": 1.9727,
+      "step": 27750
+    },
+    {
+      "epoch": 3.6502100840336134,
+      "grad_norm": 2.375,
+      "learning_rate": 0.000127837694951097,
+      "loss": 1.9782,
+      "step": 27800
+    },
+    {
+      "epoch": 3.6567752100840334,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.000127705524715834,
+      "loss": 1.9606,
+      "step": 27850
+    },
+    {
+      "epoch": 3.663340336134454,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00012757335448057097,
+      "loss": 2.0027,
+      "step": 27900
+    },
+    {
+      "epoch": 3.669905462184874,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00012744118424530795,
+      "loss": 1.9761,
+      "step": 27950
+    },
+    {
+      "epoch": 3.6764705882352944,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00012730901401004493,
+      "loss": 1.9817,
+      "step": 28000
+    },
+    {
+      "epoch": 3.6830357142857144,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00012717684377478193,
+      "loss": 1.9666,
+      "step": 28050
+    },
+    {
+      "epoch": 3.6896008403361344,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.0001270446735395189,
+      "loss": 1.9876,
+      "step": 28100
+    },
+    {
+      "epoch": 3.6961659663865545,
+      "grad_norm": 1.71875,
+      "learning_rate": 0.00012691250330425588,
+      "loss": 1.9571,
+      "step": 28150
+    },
+    {
+      "epoch": 3.7027310924369745,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00012678033306899285,
+      "loss": 1.9859,
+      "step": 28200
+    },
+    {
+      "epoch": 3.709296218487395,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00012664816283372986,
+      "loss": 2.0244,
+      "step": 28250
+    },
+    {
+      "epoch": 3.715861344537815,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00012651599259846683,
+      "loss": 2.0215,
+      "step": 28300
+    },
+    {
+      "epoch": 3.7224264705882355,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.0001263838223632038,
+      "loss": 1.962,
+      "step": 28350
+    },
+    {
+      "epoch": 3.7289915966386555,
+      "grad_norm": 2.046875,
+      "learning_rate": 0.00012625165212794078,
+      "loss": 2.0002,
+      "step": 28400
+    },
+    {
+      "epoch": 3.7355567226890756,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00012611948189267778,
+      "loss": 1.962,
+      "step": 28450
+    },
+    {
+      "epoch": 3.7421218487394956,
+      "grad_norm": 1.4921875,
+      "learning_rate": 0.00012598731165741476,
+      "loss": 2.0032,
+      "step": 28500
+    },
+    {
+      "epoch": 3.748686974789916,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00012585514142215173,
+      "loss": 1.9378,
+      "step": 28550
+    },
+    {
+      "epoch": 3.755252100840336,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.0001257229711868887,
+      "loss": 1.9797,
+      "step": 28600
+    },
+    {
+      "epoch": 3.761817226890756,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.0001255908009516257,
+      "loss": 1.9942,
+      "step": 28650
+    },
+    {
+      "epoch": 3.7683823529411766,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00012545863071636269,
+      "loss": 2.0173,
+      "step": 28700
+    },
+    {
+      "epoch": 3.7749474789915967,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00012532646048109966,
+      "loss": 2.0285,
+      "step": 28750
+    },
+    {
+      "epoch": 3.7815126050420167,
+      "grad_norm": 1.59375,
+      "learning_rate": 0.00012519429024583664,
+      "loss": 2.0068,
+      "step": 28800
+    },
+    {
+      "epoch": 3.7880777310924367,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.0001250621200105736,
+      "loss": 1.9377,
+      "step": 28850
+    },
+    {
+      "epoch": 3.794642857142857,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00012492994977531061,
+      "loss": 1.9896,
+      "step": 28900
+    },
+    {
+      "epoch": 3.8012079831932772,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001247977795400476,
+      "loss": 1.9593,
+      "step": 28950
+    },
+    {
+      "epoch": 3.8077731092436977,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00012466560930478456,
+      "loss": 1.9841,
+      "step": 29000
+    },
+    {
+      "epoch": 3.8143382352941178,
+      "grad_norm": 1.5859375,
+      "learning_rate": 0.00012453343906952154,
+      "loss": 2.0124,
+      "step": 29050
+    },
+    {
+      "epoch": 3.820903361344538,
+      "grad_norm": 1.9140625,
+      "learning_rate": 0.00012440126883425854,
+      "loss": 1.9798,
+      "step": 29100
+    },
+    {
+      "epoch": 3.827468487394958,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00012426909859899552,
+      "loss": 1.9527,
+      "step": 29150
+    },
+    {
+      "epoch": 3.8340336134453783,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.0001241369283637325,
+      "loss": 2.0337,
+      "step": 29200
+    },
+    {
+      "epoch": 3.8405987394957983,
+      "grad_norm": 1.4609375,
+      "learning_rate": 0.00012400475812846947,
+      "loss": 2.0794,
+      "step": 29250
+    },
+    {
+      "epoch": 3.8471638655462184,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00012387258789320647,
+      "loss": 1.9727,
+      "step": 29300
+    },
+    {
+      "epoch": 3.853728991596639,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00012374041765794345,
+      "loss": 2.0058,
+      "step": 29350
+    },
+    {
+      "epoch": 3.860294117647059,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00012360824742268042,
+      "loss": 1.9934,
+      "step": 29400
+    },
+    {
+      "epoch": 3.866859243697479,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.0001234760771874174,
+      "loss": 2.0031,
+      "step": 29450
+    },
+    {
+      "epoch": 3.873424369747899,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.0001233439069521544,
+      "loss": 1.9638,
+      "step": 29500
+    },
+    {
+      "epoch": 3.8799894957983194,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00012321173671689137,
+      "loss": 2.0159,
+      "step": 29550
+    },
+    {
+      "epoch": 3.8865546218487395,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00012307956648162835,
+      "loss": 2.0246,
+      "step": 29600
+    },
+    {
+      "epoch": 3.89311974789916,
+      "grad_norm": 1.765625,
+      "learning_rate": 0.00012294739624636532,
+      "loss": 1.9922,
+      "step": 29650
+    },
+    {
+      "epoch": 3.89968487394958,
+      "grad_norm": 2.671875,
+      "learning_rate": 0.00012281522601110233,
+      "loss": 2.0464,
+      "step": 29700
+    },
+    {
+      "epoch": 3.90625,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.0001226830557758393,
+      "loss": 1.9939,
+      "step": 29750
+    },
+    {
+      "epoch": 3.91281512605042,
+      "grad_norm": 2.9375,
+      "learning_rate": 0.00012255088554057628,
+      "loss": 2.0142,
+      "step": 29800
+    },
+    {
+      "epoch": 3.91938025210084,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00012241871530531325,
+      "loss": 2.0387,
+      "step": 29850
+    },
+    {
+      "epoch": 3.9259453781512605,
+      "grad_norm": 1.609375,
+      "learning_rate": 0.00012228654507005025,
+      "loss": 2.0121,
+      "step": 29900
+    },
+    {
+      "epoch": 3.9325105042016806,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00012215437483478723,
+      "loss": 2.0636,
+      "step": 29950
+    },
+    {
+      "epoch": 3.939075630252101,
+      "grad_norm": 1.7578125,
+      "learning_rate": 0.0001220222045995242,
+      "loss": 2.006,
+      "step": 30000
+    },
+    {
+      "epoch": 3.945640756302521,
+      "grad_norm": 1.8125,
+      "learning_rate": 0.00012189003436426118,
+      "loss": 2.0666,
+      "step": 30050
+    },
+    {
+      "epoch": 3.952205882352941,
+      "grad_norm": 2.4375,
+      "learning_rate": 0.00012175786412899814,
+      "loss": 1.9889,
+      "step": 30100
+    },
+    {
+      "epoch": 3.958771008403361,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00012162569389373513,
+      "loss": 1.9606,
+      "step": 30150
+    },
+    {
+      "epoch": 3.9653361344537816,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.0001214935236584721,
+      "loss": 2.0165,
+      "step": 30200
+    },
+    {
+      "epoch": 3.9719012605042017,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.0001213613534232091,
+      "loss": 2.0325,
+      "step": 30250
+    },
+    {
+      "epoch": 3.9784663865546217,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00012122918318794607,
+      "loss": 2.0117,
+      "step": 30300
+    },
+    {
+      "epoch": 3.985031512605042,
+      "grad_norm": 1.828125,
+      "learning_rate": 0.00012109701295268306,
+      "loss": 1.9916,
+      "step": 30350
+    },
+    {
+      "epoch": 3.991596638655462,
+      "grad_norm": 1.65625,
+      "learning_rate": 0.00012096484271742003,
+      "loss": 2.0431,
+      "step": 30400
+    },
+    {
+      "epoch": 3.9981617647058822,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00012083267248215702,
+      "loss": 1.9578,
+      "step": 30450
+    },
+    {
+      "epoch": 4.004726890756302,
+      "grad_norm": 2.15625,
+      "learning_rate": 0.000120700502246894,
+      "loss": 1.7744,
+      "step": 30500
+    },
+    {
+      "epoch": 4.011292016806722,
+      "grad_norm": 1.8671875,
+      "learning_rate": 0.00012056833201163099,
+      "loss": 1.7033,
+      "step": 30550
+    },
+    {
+      "epoch": 4.017857142857143,
+      "grad_norm": 2.484375,
+      "learning_rate": 0.00012043616177636796,
+      "loss": 1.7349,
+      "step": 30600
+    },
+    {
+      "epoch": 4.024422268907563,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00012030399154110495,
+      "loss": 1.7361,
+      "step": 30650
+    },
+    {
+      "epoch": 4.030987394957983,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00012017182130584192,
+      "loss": 1.7135,
+      "step": 30700
+    },
+    {
+      "epoch": 4.037552521008403,
+      "grad_norm": 4.25,
+      "learning_rate": 0.00012003965107057891,
+      "loss": 1.7177,
+      "step": 30750
+    },
+    {
+      "epoch": 4.044117647058823,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00011990748083531589,
+      "loss": 1.6554,
+      "step": 30800
+    },
+    {
+      "epoch": 4.050682773109243,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00011977531060005286,
+      "loss": 1.7107,
+      "step": 30850
+    },
+    {
+      "epoch": 4.057247899159663,
+      "grad_norm": 3.03125,
+      "learning_rate": 0.00011964314036478985,
+      "loss": 1.7729,
+      "step": 30900
+    },
+    {
+      "epoch": 4.063813025210084,
+      "grad_norm": 1.7734375,
+      "learning_rate": 0.00011951097012952683,
+      "loss": 1.7601,
+      "step": 30950
+    },
+    {
+      "epoch": 4.070378151260504,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00011937879989426382,
+      "loss": 1.7264,
+      "step": 31000
+    },
+    {
+      "epoch": 4.076943277310924,
+      "grad_norm": 2.453125,
+      "learning_rate": 0.00011924662965900079,
+      "loss": 1.7868,
+      "step": 31050
+    },
+    {
+      "epoch": 4.0835084033613445,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00011911445942373778,
+      "loss": 1.72,
+      "step": 31100
+    },
+    {
+      "epoch": 4.0900735294117645,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00011898228918847476,
+      "loss": 1.791,
+      "step": 31150
+    },
+    {
+      "epoch": 4.0966386554621845,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00011885011895321174,
+      "loss": 1.7327,
+      "step": 31200
+    },
+    {
+      "epoch": 4.1032037815126055,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00011871794871794872,
+      "loss": 1.7162,
+      "step": 31250
+    },
+    {
+      "epoch": 4.1097689075630255,
+      "grad_norm": 1.9296875,
+      "learning_rate": 0.00011858577848268571,
+      "loss": 1.7865,
+      "step": 31300
+    },
+    {
+      "epoch": 4.1163340336134455,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00011845360824742268,
+      "loss": 1.794,
+      "step": 31350
+    },
+    {
+      "epoch": 4.1228991596638656,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00011832143801215967,
+      "loss": 1.7417,
+      "step": 31400
+    },
+    {
+      "epoch": 4.129464285714286,
+      "grad_norm": 2.828125,
+      "learning_rate": 0.00011818926777689665,
+      "loss": 1.7886,
+      "step": 31450
+    },
+    {
+      "epoch": 4.136029411764706,
+      "grad_norm": 4.0625,
+      "learning_rate": 0.00011805709754163364,
+      "loss": 1.8182,
+      "step": 31500
+    },
+    {
+      "epoch": 4.142594537815126,
+      "grad_norm": 1.6796875,
+      "learning_rate": 0.00011792492730637061,
+      "loss": 1.7963,
+      "step": 31550
+    },
+    {
+      "epoch": 4.149159663865547,
+      "grad_norm": 3.265625,
+      "learning_rate": 0.0001177927570711076,
+      "loss": 1.6697,
+      "step": 31600
+    },
+    {
+      "epoch": 4.155724789915967,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00011766058683584458,
+      "loss": 1.7375,
+      "step": 31650
+    },
+    {
+      "epoch": 4.162289915966387,
+      "grad_norm": 2.28125,
+      "learning_rate": 0.00011752841660058156,
+      "loss": 1.8777,
+      "step": 31700
+    },
+    {
+      "epoch": 4.168855042016807,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00011739624636531854,
+      "loss": 1.7834,
+      "step": 31750
+    },
+    {
+      "epoch": 4.175420168067227,
+      "grad_norm": 1.515625,
+      "learning_rate": 0.00011726407613005553,
+      "loss": 1.8656,
+      "step": 31800
+    },
+    {
+      "epoch": 4.181985294117647,
+      "grad_norm": 3.203125,
+      "learning_rate": 0.0001171319058947925,
+      "loss": 1.7457,
+      "step": 31850
+    },
+    {
+      "epoch": 4.188550420168067,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00011699973565952949,
+      "loss": 1.7545,
+      "step": 31900
+    },
+    {
+      "epoch": 4.195115546218488,
+      "grad_norm": 2.5,
+      "learning_rate": 0.00011686756542426647,
+      "loss": 1.7942,
+      "step": 31950
+    },
+    {
+      "epoch": 4.201680672268908,
+      "grad_norm": 1.7265625,
+      "learning_rate": 0.00011673539518900346,
+      "loss": 1.8379,
+      "step": 32000
+    },
+    {
+      "epoch": 4.208245798319328,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00011660322495374043,
+      "loss": 1.8702,
+      "step": 32050
+    },
+    {
+      "epoch": 4.214810924369748,
+      "grad_norm": 3.84375,
+      "learning_rate": 0.0001164710547184774,
+      "loss": 1.8278,
+      "step": 32100
+    },
+    {
+      "epoch": 4.221376050420168,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.0001163388844832144,
+      "loss": 1.7786,
+      "step": 32150
+    },
+    {
+      "epoch": 4.227941176470588,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00011620671424795137,
+      "loss": 1.8027,
+      "step": 32200
+    },
+    {
+      "epoch": 4.234506302521009,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00011607454401268836,
+      "loss": 1.8067,
+      "step": 32250
+    },
+    {
+      "epoch": 4.241071428571429,
+      "grad_norm": 2.515625,
+      "learning_rate": 0.00011594237377742533,
+      "loss": 1.7544,
+      "step": 32300
+    },
+    {
+      "epoch": 4.247636554621849,
+      "grad_norm": 3.703125,
+      "learning_rate": 0.00011581020354216232,
+      "loss": 1.7908,
+      "step": 32350
+    },
+    {
+      "epoch": 4.254201680672269,
+      "grad_norm": 2.6875,
+      "learning_rate": 0.0001156780333068993,
+      "loss": 1.8138,
+      "step": 32400
+    },
+    {
+      "epoch": 4.260766806722689,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00011554586307163626,
+      "loss": 1.8029,
+      "step": 32450
+    },
+    {
+      "epoch": 4.267331932773109,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00011541369283637325,
+      "loss": 1.7928,
+      "step": 32500
+    },
+    {
+      "epoch": 4.273897058823529,
+      "grad_norm": 3.0,
+      "learning_rate": 0.00011528152260111022,
+      "loss": 1.7972,
+      "step": 32550
+    },
+    {
+      "epoch": 4.28046218487395,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00011514935236584721,
+      "loss": 1.7975,
+      "step": 32600
+    },
+    {
+      "epoch": 4.28702731092437,
+      "grad_norm": 3.015625,
+      "learning_rate": 0.00011501718213058419,
+      "loss": 1.8258,
+      "step": 32650
+    },
+    {
+      "epoch": 4.29359243697479,
+      "grad_norm": 2.34375,
+      "learning_rate": 0.00011488501189532118,
+      "loss": 1.829,
+      "step": 32700
+    },
+    {
+      "epoch": 4.30015756302521,
+      "grad_norm": 2.53125,
+      "learning_rate": 0.00011475284166005815,
+      "loss": 1.8588,
+      "step": 32750
+    },
+    {
+      "epoch": 4.30672268907563,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00011462067142479514,
+      "loss": 1.8308,
+      "step": 32800
+    },
+    {
+      "epoch": 4.31328781512605,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00011448850118953212,
+      "loss": 1.7983,
+      "step": 32850
+    },
+    {
+      "epoch": 4.319852941176471,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.0001143563309542691,
+      "loss": 1.8035,
+      "step": 32900
+    },
+    {
+      "epoch": 4.326418067226891,
+      "grad_norm": 2.96875,
+      "learning_rate": 0.00011422416071900608,
+      "loss": 1.7907,
+      "step": 32950
+    },
+    {
+      "epoch": 4.332983193277311,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00011409199048374305,
+      "loss": 1.8111,
+      "step": 33000
+    },
+    {
+      "epoch": 4.339548319327731,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00011395982024848004,
+      "loss": 1.7474,
+      "step": 33050
+    },
+    {
+      "epoch": 4.346113445378151,
+      "grad_norm": 3.078125,
+      "learning_rate": 0.00011382765001321702,
+      "loss": 1.903,
+      "step": 33100
+    },
+    {
+      "epoch": 4.352678571428571,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00011369547977795401,
+      "loss": 1.8627,
+      "step": 33150
+    },
+    {
+      "epoch": 4.359243697478991,
+      "grad_norm": 2.109375,
+      "learning_rate": 0.00011356330954269098,
+      "loss": 1.8208,
+      "step": 33200
+    },
+    {
+      "epoch": 4.365808823529412,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00011343113930742797,
+      "loss": 1.8314,
+      "step": 33250
+    },
+    {
+      "epoch": 4.372373949579832,
+      "grad_norm": 1.8984375,
+      "learning_rate": 0.00011329896907216495,
+      "loss": 1.7919,
+      "step": 33300
+    },
+    {
+      "epoch": 4.378939075630252,
+      "grad_norm": 1.9453125,
+      "learning_rate": 0.00011316679883690194,
+      "loss": 1.8027,
+      "step": 33350
+    },
+    {
+      "epoch": 4.385504201680672,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00011303462860163891,
+      "loss": 1.8715,
+      "step": 33400
+    },
+    {
+      "epoch": 4.392069327731092,
+      "grad_norm": 3.796875,
+      "learning_rate": 0.0001129024583663759,
+      "loss": 1.8567,
+      "step": 33450
+    },
+    {
+      "epoch": 4.398634453781512,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00011277028813111287,
+      "loss": 1.8811,
+      "step": 33500
+    },
+    {
+      "epoch": 4.405199579831933,
+      "grad_norm": 2.1875,
+      "learning_rate": 0.00011263811789584986,
+      "loss": 1.8409,
+      "step": 33550
+    },
+    {
+      "epoch": 4.411764705882353,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00011250594766058684,
+      "loss": 1.9408,
+      "step": 33600
+    },
+    {
+      "epoch": 4.418329831932773,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00011237377742532383,
+      "loss": 1.7322,
+      "step": 33650
+    },
+    {
+      "epoch": 4.424894957983193,
+      "grad_norm": 3.3125,
+      "learning_rate": 0.0001122416071900608,
+      "loss": 1.8245,
+      "step": 33700
+    },
+    {
+      "epoch": 4.431460084033613,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00011210943695479779,
+      "loss": 1.8147,
+      "step": 33750
+    },
+    {
+      "epoch": 4.438025210084033,
+      "grad_norm": 3.375,
+      "learning_rate": 0.00011197726671953477,
+      "loss": 1.8319,
+      "step": 33800
+    },
+    {
+      "epoch": 4.444590336134453,
+      "grad_norm": 2.0,
+      "learning_rate": 0.00011184509648427175,
+      "loss": 1.7941,
+      "step": 33850
+    },
+    {
+      "epoch": 4.451155462184874,
+      "grad_norm": 4.1875,
+      "learning_rate": 0.00011171292624900873,
+      "loss": 1.8681,
+      "step": 33900
+    },
+    {
+      "epoch": 4.457720588235294,
+      "grad_norm": 4.03125,
+      "learning_rate": 0.00011158075601374572,
+      "loss": 1.8473,
+      "step": 33950
+    },
+    {
+      "epoch": 4.464285714285714,
+      "grad_norm": 1.9609375,
+      "learning_rate": 0.0001114485857784827,
+      "loss": 1.825,
+      "step": 34000
+    },
+    {
+      "epoch": 4.4708508403361344,
+      "grad_norm": 2.84375,
+      "learning_rate": 0.00011131641554321968,
+      "loss": 1.8312,
+      "step": 34050
+    },
+    {
+      "epoch": 4.4774159663865545,
+      "grad_norm": 2.796875,
+      "learning_rate": 0.00011118424530795666,
+      "loss": 1.8738,
+      "step": 34100
+    },
+    {
+      "epoch": 4.4839810924369745,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00011105207507269363,
+      "loss": 1.8026,
+      "step": 34150
+    },
+    {
+      "epoch": 4.4905462184873945,
+      "grad_norm": 1.890625,
+      "learning_rate": 0.00011091990483743062,
+      "loss": 1.812,
+      "step": 34200
+    },
+    {
+      "epoch": 4.4971113445378155,
+      "grad_norm": 2.125,
+      "learning_rate": 0.0001107877346021676,
+      "loss": 1.852,
+      "step": 34250
+    },
+    {
+      "epoch": 4.5036764705882355,
+      "grad_norm": 3.125,
+      "learning_rate": 0.00011065556436690459,
+      "loss": 1.8035,
+      "step": 34300
+    },
+    {
+      "epoch": 4.5102415966386555,
+      "grad_norm": 3.84375,
+      "learning_rate": 0.00011052339413164156,
+      "loss": 1.8066,
+      "step": 34350
+    },
+    {
+      "epoch": 4.516806722689076,
+      "grad_norm": 2.25,
+      "learning_rate": 0.00011039122389637855,
+      "loss": 1.8587,
+      "step": 34400
+    },
+    {
+      "epoch": 4.523371848739496,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00011025905366111552,
+      "loss": 1.8111,
+      "step": 34450
+    },
+    {
+      "epoch": 4.529936974789916,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00011012688342585251,
+      "loss": 1.8182,
+      "step": 34500
+    },
+    {
+      "epoch": 4.536502100840336,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00010999471319058949,
+      "loss": 1.8458,
+      "step": 34550
+    },
+    {
+      "epoch": 4.543067226890757,
+      "grad_norm": 2.015625,
+      "learning_rate": 0.00010986254295532648,
+      "loss": 1.8681,
+      "step": 34600
+    },
+    {
+      "epoch": 4.549632352941177,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00010973037272006345,
+      "loss": 1.8308,
+      "step": 34650
+    },
+    {
+      "epoch": 4.556197478991597,
+      "grad_norm": 1.9765625,
+      "learning_rate": 0.00010959820248480044,
+      "loss": 1.8915,
+      "step": 34700
+    },
+    {
+      "epoch": 4.562762605042017,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00010946603224953742,
+      "loss": 1.8739,
+      "step": 34750
+    },
+    {
+      "epoch": 4.569327731092437,
+      "grad_norm": 2.859375,
+      "learning_rate": 0.00010933386201427438,
+      "loss": 1.8601,
+      "step": 34800
+    },
+    {
+      "epoch": 4.575892857142857,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00010920169177901137,
+      "loss": 1.7981,
+      "step": 34850
+    },
+    {
+      "epoch": 4.582457983193278,
+      "grad_norm": 2.296875,
+      "learning_rate": 0.00010906952154374834,
+      "loss": 1.8601,
+      "step": 34900
+    },
+    {
+      "epoch": 4.589023109243698,
+      "grad_norm": 1.9921875,
+      "learning_rate": 0.00010893735130848533,
+      "loss": 1.892,
+      "step": 34950
+    },
+    {
+      "epoch": 4.595588235294118,
+      "grad_norm": 2.59375,
+      "learning_rate": 0.0001088051810732223,
+      "loss": 1.8538,
+      "step": 35000
+    },
+    {
+      "epoch": 4.602153361344538,
+      "grad_norm": 2.578125,
+      "learning_rate": 0.00010867301083795928,
+      "loss": 1.8857,
+      "step": 35050
+    },
+    {
+      "epoch": 4.608718487394958,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00010854084060269627,
+      "loss": 1.8731,
+      "step": 35100
+    },
+    {
+      "epoch": 4.615283613445378,
+      "grad_norm": 2.171875,
+      "learning_rate": 0.00010840867036743325,
+      "loss": 1.9334,
+      "step": 35150
+    },
+    {
+      "epoch": 4.621848739495798,
+      "grad_norm": 2.078125,
+      "learning_rate": 0.00010827650013217023,
+      "loss": 1.9154,
+      "step": 35200
+    },
+    {
+      "epoch": 4.628413865546219,
+      "grad_norm": 1.6875,
+      "learning_rate": 0.00010814432989690721,
+      "loss": 1.8874,
+      "step": 35250
+    },
+    {
+      "epoch": 4.634978991596639,
+      "grad_norm": 3.734375,
+      "learning_rate": 0.0001080121596616442,
+      "loss": 1.8365,
+      "step": 35300
+    },
+    {
+      "epoch": 4.641544117647059,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00010787998942638117,
+      "loss": 1.9075,
+      "step": 35350
+    },
+    {
+      "epoch": 4.648109243697479,
+      "grad_norm": 3.46875,
+      "learning_rate": 0.00010774781919111816,
+      "loss": 1.7622,
+      "step": 35400
+    },
+    {
+      "epoch": 4.654674369747899,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00010761564895585514,
+      "loss": 1.8269,
+      "step": 35450
+    },
+    {
+      "epoch": 4.661239495798319,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00010748347872059213,
+      "loss": 1.8732,
+      "step": 35500
+    },
+    {
+      "epoch": 4.66780462184874,
+      "grad_norm": 2.0625,
+      "learning_rate": 0.0001073513084853291,
+      "loss": 1.9295,
+      "step": 35550
+    },
+    {
+      "epoch": 4.67436974789916,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00010721913825006609,
+      "loss": 1.8487,
+      "step": 35600
+    },
+    {
+      "epoch": 4.68093487394958,
+      "grad_norm": 2.953125,
+      "learning_rate": 0.00010708696801480307,
+      "loss": 1.8199,
+      "step": 35650
+    },
+    {
+      "epoch": 4.6875,
+      "grad_norm": 1.921875,
+      "learning_rate": 0.00010695479777954005,
+      "loss": 1.8303,
+      "step": 35700
+    },
+    {
+      "epoch": 4.69406512605042,
+      "grad_norm": 1.7890625,
+      "learning_rate": 0.00010682262754427703,
+      "loss": 1.8325,
+      "step": 35750
+    },
+    {
+      "epoch": 4.70063025210084,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00010669045730901402,
+      "loss": 1.9103,
+      "step": 35800
+    },
+    {
+      "epoch": 4.70719537815126,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00010655828707375099,
+      "loss": 1.9129,
+      "step": 35850
+    },
+    {
+      "epoch": 4.713760504201681,
+      "grad_norm": 2.609375,
+      "learning_rate": 0.00010642611683848798,
+      "loss": 1.8434,
+      "step": 35900
+    },
+    {
+      "epoch": 4.720325630252101,
+      "grad_norm": 3.234375,
+      "learning_rate": 0.00010629394660322496,
+      "loss": 1.8424,
+      "step": 35950
+    },
+    {
+      "epoch": 4.726890756302521,
+      "grad_norm": 2.421875,
+      "learning_rate": 0.00010616177636796195,
+      "loss": 1.8432,
+      "step": 36000
+    },
+    {
+      "epoch": 4.733455882352941,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00010602960613269892,
+      "loss": 1.9235,
+      "step": 36050
+    },
+    {
+      "epoch": 4.740021008403361,
+      "grad_norm": 2.71875,
+      "learning_rate": 0.00010589743589743591,
+      "loss": 1.9234,
+      "step": 36100
+    },
+    {
+      "epoch": 4.746586134453781,
+      "grad_norm": 2.90625,
+      "learning_rate": 0.00010576526566217288,
+      "loss": 1.8803,
+      "step": 36150
+    },
+    {
+      "epoch": 4.753151260504202,
+      "grad_norm": 2.40625,
+      "learning_rate": 0.00010563309542690987,
+      "loss": 1.8847,
+      "step": 36200
+    },
+    {
+      "epoch": 4.759716386554622,
+      "grad_norm": 2.65625,
+      "learning_rate": 0.00010550092519164685,
+      "loss": 1.9664,
+      "step": 36250
+    },
+    {
+      "epoch": 4.766281512605042,
+      "grad_norm": 1.53125,
+      "learning_rate": 0.00010536875495638382,
+      "loss": 1.9284,
+      "step": 36300
+    },
+    {
+      "epoch": 4.772846638655462,
+      "grad_norm": 2.03125,
+      "learning_rate": 0.00010523658472112081,
+      "loss": 1.8572,
+      "step": 36350
+    },
+    {
+      "epoch": 4.779411764705882,
+      "grad_norm": 2.78125,
+      "learning_rate": 0.00010510441448585779,
+      "loss": 1.8823,
+      "step": 36400
+    },
+    {
+      "epoch": 4.785976890756302,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00010497224425059478,
+      "loss": 1.908,
+      "step": 36450
+    },
+    {
+      "epoch": 4.792542016806722,
+      "grad_norm": 2.09375,
+      "learning_rate": 0.00010484007401533175,
+      "loss": 1.8155,
+      "step": 36500
+    },
+    {
+      "epoch": 4.799107142857143,
+      "grad_norm": 4.4375,
+      "learning_rate": 0.00010470790378006874,
+      "loss": 1.8574,
+      "step": 36550
+    },
+    {
+      "epoch": 4.805672268907563,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00010457573354480572,
+      "loss": 1.8486,
+      "step": 36600
+    },
+    {
+      "epoch": 4.812237394957983,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.0001044435633095427,
+      "loss": 1.8647,
+      "step": 36650
+    },
+    {
+      "epoch": 4.818802521008403,
+      "grad_norm": 1.96875,
+      "learning_rate": 0.00010431139307427968,
+      "loss": 1.9243,
+      "step": 36700
+    },
+    {
+      "epoch": 4.825367647058823,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00010417922283901667,
+      "loss": 1.9146,
+      "step": 36750
+    },
+    {
+      "epoch": 4.831932773109243,
+      "grad_norm": 2.359375,
+      "learning_rate": 0.00010404705260375364,
+      "loss": 1.8984,
+      "step": 36800
+    },
+    {
+      "epoch": 4.838497899159664,
+      "grad_norm": 2.21875,
+      "learning_rate": 0.00010391488236849063,
+      "loss": 1.8802,
+      "step": 36850
+    },
+    {
+      "epoch": 4.845063025210084,
+      "grad_norm": 2.125,
+      "learning_rate": 0.00010378271213322761,
+      "loss": 1.8742,
+      "step": 36900
+    },
+    {
+      "epoch": 4.851628151260504,
+      "grad_norm": 2.25,
+      "learning_rate": 0.0001036505418979646,
+      "loss": 1.884,
+      "step": 36950
+    },
+    {
+      "epoch": 4.858193277310924,
+      "grad_norm": 1.8359375,
+      "learning_rate": 0.00010351837166270157,
+      "loss": 1.9064,
+      "step": 37000
+    },
+    {
+      "epoch": 4.8647584033613445,
+      "grad_norm": 2.234375,
+      "learning_rate": 0.00010338620142743856,
+      "loss": 1.8883,
+      "step": 37050
+    },
+    {
+      "epoch": 4.8713235294117645,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00010325403119217554,
+      "loss": 1.8774,
+      "step": 37100
+    },
+    {
+      "epoch": 4.8778886554621845,
+      "grad_norm": 2.625,
+      "learning_rate": 0.0001031218609569125,
+      "loss": 1.9345,
+      "step": 37150
+    },
+    {
+      "epoch": 4.884453781512605,
+      "grad_norm": 2.890625,
+      "learning_rate": 0.00010298969072164947,
+      "loss": 1.8626,
+      "step": 37200
+    },
+    {
+      "epoch": 4.8910189075630255,
+      "grad_norm": 2.203125,
+      "learning_rate": 0.00010285752048638646,
+      "loss": 1.9422,
+      "step": 37250
+    },
+    {
+      "epoch": 4.8975840336134455,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00010272535025112344,
+      "loss": 1.8108,
+      "step": 37300
+    },
+    {
+      "epoch": 4.9041491596638656,
+      "grad_norm": 1.8203125,
+      "learning_rate": 0.00010259318001586043,
+      "loss": 1.8922,
+      "step": 37350
+    },
+    {
+      "epoch": 4.910714285714286,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.0001024610097805974,
+      "loss": 1.8552,
+      "step": 37400
+    },
+    {
+      "epoch": 4.917279411764706,
+      "grad_norm": 1.5390625,
+      "learning_rate": 0.00010232883954533439,
+      "loss": 1.8918,
+      "step": 37450
+    },
+    {
+      "epoch": 4.9238445378151265,
+      "grad_norm": 1.953125,
+      "learning_rate": 0.00010219666931007136,
+      "loss": 1.9388,
+      "step": 37500
+    },
+    {
+      "epoch": 4.930409663865547,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00010206449907480835,
+      "loss": 1.9304,
+      "step": 37550
+    },
+    {
+      "epoch": 4.936974789915967,
+      "grad_norm": 2.734375,
+      "learning_rate": 0.00010193232883954533,
+      "loss": 1.8836,
+      "step": 37600
+    },
+    {
+      "epoch": 4.943539915966387,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00010180015860428232,
+      "loss": 1.849,
+      "step": 37650
+    },
+    {
+      "epoch": 4.950105042016807,
+      "grad_norm": 2.328125,
+      "learning_rate": 0.00010166798836901929,
+      "loss": 1.8299,
+      "step": 37700
+    },
+    {
+      "epoch": 4.956670168067227,
+      "grad_norm": 2.546875,
+      "learning_rate": 0.00010153581813375628,
+      "loss": 1.8335,
+      "step": 37750
+    },
+    {
+      "epoch": 4.963235294117647,
+      "grad_norm": 2.625,
+      "learning_rate": 0.00010140364789849326,
+      "loss": 1.8908,
+      "step": 37800
+    },
+    {
+      "epoch": 4.969800420168067,
+      "grad_norm": 3.046875,
+      "learning_rate": 0.00010127147766323024,
+      "loss": 1.8803,
+      "step": 37850
+    },
+    {
+      "epoch": 4.976365546218488,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00010113930742796722,
+      "loss": 1.8904,
+      "step": 37900
+    },
+    {
+      "epoch": 4.982930672268908,
+      "grad_norm": 1.796875,
+      "learning_rate": 0.00010100713719270421,
+      "loss": 1.899,
+      "step": 37950
+    },
+    {
+      "epoch": 4.989495798319328,
+      "grad_norm": 2.265625,
+      "learning_rate": 0.00010087496695744118,
+      "loss": 1.8902,
+      "step": 38000
+    },
+    {
+      "epoch": 4.996060924369748,
+      "grad_norm": 2.140625,
+      "learning_rate": 0.00010074279672217817,
+      "loss": 1.9128,
+      "step": 38050
+    },
+    {
+      "epoch": 5.002626050420168,
+      "grad_norm": 3.109375,
+      "learning_rate": 0.00010061062648691515,
+      "loss": 1.6909,
+      "step": 38100
+    },
+    {
+      "epoch": 5.009191176470588,
+      "grad_norm": 2.5625,
+      "learning_rate": 0.00010047845625165214,
+      "loss": 1.3461,
+      "step": 38150
+    },
+    {
+      "epoch": 5.015756302521009,
+      "grad_norm": 3.46875,
+      "learning_rate": 0.00010034628601638911,
+      "loss": 1.4187,
+      "step": 38200
+    },
+    {
+      "epoch": 5.022321428571429,
+      "grad_norm": 2.375,
+      "learning_rate": 0.0001002141157811261,
+      "loss": 1.3511,
+      "step": 38250
+    },
+    {
+      "epoch": 5.028886554621849,
+      "grad_norm": 2.921875,
+      "learning_rate": 0.00010008194554586308,
+      "loss": 1.336,
+      "step": 38300
+    },
+    {
+      "epoch": 5.035451680672269,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.994977531060006e-05,
+      "loss": 1.4256,
+      "step": 38350
+    },
+    {
+      "epoch": 5.042016806722689,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.981760507533704e-05,
+      "loss": 1.3989,
+      "step": 38400
+    },
+    {
+      "epoch": 5.048581932773109,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.968543484007401e-05,
+      "loss": 1.3702,
+      "step": 38450
+    },
+    {
+      "epoch": 5.055147058823529,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.9553264604811e-05,
+      "loss": 1.3758,
+      "step": 38500
+    },
+    {
+      "epoch": 5.06171218487395,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.942109436954798e-05,
+      "loss": 1.4057,
+      "step": 38550
+    },
+    {
+      "epoch": 5.06827731092437,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.928892413428497e-05,
+      "loss": 1.4083,
+      "step": 38600
+    },
+    {
+      "epoch": 5.07484243697479,
+      "grad_norm": 2.25,
+      "learning_rate": 9.915675389902194e-05,
+      "loss": 1.3849,
+      "step": 38650
+    },
+    {
+      "epoch": 5.08140756302521,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.902458366375893e-05,
+      "loss": 1.3812,
+      "step": 38700
+    },
+    {
+      "epoch": 5.08797268907563,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.889241342849591e-05,
+      "loss": 1.4723,
+      "step": 38750
+    },
+    {
+      "epoch": 5.09453781512605,
+      "grad_norm": 2.75,
+      "learning_rate": 9.87602431932329e-05,
+      "loss": 1.3678,
+      "step": 38800
+    },
+    {
+      "epoch": 5.101102941176471,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.862807295796987e-05,
+      "loss": 1.4548,
+      "step": 38850
+    },
+    {
+      "epoch": 5.107668067226891,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.849590272270686e-05,
+      "loss": 1.4095,
+      "step": 38900
+    },
+    {
+      "epoch": 5.114233193277311,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.836373248744382e-05,
+      "loss": 1.3976,
+      "step": 38950
+    },
+    {
+      "epoch": 5.120798319327731,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.823156225218081e-05,
+      "loss": 1.4269,
+      "step": 39000
+    },
+    {
+      "epoch": 5.127363445378151,
+      "grad_norm": 2.625,
+      "learning_rate": 9.809939201691779e-05,
+      "loss": 1.4049,
+      "step": 39050
+    },
+    {
+      "epoch": 5.133928571428571,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.796722178165477e-05,
+      "loss": 1.3801,
+      "step": 39100
+    },
+    {
+      "epoch": 5.140493697478991,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.783505154639175e-05,
+      "loss": 1.4323,
+      "step": 39150
+    },
+    {
+      "epoch": 5.147058823529412,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.770288131112874e-05,
+      "loss": 1.4407,
+      "step": 39200
+    },
+    {
+      "epoch": 5.153623949579832,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.757071107586571e-05,
+      "loss": 1.4843,
+      "step": 39250
+    },
+    {
+      "epoch": 5.160189075630252,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.74385408406027e-05,
+      "loss": 1.4161,
+      "step": 39300
+    },
+    {
+      "epoch": 5.166754201680672,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.730637060533968e-05,
+      "loss": 1.426,
+      "step": 39350
+    },
+    {
+      "epoch": 5.173319327731092,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.717420037007667e-05,
+      "loss": 1.4223,
+      "step": 39400
+    },
+    {
+      "epoch": 5.179884453781512,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.704203013481364e-05,
+      "loss": 1.4874,
+      "step": 39450
+    },
+    {
+      "epoch": 5.186449579831933,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.690985989955063e-05,
+      "loss": 1.4137,
+      "step": 39500
+    },
+    {
+      "epoch": 5.193014705882353,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.67776896642876e-05,
+      "loss": 1.4493,
+      "step": 39550
+    },
+    {
+      "epoch": 5.199579831932773,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.66455194290246e-05,
+      "loss": 1.4104,
+      "step": 39600
+    },
+    {
+      "epoch": 5.206144957983193,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.651334919376157e-05,
+      "loss": 1.4584,
+      "step": 39650
+    },
+    {
+      "epoch": 5.212710084033613,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.638117895849856e-05,
+      "loss": 1.4498,
+      "step": 39700
+    },
+    {
+      "epoch": 5.219275210084033,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.624900872323553e-05,
+      "loss": 1.4353,
+      "step": 39750
+    },
+    {
+      "epoch": 5.225840336134453,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.611683848797252e-05,
+      "loss": 1.3824,
+      "step": 39800
+    },
+    {
+      "epoch": 5.232405462184874,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.59846682527095e-05,
+      "loss": 1.4526,
+      "step": 39850
+    },
+    {
+      "epoch": 5.238970588235294,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.585249801744649e-05,
+      "loss": 1.4452,
+      "step": 39900
+    },
+    {
+      "epoch": 5.245535714285714,
+      "grad_norm": 2.375,
+      "learning_rate": 9.572032778218346e-05,
+      "loss": 1.4672,
+      "step": 39950
+    },
+    {
+      "epoch": 5.2521008403361344,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.558815754692045e-05,
+      "loss": 1.455,
+      "step": 40000
+    },
+    {
+      "epoch": 5.2586659663865545,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.545598731165742e-05,
+      "loss": 1.5045,
+      "step": 40050
+    },
+    {
+      "epoch": 5.2652310924369745,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.53238170763944e-05,
+      "loss": 1.4684,
+      "step": 40100
+    },
+    {
+      "epoch": 5.2717962184873945,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.519164684113137e-05,
+      "loss": 1.3692,
+      "step": 40150
+    },
+    {
+      "epoch": 5.2783613445378155,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.505947660586836e-05,
+      "loss": 1.3653,
+      "step": 40200
+    },
+    {
+      "epoch": 5.2849264705882355,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.492730637060534e-05,
+      "loss": 1.4225,
+      "step": 40250
+    },
+    {
+      "epoch": 5.2914915966386555,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.479513613534233e-05,
+      "loss": 1.416,
+      "step": 40300
+    },
+    {
+      "epoch": 5.298056722689076,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.46629659000793e-05,
+      "loss": 1.4618,
+      "step": 40350
+    },
+    {
+      "epoch": 5.304621848739496,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.453079566481629e-05,
+      "loss": 1.4527,
+      "step": 40400
+    },
+    {
+      "epoch": 5.311186974789916,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.439862542955327e-05,
+      "loss": 1.4499,
+      "step": 40450
+    },
+    {
+      "epoch": 5.317752100840336,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.426645519429024e-05,
+      "loss": 1.4557,
+      "step": 40500
+    },
+    {
+      "epoch": 5.324317226890757,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.413428495902723e-05,
+      "loss": 1.4044,
+      "step": 40550
+    },
+    {
+      "epoch": 5.330882352941177,
+      "grad_norm": 2.375,
+      "learning_rate": 9.40021147237642e-05,
+      "loss": 1.4238,
+      "step": 40600
+    },
+    {
+      "epoch": 5.337447478991597,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.38699444885012e-05,
+      "loss": 1.5172,
+      "step": 40650
+    },
+    {
+      "epoch": 5.344012605042017,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.373777425323817e-05,
+      "loss": 1.4602,
+      "step": 40700
+    },
+    {
+      "epoch": 5.350577731092437,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.360560401797516e-05,
+      "loss": 1.435,
+      "step": 40750
+    },
+    {
+      "epoch": 5.357142857142857,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.347343378271213e-05,
+      "loss": 1.4009,
+      "step": 40800
+    },
+    {
+      "epoch": 5.363707983193278,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.334126354744912e-05,
+      "loss": 1.5033,
+      "step": 40850
+    },
+    {
+      "epoch": 5.370273109243698,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.32090933121861e-05,
+      "loss": 1.5243,
+      "step": 40900
+    },
+    {
+      "epoch": 5.376838235294118,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.307692307692309e-05,
+      "loss": 1.512,
+      "step": 40950
+    },
+    {
+      "epoch": 5.383403361344538,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.294475284166006e-05,
+      "loss": 1.4817,
+      "step": 41000
+    },
+    {
+      "epoch": 5.389968487394958,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.281258260639705e-05,
+      "loss": 1.4368,
+      "step": 41050
+    },
+    {
+      "epoch": 5.396533613445378,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.268041237113403e-05,
+      "loss": 1.4665,
+      "step": 41100
+    },
+    {
+      "epoch": 5.403098739495798,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.254824213587101e-05,
+      "loss": 1.5117,
+      "step": 41150
+    },
+    {
+      "epoch": 5.409663865546219,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.241607190060799e-05,
+      "loss": 1.4644,
+      "step": 41200
+    },
+    {
+      "epoch": 5.416228991596639,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.228390166534498e-05,
+      "loss": 1.461,
+      "step": 41250
+    },
+    {
+      "epoch": 5.422794117647059,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.215173143008194e-05,
+      "loss": 1.4493,
+      "step": 41300
+    },
+    {
+      "epoch": 5.429359243697479,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.201956119481893e-05,
+      "loss": 1.49,
+      "step": 41350
+    },
+    {
+      "epoch": 5.435924369747899,
+      "grad_norm": 5.625,
+      "learning_rate": 9.18873909595559e-05,
+      "loss": 1.4682,
+      "step": 41400
+    },
+    {
+      "epoch": 5.442489495798319,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.175522072429289e-05,
+      "loss": 1.4381,
+      "step": 41450
+    },
+    {
+      "epoch": 5.44905462184874,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.162305048902987e-05,
+      "loss": 1.4631,
+      "step": 41500
+    },
+    {
+      "epoch": 5.45561974789916,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.149088025376686e-05,
+      "loss": 1.4942,
+      "step": 41550
+    },
+    {
+      "epoch": 5.46218487394958,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.135871001850383e-05,
+      "loss": 1.4826,
+      "step": 41600
+    },
+    {
+      "epoch": 5.46875,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.122653978324082e-05,
+      "loss": 1.5446,
+      "step": 41650
+    },
+    {
+      "epoch": 5.47531512605042,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.10943695479778e-05,
+      "loss": 1.5059,
+      "step": 41700
+    },
+    {
+      "epoch": 5.48188025210084,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.096219931271478e-05,
+      "loss": 1.5444,
+      "step": 41750
+    },
+    {
+      "epoch": 5.48844537815126,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.083002907745176e-05,
+      "loss": 1.5206,
+      "step": 41800
+    },
+    {
+      "epoch": 5.495010504201681,
+      "grad_norm": 3.25,
+      "learning_rate": 9.069785884218875e-05,
+      "loss": 1.4863,
+      "step": 41850
+    },
+    {
+      "epoch": 5.501575630252101,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.056568860692572e-05,
+      "loss": 1.4104,
+      "step": 41900
+    },
+    {
+      "epoch": 5.508140756302521,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.043351837166271e-05,
+      "loss": 1.5224,
+      "step": 41950
+    },
+    {
+      "epoch": 5.514705882352941,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.030134813639969e-05,
+      "loss": 1.5038,
+      "step": 42000
+    },
+    {
+      "epoch": 5.521271008403361,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.016917790113668e-05,
+      "loss": 1.4301,
+      "step": 42050
+    },
+    {
+      "epoch": 5.527836134453781,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.003700766587365e-05,
+      "loss": 1.5132,
+      "step": 42100
+    },
+    {
+      "epoch": 5.534401260504202,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.990483743061063e-05,
+      "loss": 1.5198,
+      "step": 42150
+    },
+    {
+      "epoch": 5.540966386554622,
+      "grad_norm": 3.125,
+      "learning_rate": 8.977266719534762e-05,
+      "loss": 1.4996,
+      "step": 42200
+    },
+    {
+      "epoch": 5.547531512605042,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.964049696008459e-05,
+      "loss": 1.535,
+      "step": 42250
+    },
+    {
+      "epoch": 5.554096638655462,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.950832672482158e-05,
+      "loss": 1.4931,
+      "step": 42300
+    },
+    {
+      "epoch": 5.560661764705882,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.937615648955855e-05,
+      "loss": 1.4804,
+      "step": 42350
+    },
+    {
+      "epoch": 5.567226890756302,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.924398625429554e-05,
+      "loss": 1.5138,
+      "step": 42400
+    },
+    {
+      "epoch": 5.573792016806722,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.911181601903252e-05,
+      "loss": 1.498,
+      "step": 42450
+    },
+    {
+      "epoch": 5.580357142857143,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.89796457837695e-05,
+      "loss": 1.524,
+      "step": 42500
+    },
+    {
+      "epoch": 5.586922268907563,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.884747554850648e-05,
+      "loss": 1.4919,
+      "step": 42550
+    },
+    {
+      "epoch": 5.593487394957983,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.871530531324346e-05,
+      "loss": 1.4701,
+      "step": 42600
+    },
+    {
+      "epoch": 5.600052521008403,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.858313507798043e-05,
+      "loss": 1.5287,
+      "step": 42650
+    },
+    {
+      "epoch": 5.606617647058823,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.845096484271742e-05,
+      "loss": 1.4986,
+      "step": 42700
+    },
+    {
+      "epoch": 5.613182773109243,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.83187946074544e-05,
+      "loss": 1.604,
+      "step": 42750
+    },
+    {
+      "epoch": 5.619747899159664,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.818662437219139e-05,
+      "loss": 1.5137,
+      "step": 42800
+    },
+    {
+      "epoch": 5.626313025210084,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.805445413692836e-05,
+      "loss": 1.5157,
+      "step": 42850
+    },
+    {
+      "epoch": 5.632878151260504,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.792228390166535e-05,
+      "loss": 1.5347,
+      "step": 42900
+    },
+    {
+      "epoch": 5.639443277310924,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.779011366640232e-05,
+      "loss": 1.4842,
+      "step": 42950
+    },
+    {
+      "epoch": 5.6460084033613445,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.765794343113931e-05,
+      "loss": 1.4893,
+      "step": 43000
+    },
+    {
+      "epoch": 5.6525735294117645,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.752577319587629e-05,
+      "loss": 1.5574,
+      "step": 43050
+    },
+    {
+      "epoch": 5.6591386554621845,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.739360296061328e-05,
+      "loss": 1.5025,
+      "step": 43100
+    },
+    {
+      "epoch": 5.665703781512605,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.726143272535025e-05,
+      "loss": 1.5131,
+      "step": 43150
+    },
+    {
+      "epoch": 5.6722689075630255,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.712926249008724e-05,
+      "loss": 1.5278,
+      "step": 43200
+    },
+    {
+      "epoch": 5.6788340336134455,
+      "grad_norm": 4.1875,
+      "learning_rate": 8.699709225482422e-05,
+      "loss": 1.5521,
+      "step": 43250
+    },
+    {
+      "epoch": 5.6853991596638656,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.68649220195612e-05,
+      "loss": 1.4944,
+      "step": 43300
+    },
+    {
+      "epoch": 5.691964285714286,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.673275178429818e-05,
+      "loss": 1.5561,
+      "step": 43350
+    },
+    {
+      "epoch": 5.698529411764706,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.660058154903517e-05,
+      "loss": 1.4284,
+      "step": 43400
+    },
+    {
+      "epoch": 5.7050945378151265,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.646841131377214e-05,
+      "loss": 1.5662,
+      "step": 43450
+    },
+    {
+      "epoch": 5.711659663865547,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.633624107850913e-05,
+      "loss": 1.5686,
+      "step": 43500
+    },
+    {
+      "epoch": 5.718224789915967,
+      "grad_norm": 4.0,
+      "learning_rate": 8.620407084324611e-05,
+      "loss": 1.5144,
+      "step": 43550
+    },
+    {
+      "epoch": 5.724789915966387,
+      "grad_norm": 3.875,
+      "learning_rate": 8.60719006079831e-05,
+      "loss": 1.5455,
+      "step": 43600
+    },
+    {
+      "epoch": 5.731355042016807,
+      "grad_norm": 3.0,
+      "learning_rate": 8.593973037272007e-05,
+      "loss": 1.5595,
+      "step": 43650
+    },
+    {
+      "epoch": 5.737920168067227,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.580756013745705e-05,
+      "loss": 1.5109,
+      "step": 43700
+    },
+    {
+      "epoch": 5.744485294117647,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.567538990219402e-05,
+      "loss": 1.4854,
+      "step": 43750
+    },
+    {
+      "epoch": 5.751050420168067,
+      "grad_norm": 3.96875,
+      "learning_rate": 8.554321966693101e-05,
+      "loss": 1.5283,
+      "step": 43800
+    },
+    {
+      "epoch": 5.757615546218488,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.541104943166799e-05,
+      "loss": 1.5464,
+      "step": 43850
+    },
+    {
+      "epoch": 5.764180672268908,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.527887919640498e-05,
+      "loss": 1.5518,
+      "step": 43900
+    },
+    {
+      "epoch": 5.770745798319328,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.514670896114195e-05,
+      "loss": 1.5689,
+      "step": 43950
+    },
+    {
+      "epoch": 5.777310924369748,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.501453872587894e-05,
+      "loss": 1.5002,
+      "step": 44000
+    },
+    {
+      "epoch": 5.783876050420168,
+      "grad_norm": 2.75,
+      "learning_rate": 8.488236849061591e-05,
+      "loss": 1.5407,
+      "step": 44050
+    },
+    {
+      "epoch": 5.790441176470588,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.47501982553529e-05,
+      "loss": 1.5339,
+      "step": 44100
+    },
+    {
+      "epoch": 5.797006302521009,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.461802802008988e-05,
+      "loss": 1.5005,
+      "step": 44150
+    },
+    {
+      "epoch": 5.803571428571429,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.448585778482687e-05,
+      "loss": 1.506,
+      "step": 44200
+    },
+    {
+      "epoch": 5.810136554621849,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.435368754956384e-05,
+      "loss": 1.5579,
+      "step": 44250
+    },
+    {
+      "epoch": 5.816701680672269,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.422151731430082e-05,
+      "loss": 1.5815,
+      "step": 44300
+    },
+    {
+      "epoch": 5.823266806722689,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.40893470790378e-05,
+      "loss": 1.5633,
+      "step": 44350
+    },
+    {
+      "epoch": 5.829831932773109,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.395717684377478e-05,
+      "loss": 1.5406,
+      "step": 44400
+    },
+    {
+      "epoch": 5.836397058823529,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.382500660851177e-05,
+      "loss": 1.5523,
+      "step": 44450
+    },
+    {
+      "epoch": 5.84296218487395,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.369283637324875e-05,
+      "loss": 1.534,
+      "step": 44500
+    },
+    {
+      "epoch": 5.84952731092437,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.356066613798573e-05,
+      "loss": 1.5479,
+      "step": 44550
+    },
+    {
+      "epoch": 5.85609243697479,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.342849590272271e-05,
+      "loss": 1.5374,
+      "step": 44600
+    },
+    {
+      "epoch": 5.86265756302521,
+      "grad_norm": 3.25,
+      "learning_rate": 8.32963256674597e-05,
+      "loss": 1.5043,
+      "step": 44650
+    },
+    {
+      "epoch": 5.86922268907563,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.316415543219667e-05,
+      "loss": 1.5133,
+      "step": 44700
+    },
+    {
+      "epoch": 5.87578781512605,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.303198519693366e-05,
+      "loss": 1.5399,
+      "step": 44750
+    },
+    {
+      "epoch": 5.882352941176471,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.289981496167064e-05,
+      "loss": 1.5391,
+      "step": 44800
+    },
+    {
+      "epoch": 5.888918067226891,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.276764472640761e-05,
+      "loss": 1.5306,
+      "step": 44850
+    },
+    {
+      "epoch": 5.895483193277311,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.263547449114459e-05,
+      "loss": 1.5839,
+      "step": 44900
+    },
+    {
+      "epoch": 5.902048319327731,
+      "grad_norm": 3.421875,
+      "learning_rate": 8.250330425588158e-05,
+      "loss": 1.4878,
+      "step": 44950
+    },
+    {
+      "epoch": 5.908613445378151,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.237113402061855e-05,
+      "loss": 1.5699,
+      "step": 45000
+    },
+    {
+      "epoch": 5.915178571428571,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.223896378535554e-05,
+      "loss": 1.561,
+      "step": 45050
+    },
+    {
+      "epoch": 5.921743697478991,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.210679355009252e-05,
+      "loss": 1.5284,
+      "step": 45100
+    },
+    {
+      "epoch": 5.928308823529412,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.19746233148295e-05,
+      "loss": 1.5605,
+      "step": 45150
+    },
+    {
+      "epoch": 5.934873949579832,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.184245307956648e-05,
+      "loss": 1.5675,
+      "step": 45200
+    },
+    {
+      "epoch": 5.941439075630252,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.171028284430347e-05,
+      "loss": 1.5244,
+      "step": 45250
+    },
+    {
+      "epoch": 5.948004201680672,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.157811260904044e-05,
+      "loss": 1.5339,
+      "step": 45300
+    },
+    {
+      "epoch": 5.954569327731092,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.144594237377743e-05,
+      "loss": 1.5613,
+      "step": 45350
+    },
+    {
+      "epoch": 5.961134453781512,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.131377213851441e-05,
+      "loss": 1.5348,
+      "step": 45400
+    },
+    {
+      "epoch": 5.967699579831933,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.11816019032514e-05,
+      "loss": 1.5699,
+      "step": 45450
+    },
+    {
+      "epoch": 5.974264705882353,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.104943166798837e-05,
+      "loss": 1.5268,
+      "step": 45500
+    },
+    {
+      "epoch": 5.980829831932773,
+      "grad_norm": 3.828125,
+      "learning_rate": 8.091726143272536e-05,
+      "loss": 1.5175,
+      "step": 45550
+    },
+    {
+      "epoch": 5.987394957983193,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.078509119746234e-05,
+      "loss": 1.5399,
+      "step": 45600
+    },
+    {
+      "epoch": 5.993960084033613,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.065292096219932e-05,
+      "loss": 1.5178,
+      "step": 45650
+    },
+    {
+      "epoch": 6.000525210084033,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.05207507269363e-05,
+      "loss": 1.5124,
+      "step": 45700
+    },
+    {
+      "epoch": 6.007090336134453,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.038858049167329e-05,
+      "loss": 1.1681,
+      "step": 45750
+    },
+    {
+      "epoch": 6.013655462184874,
+      "grad_norm": 2.25,
+      "learning_rate": 8.025641025641026e-05,
+      "loss": 1.1803,
+      "step": 45800
+    },
+    {
+      "epoch": 6.020220588235294,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.012424002114725e-05,
+      "loss": 1.1269,
+      "step": 45850
+    },
+    {
+      "epoch": 6.026785714285714,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.999206978588423e-05,
+      "loss": 1.0631,
+      "step": 45900
+    },
+    {
+      "epoch": 6.0333508403361344,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.98598995506212e-05,
+      "loss": 1.155,
+      "step": 45950
+    },
+    {
+      "epoch": 6.0399159663865545,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.972772931535819e-05,
+      "loss": 1.0826,
+      "step": 46000
+    },
+    {
+      "epoch": 6.0464810924369745,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.959555908009517e-05,
+      "loss": 1.1677,
+      "step": 46050
+    },
+    {
+      "epoch": 6.0530462184873945,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.946338884483214e-05,
+      "loss": 1.089,
+      "step": 46100
+    },
+    {
+      "epoch": 6.0596113445378155,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.933121860956913e-05,
+      "loss": 1.1358,
+      "step": 46150
+    },
+    {
+      "epoch": 6.0661764705882355,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.91990483743061e-05,
+      "loss": 1.1714,
+      "step": 46200
+    },
+    {
+      "epoch": 6.0727415966386555,
+      "grad_norm": 4.375,
+      "learning_rate": 7.90668781390431e-05,
+      "loss": 1.1851,
+      "step": 46250
+    },
+    {
+      "epoch": 6.079306722689076,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.893470790378007e-05,
+      "loss": 1.104,
+      "step": 46300
+    },
+    {
+      "epoch": 6.085871848739496,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.880253766851704e-05,
+      "loss": 1.1089,
+      "step": 46350
+    },
+    {
+      "epoch": 6.092436974789916,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.867036743325403e-05,
+      "loss": 1.1587,
+      "step": 46400
+    },
+    {
+      "epoch": 6.099002100840337,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.853819719799101e-05,
+      "loss": 1.1539,
+      "step": 46450
+    },
+    {
+      "epoch": 6.105567226890757,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.8406026962728e-05,
+      "loss": 1.1214,
+      "step": 46500
+    },
+    {
+      "epoch": 6.112132352941177,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.827385672746497e-05,
+      "loss": 1.1342,
+      "step": 46550
+    },
+    {
+      "epoch": 6.118697478991597,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.814168649220196e-05,
+      "loss": 1.1118,
+      "step": 46600
+    },
+    {
+      "epoch": 6.125262605042017,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.800951625693894e-05,
+      "loss": 1.1136,
+      "step": 46650
+    },
+    {
+      "epoch": 6.131827731092437,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.787734602167593e-05,
+      "loss": 1.1687,
+      "step": 46700
+    },
+    {
+      "epoch": 6.138392857142857,
+      "grad_norm": 3.25,
+      "learning_rate": 7.77451757864129e-05,
+      "loss": 1.1565,
+      "step": 46750
+    },
+    {
+      "epoch": 6.144957983193278,
+      "grad_norm": 4.09375,
+      "learning_rate": 7.761300555114989e-05,
+      "loss": 1.0907,
+      "step": 46800
+    },
+    {
+      "epoch": 6.151523109243698,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.748083531588686e-05,
+      "loss": 1.1157,
+      "step": 46850
+    },
+    {
+      "epoch": 6.158088235294118,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.734866508062385e-05,
+      "loss": 1.1268,
+      "step": 46900
+    },
+    {
+      "epoch": 6.164653361344538,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.721649484536083e-05,
+      "loss": 1.2271,
+      "step": 46950
+    },
+    {
+      "epoch": 6.171218487394958,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.708432461009782e-05,
+      "loss": 1.1368,
+      "step": 47000
+    },
+    {
+      "epoch": 6.177783613445378,
+      "grad_norm": 3.203125,
+      "learning_rate": 7.695215437483479e-05,
+      "loss": 1.1157,
+      "step": 47050
+    },
+    {
+      "epoch": 6.184348739495798,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.681998413957178e-05,
+      "loss": 1.1223,
+      "step": 47100
+    },
+    {
+      "epoch": 6.190913865546219,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.668781390430876e-05,
+      "loss": 1.2149,
+      "step": 47150
+    },
+    {
+      "epoch": 6.197478991596639,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.655564366904573e-05,
+      "loss": 1.1906,
+      "step": 47200
+    },
+    {
+      "epoch": 6.204044117647059,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.64234734337827e-05,
+      "loss": 1.0765,
+      "step": 47250
+    },
+    {
+      "epoch": 6.210609243697479,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.62913031985197e-05,
+      "loss": 1.1369,
+      "step": 47300
+    },
+    {
+      "epoch": 6.217174369747899,
+      "grad_norm": 3.875,
+      "learning_rate": 7.615913296325667e-05,
+      "loss": 1.1365,
+      "step": 47350
+    },
+    {
+      "epoch": 6.223739495798319,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.602696272799366e-05,
+      "loss": 1.13,
+      "step": 47400
+    },
+    {
+      "epoch": 6.23030462184874,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.589479249273063e-05,
+      "loss": 1.1538,
+      "step": 47450
+    },
+    {
+      "epoch": 6.23686974789916,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.576262225746762e-05,
+      "loss": 1.1669,
+      "step": 47500
+    },
+    {
+      "epoch": 6.24343487394958,
+      "grad_norm": 4.5,
+      "learning_rate": 7.56304520222046e-05,
+      "loss": 1.1756,
+      "step": 47550
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.549828178694159e-05,
+      "loss": 1.1435,
+      "step": 47600
+    },
+    {
+      "epoch": 6.25656512605042,
+      "grad_norm": 3.484375,
+      "learning_rate": 7.536611155167856e-05,
+      "loss": 1.1327,
+      "step": 47650
+    },
+    {
+      "epoch": 6.26313025210084,
+      "grad_norm": 3.421875,
+      "learning_rate": 7.523394131641555e-05,
+      "loss": 1.1752,
+      "step": 47700
+    },
+    {
+      "epoch": 6.26969537815126,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.510177108115253e-05,
+      "loss": 1.1626,
+      "step": 47750
+    },
+    {
+      "epoch": 6.276260504201681,
+      "grad_norm": 4.375,
+      "learning_rate": 7.496960084588952e-05,
+      "loss": 1.166,
+      "step": 47800
+    },
+    {
+      "epoch": 6.282825630252101,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.483743061062649e-05,
+      "loss": 1.17,
+      "step": 47850
+    },
+    {
+      "epoch": 6.289390756302521,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.470526037536348e-05,
+      "loss": 1.2015,
+      "step": 47900
+    },
+    {
+      "epoch": 6.295955882352941,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.457309014010045e-05,
+      "loss": 1.1405,
+      "step": 47950
+    },
+    {
+      "epoch": 6.302521008403361,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.444091990483743e-05,
+      "loss": 1.1591,
+      "step": 48000
+    },
+    {
+      "epoch": 6.309086134453781,
+      "grad_norm": 3.953125,
+      "learning_rate": 7.430874966957442e-05,
+      "loss": 1.1504,
+      "step": 48050
+    },
+    {
+      "epoch": 6.315651260504202,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.41765794343114e-05,
+      "loss": 1.1835,
+      "step": 48100
+    },
+    {
+      "epoch": 6.322216386554622,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.404440919904838e-05,
+      "loss": 1.1839,
+      "step": 48150
+    },
+    {
+      "epoch": 6.328781512605042,
+      "grad_norm": 5.3125,
+      "learning_rate": 7.391223896378536e-05,
+      "loss": 1.1792,
+      "step": 48200
+    },
+    {
+      "epoch": 6.335346638655462,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.378006872852235e-05,
+      "loss": 1.1407,
+      "step": 48250
+    },
+    {
+      "epoch": 6.341911764705882,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.364789849325932e-05,
+      "loss": 1.2657,
+      "step": 48300
+    },
+    {
+      "epoch": 6.348476890756302,
+      "grad_norm": 3.125,
+      "learning_rate": 7.351572825799631e-05,
+      "loss": 1.1614,
+      "step": 48350
+    },
+    {
+      "epoch": 6.355042016806722,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.338355802273329e-05,
+      "loss": 1.2149,
+      "step": 48400
+    },
+    {
+      "epoch": 6.361607142857143,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.325138778747026e-05,
+      "loss": 1.1551,
+      "step": 48450
+    },
+    {
+      "epoch": 6.368172268907563,
+      "grad_norm": 4.125,
+      "learning_rate": 7.311921755220724e-05,
+      "loss": 1.2245,
+      "step": 48500
+    },
+    {
+      "epoch": 6.374737394957983,
+      "grad_norm": 2.75,
+      "learning_rate": 7.298704731694422e-05,
+      "loss": 1.225,
+      "step": 48550
+    },
+    {
+      "epoch": 6.381302521008403,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.28548770816812e-05,
+      "loss": 1.2405,
+      "step": 48600
+    },
+    {
+      "epoch": 6.387867647058823,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.272270684641819e-05,
+      "loss": 1.2447,
+      "step": 48650
+    },
+    {
+      "epoch": 6.394432773109243,
+      "grad_norm": 4.3125,
+      "learning_rate": 7.259053661115516e-05,
+      "loss": 1.1656,
+      "step": 48700
+    },
+    {
+      "epoch": 6.400997899159664,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.245836637589215e-05,
+      "loss": 1.2142,
+      "step": 48750
+    },
+    {
+      "epoch": 6.407563025210084,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.232619614062913e-05,
+      "loss": 1.1842,
+      "step": 48800
+    },
+    {
+      "epoch": 6.414128151260504,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.219402590536612e-05,
+      "loss": 1.1958,
+      "step": 48850
+    },
+    {
+      "epoch": 6.420693277310924,
+      "grad_norm": 3.625,
+      "learning_rate": 7.206185567010309e-05,
+      "loss": 1.1988,
+      "step": 48900
+    },
+    {
+      "epoch": 6.4272584033613445,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.192968543484008e-05,
+      "loss": 1.1995,
+      "step": 48950
+    },
+    {
+      "epoch": 6.4338235294117645,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.179751519957706e-05,
+      "loss": 1.1896,
+      "step": 49000
+    },
+    {
+      "epoch": 6.4403886554621845,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.166534496431404e-05,
+      "loss": 1.2092,
+      "step": 49050
+    },
+    {
+      "epoch": 6.4469537815126055,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.153317472905102e-05,
+      "loss": 1.168,
+      "step": 49100
+    },
+    {
+      "epoch": 6.4535189075630255,
+      "grad_norm": 3.359375,
+      "learning_rate": 7.140100449378801e-05,
+      "loss": 1.2271,
+      "step": 49150
+    },
+    {
+      "epoch": 6.4600840336134455,
+      "grad_norm": 3.71875,
+      "learning_rate": 7.126883425852498e-05,
+      "loss": 1.1761,
+      "step": 49200
+    },
+    {
+      "epoch": 6.4666491596638656,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.113666402326197e-05,
+      "loss": 1.1482,
+      "step": 49250
+    },
+    {
+      "epoch": 6.473214285714286,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.100449378799895e-05,
+      "loss": 1.1622,
+      "step": 49300
+    },
+    {
+      "epoch": 6.479779411764706,
+      "grad_norm": 3.890625,
+      "learning_rate": 7.087232355273594e-05,
+      "loss": 1.2345,
+      "step": 49350
+    },
+    {
+      "epoch": 6.486344537815126,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.074015331747291e-05,
+      "loss": 1.1928,
+      "step": 49400
+    },
+    {
+      "epoch": 6.492909663865547,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.06079830822099e-05,
+      "loss": 1.1809,
+      "step": 49450
+    },
+    {
+      "epoch": 6.499474789915967,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.047581284694688e-05,
+      "loss": 1.2192,
+      "step": 49500
+    },
+    {
+      "epoch": 6.506039915966387,
+      "grad_norm": 2.875,
+      "learning_rate": 7.034364261168386e-05,
+      "loss": 1.2011,
+      "step": 49550
+    },
+    {
+      "epoch": 6.512605042016807,
+      "grad_norm": 3.75,
+      "learning_rate": 7.021147237642083e-05,
+      "loss": 1.2188,
+      "step": 49600
+    },
+    {
+      "epoch": 6.519170168067227,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.007930214115781e-05,
+      "loss": 1.2231,
+      "step": 49650
+    },
+    {
+      "epoch": 6.525735294117647,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.994713190589479e-05,
+      "loss": 1.1578,
+      "step": 49700
+    },
+    {
+      "epoch": 6.532300420168067,
+      "grad_norm": 2.5,
+      "learning_rate": 6.981496167063178e-05,
+      "loss": 1.2147,
+      "step": 49750
+    },
+    {
+      "epoch": 6.538865546218488,
+      "grad_norm": 3.75,
+      "learning_rate": 6.968279143536875e-05,
+      "loss": 1.1989,
+      "step": 49800
+    },
+    {
+      "epoch": 6.545430672268908,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.955062120010574e-05,
+      "loss": 1.2022,
+      "step": 49850
+    },
+    {
+      "epoch": 6.551995798319328,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.941845096484272e-05,
+      "loss": 1.1972,
+      "step": 49900
+    },
+    {
+      "epoch": 6.558560924369748,
+      "grad_norm": 3.5,
+      "learning_rate": 6.92862807295797e-05,
+      "loss": 1.2239,
+      "step": 49950
+    },
+    {
+      "epoch": 6.565126050420168,
+      "grad_norm": 4.3125,
+      "learning_rate": 6.915411049431668e-05,
+      "loss": 1.1536,
+      "step": 50000
+    },
+    {
+      "epoch": 6.571691176470588,
+      "grad_norm": 3.9375,
+      "learning_rate": 6.902194025905367e-05,
+      "loss": 1.1373,
+      "step": 50050
+    },
+    {
+      "epoch": 6.578256302521009,
+      "grad_norm": 4.5,
+      "learning_rate": 6.888977002379065e-05,
+      "loss": 1.2083,
+      "step": 50100
+    },
+    {
+      "epoch": 6.584821428571429,
+      "grad_norm": 5.125,
+      "learning_rate": 6.875759978852762e-05,
+      "loss": 1.2293,
+      "step": 50150
+    },
+    {
+      "epoch": 6.591386554621849,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.862542955326461e-05,
+      "loss": 1.2328,
+      "step": 50200
+    },
+    {
+      "epoch": 6.597951680672269,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.849325931800158e-05,
+      "loss": 1.1966,
+      "step": 50250
+    },
+    {
+      "epoch": 6.604516806722689,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.836108908273857e-05,
+      "loss": 1.2417,
+      "step": 50300
+    },
+    {
+      "epoch": 6.611081932773109,
+      "grad_norm": 3.25,
+      "learning_rate": 6.822891884747555e-05,
+      "loss": 1.2,
+      "step": 50350
+    },
+    {
+      "epoch": 6.617647058823529,
+      "grad_norm": 3.453125,
+      "learning_rate": 6.809674861221254e-05,
+      "loss": 1.1948,
+      "step": 50400
+    },
+    {
+      "epoch": 6.62421218487395,
+      "grad_norm": 3.953125,
+      "learning_rate": 6.796457837694951e-05,
+      "loss": 1.2059,
+      "step": 50450
+    },
+    {
+      "epoch": 6.63077731092437,
+      "grad_norm": 3.703125,
+      "learning_rate": 6.78324081416865e-05,
+      "loss": 1.1772,
+      "step": 50500
+    },
+    {
+      "epoch": 6.63734243697479,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.770023790642348e-05,
+      "loss": 1.2675,
+      "step": 50550
+    },
+    {
+      "epoch": 6.64390756302521,
+      "grad_norm": 4.90625,
+      "learning_rate": 6.756806767116046e-05,
+      "loss": 1.1601,
+      "step": 50600
+    },
+    {
+      "epoch": 6.65047268907563,
+      "grad_norm": 3.578125,
+      "learning_rate": 6.743589743589744e-05,
+      "loss": 1.2007,
+      "step": 50650
+    },
+    {
+      "epoch": 6.65703781512605,
+      "grad_norm": 2.5,
+      "learning_rate": 6.730372720063443e-05,
+      "loss": 1.2893,
+      "step": 50700
+    },
+    {
+      "epoch": 6.663602941176471,
+      "grad_norm": 3.625,
+      "learning_rate": 6.717155696537139e-05,
+      "loss": 1.1959,
+      "step": 50750
+    },
+    {
+      "epoch": 6.670168067226891,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.703938673010838e-05,
+      "loss": 1.2391,
+      "step": 50800
+    },
+    {
+      "epoch": 6.676733193277311,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.690721649484535e-05,
+      "loss": 1.2726,
+      "step": 50850
+    },
+    {
+      "epoch": 6.683298319327731,
+      "grad_norm": 3.25,
+      "learning_rate": 6.677504625958234e-05,
+      "loss": 1.2429,
+      "step": 50900
+    },
+    {
+      "epoch": 6.689863445378151,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.664287602431932e-05,
+      "loss": 1.2426,
+      "step": 50950
+    },
+    {
+      "epoch": 6.696428571428571,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.651070578905631e-05,
+      "loss": 1.1643,
+      "step": 51000
+    },
+    {
+      "epoch": 6.702993697478991,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.637853555379328e-05,
+      "loss": 1.2403,
+      "step": 51050
+    },
+    {
+      "epoch": 6.709558823529412,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.624636531853027e-05,
+      "loss": 1.1968,
+      "step": 51100
+    },
+    {
+      "epoch": 6.716123949579832,
+      "grad_norm": 5.28125,
+      "learning_rate": 6.611419508326725e-05,
+      "loss": 1.2307,
+      "step": 51150
+    },
+    {
+      "epoch": 6.722689075630252,
+      "grad_norm": 3.765625,
+      "learning_rate": 6.598202484800423e-05,
+      "loss": 1.2227,
+      "step": 51200
+    },
+    {
+      "epoch": 6.729254201680672,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.584985461274121e-05,
+      "loss": 1.2558,
+      "step": 51250
+    },
+    {
+      "epoch": 6.735819327731092,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.57176843774782e-05,
+      "loss": 1.1961,
+      "step": 51300
+    },
+    {
+      "epoch": 6.742384453781512,
+      "grad_norm": 3.625,
+      "learning_rate": 6.558551414221517e-05,
+      "loss": 1.2159,
+      "step": 51350
+    },
+    {
+      "epoch": 6.748949579831933,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.545334390695216e-05,
+      "loss": 1.2694,
+      "step": 51400
+    },
+    {
+      "epoch": 6.755514705882353,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.532117367168914e-05,
+      "loss": 1.2875,
+      "step": 51450
+    },
+    {
+      "epoch": 6.762079831932773,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.518900343642613e-05,
+      "loss": 1.2388,
+      "step": 51500
+    },
+    {
+      "epoch": 6.768644957983193,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.50568332011631e-05,
+      "loss": 1.1447,
+      "step": 51550
+    },
+    {
+      "epoch": 6.775210084033613,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.492466296590009e-05,
+      "loss": 1.2619,
+      "step": 51600
+    },
+    {
+      "epoch": 6.781775210084033,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.479249273063707e-05,
+      "loss": 1.1767,
+      "step": 51650
+    },
+    {
+      "epoch": 6.788340336134453,
+      "grad_norm": 3.65625,
+      "learning_rate": 6.466032249537405e-05,
+      "loss": 1.2818,
+      "step": 51700
+    },
+    {
+      "epoch": 6.7949054621848735,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.452815226011103e-05,
+      "loss": 1.2318,
+      "step": 51750
+    },
+    {
+      "epoch": 6.801470588235294,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.4395982024848e-05,
+      "loss": 1.2754,
+      "step": 51800
+    },
+    {
+      "epoch": 6.808035714285714,
+      "grad_norm": 3.890625,
+      "learning_rate": 6.4263811789585e-05,
+      "loss": 1.2557,
+      "step": 51850
+    },
+    {
+      "epoch": 6.8146008403361344,
+      "grad_norm": 4.65625,
+      "learning_rate": 6.413164155432197e-05,
+      "loss": 1.2575,
+      "step": 51900
+    },
+    {
+      "epoch": 6.8211659663865545,
+      "grad_norm": 4.1875,
+      "learning_rate": 6.399947131905894e-05,
+      "loss": 1.2624,
+      "step": 51950
+    },
+    {
+      "epoch": 6.8277310924369745,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.386730108379593e-05,
+      "loss": 1.1846,
+      "step": 52000
+    },
+    {
+      "epoch": 6.834296218487395,
+      "grad_norm": 3.25,
+      "learning_rate": 6.373513084853291e-05,
+      "loss": 1.2446,
+      "step": 52050
+    },
+    {
+      "epoch": 6.8408613445378155,
+      "grad_norm": 3.6875,
+      "learning_rate": 6.36029606132699e-05,
+      "loss": 1.2239,
+      "step": 52100
+    },
+    {
+      "epoch": 6.8474264705882355,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.347079037800687e-05,
+      "loss": 1.2578,
+      "step": 52150
+    },
+    {
+      "epoch": 6.8539915966386555,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.333862014274386e-05,
+      "loss": 1.2569,
+      "step": 52200
+    },
+    {
+      "epoch": 6.860556722689076,
+      "grad_norm": 3.515625,
+      "learning_rate": 6.320644990748084e-05,
+      "loss": 1.2031,
+      "step": 52250
+    },
+    {
+      "epoch": 6.867121848739496,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.307427967221781e-05,
+      "loss": 1.2602,
+      "step": 52300
+    },
+    {
+      "epoch": 6.873686974789916,
+      "grad_norm": 5.5625,
+      "learning_rate": 6.29421094369548e-05,
+      "loss": 1.2312,
+      "step": 52350
+    },
+    {
+      "epoch": 6.880252100840336,
+      "grad_norm": 4.15625,
+      "learning_rate": 6.280993920169178e-05,
+      "loss": 1.2779,
+      "step": 52400
+    },
+    {
+      "epoch": 6.886817226890757,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.267776896642876e-05,
+      "loss": 1.2193,
+      "step": 52450
+    },
+    {
+      "epoch": 6.893382352941177,
+      "grad_norm": 3.859375,
+      "learning_rate": 6.254559873116574e-05,
+      "loss": 1.2641,
+      "step": 52500
+    },
+    {
+      "epoch": 6.899947478991597,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.241342849590273e-05,
+      "loss": 1.2719,
+      "step": 52550
+    },
+    {
+      "epoch": 6.906512605042017,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.22812582606397e-05,
+      "loss": 1.2645,
+      "step": 52600
+    },
+    {
+      "epoch": 6.913077731092437,
+      "grad_norm": 3.71875,
+      "learning_rate": 6.214908802537669e-05,
+      "loss": 1.2796,
+      "step": 52650
+    },
+    {
+      "epoch": 6.919642857142857,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.201691779011367e-05,
+      "loss": 1.2243,
+      "step": 52700
+    },
+    {
+      "epoch": 6.926207983193278,
+      "grad_norm": 4.53125,
+      "learning_rate": 6.188474755485066e-05,
+      "loss": 1.2344,
+      "step": 52750
+    },
+    {
+      "epoch": 6.932773109243698,
+      "grad_norm": 4.03125,
+      "learning_rate": 6.175257731958763e-05,
+      "loss": 1.2032,
+      "step": 52800
+    },
+    {
+      "epoch": 6.939338235294118,
+      "grad_norm": 4.78125,
+      "learning_rate": 6.162040708432462e-05,
+      "loss": 1.2038,
+      "step": 52850
+    },
+    {
+      "epoch": 6.945903361344538,
+      "grad_norm": 4.21875,
+      "learning_rate": 6.14882368490616e-05,
+      "loss": 1.2359,
+      "step": 52900
+    },
+    {
+      "epoch": 6.952468487394958,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.135606661379858e-05,
+      "loss": 1.2409,
+      "step": 52950
+    },
+    {
+      "epoch": 6.959033613445378,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.122389637853556e-05,
+      "loss": 1.2521,
+      "step": 53000
+    },
+    {
+      "epoch": 6.965598739495798,
+      "grad_norm": 3.734375,
+      "learning_rate": 6.109172614327255e-05,
+      "loss": 1.229,
+      "step": 53050
+    },
+    {
+      "epoch": 6.972163865546219,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.095955590800952e-05,
+      "loss": 1.2429,
+      "step": 53100
+    },
+    {
+      "epoch": 6.978728991596639,
+      "grad_norm": 3.296875,
+      "learning_rate": 6.08273856727465e-05,
+      "loss": 1.1811,
+      "step": 53150
+    },
+    {
+      "epoch": 6.985294117647059,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.069521543748348e-05,
+      "loss": 1.2411,
+      "step": 53200
+    },
+    {
+      "epoch": 6.991859243697479,
+      "grad_norm": 3.671875,
+      "learning_rate": 6.056304520222046e-05,
+      "loss": 1.2667,
+      "step": 53250
+    },
+    {
+      "epoch": 6.998424369747899,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.043087496695744e-05,
+      "loss": 1.2438,
+      "step": 53300
+    },
+    {
+      "epoch": 7.004989495798319,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.029870473169442e-05,
+      "loss": 1.0238,
+      "step": 53350
+    },
+    {
+      "epoch": 7.01155462184874,
+      "grad_norm": 2.875,
+      "learning_rate": 6.01665344964314e-05,
+      "loss": 0.9602,
+      "step": 53400
+    },
+    {
+      "epoch": 7.01811974789916,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.003436426116838e-05,
+      "loss": 0.9521,
+      "step": 53450
+    },
+    {
+      "epoch": 7.02468487394958,
+      "grad_norm": 3.828125,
+      "learning_rate": 5.9902194025905365e-05,
+      "loss": 0.9348,
+      "step": 53500
+    },
+    {
+      "epoch": 7.03125,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.977002379064235e-05,
+      "loss": 0.9541,
+      "step": 53550
+    },
+    {
+      "epoch": 7.03781512605042,
+      "grad_norm": 3.640625,
+      "learning_rate": 5.963785355537933e-05,
+      "loss": 1.0194,
+      "step": 53600
+    },
+    {
+      "epoch": 7.04438025210084,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.950568332011631e-05,
+      "loss": 0.9803,
+      "step": 53650
+    },
+    {
+      "epoch": 7.05094537815126,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.937351308485329e-05,
+      "loss": 1.0089,
+      "step": 53700
+    },
+    {
+      "epoch": 7.057510504201681,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.9241342849590275e-05,
+      "loss": 0.9549,
+      "step": 53750
+    },
+    {
+      "epoch": 7.064075630252101,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.910917261432726e-05,
+      "loss": 0.9472,
+      "step": 53800
+    },
+    {
+      "epoch": 7.070640756302521,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.897700237906424e-05,
+      "loss": 0.9334,
+      "step": 53850
+    },
+    {
+      "epoch": 7.077205882352941,
+      "grad_norm": 4.9375,
+      "learning_rate": 5.884483214380122e-05,
+      "loss": 0.9778,
+      "step": 53900
+    },
+    {
+      "epoch": 7.083771008403361,
+      "grad_norm": 3.796875,
+      "learning_rate": 5.87126619085382e-05,
+      "loss": 0.9484,
+      "step": 53950
+    },
+    {
+      "epoch": 7.090336134453781,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.8580491673275185e-05,
+      "loss": 0.9268,
+      "step": 54000
+    },
+    {
+      "epoch": 7.096901260504202,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.844832143801217e-05,
+      "loss": 0.992,
+      "step": 54050
+    },
+    {
+      "epoch": 7.103466386554622,
+      "grad_norm": 3.25,
+      "learning_rate": 5.831615120274915e-05,
+      "loss": 0.9315,
+      "step": 54100
+    },
+    {
+      "epoch": 7.110031512605042,
+      "grad_norm": 4.09375,
+      "learning_rate": 5.818398096748613e-05,
+      "loss": 0.9577,
+      "step": 54150
+    },
+    {
+      "epoch": 7.116596638655462,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.805181073222311e-05,
+      "loss": 0.9646,
+      "step": 54200
+    },
+    {
+      "epoch": 7.123161764705882,
+      "grad_norm": 3.59375,
+      "learning_rate": 5.7919640496960095e-05,
+      "loss": 1.0107,
+      "step": 54250
+    },
+    {
+      "epoch": 7.129726890756302,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.778747026169706e-05,
+      "loss": 0.9089,
+      "step": 54300
+    },
+    {
+      "epoch": 7.136292016806722,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.7655300026434045e-05,
+      "loss": 1.0186,
+      "step": 54350
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 3.96875,
+      "learning_rate": 5.752312979117103e-05,
+      "loss": 1.0012,
+      "step": 54400
+    },
+    {
+      "epoch": 7.149422268907563,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.739095955590801e-05,
+      "loss": 0.9624,
+      "step": 54450
+    },
+    {
+      "epoch": 7.155987394957983,
+      "grad_norm": 3.125,
+      "learning_rate": 5.725878932064499e-05,
+      "loss": 0.9607,
+      "step": 54500
+    },
+    {
+      "epoch": 7.162552521008403,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.712661908538197e-05,
+      "loss": 0.9862,
+      "step": 54550
+    },
+    {
+      "epoch": 7.169117647058823,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.6994448850118955e-05,
+      "loss": 1.022,
+      "step": 54600
+    },
+    {
+      "epoch": 7.175682773109243,
+      "grad_norm": 3.984375,
+      "learning_rate": 5.686227861485594e-05,
+      "loss": 0.9618,
+      "step": 54650
+    },
+    {
+      "epoch": 7.182247899159663,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.673010837959292e-05,
+      "loss": 1.04,
+      "step": 54700
+    },
+    {
+      "epoch": 7.188813025210084,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.65979381443299e-05,
+      "loss": 0.9641,
+      "step": 54750
+    },
+    {
+      "epoch": 7.195378151260504,
+      "grad_norm": 5.09375,
+      "learning_rate": 5.646576790906688e-05,
+      "loss": 0.9545,
+      "step": 54800
+    },
+    {
+      "epoch": 7.201943277310924,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.6333597673803865e-05,
+      "loss": 0.9333,
+      "step": 54850
+    },
+    {
+      "epoch": 7.2085084033613445,
+      "grad_norm": 4.21875,
+      "learning_rate": 5.620142743854085e-05,
+      "loss": 1.0082,
+      "step": 54900
+    },
+    {
+      "epoch": 7.2150735294117645,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.606925720327782e-05,
+      "loss": 0.9802,
+      "step": 54950
+    },
+    {
+      "epoch": 7.2216386554621845,
+      "grad_norm": 3.515625,
+      "learning_rate": 5.5937086968014804e-05,
+      "loss": 0.9939,
+      "step": 55000
+    },
+    {
+      "epoch": 7.2282037815126055,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.5804916732751786e-05,
+      "loss": 0.9994,
+      "step": 55050
+    },
+    {
+      "epoch": 7.2347689075630255,
+      "grad_norm": 4.1875,
+      "learning_rate": 5.567274649748877e-05,
+      "loss": 0.9503,
+      "step": 55100
+    },
+    {
+      "epoch": 7.2413340336134455,
+      "grad_norm": 3.609375,
+      "learning_rate": 5.554057626222575e-05,
+      "loss": 1.0269,
+      "step": 55150
+    },
+    {
+      "epoch": 7.2478991596638656,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.540840602696273e-05,
+      "loss": 0.9995,
+      "step": 55200
+    },
+    {
+      "epoch": 7.254464285714286,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.5276235791699714e-05,
+      "loss": 1.004,
+      "step": 55250
+    },
+    {
+      "epoch": 7.261029411764706,
+      "grad_norm": 3.921875,
+      "learning_rate": 5.5144065556436696e-05,
+      "loss": 0.9907,
+      "step": 55300
+    },
+    {
+      "epoch": 7.267594537815126,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.501189532117368e-05,
+      "loss": 1.0151,
+      "step": 55350
+    },
+    {
+      "epoch": 7.274159663865547,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.487972508591066e-05,
+      "loss": 1.0086,
+      "step": 55400
+    },
+    {
+      "epoch": 7.280724789915967,
+      "grad_norm": 4.1875,
+      "learning_rate": 5.474755485064764e-05,
+      "loss": 1.0091,
+      "step": 55450
+    },
+    {
+      "epoch": 7.287289915966387,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.461538461538461e-05,
+      "loss": 0.9737,
+      "step": 55500
+    },
+    {
+      "epoch": 7.293855042016807,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.448321438012159e-05,
+      "loss": 0.9781,
+      "step": 55550
+    },
+    {
+      "epoch": 7.300420168067227,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.4351044144858574e-05,
+      "loss": 0.9661,
+      "step": 55600
+    },
+    {
+      "epoch": 7.306985294117647,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.4218873909595556e-05,
+      "loss": 0.9372,
+      "step": 55650
+    },
+    {
+      "epoch": 7.313550420168067,
+      "grad_norm": 3.375,
+      "learning_rate": 5.408670367433254e-05,
+      "loss": 0.9873,
+      "step": 55700
+    },
+    {
+      "epoch": 7.320115546218488,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.395453343906952e-05,
+      "loss": 0.9819,
+      "step": 55750
+    },
+    {
+      "epoch": 7.326680672268908,
+      "grad_norm": 3.765625,
+      "learning_rate": 5.38223632038065e-05,
+      "loss": 0.9967,
+      "step": 55800
+    },
+    {
+      "epoch": 7.333245798319328,
+      "grad_norm": 4.25,
+      "learning_rate": 5.3690192968543484e-05,
+      "loss": 0.9956,
+      "step": 55850
+    },
+    {
+      "epoch": 7.339810924369748,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.3558022733280466e-05,
+      "loss": 0.9971,
+      "step": 55900
+    },
+    {
+      "epoch": 7.346376050420168,
+      "grad_norm": 3.625,
+      "learning_rate": 5.342585249801745e-05,
+      "loss": 0.9582,
+      "step": 55950
+    },
+    {
+      "epoch": 7.352941176470588,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.329368226275443e-05,
+      "loss": 0.9472,
+      "step": 56000
+    },
+    {
+      "epoch": 7.359506302521009,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.316151202749141e-05,
+      "loss": 0.9408,
+      "step": 56050
+    },
+    {
+      "epoch": 7.366071428571429,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.3029341792228394e-05,
+      "loss": 0.9915,
+      "step": 56100
+    },
+    {
+      "epoch": 7.372636554621849,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.2897171556965376e-05,
+      "loss": 0.9697,
+      "step": 56150
+    },
+    {
+      "epoch": 7.379201680672269,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.276500132170236e-05,
+      "loss": 1.0588,
+      "step": 56200
+    },
+    {
+      "epoch": 7.385766806722689,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.263283108643934e-05,
+      "loss": 1.0586,
+      "step": 56250
+    },
+    {
+      "epoch": 7.392331932773109,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.250066085117632e-05,
+      "loss": 1.0323,
+      "step": 56300
+    },
+    {
+      "epoch": 7.398897058823529,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.2368490615913304e-05,
+      "loss": 0.9774,
+      "step": 56350
+    },
+    {
+      "epoch": 7.40546218487395,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.2236320380650286e-05,
+      "loss": 1.0095,
+      "step": 56400
+    },
+    {
+      "epoch": 7.41202731092437,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.210415014538727e-05,
+      "loss": 1.0282,
+      "step": 56450
+    },
+    {
+      "epoch": 7.41859243697479,
+      "grad_norm": 3.328125,
+      "learning_rate": 5.197197991012425e-05,
+      "loss": 0.9461,
+      "step": 56500
+    },
+    {
+      "epoch": 7.42515756302521,
+      "grad_norm": 4.03125,
+      "learning_rate": 5.183980967486123e-05,
+      "loss": 1.0255,
+      "step": 56550
+    },
+    {
+      "epoch": 7.43172268907563,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.170763943959821e-05,
+      "loss": 0.9991,
+      "step": 56600
+    },
+    {
+      "epoch": 7.43828781512605,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.157546920433519e-05,
+      "loss": 1.0288,
+      "step": 56650
+    },
+    {
+      "epoch": 7.444852941176471,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.1443298969072164e-05,
+      "loss": 1.0095,
+      "step": 56700
+    },
+    {
+      "epoch": 7.451418067226891,
+      "grad_norm": 3.953125,
+      "learning_rate": 5.1311128733809146e-05,
+      "loss": 1.0206,
+      "step": 56750
+    },
+    {
+      "epoch": 7.457983193277311,
+      "grad_norm": 3.9375,
+      "learning_rate": 5.117895849854613e-05,
+      "loss": 1.0271,
+      "step": 56800
+    },
+    {
+      "epoch": 7.464548319327731,
+      "grad_norm": 4.59375,
+      "learning_rate": 5.104678826328311e-05,
+      "loss": 0.9945,
+      "step": 56850
+    },
+    {
+      "epoch": 7.471113445378151,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.091461802802009e-05,
+      "loss": 0.9645,
+      "step": 56900
+    },
+    {
+      "epoch": 7.477678571428571,
+      "grad_norm": 4.53125,
+      "learning_rate": 5.0782447792757074e-05,
+      "loss": 1.0132,
+      "step": 56950
+    },
+    {
+      "epoch": 7.484243697478991,
+      "grad_norm": 3.625,
+      "learning_rate": 5.0650277557494056e-05,
+      "loss": 1.046,
+      "step": 57000
+    },
+    {
+      "epoch": 7.490808823529412,
+      "grad_norm": 3.5625,
+      "learning_rate": 5.051810732223103e-05,
+      "loss": 1.0368,
+      "step": 57050
+    },
+    {
+      "epoch": 7.497373949579832,
+      "grad_norm": 4.0625,
+      "learning_rate": 5.038593708696801e-05,
+      "loss": 1.0087,
+      "step": 57100
+    },
+    {
+      "epoch": 7.503939075630252,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.0253766851704995e-05,
+      "loss": 0.9407,
+      "step": 57150
+    },
+    {
+      "epoch": 7.510504201680672,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.012159661644198e-05,
+      "loss": 0.9882,
+      "step": 57200
+    },
+    {
+      "epoch": 7.517069327731092,
+      "grad_norm": 4.0,
+      "learning_rate": 4.998942638117896e-05,
+      "loss": 0.9315,
+      "step": 57250
+    },
+    {
+      "epoch": 7.523634453781512,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.985725614591594e-05,
+      "loss": 1.001,
+      "step": 57300
+    },
+    {
+      "epoch": 7.530199579831933,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.972508591065292e-05,
+      "loss": 0.9481,
+      "step": 57350
+    },
+    {
+      "epoch": 7.536764705882353,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.9592915675389905e-05,
+      "loss": 1.0605,
+      "step": 57400
+    },
+    {
+      "epoch": 7.543329831932773,
+      "grad_norm": 3.125,
+      "learning_rate": 4.946074544012689e-05,
+      "loss": 1.034,
+      "step": 57450
+    },
+    {
+      "epoch": 7.549894957983193,
+      "grad_norm": 3.125,
+      "learning_rate": 4.932857520486387e-05,
+      "loss": 0.9914,
+      "step": 57500
+    },
+    {
+      "epoch": 7.556460084033613,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.9196404969600844e-05,
+      "loss": 0.9786,
+      "step": 57550
+    },
+    {
+      "epoch": 7.563025210084033,
+      "grad_norm": 3.765625,
+      "learning_rate": 4.9064234734337826e-05,
+      "loss": 0.9747,
+      "step": 57600
+    },
+    {
+      "epoch": 7.569590336134453,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.893206449907481e-05,
+      "loss": 0.9962,
+      "step": 57650
+    },
+    {
+      "epoch": 7.5761554621848735,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.879989426381179e-05,
+      "loss": 0.9514,
+      "step": 57700
+    },
+    {
+      "epoch": 7.582720588235294,
+      "grad_norm": 4.125,
+      "learning_rate": 4.866772402854877e-05,
+      "loss": 0.9623,
+      "step": 57750
+    },
+    {
+      "epoch": 7.589285714285714,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.8535553793285754e-05,
+      "loss": 0.9925,
+      "step": 57800
+    },
+    {
+      "epoch": 7.5958508403361344,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.8403383558022736e-05,
+      "loss": 0.984,
+      "step": 57850
+    },
+    {
+      "epoch": 7.6024159663865545,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.827121332275972e-05,
+      "loss": 0.9953,
+      "step": 57900
+    },
+    {
+      "epoch": 7.6089810924369745,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.81390430874967e-05,
+      "loss": 1.0407,
+      "step": 57950
+    },
+    {
+      "epoch": 7.615546218487395,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.800687285223368e-05,
+      "loss": 1.0593,
+      "step": 58000
+    },
+    {
+      "epoch": 7.6221113445378155,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.7874702616970664e-05,
+      "loss": 1.0001,
+      "step": 58050
+    },
+    {
+      "epoch": 7.6286764705882355,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.7742532381707646e-05,
+      "loss": 1.0068,
+      "step": 58100
+    },
+    {
+      "epoch": 7.6352415966386555,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.761036214644462e-05,
+      "loss": 1.0269,
+      "step": 58150
+    },
+    {
+      "epoch": 7.641806722689076,
+      "grad_norm": 4.125,
+      "learning_rate": 4.74781919111816e-05,
+      "loss": 1.0594,
+      "step": 58200
+    },
+    {
+      "epoch": 7.648371848739496,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.7346021675918585e-05,
+      "loss": 0.9949,
+      "step": 58250
+    },
+    {
+      "epoch": 7.654936974789916,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.721385144065557e-05,
+      "loss": 0.9923,
+      "step": 58300
+    },
+    {
+      "epoch": 7.661502100840336,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.708168120539255e-05,
+      "loss": 1.0112,
+      "step": 58350
+    },
+    {
+      "epoch": 7.668067226890757,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.694951097012953e-05,
+      "loss": 1.0397,
+      "step": 58400
+    },
+    {
+      "epoch": 7.674632352941177,
+      "grad_norm": 2.625,
+      "learning_rate": 4.681734073486651e-05,
+      "loss": 1.0318,
+      "step": 58450
+    },
+    {
+      "epoch": 7.681197478991597,
+      "grad_norm": 3.625,
+      "learning_rate": 4.6685170499603495e-05,
+      "loss": 1.0572,
+      "step": 58500
+    },
+    {
+      "epoch": 7.687762605042017,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.6553000264340477e-05,
+      "loss": 1.0385,
+      "step": 58550
+    },
+    {
+      "epoch": 7.694327731092437,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.642083002907746e-05,
+      "loss": 1.0306,
+      "step": 58600
+    },
+    {
+      "epoch": 7.700892857142857,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.628865979381444e-05,
+      "loss": 1.0104,
+      "step": 58650
+    },
+    {
+      "epoch": 7.707457983193278,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.6156489558551416e-05,
+      "loss": 0.9509,
+      "step": 58700
+    },
+    {
+      "epoch": 7.714023109243698,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.60243193232884e-05,
+      "loss": 1.0473,
+      "step": 58750
+    },
+    {
+      "epoch": 7.720588235294118,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.589214908802538e-05,
+      "loss": 1.0077,
+      "step": 58800
+    },
+    {
+      "epoch": 7.727153361344538,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.575997885276236e-05,
+      "loss": 1.0053,
+      "step": 58850
+    },
+    {
+      "epoch": 7.733718487394958,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.5627808617499344e-05,
+      "loss": 0.9762,
+      "step": 58900
+    },
+    {
+      "epoch": 7.740283613445378,
+      "grad_norm": 3.921875,
+      "learning_rate": 4.549563838223632e-05,
+      "loss": 1.0106,
+      "step": 58950
+    },
+    {
+      "epoch": 7.746848739495798,
+      "grad_norm": 3.875,
+      "learning_rate": 4.53634681469733e-05,
+      "loss": 0.9795,
+      "step": 59000
+    },
+    {
+      "epoch": 7.753413865546219,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.523129791171028e-05,
+      "loss": 0.9806,
+      "step": 59050
+    },
+    {
+      "epoch": 7.759978991596639,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.5099127676447265e-05,
+      "loss": 1.07,
+      "step": 59100
+    },
+    {
+      "epoch": 7.766544117647059,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.496695744118425e-05,
+      "loss": 1.0631,
+      "step": 59150
+    },
+    {
+      "epoch": 7.773109243697479,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.483478720592123e-05,
+      "loss": 1.0104,
+      "step": 59200
+    },
+    {
+      "epoch": 7.779674369747899,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.470261697065821e-05,
+      "loss": 0.9505,
+      "step": 59250
+    },
+    {
+      "epoch": 7.786239495798319,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.457044673539519e-05,
+      "loss": 1.0277,
+      "step": 59300
+    },
+    {
+      "epoch": 7.79280462184874,
+      "grad_norm": 2.5,
+      "learning_rate": 4.443827650013217e-05,
+      "loss": 1.0377,
+      "step": 59350
+    },
+    {
+      "epoch": 7.79936974789916,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.430610626486915e-05,
+      "loss": 0.9968,
+      "step": 59400
+    },
+    {
+      "epoch": 7.80593487394958,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.417393602960613e-05,
+      "loss": 1.0513,
+      "step": 59450
+    },
+    {
+      "epoch": 7.8125,
+      "grad_norm": 4.15625,
+      "learning_rate": 4.4041765794343114e-05,
+      "loss": 1.0534,
+      "step": 59500
+    },
+    {
+      "epoch": 7.81906512605042,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.3909595559080096e-05,
+      "loss": 1.0133,
+      "step": 59550
+    },
+    {
+      "epoch": 7.82563025210084,
+      "grad_norm": 3.859375,
+      "learning_rate": 4.377742532381708e-05,
+      "loss": 1.0018,
+      "step": 59600
+    },
+    {
+      "epoch": 7.83219537815126,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.364525508855406e-05,
+      "loss": 1.0268,
+      "step": 59650
+    },
+    {
+      "epoch": 7.838760504201681,
+      "grad_norm": 3.984375,
+      "learning_rate": 4.351308485329104e-05,
+      "loss": 0.9991,
+      "step": 59700
+    },
+    {
+      "epoch": 7.845325630252101,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.3380914618028024e-05,
+      "loss": 1.0332,
+      "step": 59750
+    },
+    {
+      "epoch": 7.851890756302521,
+      "grad_norm": 3.5625,
+      "learning_rate": 4.3248744382765005e-05,
+      "loss": 0.9982,
+      "step": 59800
+    },
+    {
+      "epoch": 7.858455882352941,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.311657414750199e-05,
+      "loss": 1.0299,
+      "step": 59850
+    },
+    {
+      "epoch": 7.865021008403361,
+      "grad_norm": 3.515625,
+      "learning_rate": 4.298440391223897e-05,
+      "loss": 1.0344,
+      "step": 59900
+    },
+    {
+      "epoch": 7.871586134453781,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.2852233676975945e-05,
+      "loss": 1.0037,
+      "step": 59950
+    },
+    {
+      "epoch": 7.878151260504202,
+      "grad_norm": 2.15625,
+      "learning_rate": 4.272006344171293e-05,
+      "loss": 0.9864,
+      "step": 60000
+    },
+    {
+      "epoch": 7.884716386554622,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.258789320644991e-05,
+      "loss": 1.0155,
+      "step": 60050
+    },
+    {
+      "epoch": 7.891281512605042,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.245572297118689e-05,
+      "loss": 0.9806,
+      "step": 60100
+    },
+    {
+      "epoch": 7.897846638655462,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.232355273592387e-05,
+      "loss": 1.053,
+      "step": 60150
+    },
+    {
+      "epoch": 7.904411764705882,
+      "grad_norm": 3.84375,
+      "learning_rate": 4.2191382500660855e-05,
+      "loss": 1.0841,
+      "step": 60200
+    },
+    {
+      "epoch": 7.910976890756302,
+      "grad_norm": 3.125,
+      "learning_rate": 4.2059212265397836e-05,
+      "loss": 1.0433,
+      "step": 60250
+    },
+    {
+      "epoch": 7.917542016806722,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.192704203013482e-05,
+      "loss": 1.0405,
+      "step": 60300
+    },
+    {
+      "epoch": 7.924107142857143,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.17948717948718e-05,
+      "loss": 0.9592,
+      "step": 60350
+    },
+    {
+      "epoch": 7.930672268907563,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.166270155960878e-05,
+      "loss": 0.9838,
+      "step": 60400
+    },
+    {
+      "epoch": 7.937237394957983,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.1530531324345764e-05,
+      "loss": 1.0412,
+      "step": 60450
+    },
+    {
+      "epoch": 7.943802521008403,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.139836108908274e-05,
+      "loss": 1.0022,
+      "step": 60500
+    },
+    {
+      "epoch": 7.950367647058823,
+      "grad_norm": 4.375,
+      "learning_rate": 4.126619085381972e-05,
+      "loss": 1.0335,
+      "step": 60550
+    },
+    {
+      "epoch": 7.956932773109243,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.1134020618556704e-05,
+      "loss": 0.9674,
+      "step": 60600
+    },
+    {
+      "epoch": 7.963497899159664,
+      "grad_norm": 3.53125,
+      "learning_rate": 4.1001850383293685e-05,
+      "loss": 1.0426,
+      "step": 60650
+    },
+    {
+      "epoch": 7.970063025210084,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.086968014803067e-05,
+      "loss": 0.9493,
+      "step": 60700
+    },
+    {
+      "epoch": 7.976628151260504,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.073750991276765e-05,
+      "loss": 1.0219,
+      "step": 60750
+    },
+    {
+      "epoch": 7.983193277310924,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.0605339677504625e-05,
+      "loss": 1.0364,
+      "step": 60800
+    },
+    {
+      "epoch": 7.9897584033613445,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.0473169442241607e-05,
+      "loss": 1.0049,
+      "step": 60850
+    },
+    {
+      "epoch": 7.9963235294117645,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.034099920697859e-05,
+      "loss": 1.0305,
+      "step": 60900
+    },
+    {
+      "epoch": 8.002888655462185,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.020882897171557e-05,
+      "loss": 0.9846,
+      "step": 60950
+    },
+    {
+      "epoch": 8.009453781512605,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.007665873645255e-05,
+      "loss": 0.8191,
+      "step": 61000
+    },
+    {
+      "epoch": 8.016018907563025,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.9944488501189534e-05,
+      "loss": 0.836,
+      "step": 61050
+    },
+    {
+      "epoch": 8.022584033613445,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.981231826592651e-05,
+      "loss": 0.8465,
+      "step": 61100
+    },
+    {
+      "epoch": 8.029149159663865,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.968014803066349e-05,
+      "loss": 0.8613,
+      "step": 61150
+    },
+    {
+      "epoch": 8.035714285714286,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.9547977795400474e-05,
+      "loss": 0.8724,
+      "step": 61200
+    },
+    {
+      "epoch": 8.042279411764707,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.9415807560137456e-05,
+      "loss": 0.9122,
+      "step": 61250
+    },
+    {
+      "epoch": 8.048844537815127,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.928363732487444e-05,
+      "loss": 0.8903,
+      "step": 61300
+    },
+    {
+      "epoch": 8.055409663865547,
+      "grad_norm": 3.25,
+      "learning_rate": 3.915146708961142e-05,
+      "loss": 0.8419,
+      "step": 61350
+    },
+    {
+      "epoch": 8.061974789915967,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.90192968543484e-05,
+      "loss": 0.8467,
+      "step": 61400
+    },
+    {
+      "epoch": 8.068539915966387,
+      "grad_norm": 3.34375,
+      "learning_rate": 3.8887126619085383e-05,
+      "loss": 0.8947,
+      "step": 61450
+    },
+    {
+      "epoch": 8.075105042016807,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.8754956383822365e-05,
+      "loss": 0.9266,
+      "step": 61500
+    },
+    {
+      "epoch": 8.081670168067227,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.862278614855935e-05,
+      "loss": 0.837,
+      "step": 61550
+    },
+    {
+      "epoch": 8.088235294117647,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.849061591329633e-05,
+      "loss": 0.8598,
+      "step": 61600
+    },
+    {
+      "epoch": 8.094800420168067,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.835844567803331e-05,
+      "loss": 0.8766,
+      "step": 61650
+    },
+    {
+      "epoch": 8.101365546218487,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.8226275442770287e-05,
+      "loss": 0.8423,
+      "step": 61700
+    },
+    {
+      "epoch": 8.107930672268907,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.809410520750727e-05,
+      "loss": 0.8809,
+      "step": 61750
+    },
+    {
+      "epoch": 8.114495798319327,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.796193497224425e-05,
+      "loss": 0.8501,
+      "step": 61800
+    },
+    {
+      "epoch": 8.121060924369749,
+      "grad_norm": 4.3125,
+      "learning_rate": 3.782976473698123e-05,
+      "loss": 0.8507,
+      "step": 61850
+    },
+    {
+      "epoch": 8.127626050420169,
+      "grad_norm": 3.40625,
+      "learning_rate": 3.7697594501718214e-05,
+      "loss": 0.8456,
+      "step": 61900
+    },
+    {
+      "epoch": 8.134191176470589,
+      "grad_norm": 3.125,
+      "learning_rate": 3.7565424266455196e-05,
+      "loss": 0.8371,
+      "step": 61950
+    },
+    {
+      "epoch": 8.140756302521009,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.743325403119218e-05,
+      "loss": 0.893,
+      "step": 62000
+    },
+    {
+      "epoch": 8.147321428571429,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.730108379592916e-05,
+      "loss": 0.8895,
+      "step": 62050
+    },
+    {
+      "epoch": 8.153886554621849,
+      "grad_norm": 3.65625,
+      "learning_rate": 3.716891356066614e-05,
+      "loss": 0.8391,
+      "step": 62100
+    },
+    {
+      "epoch": 8.160451680672269,
+      "grad_norm": 3.671875,
+      "learning_rate": 3.7036743325403124e-05,
+      "loss": 0.862,
+      "step": 62150
+    },
+    {
+      "epoch": 8.167016806722689,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.6904573090140106e-05,
+      "loss": 0.8852,
+      "step": 62200
+    },
+    {
+      "epoch": 8.173581932773109,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.677240285487709e-05,
+      "loss": 0.8824,
+      "step": 62250
+    },
+    {
+      "epoch": 8.180147058823529,
+      "grad_norm": 3.46875,
+      "learning_rate": 3.6640232619614063e-05,
+      "loss": 0.8212,
+      "step": 62300
+    },
+    {
+      "epoch": 8.186712184873949,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.6508062384351045e-05,
+      "loss": 0.9166,
+      "step": 62350
+    },
+    {
+      "epoch": 8.193277310924369,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.637589214908803e-05,
+      "loss": 0.8318,
+      "step": 62400
+    },
+    {
+      "epoch": 8.199842436974789,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.624372191382501e-05,
+      "loss": 0.879,
+      "step": 62450
+    },
+    {
+      "epoch": 8.206407563025211,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.611155167856199e-05,
+      "loss": 0.8354,
+      "step": 62500
+    },
+    {
+      "epoch": 8.212972689075631,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.597938144329897e-05,
+      "loss": 0.8263,
+      "step": 62550
+    },
+    {
+      "epoch": 8.219537815126051,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.5847211208035955e-05,
+      "loss": 0.9078,
+      "step": 62600
+    },
+    {
+      "epoch": 8.226102941176471,
+      "grad_norm": 4.03125,
+      "learning_rate": 3.571504097277294e-05,
+      "loss": 0.8386,
+      "step": 62650
+    },
+    {
+      "epoch": 8.232668067226891,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.558287073750991e-05,
+      "loss": 0.8261,
+      "step": 62700
+    },
+    {
+      "epoch": 8.239233193277311,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.5450700502246894e-05,
+      "loss": 0.8744,
+      "step": 62750
+    },
+    {
+      "epoch": 8.245798319327731,
+      "grad_norm": 4.40625,
+      "learning_rate": 3.5318530266983876e-05,
+      "loss": 0.9089,
+      "step": 62800
+    },
+    {
+      "epoch": 8.252363445378151,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.518636003172086e-05,
+      "loss": 0.9168,
+      "step": 62850
+    },
+    {
+      "epoch": 8.258928571428571,
+      "grad_norm": 3.484375,
+      "learning_rate": 3.5054189796457834e-05,
+      "loss": 0.8495,
+      "step": 62900
+    },
+    {
+      "epoch": 8.265493697478991,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.4922019561194816e-05,
+      "loss": 0.8767,
+      "step": 62950
+    },
+    {
+      "epoch": 8.272058823529411,
+      "grad_norm": 4.9375,
+      "learning_rate": 3.47898493259318e-05,
+      "loss": 0.9055,
+      "step": 63000
+    },
+    {
+      "epoch": 8.278623949579831,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.465767909066878e-05,
+      "loss": 0.9432,
+      "step": 63050
+    },
+    {
+      "epoch": 8.285189075630251,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.452550885540576e-05,
+      "loss": 0.9013,
+      "step": 63100
+    },
+    {
+      "epoch": 8.291754201680673,
+      "grad_norm": 3.1875,
+      "learning_rate": 3.4393338620142743e-05,
+      "loss": 0.8496,
+      "step": 63150
+    },
+    {
+      "epoch": 8.298319327731093,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.4261168384879725e-05,
+      "loss": 0.8496,
+      "step": 63200
+    },
+    {
+      "epoch": 8.304884453781513,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.412899814961671e-05,
+      "loss": 0.8911,
+      "step": 63250
+    },
+    {
+      "epoch": 8.311449579831933,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.399682791435369e-05,
+      "loss": 0.8891,
+      "step": 63300
+    },
+    {
+      "epoch": 8.318014705882353,
+      "grad_norm": 3.78125,
+      "learning_rate": 3.386465767909067e-05,
+      "loss": 0.8304,
+      "step": 63350
+    },
+    {
+      "epoch": 8.324579831932773,
+      "grad_norm": 4.375,
+      "learning_rate": 3.373248744382765e-05,
+      "loss": 0.8785,
+      "step": 63400
+    },
+    {
+      "epoch": 8.331144957983193,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.3600317208564635e-05,
+      "loss": 0.8848,
+      "step": 63450
+    },
+    {
+      "epoch": 8.337710084033613,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.346814697330161e-05,
+      "loss": 0.9199,
+      "step": 63500
+    },
+    {
+      "epoch": 8.344275210084033,
+      "grad_norm": 3.5625,
+      "learning_rate": 3.333597673803859e-05,
+      "loss": 0.8831,
+      "step": 63550
+    },
+    {
+      "epoch": 8.350840336134453,
+      "grad_norm": 3.9375,
+      "learning_rate": 3.3203806502775574e-05,
+      "loss": 0.9287,
+      "step": 63600
+    },
+    {
+      "epoch": 8.357405462184873,
+      "grad_norm": 3.75,
+      "learning_rate": 3.3071636267512556e-05,
+      "loss": 0.9415,
+      "step": 63650
+    },
+    {
+      "epoch": 8.363970588235293,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.293946603224954e-05,
+      "loss": 0.8426,
+      "step": 63700
+    },
+    {
+      "epoch": 8.370535714285714,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.280729579698652e-05,
+      "loss": 0.8968,
+      "step": 63750
+    },
+    {
+      "epoch": 8.377100840336134,
+      "grad_norm": 3.734375,
+      "learning_rate": 3.26751255617235e-05,
+      "loss": 0.8862,
+      "step": 63800
+    },
+    {
+      "epoch": 8.383665966386555,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.2542955326460484e-05,
+      "loss": 0.8493,
+      "step": 63850
+    },
+    {
+      "epoch": 8.390231092436975,
+      "grad_norm": 2.875,
+      "learning_rate": 3.2410785091197466e-05,
+      "loss": 0.919,
+      "step": 63900
+    },
+    {
+      "epoch": 8.396796218487395,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.227861485593445e-05,
+      "loss": 0.8918,
+      "step": 63950
+    },
+    {
+      "epoch": 8.403361344537815,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.214644462067143e-05,
+      "loss": 0.8576,
+      "step": 64000
+    },
+    {
+      "epoch": 8.409926470588236,
+      "grad_norm": 3.6875,
+      "learning_rate": 3.2014274385408405e-05,
+      "loss": 0.8066,
+      "step": 64050
+    },
+    {
+      "epoch": 8.416491596638656,
+      "grad_norm": 3.4375,
+      "learning_rate": 3.188210415014539e-05,
+      "loss": 0.8558,
+      "step": 64100
+    },
+    {
+      "epoch": 8.423056722689076,
+      "grad_norm": 3.625,
+      "learning_rate": 3.174993391488237e-05,
+      "loss": 0.8848,
+      "step": 64150
+    },
+    {
+      "epoch": 8.429621848739496,
+      "grad_norm": 4.0625,
+      "learning_rate": 3.161776367961935e-05,
+      "loss": 0.8849,
+      "step": 64200
+    },
+    {
+      "epoch": 8.436186974789916,
+      "grad_norm": 3.640625,
+      "learning_rate": 3.148559344435633e-05,
+      "loss": 0.9291,
+      "step": 64250
+    },
+    {
+      "epoch": 8.442752100840336,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.1353423209093315e-05,
+      "loss": 0.8496,
+      "step": 64300
+    },
+    {
+      "epoch": 8.449317226890756,
+      "grad_norm": 3.859375,
+      "learning_rate": 3.12212529738303e-05,
+      "loss": 0.831,
+      "step": 64350
+    },
+    {
+      "epoch": 8.455882352941176,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.108908273856728e-05,
+      "loss": 0.8968,
+      "step": 64400
+    },
+    {
+      "epoch": 8.462447478991596,
+      "grad_norm": 2.75,
+      "learning_rate": 3.095691250330426e-05,
+      "loss": 0.8889,
+      "step": 64450
+    },
+    {
+      "epoch": 8.469012605042018,
+      "grad_norm": 3.0,
+      "learning_rate": 3.082474226804124e-05,
+      "loss": 0.942,
+      "step": 64500
+    },
+    {
+      "epoch": 8.475577731092438,
+      "grad_norm": 5.03125,
+      "learning_rate": 3.0692572032778225e-05,
+      "loss": 0.8368,
+      "step": 64550
+    },
+    {
+      "epoch": 8.482142857142858,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.05604017975152e-05,
+      "loss": 0.8742,
+      "step": 64600
+    },
+    {
+      "epoch": 8.488707983193278,
+      "grad_norm": 3.921875,
+      "learning_rate": 3.042823156225218e-05,
+      "loss": 0.879,
+      "step": 64650
+    },
+    {
+      "epoch": 8.495273109243698,
+      "grad_norm": 2.75,
+      "learning_rate": 3.029606132698916e-05,
+      "loss": 0.8647,
+      "step": 64700
+    },
+    {
+      "epoch": 8.501838235294118,
+      "grad_norm": 3.265625,
+      "learning_rate": 3.0163891091726143e-05,
+      "loss": 0.846,
+      "step": 64750
+    },
+    {
+      "epoch": 8.508403361344538,
+      "grad_norm": 3.375,
+      "learning_rate": 3.0031720856463125e-05,
+      "loss": 0.9151,
+      "step": 64800
+    },
+    {
+      "epoch": 8.514968487394958,
+      "grad_norm": 3.625,
+      "learning_rate": 2.9899550621200107e-05,
+      "loss": 0.8529,
+      "step": 64850
+    },
+    {
+      "epoch": 8.521533613445378,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.976738038593709e-05,
+      "loss": 0.9216,
+      "step": 64900
+    },
+    {
+      "epoch": 8.528098739495798,
+      "grad_norm": 3.375,
+      "learning_rate": 2.963521015067407e-05,
+      "loss": 0.8465,
+      "step": 64950
+    },
+    {
+      "epoch": 8.534663865546218,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9503039915411053e-05,
+      "loss": 0.94,
+      "step": 65000
+    },
+    {
+      "epoch": 8.541228991596638,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.9370869680148035e-05,
+      "loss": 0.8756,
+      "step": 65050
+    },
+    {
+      "epoch": 8.547794117647058,
+      "grad_norm": 3.796875,
+      "learning_rate": 2.9238699444885013e-05,
+      "loss": 0.9175,
+      "step": 65100
+    },
+    {
+      "epoch": 8.55435924369748,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.9106529209621995e-05,
+      "loss": 0.9387,
+      "step": 65150
+    },
+    {
+      "epoch": 8.5609243697479,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.8974358974358977e-05,
+      "loss": 0.8691,
+      "step": 65200
+    },
+    {
+      "epoch": 8.56748949579832,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8842188739095956e-05,
+      "loss": 0.8827,
+      "step": 65250
+    },
+    {
+      "epoch": 8.57405462184874,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8710018503832938e-05,
+      "loss": 0.889,
+      "step": 65300
+    },
+    {
+      "epoch": 8.58061974789916,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.8577848268569916e-05,
+      "loss": 0.9372,
+      "step": 65350
+    },
+    {
+      "epoch": 8.58718487394958,
+      "grad_norm": 4.46875,
+      "learning_rate": 2.84456780333069e-05,
+      "loss": 0.8589,
+      "step": 65400
+    },
+    {
+      "epoch": 8.59375,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.831350779804388e-05,
+      "loss": 0.9059,
+      "step": 65450
+    },
+    {
+      "epoch": 8.60031512605042,
+      "grad_norm": 3.25,
+      "learning_rate": 2.8181337562780862e-05,
+      "loss": 0.86,
+      "step": 65500
+    },
+    {
+      "epoch": 8.60688025210084,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.8049167327517844e-05,
+      "loss": 0.9566,
+      "step": 65550
+    },
+    {
+      "epoch": 8.61344537815126,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.7916997092254826e-05,
+      "loss": 0.868,
+      "step": 65600
+    },
+    {
+      "epoch": 8.62001050420168,
+      "grad_norm": 3.375,
+      "learning_rate": 2.7784826856991808e-05,
+      "loss": 0.9989,
+      "step": 65650
+    },
+    {
+      "epoch": 8.6265756302521,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.765265662172879e-05,
+      "loss": 0.9477,
+      "step": 65700
+    },
+    {
+      "epoch": 8.63314075630252,
+      "grad_norm": 4.28125,
+      "learning_rate": 2.7520486386465772e-05,
+      "loss": 0.8854,
+      "step": 65750
+    },
+    {
+      "epoch": 8.639705882352942,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.7388316151202754e-05,
+      "loss": 0.8574,
+      "step": 65800
+    },
+    {
+      "epoch": 8.646271008403362,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.725614591593973e-05,
+      "loss": 0.8614,
+      "step": 65850
+    },
+    {
+      "epoch": 8.652836134453782,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.712397568067671e-05,
+      "loss": 0.8629,
+      "step": 65900
+    },
+    {
+      "epoch": 8.659401260504202,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.6991805445413693e-05,
+      "loss": 0.9085,
+      "step": 65950
+    },
+    {
+      "epoch": 8.665966386554622,
+      "grad_norm": 4.40625,
+      "learning_rate": 2.6859635210150675e-05,
+      "loss": 0.9446,
+      "step": 66000
+    },
+    {
+      "epoch": 8.672531512605042,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.6727464974887657e-05,
+      "loss": 0.8835,
+      "step": 66050
+    },
+    {
+      "epoch": 8.679096638655462,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.659529473962464e-05,
+      "loss": 0.9335,
+      "step": 66100
+    },
+    {
+      "epoch": 8.685661764705882,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.6463124504361618e-05,
+      "loss": 0.9176,
+      "step": 66150
+    },
+    {
+      "epoch": 8.692226890756302,
+      "grad_norm": 3.25,
+      "learning_rate": 2.63309542690986e-05,
+      "loss": 0.866,
+      "step": 66200
+    },
+    {
+      "epoch": 8.698792016806722,
+      "grad_norm": 3.703125,
+      "learning_rate": 2.619878403383558e-05,
+      "loss": 0.8773,
+      "step": 66250
+    },
+    {
+      "epoch": 8.705357142857142,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.6066613798572564e-05,
+      "loss": 0.8938,
+      "step": 66300
+    },
+    {
+      "epoch": 8.711922268907562,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.5934443563309546e-05,
+      "loss": 0.9595,
+      "step": 66350
+    },
+    {
+      "epoch": 8.718487394957982,
+      "grad_norm": 3.78125,
+      "learning_rate": 2.5802273328046528e-05,
+      "loss": 0.9481,
+      "step": 66400
+    },
+    {
+      "epoch": 8.725052521008404,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.5670103092783503e-05,
+      "loss": 0.853,
+      "step": 66450
+    },
+    {
+      "epoch": 8.731617647058824,
+      "grad_norm": 4.65625,
+      "learning_rate": 2.5537932857520485e-05,
+      "loss": 0.9192,
+      "step": 66500
+    },
+    {
+      "epoch": 8.738182773109244,
+      "grad_norm": 3.625,
+      "learning_rate": 2.5405762622257467e-05,
+      "loss": 0.8477,
+      "step": 66550
+    },
+    {
+      "epoch": 8.744747899159664,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.527359238699445e-05,
+      "loss": 0.9433,
+      "step": 66600
+    },
+    {
+      "epoch": 8.751313025210084,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.514142215173143e-05,
+      "loss": 0.8813,
+      "step": 66650
+    },
+    {
+      "epoch": 8.757878151260504,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.5009251916468413e-05,
+      "loss": 0.8904,
+      "step": 66700
+    },
+    {
+      "epoch": 8.764443277310924,
+      "grad_norm": 4.1875,
+      "learning_rate": 2.4877081681205395e-05,
+      "loss": 0.8608,
+      "step": 66750
+    },
+    {
+      "epoch": 8.771008403361344,
+      "grad_norm": 4.0,
+      "learning_rate": 2.4744911445942377e-05,
+      "loss": 0.9015,
+      "step": 66800
+    },
+    {
+      "epoch": 8.777573529411764,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.461274121067936e-05,
+      "loss": 0.8966,
+      "step": 66850
+    },
+    {
+      "epoch": 8.784138655462185,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.4480570975416337e-05,
+      "loss": 0.8921,
+      "step": 66900
+    },
+    {
+      "epoch": 8.790703781512605,
+      "grad_norm": 3.25,
+      "learning_rate": 2.434840074015332e-05,
+      "loss": 0.8334,
+      "step": 66950
+    },
+    {
+      "epoch": 8.797268907563025,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.42162305048903e-05,
+      "loss": 0.9139,
+      "step": 67000
+    },
+    {
+      "epoch": 8.803834033613445,
+      "grad_norm": 3.53125,
+      "learning_rate": 2.4084060269627283e-05,
+      "loss": 0.888,
+      "step": 67050
+    },
+    {
+      "epoch": 8.810399159663866,
+      "grad_norm": 3.515625,
+      "learning_rate": 2.395189003436426e-05,
+      "loss": 0.9097,
+      "step": 67100
+    },
+    {
+      "epoch": 8.816964285714286,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.3819719799101244e-05,
+      "loss": 0.8759,
+      "step": 67150
+    },
+    {
+      "epoch": 8.823529411764707,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.3687549563838222e-05,
+      "loss": 0.8913,
+      "step": 67200
+    },
+    {
+      "epoch": 8.830094537815127,
+      "grad_norm": 5.78125,
+      "learning_rate": 2.3555379328575204e-05,
+      "loss": 0.8569,
+      "step": 67250
+    },
+    {
+      "epoch": 8.836659663865547,
+      "grad_norm": 3.84375,
+      "learning_rate": 2.3423209093312186e-05,
+      "loss": 0.9132,
+      "step": 67300
+    },
+    {
+      "epoch": 8.843224789915967,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.3291038858049168e-05,
+      "loss": 0.8841,
+      "step": 67350
+    },
+    {
+      "epoch": 8.849789915966387,
+      "grad_norm": 3.625,
+      "learning_rate": 2.315886862278615e-05,
+      "loss": 0.8678,
+      "step": 67400
+    },
+    {
+      "epoch": 8.856355042016807,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.302669838752313e-05,
+      "loss": 0.8838,
+      "step": 67450
+    },
+    {
+      "epoch": 8.862920168067227,
+      "grad_norm": 3.5625,
+      "learning_rate": 2.289452815226011e-05,
+      "loss": 0.956,
+      "step": 67500
+    },
+    {
+      "epoch": 8.869485294117647,
+      "grad_norm": 3.3125,
+      "learning_rate": 2.2762357916997093e-05,
+      "loss": 0.9985,
+      "step": 67550
+    },
+    {
+      "epoch": 8.876050420168067,
+      "grad_norm": 4.0625,
+      "learning_rate": 2.2630187681734075e-05,
+      "loss": 0.9346,
+      "step": 67600
+    },
+    {
+      "epoch": 8.882615546218487,
+      "grad_norm": 4.78125,
+      "learning_rate": 2.2498017446471057e-05,
+      "loss": 0.9215,
+      "step": 67650
+    },
+    {
+      "epoch": 8.889180672268907,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.236584721120804e-05,
+      "loss": 0.9132,
+      "step": 67700
+    },
+    {
+      "epoch": 8.895745798319329,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.2233676975945017e-05,
+      "loss": 0.8704,
+      "step": 67750
+    },
+    {
+      "epoch": 8.902310924369749,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.2101506740682e-05,
+      "loss": 0.9117,
+      "step": 67800
+    },
+    {
+      "epoch": 8.908876050420169,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.196933650541898e-05,
+      "loss": 0.9104,
+      "step": 67850
+    },
+    {
+      "epoch": 8.915441176470589,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.1837166270155963e-05,
+      "loss": 0.922,
+      "step": 67900
+    },
+    {
+      "epoch": 8.922006302521009,
+      "grad_norm": 3.65625,
+      "learning_rate": 2.1704996034892945e-05,
+      "loss": 0.9083,
+      "step": 67950
+    },
+    {
+      "epoch": 8.928571428571429,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.1572825799629927e-05,
+      "loss": 0.8764,
+      "step": 68000
+    },
+    {
+      "epoch": 8.935136554621849,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.1440655564366906e-05,
+      "loss": 0.9047,
+      "step": 68050
+    },
+    {
+      "epoch": 8.941701680672269,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.1308485329103888e-05,
+      "loss": 0.9171,
+      "step": 68100
+    },
+    {
+      "epoch": 8.948266806722689,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.1176315093840866e-05,
+      "loss": 0.917,
+      "step": 68150
+    },
+    {
+      "epoch": 8.954831932773109,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.1044144858577848e-05,
+      "loss": 0.9438,
+      "step": 68200
+    },
+    {
+      "epoch": 8.961397058823529,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.091197462331483e-05,
+      "loss": 0.9877,
+      "step": 68250
+    },
+    {
+      "epoch": 8.967962184873949,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.0779804388051812e-05,
+      "loss": 0.8838,
+      "step": 68300
+    },
+    {
+      "epoch": 8.974527310924369,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.064763415278879e-05,
+      "loss": 0.8491,
+      "step": 68350
+    },
+    {
+      "epoch": 8.981092436974789,
+      "grad_norm": 4.09375,
+      "learning_rate": 2.0515463917525773e-05,
+      "loss": 0.8974,
+      "step": 68400
+    },
+    {
+      "epoch": 8.987657563025211,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.0383293682262755e-05,
+      "loss": 0.8826,
+      "step": 68450
+    },
+    {
+      "epoch": 8.994222689075631,
+      "grad_norm": 3.828125,
+      "learning_rate": 2.0251123446999737e-05,
+      "loss": 0.953,
+      "step": 68500
+    },
+    {
+      "epoch": 9.000787815126051,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.011895321173672e-05,
+      "loss": 0.9019,
+      "step": 68550
+    },
+    {
+      "epoch": 9.007352941176471,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.99867829764737e-05,
+      "loss": 0.8543,
+      "step": 68600
+    },
+    {
+      "epoch": 9.013918067226891,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.985461274121068e-05,
+      "loss": 0.8175,
+      "step": 68650
+    },
+    {
+      "epoch": 9.020483193277311,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.972244250594766e-05,
+      "loss": 0.7925,
+      "step": 68700
+    },
+    {
+      "epoch": 9.027048319327731,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9590272270684643e-05,
+      "loss": 0.8177,
+      "step": 68750
+    },
+    {
+      "epoch": 9.033613445378151,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.9458102035421625e-05,
+      "loss": 0.8351,
+      "step": 68800
+    },
+    {
+      "epoch": 9.040178571428571,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.9325931800158607e-05,
+      "loss": 0.7946,
+      "step": 68850
+    },
+    {
+      "epoch": 9.046743697478991,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.919376156489559e-05,
+      "loss": 0.836,
+      "step": 68900
+    },
+    {
+      "epoch": 9.053308823529411,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.9061591329632568e-05,
+      "loss": 0.8177,
+      "step": 68950
+    },
+    {
+      "epoch": 9.059873949579831,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.892942109436955e-05,
+      "loss": 0.8352,
+      "step": 69000
+    },
+    {
+      "epoch": 9.066439075630251,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.879725085910653e-05,
+      "loss": 0.8551,
+      "step": 69050
+    },
+    {
+      "epoch": 9.073004201680673,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.866508062384351e-05,
+      "loss": 0.807,
+      "step": 69100
+    },
+    {
+      "epoch": 9.079569327731093,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.8532910388580492e-05,
+      "loss": 0.8683,
+      "step": 69150
+    },
+    {
+      "epoch": 9.086134453781513,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.8400740153317474e-05,
+      "loss": 0.8108,
+      "step": 69200
+    },
+    {
+      "epoch": 9.092699579831933,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.8268569918054453e-05,
+      "loss": 0.8421,
+      "step": 69250
+    },
+    {
+      "epoch": 9.099264705882353,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.8136399682791435e-05,
+      "loss": 0.8363,
+      "step": 69300
+    },
+    {
+      "epoch": 9.105829831932773,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.8004229447528417e-05,
+      "loss": 0.8291,
+      "step": 69350
+    },
+    {
+      "epoch": 9.112394957983193,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.78720592122654e-05,
+      "loss": 0.8416,
+      "step": 69400
+    },
+    {
+      "epoch": 9.118960084033613,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.773988897700238e-05,
+      "loss": 0.8704,
+      "step": 69450
+    },
+    {
+      "epoch": 9.125525210084033,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.7607718741739362e-05,
+      "loss": 0.8508,
+      "step": 69500
+    },
+    {
+      "epoch": 9.132090336134453,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.747554850647634e-05,
+      "loss": 0.8543,
+      "step": 69550
+    },
+    {
+      "epoch": 9.138655462184873,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.7343378271213323e-05,
+      "loss": 0.8485,
+      "step": 69600
+    },
+    {
+      "epoch": 9.145220588235293,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.7211208035950305e-05,
+      "loss": 0.8638,
+      "step": 69650
+    },
+    {
+      "epoch": 9.151785714285714,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7079037800687287e-05,
+      "loss": 0.8383,
+      "step": 69700
+    },
+    {
+      "epoch": 9.158350840336134,
+      "grad_norm": 4.25,
+      "learning_rate": 1.694686756542427e-05,
+      "loss": 0.8408,
+      "step": 69750
+    },
+    {
+      "epoch": 9.164915966386555,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.681469733016125e-05,
+      "loss": 0.8616,
+      "step": 69800
+    },
+    {
+      "epoch": 9.171481092436975,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.668252709489823e-05,
+      "loss": 0.7984,
+      "step": 69850
+    },
+    {
+      "epoch": 9.178046218487395,
+      "grad_norm": 3.0,
+      "learning_rate": 1.655035685963521e-05,
+      "loss": 0.8265,
+      "step": 69900
+    },
+    {
+      "epoch": 9.184611344537815,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.6418186624372193e-05,
+      "loss": 0.826,
+      "step": 69950
+    },
+    {
+      "epoch": 9.191176470588236,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.6286016389109175e-05,
+      "loss": 0.8131,
+      "step": 70000
+    },
+    {
+      "epoch": 9.197741596638656,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 0.795,
+      "step": 70050
+    },
+    {
+      "epoch": 9.204306722689076,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.6021675918583136e-05,
+      "loss": 0.8655,
+      "step": 70100
+    },
+    {
+      "epoch": 9.210871848739496,
+      "grad_norm": 3.671875,
+      "learning_rate": 1.5889505683320115e-05,
+      "loss": 0.811,
+      "step": 70150
+    },
+    {
+      "epoch": 9.217436974789916,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.5757335448057097e-05,
+      "loss": 0.8409,
+      "step": 70200
+    },
+    {
+      "epoch": 9.224002100840336,
+      "grad_norm": 3.828125,
+      "learning_rate": 1.562516521279408e-05,
+      "loss": 0.8675,
+      "step": 70250
+    },
+    {
+      "epoch": 9.230567226890756,
+      "grad_norm": 3.984375,
+      "learning_rate": 1.549299497753106e-05,
+      "loss": 0.7883,
+      "step": 70300
+    },
+    {
+      "epoch": 9.237132352941176,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.5360824742268042e-05,
+      "loss": 0.8123,
+      "step": 70350
+    },
+    {
+      "epoch": 9.243697478991596,
+      "grad_norm": 3.5,
+      "learning_rate": 1.5228654507005021e-05,
+      "loss": 0.7888,
+      "step": 70400
+    },
+    {
+      "epoch": 9.250262605042018,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.5096484271742003e-05,
+      "loss": 0.8466,
+      "step": 70450
+    },
+    {
+      "epoch": 9.256827731092438,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4964314036478985e-05,
+      "loss": 0.8154,
+      "step": 70500
+    },
+    {
+      "epoch": 9.263392857142858,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.4832143801215967e-05,
+      "loss": 0.8042,
+      "step": 70550
+    },
+    {
+      "epoch": 9.269957983193278,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.4699973565952949e-05,
+      "loss": 0.8571,
+      "step": 70600
+    },
+    {
+      "epoch": 9.276523109243698,
+      "grad_norm": 3.75,
+      "learning_rate": 1.4567803330689931e-05,
+      "loss": 0.8109,
+      "step": 70650
+    },
+    {
+      "epoch": 9.283088235294118,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.443563309542691e-05,
+      "loss": 0.8494,
+      "step": 70700
+    },
+    {
+      "epoch": 9.289653361344538,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4303462860163891e-05,
+      "loss": 0.8483,
+      "step": 70750
+    },
+    {
+      "epoch": 9.296218487394958,
+      "grad_norm": 4.40625,
+      "learning_rate": 1.4171292624900873e-05,
+      "loss": 0.8561,
+      "step": 70800
+    },
+    {
+      "epoch": 9.302783613445378,
+      "grad_norm": 3.125,
+      "learning_rate": 1.4039122389637854e-05,
+      "loss": 0.8184,
+      "step": 70850
+    },
+    {
+      "epoch": 9.309348739495798,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.3906952154374836e-05,
+      "loss": 0.8726,
+      "step": 70900
+    },
+    {
+      "epoch": 9.315913865546218,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.3774781919111818e-05,
+      "loss": 0.8978,
+      "step": 70950
+    },
+    {
+      "epoch": 9.322478991596638,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.3642611683848796e-05,
+      "loss": 0.8926,
+      "step": 71000
+    },
+    {
+      "epoch": 9.329044117647058,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.3510441448585778e-05,
+      "loss": 0.8196,
+      "step": 71050
+    },
+    {
+      "epoch": 9.33560924369748,
+      "grad_norm": 3.515625,
+      "learning_rate": 1.337827121332276e-05,
+      "loss": 0.818,
+      "step": 71100
+    },
+    {
+      "epoch": 9.3421743697479,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.3246100978059742e-05,
+      "loss": 0.8168,
+      "step": 71150
+    },
+    {
+      "epoch": 9.34873949579832,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3113930742796724e-05,
+      "loss": 0.8109,
+      "step": 71200
+    },
+    {
+      "epoch": 9.35530462184874,
+      "grad_norm": 4.25,
+      "learning_rate": 1.2981760507533704e-05,
+      "loss": 0.8238,
+      "step": 71250
+    },
+    {
+      "epoch": 9.36186974789916,
+      "grad_norm": 3.75,
+      "learning_rate": 1.2849590272270685e-05,
+      "loss": 0.8574,
+      "step": 71300
+    },
+    {
+      "epoch": 9.36843487394958,
+      "grad_norm": 4.5,
+      "learning_rate": 1.2717420037007665e-05,
+      "loss": 0.8113,
+      "step": 71350
+    },
+    {
+      "epoch": 9.375,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.2585249801744647e-05,
+      "loss": 0.8139,
+      "step": 71400
+    },
+    {
+      "epoch": 9.38156512605042,
+      "grad_norm": 4.125,
+      "learning_rate": 1.2453079566481629e-05,
+      "loss": 0.8621,
+      "step": 71450
+    },
+    {
+      "epoch": 9.38813025210084,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.2320909331218611e-05,
+      "loss": 0.8902,
+      "step": 71500
+    },
+    {
+      "epoch": 9.39469537815126,
+      "grad_norm": 4.15625,
+      "learning_rate": 1.2188739095955591e-05,
+      "loss": 0.8511,
+      "step": 71550
+    },
+    {
+      "epoch": 9.40126050420168,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.2056568860692573e-05,
+      "loss": 0.8326,
+      "step": 71600
+    },
+    {
+      "epoch": 9.4078256302521,
+      "grad_norm": 4.03125,
+      "learning_rate": 1.1924398625429555e-05,
+      "loss": 0.8601,
+      "step": 71650
+    },
+    {
+      "epoch": 9.41439075630252,
+      "grad_norm": 4.25,
+      "learning_rate": 1.1792228390166535e-05,
+      "loss": 0.8633,
+      "step": 71700
+    },
+    {
+      "epoch": 9.420955882352942,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.1660058154903517e-05,
+      "loss": 0.7743,
+      "step": 71750
+    },
+    {
+      "epoch": 9.427521008403362,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.1527887919640498e-05,
+      "loss": 0.8187,
+      "step": 71800
+    },
+    {
+      "epoch": 9.434086134453782,
+      "grad_norm": 3.625,
+      "learning_rate": 1.1395717684377478e-05,
+      "loss": 0.849,
+      "step": 71850
+    },
+    {
+      "epoch": 9.440651260504202,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.126354744911446e-05,
+      "loss": 0.8291,
+      "step": 71900
+    },
+    {
+      "epoch": 9.447216386554622,
+      "grad_norm": 3.765625,
+      "learning_rate": 1.113137721385144e-05,
+      "loss": 0.8571,
+      "step": 71950
+    },
+    {
+      "epoch": 9.453781512605042,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.0999206978588422e-05,
+      "loss": 0.7721,
+      "step": 72000
+    },
+    {
+      "epoch": 9.460346638655462,
+      "grad_norm": 4.1875,
+      "learning_rate": 1.0867036743325404e-05,
+      "loss": 0.8919,
+      "step": 72050
+    },
+    {
+      "epoch": 9.466911764705882,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.0734866508062384e-05,
+      "loss": 0.837,
+      "step": 72100
+    },
+    {
+      "epoch": 9.473476890756302,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0602696272799366e-05,
+      "loss": 0.8425,
+      "step": 72150
+    },
+    {
+      "epoch": 9.480042016806722,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.0470526037536348e-05,
+      "loss": 0.8074,
+      "step": 72200
+    },
+    {
+      "epoch": 9.486607142857142,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.0338355802273329e-05,
+      "loss": 0.8214,
+      "step": 72250
+    },
+    {
+      "epoch": 9.493172268907562,
+      "grad_norm": 2.875,
+      "learning_rate": 1.0206185567010309e-05,
+      "loss": 0.8287,
+      "step": 72300
+    },
+    {
+      "epoch": 9.499737394957982,
+      "grad_norm": 3.875,
+      "learning_rate": 1.007401533174729e-05,
+      "loss": 0.8272,
+      "step": 72350
+    },
+    {
+      "epoch": 9.506302521008404,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.941845096484271e-06,
+      "loss": 0.8197,
+      "step": 72400
+    },
+    {
+      "epoch": 9.512867647058824,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.809674861221253e-06,
+      "loss": 0.8158,
+      "step": 72450
+    },
+    {
+      "epoch": 9.519432773109244,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.677504625958235e-06,
+      "loss": 0.8184,
+      "step": 72500
+    },
+    {
+      "epoch": 9.525997899159664,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.545334390695215e-06,
+      "loss": 0.8088,
+      "step": 72550
+    },
+    {
+      "epoch": 9.532563025210084,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.413164155432197e-06,
+      "loss": 0.7954,
+      "step": 72600
+    },
+    {
+      "epoch": 9.539128151260504,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.28099392016918e-06,
+      "loss": 0.813,
+      "step": 72650
+    },
+    {
+      "epoch": 9.545693277310924,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.14882368490616e-06,
+      "loss": 0.8708,
+      "step": 72700
+    },
+    {
+      "epoch": 9.552258403361344,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.01665344964314e-06,
+      "loss": 0.8545,
+      "step": 72750
+    },
+    {
+      "epoch": 9.558823529411764,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.884483214380122e-06,
+      "loss": 0.8873,
+      "step": 72800
+    },
+    {
+      "epoch": 9.565388655462185,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.752312979117102e-06,
+      "loss": 0.8071,
+      "step": 72850
+    },
+    {
+      "epoch": 9.571953781512605,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.620142743854084e-06,
+      "loss": 0.8507,
+      "step": 72900
+    },
+    {
+      "epoch": 9.578518907563025,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.487972508591066e-06,
+      "loss": 0.8495,
+      "step": 72950
+    },
+    {
+      "epoch": 9.585084033613445,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.355802273328046e-06,
+      "loss": 0.9413,
+      "step": 73000
+    },
+    {
+      "epoch": 9.591649159663866,
+      "grad_norm": 4.65625,
+      "learning_rate": 8.223632038065028e-06,
+      "loss": 0.8014,
+      "step": 73050
+    },
+    {
+      "epoch": 9.598214285714286,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.09146180280201e-06,
+      "loss": 0.8129,
+      "step": 73100
+    },
+    {
+      "epoch": 9.604779411764707,
+      "grad_norm": 4.03125,
+      "learning_rate": 7.95929156753899e-06,
+      "loss": 0.8201,
+      "step": 73150
+    },
+    {
+      "epoch": 9.611344537815127,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.827121332275973e-06,
+      "loss": 0.868,
+      "step": 73200
+    },
+    {
+      "epoch": 9.617909663865547,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.694951097012953e-06,
+      "loss": 0.8526,
+      "step": 73250
+    },
+    {
+      "epoch": 9.624474789915967,
+      "grad_norm": 2.875,
+      "learning_rate": 7.562780861749934e-06,
+      "loss": 0.876,
+      "step": 73300
+    },
+    {
+      "epoch": 9.631039915966387,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.430610626486916e-06,
+      "loss": 0.8259,
+      "step": 73350
+    },
+    {
+      "epoch": 9.637605042016807,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.298440391223897e-06,
+      "loss": 0.8365,
+      "step": 73400
+    },
+    {
+      "epoch": 9.644170168067227,
+      "grad_norm": 3.859375,
+      "learning_rate": 7.166270155960877e-06,
+      "loss": 0.8087,
+      "step": 73450
+    },
+    {
+      "epoch": 9.650735294117647,
+      "grad_norm": 4.0625,
+      "learning_rate": 7.034099920697859e-06,
+      "loss": 0.8416,
+      "step": 73500
+    },
+    {
+      "epoch": 9.657300420168067,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.901929685434841e-06,
+      "loss": 0.8645,
+      "step": 73550
+    },
+    {
+      "epoch": 9.663865546218487,
+      "grad_norm": 3.984375,
+      "learning_rate": 6.7697594501718215e-06,
+      "loss": 0.8237,
+      "step": 73600
+    },
+    {
+      "epoch": 9.670430672268907,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.637589214908803e-06,
+      "loss": 0.8426,
+      "step": 73650
+    },
+    {
+      "epoch": 9.676995798319329,
+      "grad_norm": 3.359375,
+      "learning_rate": 6.505418979645785e-06,
+      "loss": 0.8183,
+      "step": 73700
+    },
+    {
+      "epoch": 9.683560924369749,
+      "grad_norm": 3.5,
+      "learning_rate": 6.373248744382765e-06,
+      "loss": 0.8546,
+      "step": 73750
+    },
+    {
+      "epoch": 9.690126050420169,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.241078509119747e-06,
+      "loss": 0.8348,
+      "step": 73800
+    },
+    {
+      "epoch": 9.696691176470589,
+      "grad_norm": 3.625,
+      "learning_rate": 6.108908273856727e-06,
+      "loss": 0.7979,
+      "step": 73850
+    },
+    {
+      "epoch": 9.703256302521009,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.976738038593709e-06,
+      "loss": 0.8129,
+      "step": 73900
+    },
+    {
+      "epoch": 9.709821428571429,
+      "grad_norm": 3.90625,
+      "learning_rate": 5.84456780333069e-06,
+      "loss": 0.8164,
+      "step": 73950
+    },
+    {
+      "epoch": 9.716386554621849,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.712397568067671e-06,
+      "loss": 0.8302,
+      "step": 74000
+    },
+    {
+      "epoch": 9.722951680672269,
+      "grad_norm": 3.0,
+      "learning_rate": 5.580227332804653e-06,
+      "loss": 0.826,
+      "step": 74050
+    },
+    {
+      "epoch": 9.729516806722689,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.448057097541634e-06,
+      "loss": 0.8459,
+      "step": 74100
+    },
+    {
+      "epoch": 9.736081932773109,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.315886862278615e-06,
+      "loss": 0.8814,
+      "step": 74150
+    },
+    {
+      "epoch": 9.742647058823529,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.183716627015597e-06,
+      "loss": 0.8289,
+      "step": 74200
+    },
+    {
+      "epoch": 9.749212184873949,
+      "grad_norm": 6.75,
+      "learning_rate": 5.051546391752578e-06,
+      "loss": 0.8271,
+      "step": 74250
+    },
+    {
+      "epoch": 9.755777310924369,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.919376156489559e-06,
+      "loss": 0.7973,
+      "step": 74300
+    },
+    {
+      "epoch": 9.762342436974789,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.78720592122654e-06,
+      "loss": 0.9321,
+      "step": 74350
+    },
+    {
+      "epoch": 9.768907563025211,
+      "grad_norm": 2.875,
+      "learning_rate": 4.655035685963521e-06,
+      "loss": 0.8266,
+      "step": 74400
+    },
+    {
+      "epoch": 9.775472689075631,
+      "grad_norm": 3.6875,
+      "learning_rate": 4.522865450700502e-06,
+      "loss": 0.8386,
+      "step": 74450
+    },
+    {
+      "epoch": 9.782037815126051,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.3906952154374835e-06,
+      "loss": 0.8346,
+      "step": 74500
+    },
+    {
+      "epoch": 9.788602941176471,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.2585249801744654e-06,
+      "loss": 0.87,
+      "step": 74550
+    },
+    {
+      "epoch": 9.795168067226891,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.126354744911446e-06,
+      "loss": 0.7965,
+      "step": 74600
+    },
+    {
+      "epoch": 9.801733193277311,
+      "grad_norm": 4.375,
+      "learning_rate": 3.994184509648427e-06,
+      "loss": 0.8282,
+      "step": 74650
+    },
+    {
+      "epoch": 9.808298319327731,
+      "grad_norm": 3.375,
+      "learning_rate": 3.862014274385409e-06,
+      "loss": 0.8842,
+      "step": 74700
+    },
+    {
+      "epoch": 9.814863445378151,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.72984403912239e-06,
+      "loss": 0.8473,
+      "step": 74750
+    },
+    {
+      "epoch": 9.821428571428571,
+      "grad_norm": 3.84375,
+      "learning_rate": 3.5976738038593706e-06,
+      "loss": 0.8195,
+      "step": 74800
+    },
+    {
+      "epoch": 9.827993697478991,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.4655035685963526e-06,
+      "loss": 0.8491,
+      "step": 74850
+    },
+    {
+      "epoch": 9.834558823529411,
+      "grad_norm": 3.703125,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.8452,
+      "step": 74900
+    },
+    {
+      "epoch": 9.841123949579831,
+      "grad_norm": 4.46875,
+      "learning_rate": 3.2011630980703144e-06,
+      "loss": 0.8421,
+      "step": 74950
+    },
+    {
+      "epoch": 9.847689075630251,
+      "grad_norm": 3.609375,
+      "learning_rate": 3.068992862807296e-06,
+      "loss": 0.8377,
+      "step": 75000
+    },
+    {
+      "epoch": 9.854254201680671,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.936822627544277e-06,
+      "loss": 0.8778,
+      "step": 75050
+    },
+    {
+      "epoch": 9.860819327731093,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.8046523922812587e-06,
+      "loss": 0.8085,
+      "step": 75100
+    },
+    {
+      "epoch": 9.867384453781513,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.6724821570182394e-06,
+      "loss": 0.8498,
+      "step": 75150
+    },
+    {
+      "epoch": 9.873949579831933,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.540311921755221e-06,
+      "loss": 0.8532,
+      "step": 75200
+    },
+    {
+      "epoch": 9.880514705882353,
+      "grad_norm": 3.453125,
+      "learning_rate": 2.408141686492202e-06,
+      "loss": 0.8488,
+      "step": 75250
+    },
+    {
+      "epoch": 9.887079831932773,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.275971451229183e-06,
+      "loss": 0.8264,
+      "step": 75300
+    },
+    {
+      "epoch": 9.893644957983193,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.1438012159661647e-06,
+      "loss": 0.7632,
+      "step": 75350
+    },
+    {
+      "epoch": 9.900210084033613,
+      "grad_norm": 2.25,
+      "learning_rate": 2.011630980703146e-06,
+      "loss": 0.8703,
+      "step": 75400
+    },
+    {
+      "epoch": 9.906775210084033,
+      "grad_norm": 3.640625,
+      "learning_rate": 1.879460745440127e-06,
+      "loss": 0.8017,
+      "step": 75450
+    },
+    {
+      "epoch": 9.913340336134453,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7472905101771083e-06,
+      "loss": 0.8494,
+      "step": 75500
+    },
+    {
+      "epoch": 9.919905462184873,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.6151202749140896e-06,
+      "loss": 0.7825,
+      "step": 75550
+    },
+    {
+      "epoch": 9.926470588235293,
+      "grad_norm": 3.25,
+      "learning_rate": 1.4829500396510706e-06,
+      "loss": 0.8645,
+      "step": 75600
+    },
+    {
+      "epoch": 9.933035714285714,
+      "grad_norm": 4.90625,
+      "learning_rate": 1.350779804388052e-06,
+      "loss": 0.8095,
+      "step": 75650
+    },
+    {
+      "epoch": 9.939600840336134,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.218609569125033e-06,
+      "loss": 0.8369,
+      "step": 75700
+    },
+    {
+      "epoch": 9.946165966386555,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.0864393338620144e-06,
+      "loss": 0.8546,
+      "step": 75750
+    },
+    {
+      "epoch": 9.952731092436975,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.542690985989957e-07,
+      "loss": 0.8088,
+      "step": 75800
+    },
+    {
+      "epoch": 9.959296218487395,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.220988633359768e-07,
+      "loss": 0.8517,
+      "step": 75850
+    },
+    {
+      "epoch": 9.965861344537815,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.89928628072958e-07,
+      "loss": 0.9301,
+      "step": 75900
+    },
+    {
+      "epoch": 9.972426470588236,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.577583928099393e-07,
+      "loss": 0.8099,
+      "step": 75950
+    },
+    {
+      "epoch": 9.978991596638656,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.255881575469204e-07,
+      "loss": 0.8555,
+      "step": 76000
+    },
+    {
+      "epoch": 9.985556722689076,
+      "grad_norm": 4.25,
+      "learning_rate": 2.934179222839017e-07,
+      "loss": 0.8108,
+      "step": 76050
+    },
+    {
+      "epoch": 9.992121848739496,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.612476870208829e-07,
+      "loss": 0.8263,
+      "step": 76100
+    },
+    {
+      "epoch": 9.998686974789916,
+      "grad_norm": 3.640625,
+      "learning_rate": 2.907745175786413e-08,
+      "loss": 0.8353,
+      "step": 76150
+    }
+  ],
+  "logging_steps": 50,
+  "max_steps": 76160,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "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": 3.554785504833884e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}